Upload folder using huggingface_hub

.gitattributes

@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+.venv/Lib/site-packages/81d243bd2c585b0f4821__mypyc.cp314-win_amd64.pyd filter=lfs diff=lfs merge=lfs -text

.github/agents/windows-deps-repair.agent.md

@@ -0,0 +1,33 @@
+---
+description: "Use when npm install or npm ci fails on Windows with EACCES, EPERM, path syntax errors, locked node_modules, or broken .bin cleanup. Automates safe dependency repair and build validation for Vite/React/TS projects."
+name: "Windows Deps Repair"
+tools: [read, edit, search, execute]
+user-invocable: true
+---
+You are a specialist for Windows dependency repair in JavaScript/TypeScript projects.
+Your mission is to stabilize installs and unblock build/typecheck fast.
+
+## Scope
+- Focus on Windows PowerShell/npm issues.
+- Prioritize deterministic fixes over manual one-off steps.
+- Use project scripts when available, then harden them if flaky.
+
+## Constraints
+- Do not suggest destructive global resets.
+- Do not require WSL unless user explicitly asks.
+- Do not stop at diagnosis. Apply fixes and validate.
+
+## Approach
+1. Reproduce the failure and capture exact stderr patterns.
+2. Detect known causes: EACCES/EPERM locks, stale .bin files, invalid shell quoting, wrong working directory, Unicode path issues.
+3. Implement or improve an automated repair script in the project.
+4. Add npm script entrypoints for one-command execution.
+5. Validate with `npm ci` or `npm install`, then `npx tsc --noEmit` and `npm run build`.
+6. Return concise root cause, changed files, and exact commands to rerun.
+
+## Output Format
+- Root cause
+- Files changed
+- Commands executed
+- Validation result
+- Optional next hardening step

.github/workflows/ci.yml

@@ -0,0 +1,32 @@
+name: CI
+
+on:
+  push:
+    branches: [main, master]
+  pull_request:
+    branches: [main, master]
+
+jobs:
+  build:
+    runs-on: ubuntu-latest
+
+    steps:
+      - uses: actions/checkout@v4
+
+      - name: Setup Node
+        uses: actions/setup-node@v4
+        with:
+          node-version: "20"
+          cache: npm
+
+      - name: Install dependencies
+        run: npm ci
+
+      - name: TypeScript check
+        run: npx tsc --noEmit
+
+      - name: Build
+        run: npm run build
+
+      - name: Lint
+        run: npm run lint

.gitignore

@@ -0,0 +1 @@
+node_modules

.mcp.json

@@ -0,0 +1,8 @@
+{
+  "mcpServers": {
+    "cortex-debug": {
+      "type": "sse",
+      "url": "http://localhost:7580/sse"
+    }
+  }
+}

.nvmrc

@@ -0,0 +1 @@
+20

.venv/.gitignore

@@ -0,0 +1 @@
+*

.venv/Lib/site-packages/81d243bd2c585b0f4821__mypyc.cp314-win_amd64.pyd

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ab64835b63093c31975df2692756c8008e7ec190e4aa86f15a713bcc62e1d432
+size 216576

.venv/Lib/site-packages/certifi-2026.5.20.dist-info/INSTALLER

@@ -0,0 +1 @@
+pip

.venv/Lib/site-packages/certifi-2026.5.20.dist-info/METADATA

@@ -0,0 +1,78 @@
+Metadata-Version: 2.4
+Name: certifi
+Version: 2026.5.20
+Summary: Python package for providing Mozilla's CA Bundle.
+Home-page: https://github.com/certifi/python-certifi
+Author: Kenneth Reitz
+Author-email: me@kennethreitz.com
+License: MPL-2.0
+Project-URL: Source, https://github.com/certifi/python-certifi
+Classifier: Development Status :: 5 - Production/Stable
+Classifier: Intended Audience :: Developers
+Classifier: License :: OSI Approved :: Mozilla Public License 2.0 (MPL 2.0)
+Classifier: Natural Language :: English
+Classifier: Programming Language :: Python
+Classifier: Programming Language :: Python :: 3
+Classifier: Programming Language :: Python :: 3 :: Only
+Classifier: Programming Language :: Python :: 3.7
+Classifier: Programming Language :: Python :: 3.8
+Classifier: Programming Language :: Python :: 3.9
+Classifier: Programming Language :: Python :: 3.10
+Classifier: Programming Language :: Python :: 3.11
+Classifier: Programming Language :: Python :: 3.12
+Classifier: Programming Language :: Python :: 3.13
+Classifier: Programming Language :: Python :: 3.14
+Requires-Python: >=3.7
+License-File: LICENSE
+Dynamic: author
+Dynamic: author-email
+Dynamic: classifier
+Dynamic: description
+Dynamic: home-page
+Dynamic: license
+Dynamic: license-file
+Dynamic: project-url
+Dynamic: requires-python
+Dynamic: summary
+
+Certifi: Python SSL Certificates
+================================
+
+Certifi provides Mozilla's carefully curated collection of Root Certificates for
+validating the trustworthiness of SSL certificates while verifying the identity
+of TLS hosts. It has been extracted from the `Requests`_ project.
+
+Installation
+------------
+
+``certifi`` is available on PyPI. Simply install it with ``pip``::
+
+    $ pip install certifi
+
+Usage
+-----
+
+To reference the installed certificate authority (CA) bundle, you can use the
+built-in function::
+
+    >>> import certifi
+
+    >>> certifi.where()
+    '/usr/local/lib/python3.7/site-packages/certifi/cacert.pem'
+
+Or from the command line::
+
+    $ python -m certifi
+    /usr/local/lib/python3.7/site-packages/certifi/cacert.pem
+
+Enjoy!
+
+.. _`Requests`: https://requests.readthedocs.io/en/master/
+
+Addition/Removal of Certificates
+--------------------------------
+
+Certifi does not support any addition/removal or other modification of the
+CA trust store content. This project is intended to provide a reliable and
+highly portable root of trust to python deployments. Look to upstream projects
+for methods to use alternate trust.

.venv/Lib/site-packages/certifi-2026.5.20.dist-info/RECORD

@@ -0,0 +1,14 @@
+certifi-2026.5.20.dist-info/INSTALLER,sha256=zuuue4knoyJ-UwPPXg8fezS7VCrXJQrAP7zeNuwvFQg,4
+certifi-2026.5.20.dist-info/METADATA,sha256=HDiCa8lfd0TGlQFongDjoaLJOhWJiJN84y707Vwd8PY,2474
+certifi-2026.5.20.dist-info/RECORD,,
+certifi-2026.5.20.dist-info/WHEEL,sha256=aeYiig01lYGDzBgS8HxWXOg3uV61G9ijOsup-k9o1sk,91
+certifi-2026.5.20.dist-info/licenses/LICENSE,sha256=6TcW2mucDVpKHfYP5pWzcPBpVgPSH2-D8FPkLPwQyvc,989
+certifi-2026.5.20.dist-info/top_level.txt,sha256=KMu4vUCfsjLrkPbSNdgdekS-pVJzBAJFO__nI8NF6-U,8
+certifi/__init__.py,sha256=-8ljc7w7fyi5Kyi06c1_L0lG_kD_J0FvbUTa8mEpdzQ,94
+certifi/__main__.py,sha256=xBBoj905TUWBLRGANOcf7oi6e-3dMP4cEoG9OyMs11g,243
+certifi/__pycache__/__init__.cpython-314.pyc,,
+certifi/__pycache__/__main__.cpython-314.pyc,,
+certifi/__pycache__/core.cpython-314.pyc,,
+certifi/cacert.pem,sha256=fO9UDGrzvwo8fQBwvT_XquNPiqgFtAjRwnoByK6zBd8,236095
+certifi/core.py,sha256=XFXycndG5pf37ayeF8N32HUuDafsyhkVMbO4BAPWHa0,3394
+certifi/py.typed,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0

.venv/Lib/site-packages/certifi-2026.5.20.dist-info/WHEEL

@@ -0,0 +1,5 @@
+Wheel-Version: 1.0
+Generator: setuptools (82.0.1)
+Root-Is-Purelib: true
+Tag: py3-none-any
+

.venv/Lib/site-packages/certifi-2026.5.20.dist-info/licenses/LICENSE

@@ -0,0 +1,20 @@
+This package contains a modified version of ca-bundle.crt:
+
+ca-bundle.crt -- Bundle of CA Root Certificates
+
+This is a bundle of X.509 certificates of public Certificate Authorities
+(CA). These were automatically extracted from Mozilla's root certificates
+file (certdata.txt).  This file can be found in the mozilla source tree:
+https://hg.mozilla.org/mozilla-central/file/tip/security/nss/lib/ckfw/builtins/certdata.txt
+It contains the certificates in PEM format and therefore
+can be directly used with curl / libcurl / php_curl, or with
+an Apache+mod_ssl webserver for SSL client authentication.
+Just configure this file as the SSLCACertificateFile.#
+
+***** BEGIN LICENSE BLOCK *****
+This Source Code Form is subject to the terms of the Mozilla Public License,
+v. 2.0. If a copy of the MPL was not distributed with this file, You can obtain
+one at http://mozilla.org/MPL/2.0/.
+
+***** END LICENSE BLOCK *****
+@(#) $RCSfile: certdata.txt,v $ $Revision: 1.80 $ $Date: 2011/11/03 15:11:58 $

.venv/Lib/site-packages/certifi-2026.5.20.dist-info/top_level.txt

@@ -0,0 +1 @@
+certifi

.venv/Lib/site-packages/certifi/__init__.py

@@ -0,0 +1,4 @@
+from .core import contents, where
+
+__all__ = ["contents", "where"]
+__version__ = "2026.05.20"

.venv/Lib/site-packages/certifi/__main__.py

@@ -0,0 +1,12 @@
+import argparse
+
+from certifi import contents, where
+
+parser = argparse.ArgumentParser()
+parser.add_argument("-c", "--contents", action="store_true")
+args = parser.parse_args()
+
+if args.contents:
+    print(contents())
+else:
+    print(where())

.venv/Lib/site-packages/certifi/core.py

@@ -0,0 +1,83 @@
+"""
+certifi.py
+~~~~~~~~~~
+
+This module returns the installation location of cacert.pem or its contents.
+"""
+import sys
+import atexit
+
+def exit_cacert_ctx() -> None:
+    _CACERT_CTX.__exit__(None, None, None)  # type: ignore[union-attr]
+
+
+if sys.version_info >= (3, 11):
+
+    from importlib.resources import as_file, files
+
+    _CACERT_CTX = None
+    _CACERT_PATH = None
+
+    def where() -> str:
+        # This is slightly terrible, but we want to delay extracting the file
+        # in cases where we're inside of a zipimport situation until someone
+        # actually calls where(), but we don't want to re-extract the file
+        # on every call of where(), so we'll do it once then store it in a
+        # global variable.
+        global _CACERT_CTX
+        global _CACERT_PATH
+        if _CACERT_PATH is None:
+            # This is slightly janky, the importlib.resources API wants you to
+            # manage the cleanup of this file, so it doesn't actually return a
+            # path, it returns a context manager that will give you the path
+            # when you enter it and will do any cleanup when you leave it. In
+            # the common case of not needing a temporary file, it will just
+            # return the file system location and the __exit__() is a no-op.
+            #
+            # We also have to hold onto the actual context manager, because
+            # it will do the cleanup whenever it gets garbage collected, so
+            # we will also store that at the global level as well.
+            _CACERT_CTX = as_file(files("certifi").joinpath("cacert.pem"))
+            _CACERT_PATH = str(_CACERT_CTX.__enter__())
+            atexit.register(exit_cacert_ctx)
+
+        return _CACERT_PATH
+
+    def contents() -> str:
+        return files("certifi").joinpath("cacert.pem").read_text(encoding="ascii")
+
+else:
+
+    from importlib.resources import path as get_path, read_text
+
+    _CACERT_CTX = None
+    _CACERT_PATH = None
+
+    def where() -> str:
+        # This is slightly terrible, but we want to delay extracting the
+        # file in cases where we're inside of a zipimport situation until
+        # someone actually calls where(), but we don't want to re-extract
+        # the file on every call of where(), so we'll do it once then store
+        # it in a global variable.
+        global _CACERT_CTX
+        global _CACERT_PATH
+        if _CACERT_PATH is None:
+            # This is slightly janky, the importlib.resources API wants you
+            # to manage the cleanup of this file, so it doesn't actually
+            # return a path, it returns a context manager that will give
+            # you the path when you enter it and will do any cleanup when
+            # you leave it. In the common case of not needing a temporary
+            # file, it will just return the file system location and the
+            # __exit__() is a no-op.
+            #
+            # We also have to hold onto the actual context manager, because
+            # it will do the cleanup whenever it gets garbage collected, so
+            # we will also store that at the global level as well.
+            _CACERT_CTX = get_path("certifi", "cacert.pem")
+            _CACERT_PATH = str(_CACERT_CTX.__enter__())
+            atexit.register(exit_cacert_ctx)
+
+        return _CACERT_PATH
+
+    def contents() -> str:
+        return read_text("certifi", "cacert.pem", encoding="ascii")

.venv/Lib/site-packages/charset_normalizer-3.4.7.dist-info/INSTALLER

@@ -0,0 +1 @@
+pip

.venv/Lib/site-packages/charset_normalizer-3.4.7.dist-info/METADATA

@@ -0,0 +1,808 @@
+Metadata-Version: 2.4
+Name: charset-normalizer
+Version: 3.4.7
+Summary: The Real First Universal Charset Detector. Open, modern and actively maintained alternative to Chardet.
+Author-email: "Ahmed R. TAHRI" <tahri.ahmed@proton.me>
+Maintainer-email: "Ahmed R. TAHRI" <tahri.ahmed@proton.me>
+License: MIT
+Project-URL: Changelog, https://github.com/jawah/charset_normalizer/blob/master/CHANGELOG.md
+Project-URL: Documentation, https://charset-normalizer.readthedocs.io/
+Project-URL: Code, https://github.com/jawah/charset_normalizer
+Project-URL: Issue tracker, https://github.com/jawah/charset_normalizer/issues
+Keywords: encoding,charset,charset-detector,detector,normalization,unicode,chardet,detect
+Classifier: Development Status :: 5 - Production/Stable
+Classifier: Intended Audience :: Developers
+Classifier: Operating System :: OS Independent
+Classifier: Programming Language :: Python
+Classifier: Programming Language :: Python :: 3
+Classifier: Programming Language :: Python :: 3.7
+Classifier: Programming Language :: Python :: 3.8
+Classifier: Programming Language :: Python :: 3.9
+Classifier: Programming Language :: Python :: 3.10
+Classifier: Programming Language :: Python :: 3.11
+Classifier: Programming Language :: Python :: 3.12
+Classifier: Programming Language :: Python :: 3.13
+Classifier: Programming Language :: Python :: 3.14
+Classifier: Programming Language :: Python :: 3 :: Only
+Classifier: Programming Language :: Python :: Implementation :: CPython
+Classifier: Programming Language :: Python :: Implementation :: PyPy
+Classifier: Programming Language :: Python :: Free Threading :: 4 - Resilient
+Classifier: Topic :: Text Processing :: Linguistic
+Classifier: Topic :: Utilities
+Classifier: Typing :: Typed
+Requires-Python: >=3.7
+Description-Content-Type: text/markdown
+License-File: LICENSE
+Provides-Extra: unicode-backport
+Dynamic: license-file
+
+<h1 align="center">Charset Detection, for Everyone 👋</h1>
+
+<p align="center">
+  <sup>The Real First Universal Charset Detector</sup><br>
+  <a href="https://pypi.org/project/charset-normalizer">
+    <img src="https://img.shields.io/pypi/pyversions/charset_normalizer.svg?orange=blue" />
+  </a>
+  <a href="https://pepy.tech/project/charset-normalizer/">
+    <img alt="Download Count Total" src="https://static.pepy.tech/badge/charset-normalizer/month" />
+  </a>
+  <a href="https://bestpractices.coreinfrastructure.org/projects/7297">
+    <img src="https://bestpractices.coreinfrastructure.org/projects/7297/badge">
+  </a>
+</p>
+<p align="center">
+  <sup><i>Featured Packages</i></sup><br>
+  <a href="https://github.com/jawah/niquests">
+   <img alt="Static Badge" src="https://img.shields.io/badge/Niquests-Most_Advanced_HTTP_Client-cyan">
+  </a>
+  <a href="https://github.com/jawah/wassima">
+   <img alt="Static Badge" src="https://img.shields.io/badge/Wassima-Certifi_Replacement-cyan">
+  </a>
+</p>
+<p align="center">
+  <sup><i>In other language (unofficial port - by the community)</i></sup><br>
+  <a href="https://github.com/nickspring/charset-normalizer-rs">
+   <img alt="Static Badge" src="https://img.shields.io/badge/Rust-red">
+  </a>
+</p>
+
+> A library that helps you read text from an unknown charset encoding.<br /> Motivated by `chardet`,
+> I'm trying to resolve the issue by taking a new approach.
+> All IANA character set names for which the Python core library provides codecs are supported.
+> You can also register your own set of codecs, and yes, it would work as-is.
+
+<p align="center">
+  >>>>> <a href="https://charsetnormalizerweb.ousret.now.sh" target="_blank">👉 Try Me Online Now, Then Adopt Me 👈 </a> <<<<<
+</p>
+
+This project offers you an alternative to **Universal Charset Encoding Detector**, also known as **Chardet**.
+
+| Feature                                          | [Chardet](https://github.com/chardet/chardet) |                                       Charset Normalizer                                        | [cChardet](https://github.com/PyYoshi/cChardet) |
+|--------------------------------------------------|:---------------------------------------------:|:-----------------------------------------------------------------------------------------------:|:-----------------------------------------------:|
+| `Fast`                                           |                       ✅                       |                                                ✅                                                |                        ✅                        |
+| `Universal`[^1]                                  |                       ❌                       |                                                ✅                                                |                        ❌                        |
+| `Reliable` **without** distinguishable standards |                       ✅                       |                                                ✅                                                |                        ✅                        |
+| `Reliable` **with** distinguishable standards    |                       ✅                       |                                                ✅                                                |                        ✅                        |
+| `License`                                        |        _Disputed_[^2]<br>_restrictive_        |                                               MIT                                               |            MPL-1.1<br>_restrictive_             |
+| `Native Python`                                  |                       ✅                       |                                                ✅                                                |                        ❌                        |
+| `Detect spoken language`                         |                       ✅                       |                                                ✅                                                |                       N/A                       |
+| `UnicodeDecodeError Safety`                      |                       ✅                       |                                                ✅                                                |                        ❌                        |
+| `Whl Size (min)`                                 |                    500 kB                     |                                             150 kB                                              |                     ~200 kB                     |
+| `Supported Encoding`                             |                      99                       | [99](https://charset-normalizer.readthedocs.io/en/latest/user/support.html#supported-encodings) |                       40                        |
+| `Can register custom encoding`                   |                       ❌                       |                                                ✅                                                |                        ❌                        |
+
+<p align="center">
+<img src="https://i.imgflip.com/373iay.gif" alt="Reading Normalized Text" width="226"/><img src="https://media.tenor.com/images/c0180f70732a18b4965448d33adba3d0/tenor.gif" alt="Cat Reading Text" width="200"/>
+</p>
+
+[^1]: They are clearly using specific code for a specific encoding even if covering most of used one.
+[^2]: Chardet 7.0+ was relicensed from LGPL-2.1 to MIT following an AI-assisted rewrite. This relicensing is disputed on two independent grounds: **(a)** the original author [contests](https://github.com/chardet/chardet/issues/327) that the maintainer had the right to relicense, arguing the rewrite is a derivative work of the LGPL-licensed codebase since it was not a clean room implementation; **(b)** the copyright claim itself is [questionable](https://github.com/chardet/chardet/issues/334) given the code was primarily generated by an LLM, and AI-generated output may not be copyrightable under most jurisdictions. Either issue alone could undermine the MIT license. Beyond licensing, the rewrite raises questions about responsible use of AI in open source: key architectural ideas pioneered by charset-normalizer - notably decode-first validity filtering (our foundational approach since v1) and encoding pairwise similarity with the same algorithm and threshold — surfaced in chardet 7 without acknowledgment. The project also imported test files from charset-normalizer to train and benchmark against it, then claimed superior accuracy on those very files. Charset-normalizer has always been MIT-licensed, encoding-agnostic by design, and built on a verifiable human-authored history.
+
+## ⚡ Performance
+
+This package offer better performances (99th, and 95th) against Chardet. Here are some numbers.
+
+| Package                                           | Accuracy | Mean per file (ms) | File per sec (est) |
+|---------------------------------------------------|:--------:|:------------------:|:------------------:|
+| [chardet 7.1](https://github.com/chardet/chardet) |   89 %   |        3 ms        |    333 file/sec    |
+| charset-normalizer                                | **97 %** |        3 ms        |    333 file/sec    |
+
+| Package                                           | 99th percentile | 95th percentile | 50th percentile |
+|---------------------------------------------------|:---------------:|:---------------:|:---------------:|
+| [chardet 7.1](https://github.com/chardet/chardet) |      32 ms      |      17 ms      |     < 1 ms      |
+| charset-normalizer                                |      16 ms      |      10 ms      |      1 ms       |
+
+_updated as of March 2026 using CPython 3.12, Charset-Normalizer 3.4.6, and Chardet 7.1.0_
+
+~Chardet's performance on larger file (1MB+) are very poor. Expect huge difference on large payload.~ No longer the case since Chardet 7.0+
+
+> Stats are generated using 400+ files using default parameters. More details on used files, see GHA workflows.
+> And yes, these results might change at any time. The dataset can be updated to include more files.
+> The actual delays heavily depends on your CPU capabilities. The factors should remain the same.
+> Chardet claims on his documentation to have a greater accuracy than us based on the dataset they trained Chardet on(...)
+> Well, it's normal, the opposite would have been worrying. Whereas charset-normalizer don't train on anything, our solution
+> is based on a completely different algorithm, still heuristic through, it does not need weights across every encoding tables.
+
+## ✨ Installation
+
+Using pip:
+
+```sh
+pip install charset-normalizer -U
+```
+
+## 🚀 Basic Usage
+
+### CLI
+This package comes with a CLI.
+
+```
+usage: normalizer [-h] [-v] [-a] [-n] [-m] [-r] [-f] [-t THRESHOLD]
+                  file [file ...]
+
+The Real First Universal Charset Detector. Discover originating encoding used
+on text file. Normalize text to unicode.
+
+positional arguments:
+  files                 File(s) to be analysed
+
+optional arguments:
+  -h, --help            show this help message and exit
+  -v, --verbose         Display complementary information about file if any.
+                        Stdout will contain logs about the detection process.
+  -a, --with-alternative
+                        Output complementary possibilities if any. Top-level
+                        JSON WILL be a list.
+  -n, --normalize       Permit to normalize input file. If not set, program
+                        does not write anything.
+  -m, --minimal         Only output the charset detected to STDOUT. Disabling
+                        JSON output.
+  -r, --replace         Replace file when trying to normalize it instead of
+                        creating a new one.
+  -f, --force           Replace file without asking if you are sure, use this
+                        flag with caution.
+  -t THRESHOLD, --threshold THRESHOLD
+                        Define a custom maximum amount of chaos allowed in
+                        decoded content. 0. <= chaos <= 1.
+  --version             Show version information and exit.
+```
+
+```bash
+normalizer ./data/sample.1.fr.srt
+```
+
+or
+
+```bash
+python -m charset_normalizer ./data/sample.1.fr.srt
+```
+
+🎉 Since version 1.4.0 the CLI produce easily usable stdout result in JSON format.
+
+```json
+{
+    "path": "/home/default/projects/charset_normalizer/data/sample.1.fr.srt",
+    "encoding": "cp1252",
+    "encoding_aliases": [
+        "1252",
+        "windows_1252"
+    ],
+    "alternative_encodings": [
+        "cp1254",
+        "cp1256",
+        "cp1258",
+        "iso8859_14",
+        "iso8859_15",
+        "iso8859_16",
+        "iso8859_3",
+        "iso8859_9",
+        "latin_1",
+        "mbcs"
+    ],
+    "language": "French",
+    "alphabets": [
+        "Basic Latin",
+        "Latin-1 Supplement"
+    ],
+    "has_sig_or_bom": false,
+    "chaos": 0.149,
+    "coherence": 97.152,
+    "unicode_path": null,
+    "is_preferred": true
+}
+```
+
+### Python
+*Just print out normalized text*
+```python
+from charset_normalizer import from_path
+
+results = from_path('./my_subtitle.srt')
+
+print(str(results.best()))
+```
+
+*Upgrade your code without effort*
+```python
+from charset_normalizer import detect
+```
+
+The above code will behave the same as **chardet**. We ensure that we offer the best (reasonable) BC result possible.
+
+See the docs for advanced usage : [readthedocs.io](https://charset-normalizer.readthedocs.io/en/latest/)
+
+## 😇 Why
+
+When I started using Chardet, I noticed that it was not suited to my expectations, and I wanted to propose a
+reliable alternative using a completely different method. Also! I never back down on a good challenge!
+
+I **don't care** about the **originating charset** encoding, because **two different tables** can
+produce **two identical rendered string.**
+What I want is to get readable text, the best I can.
+
+In a way, **I'm brute forcing text decoding.** How cool is that ? 😎
+
+Don't confuse package **ftfy** with charset-normalizer or chardet. ftfy goal is to repair Unicode string whereas charset-normalizer to convert raw file in unknown encoding to unicode.
+
+## 🍰 How
+
+  - Discard all charset encoding table that could not fit the binary content.
+  - Measure noise, or the mess once opened (by chunks) with a corresponding charset encoding.
+  - Extract matches with the lowest mess detected.
+  - Additionally, we measure coherence / probe for a language.
+
+**Wait a minute**, what is noise/mess and coherence according to **YOU ?**
+
+*Noise :* I opened hundred of text files, **written by humans**, with the wrong encoding table. **I observed**, then
+**I established** some ground rules about **what is obvious** when **it seems like** a mess (aka. defining noise in rendered text).
+ I know that my interpretation of what is noise is probably incomplete, feel free to contribute in order to
+ improve or rewrite it.
+
+*Coherence :* For each language there is on earth, we have computed ranked letter appearance occurrences (the best we can). So I thought
+that intel is worth something here. So I use those records against decoded text to check if I can detect intelligent design.
+
+## ⚡ Known limitations
+
+  - Language detection is unreliable when text contains two or more languages sharing identical letters. (eg. HTML (english tags) + Turkish content (Sharing Latin characters))
+  - Every charset detector heavily depends on sufficient content. In common cases, do not bother run detection on very tiny content.
+
+## ⚠️ About Python EOLs
+
+**If you are running:**
+
+- Python >=2.7,<3.5: Unsupported
+- Python 3.5: charset-normalizer < 2.1
+- Python 3.6: charset-normalizer < 3.1
+
+Upgrade your Python interpreter as soon as possible.
+
+## 👤 Contributing
+
+Contributions, issues and feature requests are very much welcome.<br />
+Feel free to check [issues page](https://github.com/ousret/charset_normalizer/issues) if you want to contribute.
+
+## 📝 License
+
+Copyright © [Ahmed TAHRI @Ousret](https://github.com/Ousret).<br />
+This project is [MIT](https://github.com/Ousret/charset_normalizer/blob/master/LICENSE) licensed.
+
+Characters frequencies used in this project © 2012 [Denny Vrandečić](http://simia.net/letters/)
+
+## 💼 For Enterprise
+
+Professional support for charset-normalizer is available as part of the [Tidelift
+Subscription][1]. Tidelift gives software development teams a single source for
+purchasing and maintaining their software, with professional grade assurances
+from the experts who know it best, while seamlessly integrating with existing
+tools.
+
+[1]: https://tidelift.com/subscription/pkg/pypi-charset-normalizer?utm_source=pypi-charset-normalizer&utm_medium=readme
+
+[![OpenSSF Best Practices](https://www.bestpractices.dev/projects/7297/badge)](https://www.bestpractices.dev/projects/7297)
+
+# Changelog
+All notable changes to charset-normalizer will be documented in this file. This project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
+The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/).
+
+## [3.4.7](https://github.com/Ousret/charset_normalizer/compare/3.4.6...3.4.7) (2026-04-02)
+
+### Changed
+- Pre-built optimized version using mypy[c] v1.20.
+- Relax `setuptools` constraint to `setuptools>=68,<82.1`.
+
+### Fixed
+- Correctly remove SIG remnant in utf-7 decoded string. (#718) (#716)
+
+## [3.4.6](https://github.com/Ousret/charset_normalizer/compare/3.4.5...3.4.6) (2026-03-15)
+
+### Changed
+- Flattened the logic in `charset_normalizer.md` for higher performance. Removed `eligible(..)` and `feed(...)`
+  in favor of `feed_info(...)`.
+- Raised upper bound for mypy[c] to 1.20, for our optimized version.
+- Updated `UNICODE_RANGES_COMBINED` using Unicode blocks v17.
+
+### Fixed
+- Edge case where noise difference between two candidates can be almost insignificant. (#672)
+- CLI `--normalize` writing to wrong path when passing multiple files in. (#702)
+
+### Misc
+- Freethreaded pre-built wheels now shipped in PyPI starting with 3.14t. (#616)
+
+## [3.4.5](https://github.com/Ousret/charset_normalizer/compare/3.4.4...3.4.5) (2026-03-06)
+
+### Changed
+- Update `setuptools` constraint to `setuptools>=68,<=82`.
+- Raised upper bound of mypyc for the optional pre-built extension to v1.19.1
+
+### Fixed
+- Add explicit link to lib math in our optimized build. (#692)
+- Logger level not restored correctly for empty byte sequences. (#701)
+- TypeError when passing bytearray to from_bytes. (#703)
+
+### Misc
+- Applied safe micro-optimizations in both our noise detector and language detector.
+- Rewrote the `query_yes_no` function (inside CLI) to avoid using ambiguous licensed code.
+- Added `cd.py` submodule into mypyc optional compilation to reduce further the performance impact.
+
+## [3.4.4](https://github.com/Ousret/charset_normalizer/compare/3.4.2...3.4.4) (2025-10-13)
+
+### Changed
+- Bound `setuptools` to a specific constraint `setuptools>=68,<=81`.
+- Raised upper bound of mypyc for the optional pre-built extension to v1.18.2
+
+### Removed
+- `setuptools-scm` as a build dependency.
+
+### Misc
+- Enforced hashes in `dev-requirements.txt` and created `ci-requirements.txt` for security purposes.
+- Additional pre-built wheels for riscv64, s390x, and armv7l architectures.
+- Restore ` multiple.intoto.jsonl` in GitHub releases in addition to individual attestation file per wheel.
+
+## [3.4.3](https://github.com/Ousret/charset_normalizer/compare/3.4.2...3.4.3) (2025-08-09)
+
+### Changed
+- mypy(c) is no longer a required dependency at build time if `CHARSET_NORMALIZER_USE_MYPYC` isn't set to `1`. (#595) (#583)
+- automatically lower confidence on small bytes samples that are not Unicode in `detect` output legacy function. (#391)
+
+### Added
+- Custom build backend to overcome inability to mark mypy as an optional dependency in the build phase.
+- Support for Python 3.14
+
+### Fixed
+- sdist archive contained useless directories.
+- automatically fallback on valid UTF-16 or UTF-32 even if the md says it's noisy. (#633)
+
+### Misc
+- SBOM are automatically published to the relevant GitHub release to comply with regulatory changes.
+  Each published wheel comes with its SBOM. We choose CycloneDX as the format.
+- Prebuilt optimized wheel are no longer distributed by default for CPython 3.7 due to a change in cibuildwheel.
+
+## [3.4.2](https://github.com/Ousret/charset_normalizer/compare/3.4.1...3.4.2) (2025-05-02)
+
+### Fixed
+- Addressed the DeprecationWarning in our CLI regarding `argparse.FileType` by backporting the target class into the package. (#591)
+- Improved the overall reliability of the detector with CJK Ideographs. (#605) (#587)
+
+### Changed
+- Optional mypyc compilation upgraded to version 1.15 for Python >= 3.8
+
+## [3.4.1](https://github.com/Ousret/charset_normalizer/compare/3.4.0...3.4.1) (2024-12-24)
+
+### Changed
+- Project metadata are now stored using `pyproject.toml` instead of `setup.cfg` using setuptools as the build backend.
+- Enforce annotation delayed loading for a simpler and consistent types in the project.
+- Optional mypyc compilation upgraded to version 1.14 for Python >= 3.8
+
+### Added
+- pre-commit configuration.
+- noxfile.
+
+### Removed
+- `build-requirements.txt` as per using `pyproject.toml` native build configuration.
+- `bin/integration.py` and `bin/serve.py` in favor of downstream integration test (see noxfile).
+- `setup.cfg` in favor of `pyproject.toml` metadata configuration.
+- Unused `utils.range_scan` function.
+
+### Fixed
+- Converting content to Unicode bytes may insert `utf_8` instead of preferred `utf-8`. (#572)
+- Deprecation warning "'count' is passed as positional argument" when converting to Unicode bytes on Python 3.13+
+
+## [3.4.0](https://github.com/Ousret/charset_normalizer/compare/3.3.2...3.4.0) (2024-10-08)
+
+### Added
+- Argument `--no-preemptive` in the CLI to prevent the detector to search for hints.
+- Support for Python 3.13 (#512)
+
+### Fixed
+- Relax the TypeError exception thrown when trying to compare a CharsetMatch with anything else than a CharsetMatch.
+- Improved the general reliability of the detector based on user feedbacks. (#520) (#509) (#498) (#407) (#537)
+- Declared charset in content (preemptive detection) not changed when converting to utf-8 bytes. (#381)
+
+## [3.3.2](https://github.com/Ousret/charset_normalizer/compare/3.3.1...3.3.2) (2023-10-31)
+
+### Fixed
+- Unintentional memory usage regression when using large payload that match several encoding (#376)
+- Regression on some detection case showcased in the documentation (#371)
+
+### Added
+- Noise (md) probe that identify malformed arabic representation due to the presence of letters in isolated form (credit to my wife)
+
+## [3.3.1](https://github.com/Ousret/charset_normalizer/compare/3.3.0...3.3.1) (2023-10-22)
+
+### Changed
+- Optional mypyc compilation upgraded to version 1.6.1 for Python >= 3.8
+- Improved the general detection reliability based on reports from the community
+
+## [3.3.0](https://github.com/Ousret/charset_normalizer/compare/3.2.0...3.3.0) (2023-09-30)
+
+### Added
+- Allow to execute the CLI (e.g. normalizer) through `python -m charset_normalizer.cli` or `python -m charset_normalizer`
+- Support for 9 forgotten encoding that are supported by Python but unlisted in `encoding.aliases` as they have no alias (#323)
+
+### Removed
+- (internal) Redundant utils.is_ascii function and unused function is_private_use_only
+- (internal) charset_normalizer.assets is moved inside charset_normalizer.constant
+
+### Changed
+- (internal) Unicode code blocks in constants are updated using the latest v15.0.0 definition to improve detection
+- Optional mypyc compilation upgraded to version 1.5.1 for Python >= 3.8
+
+### Fixed
+- Unable to properly sort CharsetMatch when both chaos/noise and coherence were close due to an unreachable condition in \_\_lt\_\_ (#350)
+
+## [3.2.0](https://github.com/Ousret/charset_normalizer/compare/3.1.0...3.2.0) (2023-06-07)
+
+### Changed
+- Typehint for function `from_path` no longer enforce `PathLike` as its first argument
+- Minor improvement over the global detection reliability
+
+### Added
+- Introduce function `is_binary` that relies on main capabilities, and optimized to detect binaries
+- Propagate `enable_fallback` argument throughout `from_bytes`, `from_path`, and `from_fp` that allow a deeper control over the detection (default True)
+- Explicit support for Python 3.12
+
+### Fixed
+- Edge case detection failure where a file would contain 'very-long' camel cased word (Issue #289)
+
+## [3.1.0](https://github.com/Ousret/charset_normalizer/compare/3.0.1...3.1.0) (2023-03-06)
+
+### Added
+- Argument `should_rename_legacy` for legacy function `detect` and disregard any new arguments without errors (PR #262)
+
+### Removed
+- Support for Python 3.6 (PR #260)
+
+### Changed
+- Optional speedup provided by mypy/c 1.0.1
+
+## [3.0.1](https://github.com/Ousret/charset_normalizer/compare/3.0.0...3.0.1) (2022-11-18)
+
+### Fixed
+- Multi-bytes cutter/chunk generator did not always cut correctly (PR #233)
+
+### Changed
+- Speedup provided by mypy/c 0.990 on Python >= 3.7
+
+## [3.0.0](https://github.com/Ousret/charset_normalizer/compare/2.1.1...3.0.0) (2022-10-20)
+
+### Added
+- Extend the capability of explain=True when cp_isolation contains at most two entries (min one), will log in details of the Mess-detector results
+- Support for alternative language frequency set in charset_normalizer.assets.FREQUENCIES
+- Add parameter `language_threshold` in `from_bytes`, `from_path` and `from_fp` to adjust the minimum expected coherence ratio
+- `normalizer --version` now specify if current version provide extra speedup (meaning mypyc compilation whl)
+
+### Changed
+- Build with static metadata using 'build' frontend
+- Make the language detection stricter
+- Optional: Module `md.py` can be compiled using Mypyc to provide an extra speedup up to 4x faster than v2.1
+
+### Fixed
+- CLI with opt --normalize fail when using full path for files
+- TooManyAccentuatedPlugin induce false positive on the mess detection when too few alpha character have been fed to it
+- Sphinx warnings when generating the documentation
+
+### Removed
+- Coherence detector no longer return 'Simple English' instead return 'English'
+- Coherence detector no longer return 'Classical Chinese' instead return 'Chinese'
+- Breaking: Method `first()` and `best()` from CharsetMatch
+- UTF-7 will no longer appear as "detected" without a recognized SIG/mark (is unreliable/conflict with ASCII)
+- Breaking: Class aliases CharsetDetector, CharsetDoctor, CharsetNormalizerMatch and CharsetNormalizerMatches
+- Breaking: Top-level function `normalize`
+- Breaking: Properties `chaos_secondary_pass`, `coherence_non_latin` and `w_counter` from CharsetMatch
+- Support for the backport `unicodedata2`
+
+## [3.0.0rc1](https://github.com/Ousret/charset_normalizer/compare/3.0.0b2...3.0.0rc1) (2022-10-18)
+
+### Added
+- Extend the capability of explain=True when cp_isolation contains at most two entries (min one), will log in details of the Mess-detector results
+- Support for alternative language frequency set in charset_normalizer.assets.FREQUENCIES
+- Add parameter `language_threshold` in `from_bytes`, `from_path` and `from_fp` to adjust the minimum expected coherence ratio
+
+### Changed
+- Build with static metadata using 'build' frontend
+- Make the language detection stricter
+
+### Fixed
+- CLI with opt --normalize fail when using full path for files
+- TooManyAccentuatedPlugin induce false positive on the mess detection when too few alpha character have been fed to it
+
+### Removed
+- Coherence detector no longer return 'Simple English' instead return 'English'
+- Coherence detector no longer return 'Classical Chinese' instead return 'Chinese'
+
+## [3.0.0b2](https://github.com/Ousret/charset_normalizer/compare/3.0.0b1...3.0.0b2) (2022-08-21)
+
+### Added
+- `normalizer --version` now specify if current version provide extra speedup (meaning mypyc compilation whl)
+
+### Removed
+- Breaking: Method `first()` and `best()` from CharsetMatch
+- UTF-7 will no longer appear as "detected" without a recognized SIG/mark (is unreliable/conflict with ASCII)
+
+### Fixed
+- Sphinx warnings when generating the documentation
+
+## [3.0.0b1](https://github.com/Ousret/charset_normalizer/compare/2.1.0...3.0.0b1) (2022-08-15)
+
+### Changed
+- Optional: Module `md.py` can be compiled using Mypyc to provide an extra speedup up to 4x faster than v2.1
+
+### Removed
+- Breaking: Class aliases CharsetDetector, CharsetDoctor, CharsetNormalizerMatch and CharsetNormalizerMatches
+- Breaking: Top-level function `normalize`
+- Breaking: Properties `chaos_secondary_pass`, `coherence_non_latin` and `w_counter` from CharsetMatch
+- Support for the backport `unicodedata2`
+
+## [2.1.1](https://github.com/Ousret/charset_normalizer/compare/2.1.0...2.1.1) (2022-08-19)
+
+### Deprecated
+- Function `normalize` scheduled for removal in 3.0
+
+### Changed
+- Removed useless call to decode in fn is_unprintable (#206)
+
+### Fixed
+- Third-party library (i18n xgettext) crashing not recognizing utf_8 (PEP 263) with underscore from [@aleksandernovikov](https://github.com/aleksandernovikov) (#204)
+
+## [2.1.0](https://github.com/Ousret/charset_normalizer/compare/2.0.12...2.1.0) (2022-06-19)
+
+### Added
+- Output the Unicode table version when running the CLI with `--version` (PR #194)
+
+### Changed
+- Re-use decoded buffer for single byte character sets from [@nijel](https://github.com/nijel) (PR #175)
+- Fixing some performance bottlenecks from [@deedy5](https://github.com/deedy5) (PR #183)
+
+### Fixed
+- Workaround potential bug in cpython with Zero Width No-Break Space located in Arabic Presentation Forms-B, Unicode 1.1 not acknowledged as space (PR #175)
+- CLI default threshold aligned with the API threshold from [@oleksandr-kuzmenko](https://github.com/oleksandr-kuzmenko) (PR #181)
+
+### Removed
+- Support for Python 3.5 (PR #192)
+
+### Deprecated
+- Use of backport unicodedata from `unicodedata2` as Python is quickly catching up, scheduled for removal in 3.0 (PR #194)
+
+## [2.0.12](https://github.com/Ousret/charset_normalizer/compare/2.0.11...2.0.12) (2022-02-12)
+
+### Fixed
+- ASCII miss-detection on rare cases (PR #170)
+
+## [2.0.11](https://github.com/Ousret/charset_normalizer/compare/2.0.10...2.0.11) (2022-01-30)
+
+### Added
+- Explicit support for Python 3.11 (PR #164)
+
+### Changed
+- The logging behavior have been completely reviewed, now using only TRACE and DEBUG levels (PR #163 #165)
+
+## [2.0.10](https://github.com/Ousret/charset_normalizer/compare/2.0.9...2.0.10) (2022-01-04)
+
+### Fixed
+- Fallback match entries might lead to UnicodeDecodeError for large bytes sequence (PR #154)
+
+### Changed
+- Skipping the language-detection (CD) on ASCII (PR #155)
+
+## [2.0.9](https://github.com/Ousret/charset_normalizer/compare/2.0.8...2.0.9) (2021-12-03)
+
+### Changed
+- Moderating the logging impact (since 2.0.8) for specific environments (PR #147)
+
+### Fixed
+- Wrong logging level applied when setting kwarg `explain` to True (PR #146)
+
+## [2.0.8](https://github.com/Ousret/charset_normalizer/compare/2.0.7...2.0.8) (2021-11-24)
+### Changed
+- Improvement over Vietnamese detection (PR #126)
+- MD improvement on trailing data and long foreign (non-pure latin) data (PR #124)
+- Efficiency improvements in cd/alphabet_languages from [@adbar](https://github.com/adbar) (PR #122)
+- call sum() without an intermediary list following PEP 289 recommendations from [@adbar](https://github.com/adbar) (PR #129)
+- Code style as refactored by Sourcery-AI (PR #131)
+- Minor adjustment on the MD around european words (PR #133)
+- Remove and replace SRTs from assets / tests (PR #139)
+- Initialize the library logger with a `NullHandler` by default from [@nmaynes](https://github.com/nmaynes) (PR #135)
+- Setting kwarg `explain` to True will add provisionally (bounded to function lifespan) a specific stream handler (PR #135)
+
+### Fixed
+- Fix large (misleading) sequence giving UnicodeDecodeError (PR #137)
+- Avoid using too insignificant chunk (PR #137)
+
+### Added
+- Add and expose function `set_logging_handler` to configure a specific StreamHandler from [@nmaynes](https://github.com/nmaynes) (PR #135)
+- Add `CHANGELOG.md` entries, format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/) (PR #141)
+
+## [2.0.7](https://github.com/Ousret/charset_normalizer/compare/2.0.6...2.0.7) (2021-10-11)
+### Added
+- Add support for Kazakh (Cyrillic) language detection (PR #109)
+
+### Changed
+- Further, improve inferring the language from a given single-byte code page (PR #112)
+- Vainly trying to leverage PEP263 when PEP3120 is not supported (PR #116)
+- Refactoring for potential performance improvements in loops from [@adbar](https://github.com/adbar) (PR #113)
+- Various detection improvement (MD+CD) (PR #117)
+
+### Removed
+- Remove redundant logging entry about detected language(s) (PR #115)
+
+### Fixed
+- Fix a minor inconsistency between Python 3.5 and other versions regarding language detection (PR #117 #102)
+
+## [2.0.6](https://github.com/Ousret/charset_normalizer/compare/2.0.5...2.0.6) (2021-09-18)
+### Fixed
+- Unforeseen regression with the loss of the backward-compatibility with some older minor of Python 3.5.x (PR #100)
+- Fix CLI crash when using --minimal output in certain cases (PR #103)
+
+### Changed
+- Minor improvement to the detection efficiency (less than 1%) (PR #106 #101)
+
+## [2.0.5](https://github.com/Ousret/charset_normalizer/compare/2.0.4...2.0.5) (2021-09-14)
+### Changed
+- The project now comply with: flake8, mypy, isort and black to ensure a better overall quality (PR #81)
+- The BC-support with v1.x was improved, the old staticmethods are restored (PR #82)
+- The Unicode detection is slightly improved (PR #93)
+- Add syntax sugar \_\_bool\_\_ for results CharsetMatches list-container (PR #91)
+
+### Removed
+- The project no longer raise warning on tiny content given for detection, will be simply logged as warning instead (PR #92)
+
+### Fixed
+- In some rare case, the chunks extractor could cut in the middle of a multi-byte character and could mislead the mess detection (PR #95)
+- Some rare 'space' characters could trip up the UnprintablePlugin/Mess detection (PR #96)
+- The MANIFEST.in was not exhaustive (PR #78)
+
+## [2.0.4](https://github.com/Ousret/charset_normalizer/compare/2.0.3...2.0.4) (2021-07-30)
+### Fixed
+- The CLI no longer raise an unexpected exception when no encoding has been found (PR #70)
+- Fix accessing the 'alphabets' property when the payload contains surrogate characters (PR #68)
+- The logger could mislead (explain=True) on detected languages and the impact of one MBCS match (PR #72)
+- Submatch factoring could be wrong in rare edge cases (PR #72)
+- Multiple files given to the CLI were ignored when publishing results to STDOUT. (After the first path) (PR #72)
+- Fix line endings from CRLF to LF for certain project files (PR #67)
+
+### Changed
+- Adjust the MD to lower the sensitivity, thus improving the global detection reliability (PR #69 #76)
+- Allow fallback on specified encoding if any (PR #71)
+
+## [2.0.3](https://github.com/Ousret/charset_normalizer/compare/2.0.2...2.0.3) (2021-07-16)
+### Changed
+- Part of the detection mechanism has been improved to be less sensitive, resulting in more accurate detection results. Especially ASCII. (PR #63)
+- According to the community wishes, the detection will fall back on ASCII or UTF-8 in a last-resort case. (PR #64)
+
+## [2.0.2](https://github.com/Ousret/charset_normalizer/compare/2.0.1...2.0.2) (2021-07-15)
+### Fixed
+- Empty/Too small JSON payload miss-detection fixed. Report from [@tseaver](https://github.com/tseaver) (PR #59)
+
+### Changed
+- Don't inject unicodedata2 into sys.modules from [@akx](https://github.com/akx) (PR #57)
+
+## [2.0.1](https://github.com/Ousret/charset_normalizer/compare/2.0.0...2.0.1) (2021-07-13)
+### Fixed
+- Make it work where there isn't a filesystem available, dropping assets frequencies.json. Report from [@sethmlarson](https://github.com/sethmlarson). (PR #55)
+- Using explain=False permanently disable the verbose output in the current runtime (PR #47)
+- One log entry (language target preemptive) was not show in logs when using explain=True (PR #47)
+- Fix undesired exception (ValueError) on getitem of instance CharsetMatches (PR #52)
+
+### Changed
+- Public function normalize default args values were not aligned with from_bytes (PR #53)
+
+### Added
+- You may now use charset aliases in cp_isolation and cp_exclusion arguments (PR #47)
+
+## [2.0.0](https://github.com/Ousret/charset_normalizer/compare/1.4.1...2.0.0) (2021-07-02)
+### Changed
+- 4x to 5 times faster than the previous 1.4.0 release. At least 2x faster than Chardet.
+- Accent has been made on UTF-8 detection, should perform rather instantaneous.
+- The backward compatibility with Chardet has been greatly improved. The legacy detect function returns an identical charset name whenever possible.
+- The detection mechanism has been slightly improved, now Turkish content is detected correctly (most of the time)
+- The program has been rewritten to ease the readability and maintainability. (+Using static typing)+
+- utf_7 detection has been reinstated.
+
+### Removed
+- This package no longer require anything when used with Python 3.5 (Dropped cached_property)
+- Removed support for these languages: Catalan, Esperanto, Kazakh, Baque, Volapük, Azeri, Galician, Nynorsk, Macedonian, and Serbocroatian.
+- The exception hook on UnicodeDecodeError has been removed.
+
+### Deprecated
+- Methods coherence_non_latin, w_counter, chaos_secondary_pass of the class CharsetMatch are now deprecated and scheduled for removal in v3.0
+
+### Fixed
+- The CLI output used the relative path of the file(s). Should be absolute.
+
+## [1.4.1](https://github.com/Ousret/charset_normalizer/compare/1.4.0...1.4.1) (2021-05-28)
+### Fixed
+- Logger configuration/usage no longer conflict with others (PR #44)
+
+## [1.4.0](https://github.com/Ousret/charset_normalizer/compare/1.3.9...1.4.0) (2021-05-21)
+### Removed
+- Using standard logging instead of using the package loguru.
+- Dropping nose test framework in favor of the maintained pytest.
+- Choose to not use dragonmapper package to help with gibberish Chinese/CJK text.
+- Require cached_property only for Python 3.5 due to constraint. Dropping for every other interpreter version.
+- Stop support for UTF-7 that does not contain a SIG.
+- Dropping PrettyTable, replaced with pure JSON output in CLI.
+
+### Fixed
+- BOM marker in a CharsetNormalizerMatch instance could be False in rare cases even if obviously present. Due to the sub-match factoring process.
+- Not searching properly for the BOM when trying utf32/16 parent codec.
+
+### Changed
+- Improving the package final size by compressing frequencies.json.
+- Huge improvement over the larges payload.
+
+### Added
+- CLI now produces JSON consumable output.
+- Return ASCII if given sequences fit. Given reasonable confidence.
+
+## [1.3.9](https://github.com/Ousret/charset_normalizer/compare/1.3.8...1.3.9) (2021-05-13)
+
+### Fixed
+- In some very rare cases, you may end up getting encode/decode errors due to a bad bytes payload (PR #40)
+
+## [1.3.8](https://github.com/Ousret/charset_normalizer/compare/1.3.7...1.3.8) (2021-05-12)
+
+### Fixed
+- Empty given payload for detection may cause an exception if trying to access the `alphabets` property. (PR #39)
+
+## [1.3.7](https://github.com/Ousret/charset_normalizer/compare/1.3.6...1.3.7) (2021-05-12)
+
+### Fixed
+- The legacy detect function should return UTF-8-SIG if sig is present in the payload. (PR #38)
+
+## [1.3.6](https://github.com/Ousret/charset_normalizer/compare/1.3.5...1.3.6) (2021-02-09)
+
+### Changed
+- Amend the previous release to allow prettytable 2.0 (PR #35)
+
+## [1.3.5](https://github.com/Ousret/charset_normalizer/compare/1.3.4...1.3.5) (2021-02-08)
+
+### Fixed
+- Fix error while using the package with a python pre-release interpreter (PR #33)
+
+### Changed
+- Dependencies refactoring, constraints revised.
+
+### Added
+- Add python 3.9 and 3.10 to the supported interpreters
+
+MIT License
+
+Copyright (c) 2025 TAHRI Ahmed R.
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

.venv/Lib/site-packages/charset_normalizer-3.4.7.dist-info/RECORD

@@ -0,0 +1,36 @@
+../../Scripts/normalizer.exe,sha256=qTlnspDfyZpwuu4TuqkTqANl1ub6baifR7qWfftCfBc,108410
+81d243bd2c585b0f4821__mypyc.cp314-win_amd64.pyd,sha256=q2SDW2MJPDGXXfJpJ1bIAI5-wZDkqobxWnE7zGLh1DI,216576
+charset_normalizer-3.4.7.dist-info/INSTALLER,sha256=zuuue4knoyJ-UwPPXg8fezS7VCrXJQrAP7zeNuwvFQg,4
+charset_normalizer-3.4.7.dist-info/METADATA,sha256=0llkYJSRRJSKgCIy09lSxe0sJGV1SBWfRvNSjKX5-rs,41739
+charset_normalizer-3.4.7.dist-info/RECORD,,
+charset_normalizer-3.4.7.dist-info/WHEEL,sha256=jhxrRP3tdEfm7toB1X7dl5x6f0bu6Tve2cWwIdkbL8U,101
+charset_normalizer-3.4.7.dist-info/entry_points.txt,sha256=ADSTKrkXZ3hhdOVFi6DcUEHQRS0xfxDIE_pEz4wLIXA,65
+charset_normalizer-3.4.7.dist-info/licenses/LICENSE,sha256=GFd0hdNwTxpHne2OVzwJds_tMV_S_ReYP6mI2kwvcNE,1092
+charset_normalizer-3.4.7.dist-info/top_level.txt,sha256=c_vZbitqecT2GfK3zdxSTLCn8C-6pGnHQY5o_5Y32M0,47
+charset_normalizer/__init__.py,sha256=0NT8MHi7SKq3juMqYfOdrkzjisK0L73lneNHH4qaUAs,1638
+charset_normalizer/__main__.py,sha256=2sj_BS6H0sU25C1bMqz9DVwa6kOK9lchSEbSU-_iu7M,115
+charset_normalizer/__pycache__/__init__.cpython-314.pyc,,
+charset_normalizer/__pycache__/__main__.cpython-314.pyc,,
+charset_normalizer/__pycache__/api.cpython-314.pyc,,
+charset_normalizer/__pycache__/cd.cpython-314.pyc,,
+charset_normalizer/__pycache__/constant.cpython-314.pyc,,
+charset_normalizer/__pycache__/legacy.cpython-314.pyc,,
+charset_normalizer/__pycache__/md.cpython-314.pyc,,
+charset_normalizer/__pycache__/models.cpython-314.pyc,,
+charset_normalizer/__pycache__/utils.cpython-314.pyc,,
+charset_normalizer/__pycache__/version.cpython-314.pyc,,
+charset_normalizer/api.py,sha256=mHkwxdTIzOWLNDsADQ0IpedjW8MY0VNWan1MMexenfc,39832
+charset_normalizer/cd.cp314-win_amd64.pyd,sha256=2jzngNbViujGkguZ-Ab2TVihsz9KATgoYbAbUFSRmfg,10752
+charset_normalizer/cd.py,sha256=oxPRzNqGqGMm6V4Os7lI2jQK_Pe8vgzEOWtIH4FWYhQ,15628
+charset_normalizer/cli/__init__.py,sha256=d9MUx-1V_qD3x9igIy4JT4oC5CU0yjulk7QyZWeRFhg,144
+charset_normalizer/cli/__main__.py,sha256=8TfOJ7pE3_USMlquyzg1TnEvSWdshj3Vmycsmsd0EJA,12302
+charset_normalizer/cli/__pycache__/__init__.cpython-314.pyc,,
+charset_normalizer/cli/__pycache__/__main__.cpython-314.pyc,,
+charset_normalizer/constant.py,sha256=CA94LLkydl625rnSViIv_egUOpzZx1mAAUmCu3QN5mQ,46481
+charset_normalizer/legacy.py,sha256=MCz6fcVj-h_VG9lMgnMUibESu4iAQqznE6HbMdleQIM,2740
+charset_normalizer/md.cp314-win_amd64.pyd,sha256=jj4KcZpOyDX72bMShjaglpL-Fz9OAcUnZTv9t71yOAg,10752
+charset_normalizer/md.py,sha256=_Sv-zOZTD64bz7LsMcs5GEOBdJdvYreOiVjtUiEcVuU,31377
+charset_normalizer/models.py,sha256=wzkS2p-SXFmfOcGhs5Xt_eke1OwxthBGCXwSkAMG2_c,13190
+charset_normalizer/py.typed,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
+charset_normalizer/utils.py,sha256=NzMBuvIt5IZ5_sYqboqJs7L1X_uWhwz5RfPWBYaDPX0,12704
+charset_normalizer/version.py,sha256=R38VXV25CgZdFHNOs5FpuL29tVaU302SYUBvWpWCVNw,123

.venv/Lib/site-packages/charset_normalizer-3.4.7.dist-info/WHEEL

@@ -0,0 +1,5 @@
+Wheel-Version: 1.0
+Generator: setuptools (82.0.1)
+Root-Is-Purelib: false
+Tag: cp314-cp314-win_amd64
+

.venv/Lib/site-packages/charset_normalizer-3.4.7.dist-info/entry_points.txt

@@ -0,0 +1,2 @@
+[console_scripts]
+normalizer = charset_normalizer.cli:cli_detect

.venv/Lib/site-packages/charset_normalizer-3.4.7.dist-info/licenses/LICENSE

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2025 TAHRI Ahmed R.
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

.venv/Lib/site-packages/charset_normalizer-3.4.7.dist-info/top_level.txt

@@ -0,0 +1,2 @@
+81d243bd2c585b0f4821__mypyc
+charset_normalizer

.venv/Lib/site-packages/charset_normalizer/__init__.py

@@ -0,0 +1,48 @@
+"""
+Charset-Normalizer
+~~~~~~~~~~~~~~
+The Real First Universal Charset Detector.
+A library that helps you read text from an unknown charset encoding.
+Motivated by chardet, This package is trying to resolve the issue by taking a new approach.
+All IANA character set names for which the Python core library provides codecs are supported.
+
+Basic usage:
+   >>> from charset_normalizer import from_bytes
+   >>> results = from_bytes('Bсеки човек има право на образование. Oбразованието!'.encode('utf_8'))
+   >>> best_guess = results.best()
+   >>> str(best_guess)
+   'Bсеки човек има право на образование. Oбразованието!'
+
+Others methods and usages are available - see the full documentation
+at <https://github.com/Ousret/charset_normalizer>.
+:copyright: (c) 2021 by Ahmed TAHRI
+:license: MIT, see LICENSE for more details.
+"""
+
+from __future__ import annotations
+
+import logging
+
+from .api import from_bytes, from_fp, from_path, is_binary
+from .legacy import detect
+from .models import CharsetMatch, CharsetMatches
+from .utils import set_logging_handler
+from .version import VERSION, __version__
+
+__all__ = (
+    "from_fp",
+    "from_path",
+    "from_bytes",
+    "is_binary",
+    "detect",
+    "CharsetMatch",
+    "CharsetMatches",
+    "__version__",
+    "VERSION",
+    "set_logging_handler",
+)
+
+# Attach a NullHandler to the top level logger by default
+# https://docs.python.org/3.3/howto/logging.html#configuring-logging-for-a-library
+
+logging.getLogger("charset_normalizer").addHandler(logging.NullHandler())

.venv/Lib/site-packages/charset_normalizer/__main__.py

@@ -0,0 +1,6 @@
+from __future__ import annotations
+
+from .cli import cli_detect
+
+if __name__ == "__main__":
+    cli_detect()

.venv/Lib/site-packages/charset_normalizer/api.py

@@ -0,0 +1,988 @@
+from __future__ import annotations
+
+import logging
+from os import PathLike
+from typing import BinaryIO
+
+from .cd import (
+    coherence_ratio,
+    encoding_languages,
+    mb_encoding_languages,
+    merge_coherence_ratios,
+)
+from .constant import (
+    IANA_SUPPORTED,
+    IANA_SUPPORTED_SIMILAR,
+    TOO_BIG_SEQUENCE,
+    TOO_SMALL_SEQUENCE,
+    TRACE,
+)
+from .md import mess_ratio
+from .models import CharsetMatch, CharsetMatches
+from .utils import (
+    any_specified_encoding,
+    cut_sequence_chunks,
+    iana_name,
+    identify_sig_or_bom,
+    is_multi_byte_encoding,
+    should_strip_sig_or_bom,
+)
+
+logger = logging.getLogger("charset_normalizer")
+explain_handler = logging.StreamHandler()
+explain_handler.setFormatter(
+    logging.Formatter("%(asctime)s | %(levelname)s | %(message)s")
+)
+
+# Pre-compute a reordered encoding list: multibyte first, then single-byte.
+# This allows the mb_definitive_match optimization to fire earlier, skipping
+# all single-byte encodings for genuine CJK content. Multibyte codecs
+# hard-fail (UnicodeDecodeError) on single-byte data almost instantly, so
+# testing them first costs negligible time for non-CJK files.
+_mb_supported: list[str] = []
+_sb_supported: list[str] = []
+
+for _supported_enc in IANA_SUPPORTED:
+    try:
+        if is_multi_byte_encoding(_supported_enc):
+            _mb_supported.append(_supported_enc)
+        else:
+            _sb_supported.append(_supported_enc)
+    except ImportError:
+        _sb_supported.append(_supported_enc)
+
+IANA_SUPPORTED_MB_FIRST: list[str] = _mb_supported + _sb_supported
+
+
+def from_bytes(
+    sequences: bytes | bytearray,
+    steps: int = 5,
+    chunk_size: int = 512,
+    threshold: float = 0.2,
+    cp_isolation: list[str] | None = None,
+    cp_exclusion: list[str] | None = None,
+    preemptive_behaviour: bool = True,
+    explain: bool = False,
+    language_threshold: float = 0.1,
+    enable_fallback: bool = True,
+) -> CharsetMatches:
+    """
+    Given a raw bytes sequence, return the best possibles charset usable to render str objects.
+    If there is no results, it is a strong indicator that the source is binary/not text.
+    By default, the process will extract 5 blocks of 512o each to assess the mess and coherence of a given sequence.
+    And will give up a particular code page after 20% of measured mess. Those criteria are customizable at will.
+
+    The preemptive behavior DOES NOT replace the traditional detection workflow, it prioritize a particular code page
+    but never take it for granted. Can improve the performance.
+
+    You may want to focus your attention to some code page or/and not others, use cp_isolation and cp_exclusion for that
+    purpose.
+
+    This function will strip the SIG in the payload/sequence every time except on UTF-16, UTF-32.
+    By default the library does not setup any handler other than the NullHandler, if you choose to set the 'explain'
+    toggle to True it will alter the logger configuration to add a StreamHandler that is suitable for debugging.
+    Custom logging format and handler can be set manually.
+    """
+
+    if not isinstance(sequences, (bytearray, bytes)):
+        raise TypeError(
+            "Expected object of type bytes or bytearray, got: {}".format(
+                type(sequences)
+            )
+        )
+
+    if explain:
+        previous_logger_level: int = logger.level
+        logger.addHandler(explain_handler)
+        logger.setLevel(TRACE)
+
+    length: int = len(sequences)
+
+    if length == 0:
+        logger.debug("Encoding detection on empty bytes, assuming utf_8 intention.")
+        if explain:  # Defensive: ensure exit path clean handler
+            logger.removeHandler(explain_handler)
+            logger.setLevel(previous_logger_level)
+        return CharsetMatches([CharsetMatch(sequences, "utf_8", 0.0, False, [], "")])
+
+    if cp_isolation is not None:
+        logger.log(
+            TRACE,
+            "cp_isolation is set. use this flag for debugging purpose. "
+            "limited list of encoding allowed : %s.",
+            ", ".join(cp_isolation),
+        )
+        cp_isolation = [iana_name(cp, False) for cp in cp_isolation]
+    else:
+        cp_isolation = []
+
+    if cp_exclusion is not None:
+        logger.log(
+            TRACE,
+            "cp_exclusion is set. use this flag for debugging purpose. "
+            "limited list of encoding excluded : %s.",
+            ", ".join(cp_exclusion),
+        )
+        cp_exclusion = [iana_name(cp, False) for cp in cp_exclusion]
+    else:
+        cp_exclusion = []
+
+    if length <= (chunk_size * steps):
+        logger.log(
+            TRACE,
+            "override steps (%i) and chunk_size (%i) as content does not fit (%i byte(s) given) parameters.",
+            steps,
+            chunk_size,
+            length,
+        )
+        steps = 1
+        chunk_size = length
+
+    if steps > 1 and length / steps < chunk_size:
+        chunk_size = int(length / steps)
+
+    is_too_small_sequence: bool = len(sequences) < TOO_SMALL_SEQUENCE
+    is_too_large_sequence: bool = len(sequences) >= TOO_BIG_SEQUENCE
+
+    if is_too_small_sequence:
+        logger.log(
+            TRACE,
+            "Trying to detect encoding from a tiny portion of ({}) byte(s).".format(
+                length
+            ),
+        )
+    elif is_too_large_sequence:
+        logger.log(
+            TRACE,
+            "Using lazy str decoding because the payload is quite large, ({}) byte(s).".format(
+                length
+            ),
+        )
+
+    prioritized_encodings: list[str] = []
+
+    specified_encoding: str | None = (
+        any_specified_encoding(sequences) if preemptive_behaviour else None
+    )
+
+    if specified_encoding is not None:
+        prioritized_encodings.append(specified_encoding)
+        logger.log(
+            TRACE,
+            "Detected declarative mark in sequence. Priority +1 given for %s.",
+            specified_encoding,
+        )
+
+    tested: set[str] = set()
+    tested_but_hard_failure: list[str] = []
+    tested_but_soft_failure: list[str] = []
+    soft_failure_skip: set[str] = set()
+    success_fast_tracked: set[str] = set()
+
+    # Cache for decoded payload deduplication: hash(decoded_payload) -> (mean_mess_ratio, cd_ratios_merged, passed)
+    # When multiple encodings decode to the exact same string, we can skip the expensive
+    # mess_ratio and coherence_ratio analysis and reuse the results from the first encoding.
+    payload_result_cache: dict[int, tuple[float, list[tuple[str, float]], bool]] = {}
+
+    # When a definitive result (chaos=0.0 and good coherence) is found after testing
+    # the prioritized encodings (ascii, utf_8), we can significantly reduce the remaining
+    # work. Encodings that target completely different language families (e.g., Cyrillic
+    # when the definitive match is Latin) are skipped entirely.
+    # Additionally, for same-family encodings that pass chaos probing, we reuse the
+    # definitive match's coherence ratios instead of recomputing them — a major savings
+    # since coherence_ratio accounts for ~30% of total time on slow Latin files.
+    definitive_match_found: bool = False
+    definitive_target_languages: set[str] = set()
+    # After the definitive match fires, we cap the number of additional same-family
+    # single-byte encodings that pass chaos probing. Once we've accumulated enough
+    # good candidates (N), further same-family SB encodings are unlikely to produce
+    # a better best() result and just waste mess_ratio + coherence_ratio time.
+    # The first encoding to trigger the definitive match is NOT counted (it's already in).
+    post_definitive_sb_success_count: int = 0
+    POST_DEFINITIVE_SB_CAP: int = 7
+
+    # When a non-UTF multibyte encoding passes chaos probing with significant multibyte
+    # content (decoded length < 98% of raw length), skip all remaining single-byte encodings.
+    # Rationale: multi-byte decoders (CJK) have strict byte-sequence validation — if they
+    # decode without error AND pass chaos probing with substantial multibyte content, the
+    # data is genuinely multibyte encoded. Single-byte encodings will always decode (every
+    # byte maps to something) but waste time on mess_ratio before failing.
+    # The 98% threshold prevents false triggers on files that happen to have a few valid
+    # multibyte pairs (e.g., cp424/_ude_1.txt where big5 decodes with 99% ratio).
+    mb_definitive_match_found: bool = False
+
+    fallback_ascii: CharsetMatch | None = None
+    fallback_u8: CharsetMatch | None = None
+    fallback_specified: CharsetMatch | None = None
+
+    results: CharsetMatches = CharsetMatches()
+
+    early_stop_results: CharsetMatches = CharsetMatches()
+
+    sig_encoding, sig_payload = identify_sig_or_bom(sequences)
+
+    if sig_encoding is not None:
+        prioritized_encodings.append(sig_encoding)
+        logger.log(
+            TRACE,
+            "Detected a SIG or BOM mark on first %i byte(s). Priority +1 given for %s.",
+            len(sig_payload),
+            sig_encoding,
+        )
+
+    prioritized_encodings.append("ascii")
+
+    if "utf_8" not in prioritized_encodings:
+        prioritized_encodings.append("utf_8")
+
+    for encoding_iana in prioritized_encodings + IANA_SUPPORTED_MB_FIRST:
+        if cp_isolation and encoding_iana not in cp_isolation:
+            continue
+
+        if cp_exclusion and encoding_iana in cp_exclusion:
+            continue
+
+        if encoding_iana in tested:
+            continue
+
+        tested.add(encoding_iana)
+
+        decoded_payload: str | None = None
+        bom_or_sig_available: bool = sig_encoding == encoding_iana
+        strip_sig_or_bom: bool = bom_or_sig_available and should_strip_sig_or_bom(
+            encoding_iana
+        )
+
+        if encoding_iana in {"utf_16", "utf_32"} and not bom_or_sig_available:
+            logger.log(
+                TRACE,
+                "Encoding %s won't be tested as-is because it require a BOM. Will try some sub-encoder LE/BE.",
+                encoding_iana,
+            )
+            continue
+        if encoding_iana in {"utf_7"} and not bom_or_sig_available:
+            logger.log(
+                TRACE,
+                "Encoding %s won't be tested as-is because detection is unreliable without BOM/SIG.",
+                encoding_iana,
+            )
+            continue
+
+        # Skip encodings similar to ones that already soft-failed (high mess ratio).
+        # Checked BEFORE the expensive decode attempt.
+        if encoding_iana in soft_failure_skip:
+            logger.log(
+                TRACE,
+                "%s is deemed too similar to a code page that was already considered unsuited. Continuing!",
+                encoding_iana,
+            )
+            continue
+
+        # Skip encodings that were already fast-tracked from a similar successful encoding.
+        if encoding_iana in success_fast_tracked:
+            logger.log(
+                TRACE,
+                "Skipping %s: already fast-tracked from a similar successful encoding.",
+                encoding_iana,
+            )
+            continue
+
+        try:
+            is_multi_byte_decoder: bool = is_multi_byte_encoding(encoding_iana)
+        except (ModuleNotFoundError, ImportError):  # Defensive:
+            logger.log(
+                TRACE,
+                "Encoding %s does not provide an IncrementalDecoder",
+                encoding_iana,
+            )
+            continue
+
+        # When we've already found a definitive match (chaos=0.0 with good coherence)
+        # after testing the prioritized encodings, skip encodings that target
+        # completely different language families. This avoids running expensive
+        # mess_ratio + coherence_ratio on clearly unrelated candidates (e.g., Cyrillic
+        # when the definitive match is Latin-based).
+        if definitive_match_found:
+            if not is_multi_byte_decoder:
+                enc_languages = set(encoding_languages(encoding_iana))
+            else:
+                enc_languages = set(mb_encoding_languages(encoding_iana))
+            if not enc_languages.intersection(definitive_target_languages):
+                logger.log(
+                    TRACE,
+                    "Skipping %s: definitive match already found, this encoding targets different languages (%s vs %s).",
+                    encoding_iana,
+                    enc_languages,
+                    definitive_target_languages,
+                )
+                continue
+
+        # After the definitive match, cap the number of additional same-family
+        # single-byte encodings that pass chaos probing. This avoids testing the
+        # tail of rare, low-value same-family encodings (mac_iceland, cp860, etc.)
+        # that almost never change best() but each cost ~1-2ms of mess_ratio + coherence.
+        if (
+            definitive_match_found
+            and not is_multi_byte_decoder
+            and post_definitive_sb_success_count >= POST_DEFINITIVE_SB_CAP
+        ):
+            logger.log(
+                TRACE,
+                "Skipping %s: already accumulated %d same-family results after definitive match (cap=%d).",
+                encoding_iana,
+                post_definitive_sb_success_count,
+                POST_DEFINITIVE_SB_CAP,
+            )
+            continue
+
+        # When a multibyte encoding with significant multibyte content has already
+        # passed chaos probing, skip all single-byte encodings. They will either fail
+        # chaos probing (wasting mess_ratio time) or produce inferior results.
+        if mb_definitive_match_found and not is_multi_byte_decoder:
+            logger.log(
+                TRACE,
+                "Skipping single-byte %s: multi-byte definitive match already found.",
+                encoding_iana,
+            )
+            continue
+
+        try:
+            if is_too_large_sequence and is_multi_byte_decoder is False:
+                str(
+                    (
+                        sequences[: int(50e4)]
+                        if strip_sig_or_bom is False
+                        else sequences[len(sig_payload) : int(50e4)]
+                    ),
+                    encoding=encoding_iana,
+                )
+            else:
+                # UTF-7 BOM is encoded in modified Base64 whose byte boundary
+                # can overlap with the next character. Stripping raw SIG bytes
+                # before decoding may leave stray bytes that decode as garbage.
+                # Decode the full sequence and remove the leading BOM char instead.
+                # see https://github.com/jawah/charset_normalizer/issues/718
+                # and https://github.com/jawah/charset_normalizer/issues/716
+                if encoding_iana == "utf_7" and bom_or_sig_available:
+                    decoded_payload = str(
+                        sequences,
+                        encoding=encoding_iana,
+                    )
+                    if decoded_payload and decoded_payload[0] == "\ufeff":
+                        decoded_payload = decoded_payload[1:]
+                else:
+                    decoded_payload = str(
+                        (
+                            sequences
+                            if strip_sig_or_bom is False
+                            else sequences[len(sig_payload) :]
+                        ),
+                        encoding=encoding_iana,
+                    )
+        except (UnicodeDecodeError, LookupError) as e:
+            if not isinstance(e, LookupError):
+                logger.log(
+                    TRACE,
+                    "Code page %s does not fit given bytes sequence at ALL. %s",
+                    encoding_iana,
+                    str(e),
+                )
+            tested_but_hard_failure.append(encoding_iana)
+            continue
+
+        r_ = range(
+            0 if not bom_or_sig_available else len(sig_payload),
+            length,
+            int(length / steps),
+        )
+
+        multi_byte_bonus: bool = (
+            is_multi_byte_decoder
+            and decoded_payload is not None
+            and len(decoded_payload) < length
+        )
+
+        if multi_byte_bonus:
+            logger.log(
+                TRACE,
+                "Code page %s is a multi byte encoding table and it appear that at least one character "
+                "was encoded using n-bytes.",
+                encoding_iana,
+            )
+
+        # Payload-hash deduplication: if another encoding already decoded to the
+        # exact same string, reuse its mess_ratio and coherence results entirely.
+        # This is strictly more general than the old IANA_SUPPORTED_SIMILAR approach
+        # because it catches ALL identical decoding, not just pre-mapped ones.
+        if decoded_payload is not None and not is_multi_byte_decoder:
+            payload_hash: int = hash(decoded_payload)
+            cached = payload_result_cache.get(payload_hash)
+            if cached is not None:
+                cached_mess, cached_cd, cached_passed = cached
+                if cached_passed:
+                    # The previous encoding with identical output passed chaos probing.
+                    fast_match = CharsetMatch(
+                        sequences,
+                        encoding_iana,
+                        cached_mess,
+                        bom_or_sig_available,
+                        cached_cd,
+                        (
+                            decoded_payload
+                            if (
+                                is_too_large_sequence is False
+                                or encoding_iana
+                                in [specified_encoding, "ascii", "utf_8"]
+                            )
+                            else None
+                        ),
+                        preemptive_declaration=specified_encoding,
+                    )
+                    results.append(fast_match)
+                    success_fast_tracked.add(encoding_iana)
+                    logger.log(
+                        TRACE,
+                        "%s fast-tracked (identical decoded payload to a prior encoding, chaos=%f %%).",
+                        encoding_iana,
+                        round(cached_mess * 100, ndigits=3),
+                    )
+
+                    if (
+                        encoding_iana in [specified_encoding, "ascii", "utf_8"]
+                        and cached_mess < 0.1
+                    ):
+                        if cached_mess == 0.0:
+                            logger.debug(
+                                "Encoding detection: %s is most likely the one.",
+                                fast_match.encoding,
+                            )
+                            if explain:
+                                logger.removeHandler(explain_handler)
+                                logger.setLevel(previous_logger_level)
+                            return CharsetMatches([fast_match])
+                        early_stop_results.append(fast_match)
+
+                    if (
+                        len(early_stop_results)
+                        and (specified_encoding is None or specified_encoding in tested)
+                        and "ascii" in tested
+                        and "utf_8" in tested
+                    ):
+                        probable_result: CharsetMatch = early_stop_results.best()  # type: ignore[assignment]
+                        logger.debug(
+                            "Encoding detection: %s is most likely the one.",
+                            probable_result.encoding,
+                        )
+                        if explain:
+                            logger.removeHandler(explain_handler)
+                            logger.setLevel(previous_logger_level)
+                        return CharsetMatches([probable_result])
+
+                    continue
+                else:
+                    # The previous encoding with identical output failed chaos probing.
+                    tested_but_soft_failure.append(encoding_iana)
+                    logger.log(
+                        TRACE,
+                        "%s fast-skipped (identical decoded payload to a prior encoding that failed chaos probing).",
+                        encoding_iana,
+                    )
+                    # Prepare fallbacks for special encodings even when skipped.
+                    if enable_fallback and encoding_iana in [
+                        "ascii",
+                        "utf_8",
+                        specified_encoding,
+                        "utf_16",
+                        "utf_32",
+                    ]:
+                        fallback_entry = CharsetMatch(
+                            sequences,
+                            encoding_iana,
+                            threshold,
+                            bom_or_sig_available,
+                            [],
+                            decoded_payload,
+                            preemptive_declaration=specified_encoding,
+                        )
+                        if encoding_iana == specified_encoding:
+                            fallback_specified = fallback_entry
+                        elif encoding_iana == "ascii":
+                            fallback_ascii = fallback_entry
+                        else:
+                            fallback_u8 = fallback_entry
+                    continue
+
+        max_chunk_gave_up: int = int(len(r_) / 4)
+
+        max_chunk_gave_up = max(max_chunk_gave_up, 2)
+        early_stop_count: int = 0
+        lazy_str_hard_failure = False
+
+        md_chunks: list[str] = []
+        md_ratios = []
+
+        try:
+            for chunk in cut_sequence_chunks(
+                sequences,
+                encoding_iana,
+                r_,
+                chunk_size,
+                bom_or_sig_available,
+                strip_sig_or_bom,
+                sig_payload,
+                is_multi_byte_decoder,
+                decoded_payload,
+            ):
+                md_chunks.append(chunk)
+
+                md_ratios.append(
+                    mess_ratio(
+                        chunk,
+                        threshold,
+                        explain is True and 1 <= len(cp_isolation) <= 2,
+                    )
+                )
+
+                if md_ratios[-1] >= threshold:
+                    early_stop_count += 1
+
+                if (early_stop_count >= max_chunk_gave_up) or (
+                    bom_or_sig_available and strip_sig_or_bom is False
+                ):
+                    break
+        except (
+            UnicodeDecodeError
+        ) as e:  # Lazy str loading may have missed something there
+            logger.log(
+                TRACE,
+                "LazyStr Loading: After MD chunk decode, code page %s does not fit given bytes sequence at ALL. %s",
+                encoding_iana,
+                str(e),
+            )
+            early_stop_count = max_chunk_gave_up
+            lazy_str_hard_failure = True
+
+        # We might want to check the sequence again with the whole content
+        # Only if initial MD tests passes
+        if (
+            not lazy_str_hard_failure
+            and is_too_large_sequence
+            and not is_multi_byte_decoder
+        ):
+            try:
+                sequences[int(50e3) :].decode(encoding_iana, errors="strict")
+            except UnicodeDecodeError as e:
+                logger.log(
+                    TRACE,
+                    "LazyStr Loading: After final lookup, code page %s does not fit given bytes sequence at ALL. %s",
+                    encoding_iana,
+                    str(e),
+                )
+                tested_but_hard_failure.append(encoding_iana)
+                continue
+
+        mean_mess_ratio: float = sum(md_ratios) / len(md_ratios) if md_ratios else 0.0
+        if mean_mess_ratio >= threshold or early_stop_count >= max_chunk_gave_up:
+            tested_but_soft_failure.append(encoding_iana)
+            if encoding_iana in IANA_SUPPORTED_SIMILAR:
+                soft_failure_skip.update(IANA_SUPPORTED_SIMILAR[encoding_iana])
+            # Cache this soft-failure so identical decoding from other encodings
+            # can be skipped immediately.
+            if decoded_payload is not None and not is_multi_byte_decoder:
+                payload_result_cache.setdefault(
+                    hash(decoded_payload), (mean_mess_ratio, [], False)
+                )
+            logger.log(
+                TRACE,
+                "%s was excluded because of initial chaos probing. Gave up %i time(s). "
+                "Computed mean chaos is %f %%.",
+                encoding_iana,
+                early_stop_count,
+                round(mean_mess_ratio * 100, ndigits=3),
+            )
+            # Preparing those fallbacks in case we got nothing.
+            if (
+                enable_fallback
+                and encoding_iana
+                in ["ascii", "utf_8", specified_encoding, "utf_16", "utf_32"]
+                and not lazy_str_hard_failure
+            ):
+                fallback_entry = CharsetMatch(
+                    sequences,
+                    encoding_iana,
+                    threshold,
+                    bom_or_sig_available,
+                    [],
+                    decoded_payload,
+                    preemptive_declaration=specified_encoding,
+                )
+                if encoding_iana == specified_encoding:
+                    fallback_specified = fallback_entry
+                elif encoding_iana == "ascii":
+                    fallback_ascii = fallback_entry
+                else:
+                    fallback_u8 = fallback_entry
+            continue
+
+        logger.log(
+            TRACE,
+            "%s passed initial chaos probing. Mean measured chaos is %f %%",
+            encoding_iana,
+            round(mean_mess_ratio * 100, ndigits=3),
+        )
+
+        if not is_multi_byte_decoder:
+            target_languages: list[str] = encoding_languages(encoding_iana)
+        else:
+            target_languages = mb_encoding_languages(encoding_iana)
+
+        if target_languages:
+            logger.log(
+                TRACE,
+                "{} should target any language(s) of {}".format(
+                    encoding_iana, str(target_languages)
+                ),
+            )
+
+        cd_ratios = []
+
+        # Run coherence detection on all chunks. We previously tried limiting to
+        # 1-2 chunks for post-definitive encodings to save time, but this caused
+        # coverage regressions by producing unrepresentative coherence scores.
+        # The SB cap and language-family skip optimizations provide sufficient
+        # speedup without sacrificing coherence accuracy.
+        if encoding_iana != "ascii":
+            # We shall skip the CD when its about ASCII
+            # Most of the time its not relevant to run "language-detection" on it.
+            for chunk in md_chunks:
+                chunk_languages = coherence_ratio(
+                    chunk,
+                    language_threshold,
+                    ",".join(target_languages) if target_languages else None,
+                )
+
+                cd_ratios.append(chunk_languages)
+            cd_ratios_merged = merge_coherence_ratios(cd_ratios)
+        else:
+            cd_ratios_merged = merge_coherence_ratios(cd_ratios)
+
+        if cd_ratios_merged:
+            logger.log(
+                TRACE,
+                "We detected language {} using {}".format(
+                    cd_ratios_merged, encoding_iana
+                ),
+            )
+
+        current_match = CharsetMatch(
+            sequences,
+            encoding_iana,
+            mean_mess_ratio,
+            bom_or_sig_available,
+            cd_ratios_merged,
+            (
+                decoded_payload
+                if (
+                    is_too_large_sequence is False
+                    or encoding_iana in [specified_encoding, "ascii", "utf_8"]
+                )
+                else None
+            ),
+            preemptive_declaration=specified_encoding,
+        )
+
+        results.append(current_match)
+
+        # Cache the successful result for payload-hash deduplication.
+        if decoded_payload is not None and not is_multi_byte_decoder:
+            payload_result_cache.setdefault(
+                hash(decoded_payload),
+                (mean_mess_ratio, cd_ratios_merged, True),
+            )
+
+        # Count post-definitive same-family SB successes for the early termination cap.
+        # Only count low-mess encodings (< 2%) toward the cap. High-mess encodings are
+        # marginal results that shouldn't prevent better-quality candidates from being
+        # tested. For example, iso8859_4 (mess=0%) should not be skipped just because
+        # 7 high-mess Latin encodings (cp1252 at 8%, etc.) were tried first.
+        if (
+            definitive_match_found
+            and not is_multi_byte_decoder
+            and mean_mess_ratio < 0.02
+        ):
+            post_definitive_sb_success_count += 1
+
+        if (
+            encoding_iana in [specified_encoding, "ascii", "utf_8"]
+            and mean_mess_ratio < 0.1
+        ):
+            # If md says nothing to worry about, then... stop immediately!
+            if mean_mess_ratio == 0.0:
+                logger.debug(
+                    "Encoding detection: %s is most likely the one.",
+                    current_match.encoding,
+                )
+                if explain:  # Defensive: ensure exit path clean handler
+                    logger.removeHandler(explain_handler)
+                    logger.setLevel(previous_logger_level)
+                return CharsetMatches([current_match])
+
+            early_stop_results.append(current_match)
+
+        if (
+            len(early_stop_results)
+            and (specified_encoding is None or specified_encoding in tested)
+            and "ascii" in tested
+            and "utf_8" in tested
+        ):
+            probable_result = early_stop_results.best()  # type: ignore[assignment]
+            logger.debug(
+                "Encoding detection: %s is most likely the one.",
+                probable_result.encoding,  # type: ignore[union-attr]
+            )
+            if explain:  # Defensive: ensure exit path clean handler
+                logger.removeHandler(explain_handler)
+                logger.setLevel(previous_logger_level)
+
+            return CharsetMatches([probable_result])
+
+        # Once we find a result with good coherence (>= 0.5) after testing the
+        # prioritized encodings (ascii, utf_8), activate "definitive mode": skip
+        # encodings that target completely different language families. This avoids
+        # running expensive mess_ratio + coherence_ratio on clearly unrelated
+        # candidates (e.g., Cyrillic encodings when the match is Latin-based).
+        # We require coherence >= 0.5 to avoid false positives (e.g., cp1251 decoding
+        # Hebrew text with 0.0 chaos but wrong language detection at coherence 0.33).
+        if not definitive_match_found and not is_multi_byte_decoder:
+            best_coherence = (
+                max((v for _, v in cd_ratios_merged), default=0.0)
+                if cd_ratios_merged
+                else 0.0
+            )
+            if best_coherence >= 0.5 and "ascii" in tested and "utf_8" in tested:
+                definitive_match_found = True
+                definitive_target_languages.update(target_languages)
+                logger.log(
+                    TRACE,
+                    "Definitive match found: %s (chaos=%.3f, coherence=%.2f). Encodings targeting different language families will be skipped.",
+                    encoding_iana,
+                    mean_mess_ratio,
+                    best_coherence,
+                )
+
+        # When a non-UTF multibyte encoding passes chaos probing with significant
+        # multibyte content (decoded < 98% of raw), activate mb_definitive_match.
+        # This skips all remaining single-byte encodings which would either soft-fail
+        # (running expensive mess_ratio for nothing) or produce inferior results.
+        if (
+            not mb_definitive_match_found
+            and is_multi_byte_decoder
+            and multi_byte_bonus
+            and decoded_payload is not None
+            and len(decoded_payload) < length * 0.98
+            and encoding_iana
+            not in {
+                "utf_8",
+                "utf_8_sig",
+                "utf_16",
+                "utf_16_be",
+                "utf_16_le",
+                "utf_32",
+                "utf_32_be",
+                "utf_32_le",
+                "utf_7",
+            }
+            and "ascii" in tested
+            and "utf_8" in tested
+        ):
+            mb_definitive_match_found = True
+            logger.log(
+                TRACE,
+                "Multi-byte definitive match: %s (chaos=%.3f, decoded=%d/%d=%.1f%%). Single-byte encodings will be skipped.",
+                encoding_iana,
+                mean_mess_ratio,
+                len(decoded_payload),
+                length,
+                len(decoded_payload) / length * 100,
+            )
+
+        if encoding_iana == sig_encoding:
+            logger.debug(
+                "Encoding detection: %s is most likely the one as we detected a BOM or SIG within "
+                "the beginning of the sequence.",
+                encoding_iana,
+            )
+            if explain:  # Defensive: ensure exit path clean handler
+                logger.removeHandler(explain_handler)
+                logger.setLevel(previous_logger_level)
+            return CharsetMatches([results[encoding_iana]])
+
+    if len(results) == 0:
+        if fallback_u8 or fallback_ascii or fallback_specified:
+            logger.log(
+                TRACE,
+                "Nothing got out of the detection process. Using ASCII/UTF-8/Specified fallback.",
+            )
+
+        if fallback_specified:
+            logger.debug(
+                "Encoding detection: %s will be used as a fallback match",
+                fallback_specified.encoding,
+            )
+            results.append(fallback_specified)
+        elif (
+            (fallback_u8 and fallback_ascii is None)
+            or (
+                fallback_u8
+                and fallback_ascii
+                and fallback_u8.fingerprint != fallback_ascii.fingerprint
+            )
+            or (fallback_u8 is not None)
+        ):
+            logger.debug("Encoding detection: utf_8 will be used as a fallback match")
+            results.append(fallback_u8)
+        elif fallback_ascii:
+            logger.debug("Encoding detection: ascii will be used as a fallback match")
+            results.append(fallback_ascii)
+
+    if results:
+        logger.debug(
+            "Encoding detection: Found %s as plausible (best-candidate) for content. With %i alternatives.",
+            results.best().encoding,  # type: ignore
+            len(results) - 1,
+        )
+    else:
+        logger.debug("Encoding detection: Unable to determine any suitable charset.")
+
+    if explain:
+        logger.removeHandler(explain_handler)
+        logger.setLevel(previous_logger_level)
+
+    return results
+
+
+def from_fp(
+    fp: BinaryIO,
+    steps: int = 5,
+    chunk_size: int = 512,
+    threshold: float = 0.20,
+    cp_isolation: list[str] | None = None,
+    cp_exclusion: list[str] | None = None,
+    preemptive_behaviour: bool = True,
+    explain: bool = False,
+    language_threshold: float = 0.1,
+    enable_fallback: bool = True,
+) -> CharsetMatches:
+    """
+    Same thing than the function from_bytes but using a file pointer that is already ready.
+    Will not close the file pointer.
+    """
+    return from_bytes(
+        fp.read(),
+        steps,
+        chunk_size,
+        threshold,
+        cp_isolation,
+        cp_exclusion,
+        preemptive_behaviour,
+        explain,
+        language_threshold,
+        enable_fallback,
+    )
+
+
+def from_path(
+    path: str | bytes | PathLike,  # type: ignore[type-arg]
+    steps: int = 5,
+    chunk_size: int = 512,
+    threshold: float = 0.20,
+    cp_isolation: list[str] | None = None,
+    cp_exclusion: list[str] | None = None,
+    preemptive_behaviour: bool = True,
+    explain: bool = False,
+    language_threshold: float = 0.1,
+    enable_fallback: bool = True,
+) -> CharsetMatches:
+    """
+    Same thing than the function from_bytes but with one extra step. Opening and reading given file path in binary mode.
+    Can raise IOError.
+    """
+    with open(path, "rb") as fp:
+        return from_fp(
+            fp,
+            steps,
+            chunk_size,
+            threshold,
+            cp_isolation,
+            cp_exclusion,
+            preemptive_behaviour,
+            explain,
+            language_threshold,
+            enable_fallback,
+        )
+
+
+def is_binary(
+    fp_or_path_or_payload: PathLike | str | BinaryIO | bytes,  # type: ignore[type-arg]
+    steps: int = 5,
+    chunk_size: int = 512,
+    threshold: float = 0.20,
+    cp_isolation: list[str] | None = None,
+    cp_exclusion: list[str] | None = None,
+    preemptive_behaviour: bool = True,
+    explain: bool = False,
+    language_threshold: float = 0.1,
+    enable_fallback: bool = False,
+) -> bool:
+    """
+    Detect if the given input (file, bytes, or path) points to a binary file. aka. not a string.
+    Based on the same main heuristic algorithms and default kwargs at the sole exception that fallbacks match
+    are disabled to be stricter around ASCII-compatible but unlikely to be a string.
+    """
+    if isinstance(fp_or_path_or_payload, (str, PathLike)):
+        guesses = from_path(
+            fp_or_path_or_payload,
+            steps=steps,
+            chunk_size=chunk_size,
+            threshold=threshold,
+            cp_isolation=cp_isolation,
+            cp_exclusion=cp_exclusion,
+            preemptive_behaviour=preemptive_behaviour,
+            explain=explain,
+            language_threshold=language_threshold,
+            enable_fallback=enable_fallback,
+        )
+    elif isinstance(
+        fp_or_path_or_payload,
+        (
+            bytes,
+            bytearray,
+        ),
+    ):
+        guesses = from_bytes(
+            fp_or_path_or_payload,
+            steps=steps,
+            chunk_size=chunk_size,
+            threshold=threshold,
+            cp_isolation=cp_isolation,
+            cp_exclusion=cp_exclusion,
+            preemptive_behaviour=preemptive_behaviour,
+            explain=explain,
+            language_threshold=language_threshold,
+            enable_fallback=enable_fallback,
+        )
+    else:
+        guesses = from_fp(
+            fp_or_path_or_payload,
+            steps=steps,
+            chunk_size=chunk_size,
+            threshold=threshold,
+            cp_isolation=cp_isolation,
+            cp_exclusion=cp_exclusion,
+            preemptive_behaviour=preemptive_behaviour,
+            explain=explain,
+            language_threshold=language_threshold,
+            enable_fallback=enable_fallback,
+        )
+
+    return not guesses

.venv/Lib/site-packages/charset_normalizer/cd.py

@@ -0,0 +1,454 @@
+from __future__ import annotations
+
+import importlib
+from codecs import IncrementalDecoder
+from collections import Counter
+from functools import lru_cache
+from typing import Counter as TypeCounter
+
+from .constant import (
+    FREQUENCIES,
+    KO_NAMES,
+    LANGUAGE_SUPPORTED_COUNT,
+    TOO_SMALL_SEQUENCE,
+    ZH_NAMES,
+    _FREQUENCIES_SET,
+    _FREQUENCIES_RANK,
+)
+from .md import is_suspiciously_successive_range
+from .models import CoherenceMatches
+from .utils import (
+    is_accentuated,
+    is_latin,
+    is_multi_byte_encoding,
+    is_unicode_range_secondary,
+    unicode_range,
+)
+
+
+def encoding_unicode_range(iana_name: str) -> list[str]:
+    """
+    Return associated unicode ranges in a single byte code page.
+    """
+    if is_multi_byte_encoding(iana_name):
+        raise OSError(  # Defensive:
+            "Function not supported on multi-byte code page"
+        )
+
+    decoder = importlib.import_module(f"encodings.{iana_name}").IncrementalDecoder
+
+    p: IncrementalDecoder = decoder(errors="ignore")
+    seen_ranges: dict[str, int] = {}
+    character_count: int = 0
+
+    for i in range(0x40, 0xFF):
+        chunk: str = p.decode(bytes([i]))
+
+        if chunk:
+            character_range: str | None = unicode_range(chunk)
+
+            if character_range is None:
+                continue
+
+            if is_unicode_range_secondary(character_range) is False:
+                if character_range not in seen_ranges:
+                    seen_ranges[character_range] = 0
+                seen_ranges[character_range] += 1
+                character_count += 1
+
+    return sorted(
+        [
+            character_range
+            for character_range in seen_ranges
+            if seen_ranges[character_range] / character_count >= 0.15
+        ]
+    )
+
+
+def unicode_range_languages(primary_range: str) -> list[str]:
+    """
+    Return inferred languages used with a unicode range.
+    """
+    languages: list[str] = []
+
+    for language, characters in FREQUENCIES.items():
+        for character in characters:
+            if unicode_range(character) == primary_range:
+                languages.append(language)
+                break
+
+    return languages
+
+
+@lru_cache()
+def encoding_languages(iana_name: str) -> list[str]:
+    """
+    Single-byte encoding language association. Some code page are heavily linked to particular language(s).
+    This function does the correspondence.
+    """
+    unicode_ranges: list[str] = encoding_unicode_range(iana_name)
+    primary_range: str | None = None
+
+    for specified_range in unicode_ranges:
+        if "Latin" not in specified_range:
+            primary_range = specified_range
+            break
+
+    if primary_range is None:
+        return ["Latin Based"]
+
+    return unicode_range_languages(primary_range)
+
+
+@lru_cache()
+def mb_encoding_languages(iana_name: str) -> list[str]:
+    """
+    Multi-byte encoding language association. Some code page are heavily linked to particular language(s).
+    This function does the correspondence.
+    """
+    if (
+        iana_name.startswith("shift_")
+        or iana_name.startswith("iso2022_jp")
+        or iana_name.startswith("euc_j")
+        or iana_name == "cp932"
+    ):
+        return ["Japanese"]
+    if iana_name.startswith("gb") or iana_name in ZH_NAMES:
+        return ["Chinese"]
+    if iana_name.startswith("iso2022_kr") or iana_name in KO_NAMES:
+        return ["Korean"]
+
+    return []
+
+
+@lru_cache(maxsize=LANGUAGE_SUPPORTED_COUNT)
+def get_target_features(language: str) -> tuple[bool, bool]:
+    """
+    Determine main aspects from a supported language if it contains accents and if is pure Latin.
+    """
+    target_have_accents: bool = False
+    target_pure_latin: bool = True
+
+    for character in FREQUENCIES[language]:
+        if not target_have_accents and is_accentuated(character):
+            target_have_accents = True
+        if target_pure_latin and is_latin(character) is False:
+            target_pure_latin = False
+
+    return target_have_accents, target_pure_latin
+
+
+def alphabet_languages(
+    characters: list[str], ignore_non_latin: bool = False
+) -> list[str]:
+    """
+    Return associated languages associated to given characters.
+    """
+    languages: list[tuple[str, float]] = []
+
+    characters_set: frozenset[str] = frozenset(characters)
+    source_have_accents = any(is_accentuated(character) for character in characters)
+
+    for language, language_characters in FREQUENCIES.items():
+        target_have_accents, target_pure_latin = get_target_features(language)
+
+        if ignore_non_latin and target_pure_latin is False:
+            continue
+
+        if target_have_accents is False and source_have_accents:
+            continue
+
+        character_count: int = len(language_characters)
+
+        character_match_count: int = len(_FREQUENCIES_SET[language] & characters_set)
+
+        ratio: float = character_match_count / character_count
+
+        if ratio >= 0.2:
+            languages.append((language, ratio))
+
+    languages = sorted(languages, key=lambda x: x[1], reverse=True)
+
+    return [compatible_language[0] for compatible_language in languages]
+
+
+def characters_popularity_compare(
+    language: str, ordered_characters: list[str]
+) -> float:
+    """
+    Determine if a ordered characters list (by occurrence from most appearance to rarest) match a particular language.
+    The result is a ratio between 0. (absolutely no correspondence) and 1. (near perfect fit).
+    Beware that is function is not strict on the match in order to ease the detection. (Meaning close match is 1.)
+    """
+    if language not in FREQUENCIES:
+        raise ValueError(f"{language} not available")  # Defensive:
+
+    character_approved_count: int = 0
+    frequencies_language_set: frozenset[str] = _FREQUENCIES_SET[language]
+    lang_rank: dict[str, int] = _FREQUENCIES_RANK[language]
+
+    ordered_characters_count: int = len(ordered_characters)
+    target_language_characters_count: int = len(FREQUENCIES[language])
+
+    large_alphabet: bool = target_language_characters_count > 26
+
+    expected_projection_ratio: float = (
+        target_language_characters_count / ordered_characters_count
+    )
+
+    # Pre-built rank dict for ordered_characters (avoids repeated list slicing).
+    ordered_rank: dict[str, int] = {
+        char: rank for rank, char in enumerate(ordered_characters)
+    }
+
+    # Pre-compute characters common to both orderings.
+    # Avoids repeated `c in ordered_rank` dict lookups in the inner counts.
+    common_chars: list[tuple[int, int]] = [
+        (lr, ordered_rank[c]) for c, lr in lang_rank.items() if c in ordered_rank
+    ]
+
+    # Pre-extract lr and orr arrays for faster iteration in the inner loop.
+    # Plain integer loops with local arrays are much faster under mypyc than
+    # generator expression sums over a list of tuples.
+    common_count: int = len(common_chars)
+    common_lr: list[int] = [p[0] for p in common_chars]
+    common_orr: list[int] = [p[1] for p in common_chars]
+
+    for character, character_rank in zip(
+        ordered_characters, range(0, ordered_characters_count)
+    ):
+        if character not in frequencies_language_set:
+            continue
+
+        character_rank_in_language: int = lang_rank[character]
+        character_rank_projection: int = int(character_rank * expected_projection_ratio)
+
+        if (
+            large_alphabet is False
+            and abs(character_rank_projection - character_rank_in_language) > 4
+        ):
+            continue
+
+        if (
+            large_alphabet is True
+            and abs(character_rank_projection - character_rank_in_language)
+            < target_language_characters_count / 3
+        ):
+            character_approved_count += 1
+            continue
+
+        # Count how many characters appear "before" in both orderings,
+        # and how many appear "at or after" in both orderings.
+        # Single pass over pre-extracted arrays — much faster under mypyc
+        # than two generator expression sums.
+        before_match_count: int = 0
+        after_match_count: int = 0
+        for i in range(common_count):
+            lr_i: int = common_lr[i]
+            orr_i: int = common_orr[i]
+            if lr_i < character_rank_in_language:
+                if orr_i < character_rank:
+                    before_match_count += 1
+            else:
+                if orr_i >= character_rank:
+                    after_match_count += 1
+
+        after_len: int = target_language_characters_count - character_rank_in_language
+
+        if character_rank_in_language == 0 and before_match_count <= 4:
+            character_approved_count += 1
+            continue
+
+        if after_len == 0 and after_match_count <= 4:
+            character_approved_count += 1
+            continue
+
+        if (
+            character_rank_in_language > 0
+            and before_match_count / character_rank_in_language >= 0.4
+        ) or (after_len > 0 and after_match_count / after_len >= 0.4):
+            character_approved_count += 1
+            continue
+
+    return character_approved_count / len(ordered_characters)
+
+
+def alpha_unicode_split(decoded_sequence: str) -> list[str]:
+    """
+    Given a decoded text sequence, return a list of str. Unicode range / alphabet separation.
+    Ex. a text containing English/Latin with a bit a Hebrew will return two items in the resulting list;
+    One containing the latin letters and the other hebrew.
+    """
+    layers: dict[str, list[str]] = {}
+
+    # Fast path: track single-layer key to skip dict iteration for single-script text.
+    single_layer_key: str | None = None
+    multi_layer: bool = False
+
+    # Cache the last character_range and its resolved layer to avoid repeated
+    # is_suspiciously_successive_range calls for consecutive same-range chars.
+    prev_character_range: str | None = None
+    prev_layer_target: str | None = None
+
+    for character in decoded_sequence:
+        if character.isalpha() is False:
+            continue
+
+        # ASCII fast-path: a-z and A-Z are always "Basic Latin".
+        # Avoids unicode_range() function call overhead for the most common case.
+        character_ord: int = ord(character)
+        if character_ord < 128:
+            character_range: str | None = "Basic Latin"
+        else:
+            character_range = unicode_range(character)
+
+        if character_range is None:
+            continue
+
+        # Fast path: same range as previous character → reuse cached layer target.
+        if character_range == prev_character_range:
+            if prev_layer_target is not None:
+                layers[prev_layer_target].append(character)
+            continue
+
+        layer_target_range: str | None = None
+
+        if multi_layer:
+            for discovered_range in layers:
+                if (
+                    is_suspiciously_successive_range(discovered_range, character_range)
+                    is False
+                ):
+                    layer_target_range = discovered_range
+                    break
+        elif single_layer_key is not None:
+            if (
+                is_suspiciously_successive_range(single_layer_key, character_range)
+                is False
+            ):
+                layer_target_range = single_layer_key
+
+        if layer_target_range is None:
+            layer_target_range = character_range
+
+        if layer_target_range not in layers:
+            layers[layer_target_range] = []
+            if single_layer_key is None:
+                single_layer_key = layer_target_range
+            else:
+                multi_layer = True
+
+        layers[layer_target_range].append(character)
+
+        # Cache for next iteration
+        prev_character_range = character_range
+        prev_layer_target = layer_target_range
+
+    return ["".join(chars).lower() for chars in layers.values()]
+
+
+def merge_coherence_ratios(results: list[CoherenceMatches]) -> CoherenceMatches:
+    """
+    This function merge results previously given by the function coherence_ratio.
+    The return type is the same as coherence_ratio.
+    """
+    per_language_ratios: dict[str, list[float]] = {}
+    for result in results:
+        for sub_result in result:
+            language, ratio = sub_result
+            if language not in per_language_ratios:
+                per_language_ratios[language] = [ratio]
+                continue
+            per_language_ratios[language].append(ratio)
+
+    merge = [
+        (
+            language,
+            round(
+                sum(per_language_ratios[language]) / len(per_language_ratios[language]),
+                4,
+            ),
+        )
+        for language in per_language_ratios
+    ]
+
+    return sorted(merge, key=lambda x: x[1], reverse=True)
+
+
+def filter_alt_coherence_matches(results: CoherenceMatches) -> CoherenceMatches:
+    """
+    We shall NOT return "English—" in CoherenceMatches because it is an alternative
+    of "English". This function only keeps the best match and remove the em-dash in it.
+    """
+    index_results: dict[str, list[float]] = dict()
+
+    for result in results:
+        language, ratio = result
+        no_em_name: str = language.replace("—", "")
+
+        if no_em_name not in index_results:
+            index_results[no_em_name] = []
+
+        index_results[no_em_name].append(ratio)
+
+    if any(len(index_results[e]) > 1 for e in index_results):
+        filtered_results: CoherenceMatches = []
+
+        for language in index_results:
+            filtered_results.append((language, max(index_results[language])))
+
+        return filtered_results
+
+    return results
+
+
+@lru_cache(maxsize=2048)
+def coherence_ratio(
+    decoded_sequence: str, threshold: float = 0.1, lg_inclusion: str | None = None
+) -> CoherenceMatches:
+    """
+    Detect ANY language that can be identified in given sequence. The sequence will be analysed by layers.
+    A layer = Character extraction by alphabets/ranges.
+    """
+
+    results: list[tuple[str, float]] = []
+    ignore_non_latin: bool = False
+
+    sufficient_match_count: int = 0
+
+    lg_inclusion_list = lg_inclusion.split(",") if lg_inclusion is not None else []
+    if "Latin Based" in lg_inclusion_list:
+        ignore_non_latin = True
+        lg_inclusion_list.remove("Latin Based")
+
+    for layer in alpha_unicode_split(decoded_sequence):
+        sequence_frequencies: TypeCounter[str] = Counter(layer)
+        most_common = sequence_frequencies.most_common()
+
+        character_count: int = len(layer)
+
+        if character_count <= TOO_SMALL_SEQUENCE:
+            continue
+
+        popular_character_ordered: list[str] = [c for c, o in most_common]
+
+        for language in lg_inclusion_list or alphabet_languages(
+            popular_character_ordered, ignore_non_latin
+        ):
+            ratio: float = characters_popularity_compare(
+                language, popular_character_ordered
+            )
+
+            if ratio < threshold:
+                continue
+            elif ratio >= 0.8:
+                sufficient_match_count += 1
+
+            results.append((language, round(ratio, 4)))
+
+            if sufficient_match_count >= 3:
+                break
+
+    return sorted(
+        filter_alt_coherence_matches(results), key=lambda x: x[1], reverse=True
+    )

.venv/Lib/site-packages/charset_normalizer/cli/__init__.py

@@ -0,0 +1,8 @@
+from __future__ import annotations
+
+from .__main__ import cli_detect, query_yes_no
+
+__all__ = (
+    "cli_detect",
+    "query_yes_no",
+)

.venv/Lib/site-packages/charset_normalizer/cli/__main__.py

@@ -0,0 +1,362 @@
+from __future__ import annotations
+
+import argparse
+import sys
+import typing
+from json import dumps
+from os.path import abspath, basename, dirname, join, realpath
+from platform import python_version
+from unicodedata import unidata_version
+
+import charset_normalizer.md as md_module
+from charset_normalizer import from_fp
+from charset_normalizer.models import CliDetectionResult
+from charset_normalizer.version import __version__
+
+
+def query_yes_no(question: str, default: str = "yes") -> bool:  # Defensive:
+    """Ask a yes/no question via input() and return the answer as a bool."""
+    prompt = " [Y/n] " if default == "yes" else " [y/N] "
+
+    while True:
+        choice = input(question + prompt).strip().lower()
+        if not choice:
+            return default == "yes"
+        if choice in ("y", "yes"):
+            return True
+        if choice in ("n", "no"):
+            return False
+        print("Please respond with 'y' or 'n'.")
+
+
+class FileType:
+    """Factory for creating file object types
+
+    Instances of FileType are typically passed as type= arguments to the
+    ArgumentParser add_argument() method.
+
+    Keyword Arguments:
+        - mode -- A string indicating how the file is to be opened. Accepts the
+            same values as the builtin open() function.
+        - bufsize -- The file's desired buffer size. Accepts the same values as
+            the builtin open() function.
+        - encoding -- The file's encoding. Accepts the same values as the
+            builtin open() function.
+        - errors -- A string indicating how encoding and decoding errors are to
+            be handled. Accepts the same value as the builtin open() function.
+
+    Backported from CPython 3.12
+    """
+
+    def __init__(
+        self,
+        mode: str = "r",
+        bufsize: int = -1,
+        encoding: str | None = None,
+        errors: str | None = None,
+    ):
+        self._mode = mode
+        self._bufsize = bufsize
+        self._encoding = encoding
+        self._errors = errors
+
+    def __call__(self, string: str) -> typing.IO:  # type: ignore[type-arg]
+        # the special argument "-" means sys.std{in,out}
+        if string == "-":
+            if "r" in self._mode:
+                return sys.stdin.buffer if "b" in self._mode else sys.stdin
+            elif any(c in self._mode for c in "wax"):
+                return sys.stdout.buffer if "b" in self._mode else sys.stdout
+            else:
+                msg = f'argument "-" with mode {self._mode}'
+                raise ValueError(msg)
+
+        # all other arguments are used as file names
+        try:
+            return open(string, self._mode, self._bufsize, self._encoding, self._errors)
+        except OSError as e:
+            message = f"can't open '{string}': {e}"
+            raise argparse.ArgumentTypeError(message)
+
+    def __repr__(self) -> str:
+        args = self._mode, self._bufsize
+        kwargs = [("encoding", self._encoding), ("errors", self._errors)]
+        args_str = ", ".join(
+            [repr(arg) for arg in args if arg != -1]
+            + [f"{kw}={arg!r}" for kw, arg in kwargs if arg is not None]
+        )
+        return f"{type(self).__name__}({args_str})"
+
+
+def cli_detect(argv: list[str] | None = None) -> int:
+    """
+    CLI assistant using ARGV and ArgumentParser
+    :param argv:
+    :return: 0 if everything is fine, anything else equal trouble
+    """
+    parser = argparse.ArgumentParser(
+        description="The Real First Universal Charset Detector. "
+        "Discover originating encoding used on text file. "
+        "Normalize text to unicode."
+    )
+
+    parser.add_argument(
+        "files", type=FileType("rb"), nargs="+", help="File(s) to be analysed"
+    )
+    parser.add_argument(
+        "-v",
+        "--verbose",
+        action="store_true",
+        default=False,
+        dest="verbose",
+        help="Display complementary information about file if any. "
+        "Stdout will contain logs about the detection process.",
+    )
+    parser.add_argument(
+        "-a",
+        "--with-alternative",
+        action="store_true",
+        default=False,
+        dest="alternatives",
+        help="Output complementary possibilities if any. Top-level JSON WILL be a list.",
+    )
+    parser.add_argument(
+        "-n",
+        "--normalize",
+        action="store_true",
+        default=False,
+        dest="normalize",
+        help="Permit to normalize input file. If not set, program does not write anything.",
+    )
+    parser.add_argument(
+        "-m",
+        "--minimal",
+        action="store_true",
+        default=False,
+        dest="minimal",
+        help="Only output the charset detected to STDOUT. Disabling JSON output.",
+    )
+    parser.add_argument(
+        "-r",
+        "--replace",
+        action="store_true",
+        default=False,
+        dest="replace",
+        help="Replace file when trying to normalize it instead of creating a new one.",
+    )
+    parser.add_argument(
+        "-f",
+        "--force",
+        action="store_true",
+        default=False,
+        dest="force",
+        help="Replace file without asking if you are sure, use this flag with caution.",
+    )
+    parser.add_argument(
+        "-i",
+        "--no-preemptive",
+        action="store_true",
+        default=False,
+        dest="no_preemptive",
+        help="Disable looking at a charset declaration to hint the detector.",
+    )
+    parser.add_argument(
+        "-t",
+        "--threshold",
+        action="store",
+        default=0.2,
+        type=float,
+        dest="threshold",
+        help="Define a custom maximum amount of noise allowed in decoded content. 0. <= noise <= 1.",
+    )
+    parser.add_argument(
+        "--version",
+        action="version",
+        version="Charset-Normalizer {} - Python {} - Unicode {} - SpeedUp {}".format(
+            __version__,
+            python_version(),
+            unidata_version,
+            "OFF" if md_module.__file__.lower().endswith(".py") else "ON",
+        ),
+        help="Show version information and exit.",
+    )
+
+    args = parser.parse_args(argv)
+
+    if args.replace is True and args.normalize is False:
+        if args.files:
+            for my_file in args.files:
+                my_file.close()
+        print("Use --replace in addition of --normalize only.", file=sys.stderr)
+        return 1
+
+    if args.force is True and args.replace is False:
+        if args.files:
+            for my_file in args.files:
+                my_file.close()
+        print("Use --force in addition of --replace only.", file=sys.stderr)
+        return 1
+
+    if args.threshold < 0.0 or args.threshold > 1.0:
+        if args.files:
+            for my_file in args.files:
+                my_file.close()
+        print("--threshold VALUE should be between 0. AND 1.", file=sys.stderr)
+        return 1
+
+    x_ = []
+
+    for my_file in args.files:
+        matches = from_fp(
+            my_file,
+            threshold=args.threshold,
+            explain=args.verbose,
+            preemptive_behaviour=args.no_preemptive is False,
+        )
+
+        best_guess = matches.best()
+
+        if best_guess is None:
+            print(
+                'Unable to identify originating encoding for "{}". {}'.format(
+                    my_file.name,
+                    (
+                        "Maybe try increasing maximum amount of chaos."
+                        if args.threshold < 1.0
+                        else ""
+                    ),
+                ),
+                file=sys.stderr,
+            )
+            x_.append(
+                CliDetectionResult(
+                    abspath(my_file.name),
+                    None,
+                    [],
+                    [],
+                    "Unknown",
+                    [],
+                    False,
+                    1.0,
+                    0.0,
+                    None,
+                    True,
+                )
+            )
+        else:
+            cli_result = CliDetectionResult(
+                abspath(my_file.name),
+                best_guess.encoding,
+                best_guess.encoding_aliases,
+                [
+                    cp
+                    for cp in best_guess.could_be_from_charset
+                    if cp != best_guess.encoding
+                ],
+                best_guess.language,
+                best_guess.alphabets,
+                best_guess.bom,
+                best_guess.percent_chaos,
+                best_guess.percent_coherence,
+                None,
+                True,
+            )
+            x_.append(cli_result)
+
+            if len(matches) > 1 and args.alternatives:
+                for el in matches:
+                    if el != best_guess:
+                        x_.append(
+                            CliDetectionResult(
+                                abspath(my_file.name),
+                                el.encoding,
+                                el.encoding_aliases,
+                                [
+                                    cp
+                                    for cp in el.could_be_from_charset
+                                    if cp != el.encoding
+                                ],
+                                el.language,
+                                el.alphabets,
+                                el.bom,
+                                el.percent_chaos,
+                                el.percent_coherence,
+                                None,
+                                False,
+                            )
+                        )
+
+            if args.normalize is True:
+                if best_guess.encoding.startswith("utf") is True:
+                    print(
+                        '"{}" file does not need to be normalized, as it already came from unicode.'.format(
+                            my_file.name
+                        ),
+                        file=sys.stderr,
+                    )
+                    if my_file.closed is False:
+                        my_file.close()
+                    continue
+
+                dir_path = dirname(realpath(my_file.name))
+                file_name = basename(realpath(my_file.name))
+
+                o_: list[str] = file_name.split(".")
+
+                if args.replace is False:
+                    o_.insert(-1, best_guess.encoding)
+                    if my_file.closed is False:
+                        my_file.close()
+                elif (
+                    args.force is False
+                    and query_yes_no(
+                        'Are you sure to normalize "{}" by replacing it ?'.format(
+                            my_file.name
+                        ),
+                        "no",
+                    )
+                    is False
+                ):
+                    if my_file.closed is False:
+                        my_file.close()
+                    continue
+
+                try:
+                    cli_result.unicode_path = join(dir_path, ".".join(o_))
+
+                    with open(cli_result.unicode_path, "wb") as fp:
+                        fp.write(best_guess.output())
+                except OSError as e:  # Defensive:
+                    print(str(e), file=sys.stderr)
+                    if my_file.closed is False:
+                        my_file.close()
+                    return 2
+
+        if my_file.closed is False:
+            my_file.close()
+
+    if args.minimal is False:
+        print(
+            dumps(
+                [el.__dict__ for el in x_] if len(x_) > 1 else x_[0].__dict__,
+                ensure_ascii=True,
+                indent=4,
+            )
+        )
+    else:
+        for my_file in args.files:
+            print(
+                ", ".join(
+                    [
+                        el.encoding or "undefined"
+                        for el in x_
+                        if el.path == abspath(my_file.name)
+                    ]
+                )
+            )
+
+    return 0
+
+
+if __name__ == "__main__":  # Defensive:
+    cli_detect()

.venv/Lib/site-packages/charset_normalizer/constant.py

@@ -0,0 +1,2050 @@
+from __future__ import annotations
+
+from codecs import BOM_UTF8, BOM_UTF16_BE, BOM_UTF16_LE, BOM_UTF32_BE, BOM_UTF32_LE
+from encodings.aliases import aliases
+from re import IGNORECASE
+from re import compile as re_compile
+
+# Contain for each eligible encoding a list of/item bytes SIG/BOM
+ENCODING_MARKS: dict[str, bytes | list[bytes]] = {
+    "utf_8": BOM_UTF8,
+    "utf_7": [
+        b"\x2b\x2f\x76\x38\x2d",
+        b"\x2b\x2f\x76\x38",
+        b"\x2b\x2f\x76\x39",
+        b"\x2b\x2f\x76\x2b",
+        b"\x2b\x2f\x76\x2f",
+    ],
+    "gb18030": b"\x84\x31\x95\x33",
+    "utf_32": [BOM_UTF32_BE, BOM_UTF32_LE],
+    "utf_16": [BOM_UTF16_BE, BOM_UTF16_LE],
+}
+
+TOO_SMALL_SEQUENCE: int = 32
+TOO_BIG_SEQUENCE: int = int(10e6)
+
+UTF8_MAXIMAL_ALLOCATION: int = 1_112_064
+
+# Up-to-date Unicode ucd/17.0.0
+UNICODE_RANGES_COMBINED: dict[str, range] = {
+    "Control character": range(32),
+    "Basic Latin": range(32, 128),
+    "Latin-1 Supplement": range(128, 256),
+    "Latin Extended-A": range(256, 384),
+    "Latin Extended-B": range(384, 592),
+    "IPA Extensions": range(592, 688),
+    "Spacing Modifier Letters": range(688, 768),
+    "Combining Diacritical Marks": range(768, 880),
+    "Greek and Coptic": range(880, 1024),
+    "Cyrillic": range(1024, 1280),
+    "Cyrillic Supplement": range(1280, 1328),
+    "Armenian": range(1328, 1424),
+    "Hebrew": range(1424, 1536),
+    "Arabic": range(1536, 1792),
+    "Syriac": range(1792, 1872),
+    "Arabic Supplement": range(1872, 1920),
+    "Thaana": range(1920, 1984),
+    "NKo": range(1984, 2048),
+    "Samaritan": range(2048, 2112),
+    "Mandaic": range(2112, 2144),
+    "Syriac Supplement": range(2144, 2160),
+    "Arabic Extended-B": range(2160, 2208),
+    "Arabic Extended-A": range(2208, 2304),
+    "Devanagari": range(2304, 2432),
+    "Bengali": range(2432, 2560),
+    "Gurmukhi": range(2560, 2688),
+    "Gujarati": range(2688, 2816),
+    "Oriya": range(2816, 2944),
+    "Tamil": range(2944, 3072),
+    "Telugu": range(3072, 3200),
+    "Kannada": range(3200, 3328),
+    "Malayalam": range(3328, 3456),
+    "Sinhala": range(3456, 3584),
+    "Thai": range(3584, 3712),
+    "Lao": range(3712, 3840),
+    "Tibetan": range(3840, 4096),
+    "Myanmar": range(4096, 4256),
+    "Georgian": range(4256, 4352),
+    "Hangul Jamo": range(4352, 4608),
+    "Ethiopic": range(4608, 4992),
+    "Ethiopic Supplement": range(4992, 5024),
+    "Cherokee": range(5024, 5120),
+    "Unified Canadian Aboriginal Syllabics": range(5120, 5760),
+    "Ogham": range(5760, 5792),
+    "Runic": range(5792, 5888),
+    "Tagalog": range(5888, 5920),
+    "Hanunoo": range(5920, 5952),
+    "Buhid": range(5952, 5984),
+    "Tagbanwa": range(5984, 6016),
+    "Khmer": range(6016, 6144),
+    "Mongolian": range(6144, 6320),
+    "Unified Canadian Aboriginal Syllabics Extended": range(6320, 6400),
+    "Limbu": range(6400, 6480),
+    "Tai Le": range(6480, 6528),
+    "New Tai Lue": range(6528, 6624),
+    "Khmer Symbols": range(6624, 6656),
+    "Buginese": range(6656, 6688),
+    "Tai Tham": range(6688, 6832),
+    "Combining Diacritical Marks Extended": range(6832, 6912),
+    "Balinese": range(6912, 7040),
+    "Sundanese": range(7040, 7104),
+    "Batak": range(7104, 7168),
+    "Lepcha": range(7168, 7248),
+    "Ol Chiki": range(7248, 7296),
+    "Cyrillic Extended-C": range(7296, 7312),
+    "Georgian Extended": range(7312, 7360),
+    "Sundanese Supplement": range(7360, 7376),
+    "Vedic Extensions": range(7376, 7424),
+    "Phonetic Extensions": range(7424, 7552),
+    "Phonetic Extensions Supplement": range(7552, 7616),
+    "Combining Diacritical Marks Supplement": range(7616, 7680),
+    "Latin Extended Additional": range(7680, 7936),
+    "Greek Extended": range(7936, 8192),
+    "General Punctuation": range(8192, 8304),
+    "Superscripts and Subscripts": range(8304, 8352),
+    "Currency Symbols": range(8352, 8400),
+    "Combining Diacritical Marks for Symbols": range(8400, 8448),
+    "Letterlike Symbols": range(8448, 8528),
+    "Number Forms": range(8528, 8592),
+    "Arrows": range(8592, 8704),
+    "Mathematical Operators": range(8704, 8960),
+    "Miscellaneous Technical": range(8960, 9216),
+    "Control Pictures": range(9216, 9280),
+    "Optical Character Recognition": range(9280, 9312),
+    "Enclosed Alphanumerics": range(9312, 9472),
+    "Box Drawing": range(9472, 9600),
+    "Block Elements": range(9600, 9632),
+    "Geometric Shapes": range(9632, 9728),
+    "Miscellaneous Symbols": range(9728, 9984),
+    "Dingbats": range(9984, 10176),
+    "Miscellaneous Mathematical Symbols-A": range(10176, 10224),
+    "Supplemental Arrows-A": range(10224, 10240),
+    "Braille Patterns": range(10240, 10496),
+    "Supplemental Arrows-B": range(10496, 10624),
+    "Miscellaneous Mathematical Symbols-B": range(10624, 10752),
+    "Supplemental Mathematical Operators": range(10752, 11008),
+    "Miscellaneous Symbols and Arrows": range(11008, 11264),
+    "Glagolitic": range(11264, 11360),
+    "Latin Extended-C": range(11360, 11392),
+    "Coptic": range(11392, 11520),
+    "Georgian Supplement": range(11520, 11568),
+    "Tifinagh": range(11568, 11648),
+    "Ethiopic Extended": range(11648, 11744),
+    "Cyrillic Extended-A": range(11744, 11776),
+    "Supplemental Punctuation": range(11776, 11904),
+    "CJK Radicals Supplement": range(11904, 12032),
+    "Kangxi Radicals": range(12032, 12256),
+    "Ideographic Description Characters": range(12272, 12288),
+    "CJK Symbols and Punctuation": range(12288, 12352),
+    "Hiragana": range(12352, 12448),
+    "Katakana": range(12448, 12544),
+    "Bopomofo": range(12544, 12592),
+    "Hangul Compatibility Jamo": range(12592, 12688),
+    "Kanbun": range(12688, 12704),
+    "Bopomofo Extended": range(12704, 12736),
+    "CJK Strokes": range(12736, 12784),
+    "Katakana Phonetic Extensions": range(12784, 12800),
+    "Enclosed CJK Letters and Months": range(12800, 13056),
+    "CJK Compatibility": range(13056, 13312),
+    "CJK Unified Ideographs Extension A": range(13312, 19904),
+    "Yijing Hexagram Symbols": range(19904, 19968),
+    "CJK Unified Ideographs": range(19968, 40960),
+    "Yi Syllables": range(40960, 42128),
+    "Yi Radicals": range(42128, 42192),
+    "Lisu": range(42192, 42240),
+    "Vai": range(42240, 42560),
+    "Cyrillic Extended-B": range(42560, 42656),
+    "Bamum": range(42656, 42752),
+    "Modifier Tone Letters": range(42752, 42784),
+    "Latin Extended-D": range(42784, 43008),
+    "Syloti Nagri": range(43008, 43056),
+    "Common Indic Number Forms": range(43056, 43072),
+    "Phags-pa": range(43072, 43136),
+    "Saurashtra": range(43136, 43232),
+    "Devanagari Extended": range(43232, 43264),
+    "Kayah Li": range(43264, 43312),
+    "Rejang": range(43312, 43360),
+    "Hangul Jamo Extended-A": range(43360, 43392),
+    "Javanese": range(43392, 43488),
+    "Myanmar Extended-B": range(43488, 43520),
+    "Cham": range(43520, 43616),
+    "Myanmar Extended-A": range(43616, 43648),
+    "Tai Viet": range(43648, 43744),
+    "Meetei Mayek Extensions": range(43744, 43776),
+    "Ethiopic Extended-A": range(43776, 43824),
+    "Latin Extended-E": range(43824, 43888),
+    "Cherokee Supplement": range(43888, 43968),
+    "Meetei Mayek": range(43968, 44032),
+    "Hangul Syllables": range(44032, 55216),
+    "Hangul Jamo Extended-B": range(55216, 55296),
+    "High Surrogates": range(55296, 56192),
+    "High Private Use Surrogates": range(56192, 56320),
+    "Low Surrogates": range(56320, 57344),
+    "Private Use Area": range(57344, 63744),
+    "CJK Compatibility Ideographs": range(63744, 64256),
+    "Alphabetic Presentation Forms": range(64256, 64336),
+    "Arabic Presentation Forms-A": range(64336, 65024),
+    "Variation Selectors": range(65024, 65040),
+    "Vertical Forms": range(65040, 65056),
+    "Combining Half Marks": range(65056, 65072),
+    "CJK Compatibility Forms": range(65072, 65104),
+    "Small Form Variants": range(65104, 65136),
+    "Arabic Presentation Forms-B": range(65136, 65280),
+    "Halfwidth and Fullwidth Forms": range(65280, 65520),
+    "Specials": range(65520, 65536),
+    "Linear B Syllabary": range(65536, 65664),
+    "Linear B Ideograms": range(65664, 65792),
+    "Aegean Numbers": range(65792, 65856),
+    "Ancient Greek Numbers": range(65856, 65936),
+    "Ancient Symbols": range(65936, 66000),
+    "Phaistos Disc": range(66000, 66048),
+    "Lycian": range(66176, 66208),
+    "Carian": range(66208, 66272),
+    "Coptic Epact Numbers": range(66272, 66304),
+    "Old Italic": range(66304, 66352),
+    "Gothic": range(66352, 66384),
+    "Old Permic": range(66384, 66432),
+    "Ugaritic": range(66432, 66464),
+    "Old Persian": range(66464, 66528),
+    "Deseret": range(66560, 66640),
+    "Shavian": range(66640, 66688),
+    "Osmanya": range(66688, 66736),
+    "Osage": range(66736, 66816),
+    "Elbasan": range(66816, 66864),
+    "Caucasian Albanian": range(66864, 66928),
+    "Vithkuqi": range(66928, 67008),
+    "Todhri": range(67008, 67072),
+    "Linear A": range(67072, 67456),
+    "Latin Extended-F": range(67456, 67520),
+    "Cypriot Syllabary": range(67584, 67648),
+    "Imperial Aramaic": range(67648, 67680),
+    "Palmyrene": range(67680, 67712),
+    "Nabataean": range(67712, 67760),
+    "Hatran": range(67808, 67840),
+    "Phoenician": range(67840, 67872),
+    "Lydian": range(67872, 67904),
+    "Sidetic": range(67904, 67936),
+    "Meroitic Hieroglyphs": range(67968, 68000),
+    "Meroitic Cursive": range(68000, 68096),
+    "Kharoshthi": range(68096, 68192),
+    "Old South Arabian": range(68192, 68224),
+    "Old North Arabian": range(68224, 68256),
+    "Manichaean": range(68288, 68352),
+    "Avestan": range(68352, 68416),
+    "Inscriptional Parthian": range(68416, 68448),
+    "Inscriptional Pahlavi": range(68448, 68480),
+    "Psalter Pahlavi": range(68480, 68528),
+    "Old Turkic": range(68608, 68688),
+    "Old Hungarian": range(68736, 68864),
+    "Hanifi Rohingya": range(68864, 68928),
+    "Garay": range(68928, 69008),
+    "Rumi Numeral Symbols": range(69216, 69248),
+    "Yezidi": range(69248, 69312),
+    "Arabic Extended-C": range(69312, 69376),
+    "Old Sogdian": range(69376, 69424),
+    "Sogdian": range(69424, 69488),
+    "Old Uyghur": range(69488, 69552),
+    "Chorasmian": range(69552, 69600),
+    "Elymaic": range(69600, 69632),
+    "Brahmi": range(69632, 69760),
+    "Kaithi": range(69760, 69840),
+    "Sora Sompeng": range(69840, 69888),
+    "Chakma": range(69888, 69968),
+    "Mahajani": range(69968, 70016),
+    "Sharada": range(70016, 70112),
+    "Sinhala Archaic Numbers": range(70112, 70144),
+    "Khojki": range(70144, 70224),
+    "Multani": range(70272, 70320),
+    "Khudawadi": range(70320, 70400),
+    "Grantha": range(70400, 70528),
+    "Tulu-Tigalari": range(70528, 70656),
+    "Newa": range(70656, 70784),
+    "Tirhuta": range(70784, 70880),
+    "Siddham": range(71040, 71168),
+    "Modi": range(71168, 71264),
+    "Mongolian Supplement": range(71264, 71296),
+    "Takri": range(71296, 71376),
+    "Myanmar Extended-C": range(71376, 71424),
+    "Ahom": range(71424, 71504),
+    "Dogra": range(71680, 71760),
+    "Warang Citi": range(71840, 71936),
+    "Dives Akuru": range(71936, 72032),
+    "Nandinagari": range(72096, 72192),
+    "Zanabazar Square": range(72192, 72272),
+    "Soyombo": range(72272, 72368),
+    "Unified Canadian Aboriginal Syllabics Extended-A": range(72368, 72384),
+    "Pau Cin Hau": range(72384, 72448),
+    "Devanagari Extended-A": range(72448, 72544),
+    "Sharada Supplement": range(72544, 72576),
+    "Sunuwar": range(72640, 72704),
+    "Bhaiksuki": range(72704, 72816),
+    "Marchen": range(72816, 72896),
+    "Masaram Gondi": range(72960, 73056),
+    "Gunjala Gondi": range(73056, 73136),
+    "Tolong Siki": range(73136, 73200),
+    "Makasar": range(73440, 73472),
+    "Kawi": range(73472, 73568),
+    "Lisu Supplement": range(73648, 73664),
+    "Tamil Supplement": range(73664, 73728),
+    "Cuneiform": range(73728, 74752),
+    "Cuneiform Numbers and Punctuation": range(74752, 74880),
+    "Early Dynastic Cuneiform": range(74880, 75088),
+    "Cypro-Minoan": range(77712, 77824),
+    "Egyptian Hieroglyphs": range(77824, 78896),
+    "Egyptian Hieroglyph Format Controls": range(78896, 78944),
+    "Egyptian Hieroglyphs Extended-A": range(78944, 82944),
+    "Anatolian Hieroglyphs": range(82944, 83584),
+    "Gurung Khema": range(90368, 90432),
+    "Bamum Supplement": range(92160, 92736),
+    "Mro": range(92736, 92784),
+    "Tangsa": range(92784, 92880),
+    "Bassa Vah": range(92880, 92928),
+    "Pahawh Hmong": range(92928, 93072),
+    "Kirat Rai": range(93504, 93568),
+    "Medefaidrin": range(93760, 93856),
+    "Beria Erfe": range(93856, 93920),
+    "Miao": range(93952, 94112),
+    "Ideographic Symbols and Punctuation": range(94176, 94208),
+    "Tangut": range(94208, 100352),
+    "Tangut Components": range(100352, 101120),
+    "Khitan Small Script": range(101120, 101632),
+    "Tangut Supplement": range(101632, 101760),
+    "Tangut Components Supplement": range(101760, 101888),
+    "Kana Extended-B": range(110576, 110592),
+    "Kana Supplement": range(110592, 110848),
+    "Kana Extended-A": range(110848, 110896),
+    "Small Kana Extension": range(110896, 110960),
+    "Nushu": range(110960, 111360),
+    "Duployan": range(113664, 113824),
+    "Shorthand Format Controls": range(113824, 113840),
+    "Symbols for Legacy Computing Supplement": range(117760, 118464),
+    "Miscellaneous Symbols Supplement": range(118464, 118528),
+    "Znamenny Musical Notation": range(118528, 118736),
+    "Byzantine Musical Symbols": range(118784, 119040),
+    "Musical Symbols": range(119040, 119296),
+    "Ancient Greek Musical Notation": range(119296, 119376),
+    "Kaktovik Numerals": range(119488, 119520),
+    "Mayan Numerals": range(119520, 119552),
+    "Tai Xuan Jing Symbols": range(119552, 119648),
+    "Counting Rod Numerals": range(119648, 119680),
+    "Mathematical Alphanumeric Symbols": range(119808, 120832),
+    "Sutton SignWriting": range(120832, 121520),
+    "Latin Extended-G": range(122624, 122880),
+    "Glagolitic Supplement": range(122880, 122928),
+    "Cyrillic Extended-D": range(122928, 123024),
+    "Nyiakeng Puachue Hmong": range(123136, 123216),
+    "Toto": range(123536, 123584),
+    "Wancho": range(123584, 123648),
+    "Nag Mundari": range(124112, 124160),
+    "Ol Onal": range(124368, 124416),
+    "Tai Yo": range(124608, 124672),
+    "Ethiopic Extended-B": range(124896, 124928),
+    "Mende Kikakui": range(124928, 125152),
+    "Adlam": range(125184, 125280),
+    "Indic Siyaq Numbers": range(126064, 126144),
+    "Ottoman Siyaq Numbers": range(126208, 126288),
+    "Arabic Mathematical Alphabetic Symbols": range(126464, 126720),
+    "Mahjong Tiles": range(126976, 127024),
+    "Domino Tiles": range(127024, 127136),
+    "Playing Cards": range(127136, 127232),
+    "Enclosed Alphanumeric Supplement": range(127232, 127488),
+    "Enclosed Ideographic Supplement": range(127488, 127744),
+    "Miscellaneous Symbols and Pictographs": range(127744, 128512),
+    "Emoticons": range(128512, 128592),
+    "Ornamental Dingbats": range(128592, 128640),
+    "Transport and Map Symbols": range(128640, 128768),
+    "Alchemical Symbols": range(128768, 128896),
+    "Geometric Shapes Extended": range(128896, 129024),
+    "Supplemental Arrows-C": range(129024, 129280),
+    "Supplemental Symbols and Pictographs": range(129280, 129536),
+    "Chess Symbols": range(129536, 129648),
+    "Symbols and Pictographs Extended-A": range(129648, 129792),
+    "Symbols for Legacy Computing": range(129792, 130048),
+    "CJK Unified Ideographs Extension B": range(131072, 173792),
+    "CJK Unified Ideographs Extension C": range(173824, 177984),
+    "CJK Unified Ideographs Extension D": range(177984, 178208),
+    "CJK Unified Ideographs Extension E": range(178208, 183984),
+    "CJK Unified Ideographs Extension F": range(183984, 191472),
+    "CJK Unified Ideographs Extension I": range(191472, 192096),
+    "CJK Compatibility Ideographs Supplement": range(194560, 195104),
+    "CJK Unified Ideographs Extension G": range(196608, 201552),
+    "CJK Unified Ideographs Extension H": range(201552, 205744),
+    "CJK Unified Ideographs Extension J": range(205744, 210048),
+    "Tags": range(917504, 917632),
+    "Variation Selectors Supplement": range(917760, 918000),
+    "Supplementary Private Use Area-A": range(983040, 1048576),
+    "Supplementary Private Use Area-B": range(1048576, 1114112),
+}
+
+
+UNICODE_SECONDARY_RANGE_KEYWORD: list[str] = [
+    "Supplement",
+    "Extended",
+    "Extensions",
+    "Modifier",
+    "Marks",
+    "Punctuation",
+    "Symbols",
+    "Forms",
+    "Operators",
+    "Miscellaneous",
+    "Drawing",
+    "Block",
+    "Shapes",
+    "Supplemental",
+    "Tags",
+]
+
+RE_POSSIBLE_ENCODING_INDICATION = re_compile(
+    r"(?:(?:encoding)|(?:charset)|(?:coding))(?:[\:= ]{1,10})(?:[\"\']?)([a-zA-Z0-9\-_]+)(?:[\"\']?)",
+    IGNORECASE,
+)
+
+IANA_NO_ALIASES = [
+    "cp720",
+    "cp737",
+    "cp856",
+    "cp874",
+    "cp875",
+    "cp1006",
+    "koi8_r",
+    "koi8_t",
+    "koi8_u",
+]
+
+IANA_SUPPORTED: list[str] = sorted(
+    filter(
+        lambda x: x.endswith("_codec") is False
+        and x not in {"rot_13", "tactis", "mbcs"},
+        list(set(aliases.values())) + IANA_NO_ALIASES,
+    )
+)
+
+IANA_SUPPORTED_COUNT: int = len(IANA_SUPPORTED)
+
+# pre-computed code page that are similar using the function cp_similarity.
+IANA_SUPPORTED_SIMILAR: dict[str, list[str]] = {
+    "cp037": ["cp1026", "cp1140", "cp273", "cp500"],
+    "cp1026": ["cp037", "cp1140", "cp273", "cp500"],
+    "cp1125": ["cp866"],
+    "cp1140": ["cp037", "cp1026", "cp273", "cp500"],
+    "cp1250": ["iso8859_2"],
+    "cp1251": ["kz1048", "ptcp154"],
+    "cp1252": ["iso8859_15", "iso8859_9", "latin_1"],
+    "cp1253": ["iso8859_7"],
+    "cp1254": ["iso8859_15", "iso8859_9", "latin_1"],
+    "cp1257": ["iso8859_13"],
+    "cp273": ["cp037", "cp1026", "cp1140", "cp500"],
+    "cp437": ["cp850", "cp858", "cp860", "cp861", "cp862", "cp863", "cp865"],
+    "cp500": ["cp037", "cp1026", "cp1140", "cp273"],
+    "cp850": ["cp437", "cp857", "cp858", "cp865"],
+    "cp857": ["cp850", "cp858", "cp865"],
+    "cp858": ["cp437", "cp850", "cp857", "cp865"],
+    "cp860": ["cp437", "cp861", "cp862", "cp863", "cp865"],
+    "cp861": ["cp437", "cp860", "cp862", "cp863", "cp865"],
+    "cp862": ["cp437", "cp860", "cp861", "cp863", "cp865"],
+    "cp863": ["cp437", "cp860", "cp861", "cp862", "cp865"],
+    "cp865": ["cp437", "cp850", "cp857", "cp858", "cp860", "cp861", "cp862", "cp863"],
+    "cp866": ["cp1125"],
+    "iso8859_10": ["iso8859_14", "iso8859_15", "iso8859_4", "iso8859_9", "latin_1"],
+    "iso8859_11": ["tis_620"],
+    "iso8859_13": ["cp1257"],
+    "iso8859_14": [
+        "iso8859_10",
+        "iso8859_15",
+        "iso8859_16",
+        "iso8859_3",
+        "iso8859_9",
+        "latin_1",
+    ],
+    "iso8859_15": [
+        "cp1252",
+        "cp1254",
+        "iso8859_10",
+        "iso8859_14",
+        "iso8859_16",
+        "iso8859_3",
+        "iso8859_9",
+        "latin_1",
+    ],
+    "iso8859_16": [
+        "iso8859_14",
+        "iso8859_15",
+        "iso8859_2",
+        "iso8859_3",
+        "iso8859_9",
+        "latin_1",
+    ],
+    "iso8859_2": ["cp1250", "iso8859_16", "iso8859_4"],
+    "iso8859_3": ["iso8859_14", "iso8859_15", "iso8859_16", "iso8859_9", "latin_1"],
+    "iso8859_4": ["iso8859_10", "iso8859_2", "iso8859_9", "latin_1"],
+    "iso8859_7": ["cp1253"],
+    "iso8859_9": [
+        "cp1252",
+        "cp1254",
+        "cp1258",
+        "iso8859_10",
+        "iso8859_14",
+        "iso8859_15",
+        "iso8859_16",
+        "iso8859_3",
+        "iso8859_4",
+        "latin_1",
+    ],
+    "kz1048": ["cp1251", "ptcp154"],
+    "latin_1": [
+        "cp1252",
+        "cp1254",
+        "cp1258",
+        "iso8859_10",
+        "iso8859_14",
+        "iso8859_15",
+        "iso8859_16",
+        "iso8859_3",
+        "iso8859_4",
+        "iso8859_9",
+    ],
+    "mac_iceland": ["mac_roman", "mac_turkish"],
+    "mac_roman": ["mac_iceland", "mac_turkish"],
+    "mac_turkish": ["mac_iceland", "mac_roman"],
+    "ptcp154": ["cp1251", "kz1048"],
+    "tis_620": ["iso8859_11"],
+}
+
+
+CHARDET_CORRESPONDENCE: dict[str, str] = {
+    "iso2022_kr": "ISO-2022-KR",
+    "iso2022_jp": "ISO-2022-JP",
+    "euc_kr": "EUC-KR",
+    "tis_620": "TIS-620",
+    "utf_32": "UTF-32",
+    "euc_jp": "EUC-JP",
+    "koi8_r": "KOI8-R",
+    "iso8859_1": "ISO-8859-1",
+    "iso8859_2": "ISO-8859-2",
+    "iso8859_5": "ISO-8859-5",
+    "iso8859_6": "ISO-8859-6",
+    "iso8859_7": "ISO-8859-7",
+    "iso8859_8": "ISO-8859-8",
+    "utf_16": "UTF-16",
+    "cp855": "IBM855",
+    "mac_cyrillic": "MacCyrillic",
+    "gb2312": "GB2312",
+    "gb18030": "GB18030",
+    "cp932": "CP932",
+    "cp866": "IBM866",
+    "utf_8": "utf-8",
+    "utf_8_sig": "UTF-8-SIG",
+    "shift_jis": "SHIFT_JIS",
+    "big5": "Big5",
+    "cp1250": "windows-1250",
+    "cp1251": "windows-1251",
+    "cp1252": "Windows-1252",
+    "cp1253": "windows-1253",
+    "cp1255": "windows-1255",
+    "cp1256": "windows-1256",
+    "cp1254": "Windows-1254",
+    "cp949": "CP949",
+}
+
+
+COMMON_SAFE_ASCII_CHARACTERS: frozenset[str] = frozenset(
+    {
+        "<",
+        ">",
+        "=",
+        ":",
+        "/",
+        "&",
+        ";",
+        "{",
+        "}",
+        "[",
+        "]",
+        ",",
+        "|",
+        '"',
+        "-",
+        "(",
+        ")",
+    }
+)
+
+# Sample character sets — replace with full lists if needed
+COMMON_CHINESE_CHARACTERS = "的一是在不了有和人这中大为上个国我以要他时来用们生到作地于出就分对成会可主发年动同工也能下过子说产种面而方后多定行学法所民得经十三之进着等部度家电力里如水化高自二理起小物现实加量都两体制机当使点从业本去把性好应开它合还因由其些然前外天政四日那社义事平形相全表间样与关各重新线内数正心反你明看原又么利比或但质气第向道命此变条只没结解问意建月公无系军很情者最立代想已通并提直题党程展五果料象员革位入常文总次品式活设及管特件长求老头基资边流路级少图山统接知较将组见计别她手角期根论运农指几九区强放决西被干做必战先回则任取据处队南给色光门即保治北造百规热领七海口东导器压志世金增争济阶油思术极交受联什认六共权收证改清己美再采转更单风切打白教速花带安场身车例真务具万每目至达走积示议声报斗完类八离华名确才科张信马节话米整空元况今集温传土许步群广石记需段研界拉林律叫且究观越织装影算低持音众书布复容儿须际商非验连断深难近矿千周委素技备半办青省列习响约支般史感劳便团往酸历市克何除消构府太准精值号率族维划选标写存候毛亲快效斯院查江型眼王按格养易置派层片始却专状育厂京识适属圆包火住调满县局照参红细引听该铁价严龙飞"
+
+COMMON_JAPANESE_CHARACTERS = "日一国年大十二本中長出三時行見月分後前生五間上東四今金九入学高円子外八六下来気小七山話女北午百書先名川千水半男西電校語土木聞食車何南万毎白天母火右読友左休父雨"
+
+COMMON_KOREAN_CHARACTERS = "一二三四五六七八九十百千萬上下左右中人女子大小山川日月火水木金土父母天地國名年時文校學生"
+
+# Combine all into a frozenset
+COMMON_CJK_CHARACTERS = frozenset(
+    "".join(
+        [
+            COMMON_CHINESE_CHARACTERS,
+            COMMON_JAPANESE_CHARACTERS,
+            COMMON_KOREAN_CHARACTERS,
+        ]
+    )
+)
+
+KO_NAMES: frozenset[str] = frozenset({"johab", "cp949", "euc_kr"})
+ZH_NAMES: frozenset[str] = frozenset({"big5", "cp950", "big5hkscs", "hz"})
+
+# Logging LEVEL below DEBUG
+TRACE: int = 5
+
+
+# Language label that contain the em dash "—"
+# character are to be considered alternative seq to origin
+FREQUENCIES: dict[str, list[str]] = {
+    "English": [
+        "e",
+        "a",
+        "t",
+        "i",
+        "o",
+        "n",
+        "s",
+        "r",
+        "h",
+        "l",
+        "d",
+        "c",
+        "u",
+        "m",
+        "f",
+        "p",
+        "g",
+        "w",
+        "y",
+        "b",
+        "v",
+        "k",
+        "x",
+        "j",
+        "z",
+        "q",
+    ],
+    "English—": [
+        "e",
+        "a",
+        "t",
+        "i",
+        "o",
+        "n",
+        "s",
+        "r",
+        "h",
+        "l",
+        "d",
+        "c",
+        "m",
+        "u",
+        "f",
+        "p",
+        "g",
+        "w",
+        "b",
+        "y",
+        "v",
+        "k",
+        "j",
+        "x",
+        "z",
+        "q",
+    ],
+    "German": [
+        "e",
+        "n",
+        "i",
+        "r",
+        "s",
+        "t",
+        "a",
+        "d",
+        "h",
+        "u",
+        "l",
+        "g",
+        "o",
+        "c",
+        "m",
+        "b",
+        "f",
+        "k",
+        "w",
+        "z",
+        "p",
+        "v",
+        "ü",
+        "ä",
+        "ö",
+        "j",
+    ],
+    "French": [
+        "e",
+        "a",
+        "s",
+        "n",
+        "i",
+        "t",
+        "r",
+        "l",
+        "u",
+        "o",
+        "d",
+        "c",
+        "p",
+        "m",
+        "é",
+        "v",
+        "g",
+        "f",
+        "b",
+        "h",
+        "q",
+        "à",
+        "x",
+        "è",
+        "y",
+        "j",
+    ],
+    "Dutch": [
+        "e",
+        "n",
+        "a",
+        "i",
+        "r",
+        "t",
+        "o",
+        "d",
+        "s",
+        "l",
+        "g",
+        "h",
+        "v",
+        "m",
+        "u",
+        "k",
+        "c",
+        "p",
+        "b",
+        "w",
+        "j",
+        "z",
+        "f",
+        "y",
+        "x",
+        "ë",
+    ],
+    "Italian": [
+        "e",
+        "i",
+        "a",
+        "o",
+        "n",
+        "l",
+        "t",
+        "r",
+        "s",
+        "c",
+        "d",
+        "u",
+        "p",
+        "m",
+        "g",
+        "v",
+        "f",
+        "b",
+        "z",
+        "h",
+        "q",
+        "è",
+        "à",
+        "k",
+        "y",
+        "ò",
+    ],
+    "Polish": [
+        "a",
+        "i",
+        "o",
+        "e",
+        "n",
+        "r",
+        "z",
+        "w",
+        "s",
+        "c",
+        "t",
+        "k",
+        "y",
+        "d",
+        "p",
+        "m",
+        "u",
+        "l",
+        "j",
+        "ł",
+        "g",
+        "b",
+        "h",
+        "ą",
+        "ę",
+        "ó",
+    ],
+    "Spanish": [
+        "e",
+        "a",
+        "o",
+        "n",
+        "s",
+        "r",
+        "i",
+        "l",
+        "d",
+        "t",
+        "c",
+        "u",
+        "m",
+        "p",
+        "b",
+        "g",
+        "v",
+        "f",
+        "y",
+        "ó",
+        "h",
+        "q",
+        "í",
+        "j",
+        "z",
+        "á",
+    ],
+    "Russian": [
+        "о",
+        "е",
+        "а",
+        "и",
+        "н",
+        "т",
+        "с",
+        "р",
+        "в",
+        "л",
+        "к",
+        "м",
+        "д",
+        "п",
+        "у",
+        "г",
+        "я",
+        "ы",
+        "з",
+        "б",
+        "й",
+        "ь",
+        "ч",
+        "х",
+        "ж",
+        "ц",
+    ],
+    # Jap-Kanji
+    "Japanese": [
+        "日",
+        "一",
+        "人",
+        "年",
+        "大",
+        "十",
+        "二",
+        "本",
+        "中",
+        "長",
+        "出",
+        "三",
+        "時",
+        "行",
+        "見",
+        "月",
+        "分",
+        "後",
+        "前",
+        "生",
+        "五",
+        "間",
+        "上",
+        "東",
+        "四",
+        "今",
+        "金",
+        "九",
+        "入",
+        "学",
+        "高",
+        "円",
+        "子",
+        "外",
+        "八",
+        "六",
+        "下",
+        "来",
+        "気",
+        "小",
+        "七",
+        "山",
+        "話",
+        "女",
+        "北",
+        "午",
+        "百",
+        "書",
+        "先",
+        "名",
+        "川",
+        "千",
+        "水",
+        "半",
+        "男",
+        "西",
+        "電",
+        "校",
+        "語",
+        "土",
+        "木",
+        "聞",
+        "食",
+        "車",
+        "何",
+        "南",
+        "万",
+        "毎",
+        "白",
+        "天",
+        "母",
+        "火",
+        "右",
+        "読",
+        "友",
+        "左",
+        "休",
+        "父",
+        "雨",
+    ],
+    # Jap-Katakana
+    "Japanese—": [
+        "ー",
+        "ン",
+        "ス",
+        "・",
+        "ル",
+        "ト",
+        "リ",
+        "イ",
+        "ア",
+        "ラ",
+        "ッ",
+        "ク",
+        "ド",
+        "シ",
+        "レ",
+        "ジ",
+        "タ",
+        "フ",
+        "ロ",
+        "カ",
+        "テ",
+        "マ",
+        "ィ",
+        "グ",
+        "バ",
+        "ム",
+        "プ",
+        "オ",
+        "コ",
+        "デ",
+        "ニ",
+        "ウ",
+        "メ",
+        "サ",
+        "ビ",
+        "ナ",
+        "ブ",
+        "ャ",
+        "エ",
+        "ュ",
+        "チ",
+        "キ",
+        "ズ",
+        "ダ",
+        "パ",
+        "ミ",
+        "ェ",
+        "ョ",
+        "ハ",
+        "セ",
+        "ベ",
+        "ガ",
+        "モ",
+        "ツ",
+        "ネ",
+        "ボ",
+        "ソ",
+        "ノ",
+        "ァ",
+        "ヴ",
+        "ワ",
+        "ポ",
+        "ペ",
+        "ピ",
+        "ケ",
+        "ゴ",
+        "ギ",
+        "ザ",
+        "ホ",
+        "ゲ",
+        "ォ",
+        "ヤ",
+        "ヒ",
+        "ユ",
+        "ヨ",
+        "ヘ",
+        "ゼ",
+        "ヌ",
+        "ゥ",
+        "ゾ",
+        "ヶ",
+        "ヂ",
+        "ヲ",
+        "ヅ",
+        "ヵ",
+        "ヱ",
+        "ヰ",
+        "ヮ",
+        "ヽ",
+        "゠",
+        "ヾ",
+        "ヷ",
+        "ヿ",
+        "ヸ",
+        "ヹ",
+        "ヺ",
+    ],
+    # Jap-Hiragana
+    "Japanese——": [
+        "の",
+        "に",
+        "る",
+        "た",
+        "と",
+        "は",
+        "し",
+        "い",
+        "を",
+        "で",
+        "て",
+        "が",
+        "な",
+        "れ",
+        "か",
+        "ら",
+        "さ",
+        "っ",
+        "り",
+        "す",
+        "あ",
+        "も",
+        "こ",
+        "ま",
+        "う",
+        "く",
+        "よ",
+        "き",
+        "ん",
+        "め",
+        "お",
+        "け",
+        "そ",
+        "つ",
+        "だ",
+        "や",
+        "え",
+        "ど",
+        "わ",
+        "ち",
+        "み",
+        "せ",
+        "じ",
+        "ば",
+        "へ",
+        "び",
+        "ず",
+        "ろ",
+        "ほ",
+        "げ",
+        "む",
+        "べ",
+        "ひ",
+        "ょ",
+        "ゆ",
+        "ぶ",
+        "ご",
+        "ゃ",
+        "ね",
+        "ふ",
+        "ぐ",
+        "ぎ",
+        "ぼ",
+        "ゅ",
+        "づ",
+        "ざ",
+        "ぞ",
+        "ぬ",
+        "ぜ",
+        "ぱ",
+        "ぽ",
+        "ぷ",
+        "ぴ",
+        "ぃ",
+        "ぁ",
+        "ぇ",
+        "ぺ",
+        "ゞ",
+        "ぢ",
+        "ぉ",
+        "ぅ",
+        "ゐ",
+        "ゝ",
+        "ゑ",
+        "゛",
+        "゜",
+        "ゎ",
+        "ゔ",
+        "゚",
+        "ゟ",
+        "゙",
+        "ゕ",
+        "ゖ",
+    ],
+    "Portuguese": [
+        "a",
+        "e",
+        "o",
+        "s",
+        "i",
+        "r",
+        "d",
+        "n",
+        "t",
+        "m",
+        "u",
+        "c",
+        "l",
+        "p",
+        "g",
+        "v",
+        "b",
+        "f",
+        "h",
+        "ã",
+        "q",
+        "é",
+        "ç",
+        "á",
+        "z",
+        "í",
+    ],
+    "Swedish": [
+        "e",
+        "a",
+        "n",
+        "r",
+        "t",
+        "s",
+        "i",
+        "l",
+        "d",
+        "o",
+        "m",
+        "k",
+        "g",
+        "v",
+        "h",
+        "f",
+        "u",
+        "p",
+        "ä",
+        "c",
+        "b",
+        "ö",
+        "å",
+        "y",
+        "j",
+        "x",
+    ],
+    "Chinese": [
+        "的",
+        "一",
+        "是",
+        "不",
+        "了",
+        "在",
+        "人",
+        "有",
+        "我",
+        "他",
+        "这",
+        "个",
+        "们",
+        "中",
+        "来",
+        "上",
+        "大",
+        "为",
+        "和",
+        "国",
+        "地",
+        "到",
+        "以",
+        "说",
+        "时",
+        "要",
+        "就",
+        "出",
+        "会",
+        "可",
+        "也",
+        "你",
+        "对",
+        "生",
+        "能",
+        "而",
+        "子",
+        "那",
+        "得",
+        "于",
+        "着",
+        "下",
+        "自",
+        "之",
+        "年",
+        "过",
+        "发",
+        "后",
+        "作",
+        "里",
+        "用",
+        "道",
+        "行",
+        "所",
+        "然",
+        "家",
+        "种",
+        "事",
+        "成",
+        "方",
+        "多",
+        "经",
+        "么",
+        "去",
+        "法",
+        "学",
+        "如",
+        "都",
+        "同",
+        "现",
+        "当",
+        "没",
+        "动",
+        "面",
+        "起",
+        "看",
+        "定",
+        "天",
+        "分",
+        "还",
+        "进",
+        "好",
+        "小",
+        "部",
+        "其",
+        "些",
+        "主",
+        "样",
+        "理",
+        "心",
+        "她",
+        "本",
+        "前",
+        "开",
+        "但",
+        "因",
+        "只",
+        "从",
+        "想",
+        "实",
+    ],
+    "Ukrainian": [
+        "о",
+        "а",
+        "н",
+        "і",
+        "и",
+        "р",
+        "в",
+        "т",
+        "е",
+        "с",
+        "к",
+        "л",
+        "у",
+        "д",
+        "м",
+        "п",
+        "з",
+        "я",
+        "ь",
+        "б",
+        "г",
+        "й",
+        "ч",
+        "х",
+        "ц",
+        "ї",
+    ],
+    "Norwegian": [
+        "e",
+        "r",
+        "n",
+        "t",
+        "a",
+        "s",
+        "i",
+        "o",
+        "l",
+        "d",
+        "g",
+        "k",
+        "m",
+        "v",
+        "f",
+        "p",
+        "u",
+        "b",
+        "h",
+        "å",
+        "y",
+        "j",
+        "ø",
+        "c",
+        "æ",
+        "w",
+    ],
+    "Finnish": [
+        "a",
+        "i",
+        "n",
+        "t",
+        "e",
+        "s",
+        "l",
+        "o",
+        "u",
+        "k",
+        "ä",
+        "m",
+        "r",
+        "v",
+        "j",
+        "h",
+        "p",
+        "y",
+        "d",
+        "ö",
+        "g",
+        "c",
+        "b",
+        "f",
+        "w",
+        "z",
+    ],
+    "Vietnamese": [
+        "n",
+        "h",
+        "t",
+        "i",
+        "c",
+        "g",
+        "a",
+        "o",
+        "u",
+        "m",
+        "l",
+        "r",
+        "à",
+        "đ",
+        "s",
+        "e",
+        "v",
+        "p",
+        "b",
+        "y",
+        "ư",
+        "d",
+        "á",
+        "k",
+        "ộ",
+        "ế",
+    ],
+    "Czech": [
+        "o",
+        "e",
+        "a",
+        "n",
+        "t",
+        "s",
+        "i",
+        "l",
+        "v",
+        "r",
+        "k",
+        "d",
+        "u",
+        "m",
+        "p",
+        "í",
+        "c",
+        "h",
+        "z",
+        "á",
+        "y",
+        "j",
+        "b",
+        "ě",
+        "é",
+        "ř",
+    ],
+    "Hungarian": [
+        "e",
+        "a",
+        "t",
+        "l",
+        "s",
+        "n",
+        "k",
+        "r",
+        "i",
+        "o",
+        "z",
+        "á",
+        "é",
+        "g",
+        "m",
+        "b",
+        "y",
+        "v",
+        "d",
+        "h",
+        "u",
+        "p",
+        "j",
+        "ö",
+        "f",
+        "c",
+    ],
+    "Korean": [
+        "이",
+        "다",
+        "에",
+        "의",
+        "는",
+        "로",
+        "하",
+        "을",
+        "가",
+        "고",
+        "지",
+        "서",
+        "한",
+        "은",
+        "기",
+        "으",
+        "년",
+        "대",
+        "사",
+        "시",
+        "를",
+        "리",
+        "도",
+        "인",
+        "스",
+        "일",
+    ],
+    "Indonesian": [
+        "a",
+        "n",
+        "e",
+        "i",
+        "r",
+        "t",
+        "u",
+        "s",
+        "d",
+        "k",
+        "m",
+        "l",
+        "g",
+        "p",
+        "b",
+        "o",
+        "h",
+        "y",
+        "j",
+        "c",
+        "w",
+        "f",
+        "v",
+        "z",
+        "x",
+        "q",
+    ],
+    "Turkish": [
+        "a",
+        "e",
+        "i",
+        "n",
+        "r",
+        "l",
+        "ı",
+        "k",
+        "d",
+        "t",
+        "s",
+        "m",
+        "y",
+        "u",
+        "o",
+        "b",
+        "ü",
+        "ş",
+        "v",
+        "g",
+        "z",
+        "h",
+        "c",
+        "p",
+        "ç",
+        "ğ",
+    ],
+    "Romanian": [
+        "e",
+        "i",
+        "a",
+        "r",
+        "n",
+        "t",
+        "u",
+        "l",
+        "o",
+        "c",
+        "s",
+        "d",
+        "p",
+        "m",
+        "ă",
+        "f",
+        "v",
+        "î",
+        "g",
+        "b",
+        "ș",
+        "ț",
+        "z",
+        "h",
+        "â",
+        "j",
+    ],
+    "Farsi": [
+        "ا",
+        "ی",
+        "ر",
+        "د",
+        "ن",
+        "ه",
+        "و",
+        "م",
+        "ت",
+        "ب",
+        "س",
+        "ل",
+        "ک",
+        "ش",
+        "ز",
+        "ف",
+        "گ",
+        "ع",
+        "خ",
+        "ق",
+        "ج",
+        "آ",
+        "پ",
+        "ح",
+        "ط",
+        "ص",
+    ],
+    "Arabic": [
+        "ا",
+        "ل",
+        "ي",
+        "م",
+        "و",
+        "ن",
+        "ر",
+        "ت",
+        "ب",
+        "ة",
+        "ع",
+        "د",
+        "س",
+        "ف",
+        "ه",
+        "ك",
+        "ق",
+        "أ",
+        "ح",
+        "ج",
+        "ش",
+        "ط",
+        "ص",
+        "ى",
+        "خ",
+        "إ",
+    ],
+    "Danish": [
+        "e",
+        "r",
+        "n",
+        "t",
+        "a",
+        "i",
+        "s",
+        "d",
+        "l",
+        "o",
+        "g",
+        "m",
+        "k",
+        "f",
+        "v",
+        "u",
+        "b",
+        "h",
+        "p",
+        "å",
+        "y",
+        "ø",
+        "æ",
+        "c",
+        "j",
+        "w",
+    ],
+    "Serbian": [
+        "а",
+        "и",
+        "о",
+        "е",
+        "н",
+        "р",
+        "с",
+        "у",
+        "т",
+        "к",
+        "ј",
+        "в",
+        "д",
+        "м",
+        "п",
+        "л",
+        "г",
+        "з",
+        "б",
+        "a",
+        "i",
+        "e",
+        "o",
+        "n",
+        "ц",
+        "ш",
+    ],
+    "Lithuanian": [
+        "i",
+        "a",
+        "s",
+        "o",
+        "r",
+        "e",
+        "t",
+        "n",
+        "u",
+        "k",
+        "m",
+        "l",
+        "p",
+        "v",
+        "d",
+        "j",
+        "g",
+        "ė",
+        "b",
+        "y",
+        "ų",
+        "š",
+        "ž",
+        "c",
+        "ą",
+        "į",
+    ],
+    "Slovene": [
+        "e",
+        "a",
+        "i",
+        "o",
+        "n",
+        "r",
+        "s",
+        "l",
+        "t",
+        "j",
+        "v",
+        "k",
+        "d",
+        "p",
+        "m",
+        "u",
+        "z",
+        "b",
+        "g",
+        "h",
+        "č",
+        "c",
+        "š",
+        "ž",
+        "f",
+        "y",
+    ],
+    "Slovak": [
+        "o",
+        "a",
+        "e",
+        "n",
+        "i",
+        "r",
+        "v",
+        "t",
+        "s",
+        "l",
+        "k",
+        "d",
+        "m",
+        "p",
+        "u",
+        "c",
+        "h",
+        "j",
+        "b",
+        "z",
+        "á",
+        "y",
+        "ý",
+        "í",
+        "č",
+        "é",
+    ],
+    "Hebrew": [
+        "י",
+        "ו",
+        "ה",
+        "ל",
+        "ר",
+        "ב",
+        "ת",
+        "מ",
+        "א",
+        "ש",
+        "נ",
+        "ע",
+        "ם",
+        "ד",
+        "ק",
+        "ח",
+        "פ",
+        "ס",
+        "כ",
+        "ג",
+        "ט",
+        "צ",
+        "ן",
+        "ז",
+        "ך",
+    ],
+    "Bulgarian": [
+        "а",
+        "и",
+        "о",
+        "е",
+        "н",
+        "т",
+        "р",
+        "с",
+        "в",
+        "л",
+        "к",
+        "д",
+        "п",
+        "м",
+        "з",
+        "г",
+        "я",
+        "ъ",
+        "у",
+        "б",
+        "ч",
+        "ц",
+        "й",
+        "ж",
+        "щ",
+        "х",
+    ],
+    "Croatian": [
+        "a",
+        "i",
+        "o",
+        "e",
+        "n",
+        "r",
+        "j",
+        "s",
+        "t",
+        "u",
+        "k",
+        "l",
+        "v",
+        "d",
+        "m",
+        "p",
+        "g",
+        "z",
+        "b",
+        "c",
+        "č",
+        "h",
+        "š",
+        "ž",
+        "ć",
+        "f",
+    ],
+    "Hindi": [
+        "क",
+        "र",
+        "स",
+        "न",
+        "त",
+        "म",
+        "ह",
+        "प",
+        "य",
+        "ल",
+        "व",
+        "ज",
+        "द",
+        "ग",
+        "ब",
+        "श",
+        "ट",
+        "अ",
+        "ए",
+        "थ",
+        "भ",
+        "ड",
+        "च",
+        "ध",
+        "ष",
+        "इ",
+    ],
+    "Estonian": [
+        "a",
+        "i",
+        "e",
+        "s",
+        "t",
+        "l",
+        "u",
+        "n",
+        "o",
+        "k",
+        "r",
+        "d",
+        "m",
+        "v",
+        "g",
+        "p",
+        "j",
+        "h",
+        "ä",
+        "b",
+        "õ",
+        "ü",
+        "f",
+        "c",
+        "ö",
+        "y",
+    ],
+    "Thai": [
+        "า",
+        "น",
+        "ร",
+        "อ",
+        "ก",
+        "เ",
+        "ง",
+        "ม",
+        "ย",
+        "ล",
+        "ว",
+        "ด",
+        "ท",
+        "ส",
+        "ต",
+        "ะ",
+        "ป",
+        "บ",
+        "ค",
+        "ห",
+        "แ",
+        "จ",
+        "พ",
+        "ช",
+        "ข",
+        "ใ",
+    ],
+    "Greek": [
+        "α",
+        "τ",
+        "ο",
+        "ι",
+        "ε",
+        "ν",
+        "ρ",
+        "σ",
+        "κ",
+        "η",
+        "π",
+        "ς",
+        "υ",
+        "μ",
+        "λ",
+        "ί",
+        "ό",
+        "ά",
+        "γ",
+        "έ",
+        "δ",
+        "ή",
+        "ω",
+        "χ",
+        "θ",
+        "ύ",
+    ],
+    "Tamil": [
+        "க",
+        "த",
+        "ப",
+        "ட",
+        "ர",
+        "ம",
+        "ல",
+        "ன",
+        "வ",
+        "ற",
+        "ய",
+        "ள",
+        "ச",
+        "ந",
+        "இ",
+        "ண",
+        "அ",
+        "ஆ",
+        "ழ",
+        "ங",
+        "எ",
+        "உ",
+        "ஒ",
+        "ஸ",
+    ],
+    "Kazakh": [
+        "а",
+        "ы",
+        "е",
+        "н",
+        "т",
+        "р",
+        "л",
+        "і",
+        "д",
+        "с",
+        "м",
+        "қ",
+        "к",
+        "о",
+        "б",
+        "и",
+        "у",
+        "ғ",
+        "ж",
+        "ң",
+        "з",
+        "ш",
+        "й",
+        "п",
+        "г",
+        "ө",
+    ],
+}
+
+LANGUAGE_SUPPORTED_COUNT: int = len(FREQUENCIES)
+
+# Bit flags for unified character classification.
+# A single unicodedata.name() call sets all relevant flags at once.
+_LATIN: int = 1
+_ACCENTUATED: int = 1 << 1
+_CJK: int = 1 << 2
+_HANGUL: int = 1 << 3
+_KATAKANA: int = 1 << 4
+_HIRAGANA: int = 1 << 5
+_THAI: int = 1 << 6
+_ARABIC: int = 1 << 7
+_ARABIC_ISOLATED_FORM: int = 1 << 8
+
+_ACCENT_KEYWORDS: tuple[str, ...] = (
+    "WITH GRAVE",
+    "WITH ACUTE",
+    "WITH CEDILLA",
+    "WITH DIAERESIS",
+    "WITH CIRCUMFLEX",
+    "WITH TILDE",
+    "WITH MACRON",
+    "WITH RING ABOVE",
+)
+
+# Pre-built lookup structures for FREQUENCIES (computed once at import time).
+# character -> rank mapping per language (replaces list .index() calls).
+_FREQUENCIES_RANK: dict[str, dict[str, int]] = {
+    lang: {char: rank for rank, char in enumerate(chars)}
+    for lang, chars in FREQUENCIES.items()
+}
+
+# frozenset per language (avoids rebuilding set() per call).
+_FREQUENCIES_SET: dict[str, frozenset[str]] = {
+    lang: frozenset(chars) for lang, chars in FREQUENCIES.items()
+}

.venv/Lib/site-packages/charset_normalizer/legacy.py

@@ -0,0 +1,79 @@
+from __future__ import annotations
+
+from typing import TYPE_CHECKING, Any
+from warnings import warn
+
+from .api import from_bytes
+from .constant import CHARDET_CORRESPONDENCE, TOO_SMALL_SEQUENCE
+
+if TYPE_CHECKING:
+    from typing import TypedDict
+
+    class ResultDict(TypedDict):
+        encoding: str | None
+        language: str
+        confidence: float | None
+
+
+def detect(
+    byte_str: bytes, should_rename_legacy: bool = False, **kwargs: Any
+) -> ResultDict:
+    """
+    chardet legacy method
+    Detect the encoding of the given byte string. It should be mostly backward-compatible.
+    Encoding name will match Chardet own writing whenever possible. (Not on encoding name unsupported by it)
+    This function is deprecated and should be used to migrate your project easily, consult the documentation for
+    further information. Not planned for removal.
+
+    :param byte_str:     The byte sequence to examine.
+    :param should_rename_legacy:  Should we rename legacy encodings
+                                  to their more modern equivalents?
+    """
+    if len(kwargs):
+        warn(
+            f"charset-normalizer disregard arguments '{','.join(list(kwargs.keys()))}' in legacy function detect()"
+        )
+
+    if not isinstance(byte_str, (bytearray, bytes)):
+        raise TypeError(  # pragma: nocover
+            f"Expected object of type bytes or bytearray, got: {type(byte_str)}"
+        )
+
+    if isinstance(byte_str, bytearray):
+        byte_str = bytes(byte_str)
+
+    r = from_bytes(byte_str).best()
+
+    encoding = r.encoding if r is not None else None
+    language = r.language if r is not None and r.language != "Unknown" else ""
+    confidence = 1.0 - r.chaos if r is not None else None
+
+    # automatically lower confidence
+    # on small bytes samples.
+    # https://github.com/jawah/charset_normalizer/issues/391
+    if (
+        confidence is not None
+        and confidence >= 0.9
+        and encoding
+        not in {
+            "utf_8",
+            "ascii",
+        }
+        and r.bom is False  # type: ignore[union-attr]
+        and len(byte_str) < TOO_SMALL_SEQUENCE
+    ):
+        confidence -= 0.2
+
+    # Note: CharsetNormalizer does not return 'UTF-8-SIG' as the sig get stripped in the detection/normalization process
+    # but chardet does return 'utf-8-sig' and it is a valid codec name.
+    if r is not None and encoding == "utf_8" and r.bom:
+        encoding += "_sig"
+
+    if should_rename_legacy is False and encoding in CHARDET_CORRESPONDENCE:
+        encoding = CHARDET_CORRESPONDENCE[encoding]
+
+    return {
+        "encoding": encoding,
+        "language": language,
+        "confidence": confidence,
+    }

.venv/Lib/site-packages/charset_normalizer/md.py

@@ -0,0 +1,936 @@
+from __future__ import annotations
+
+import sys
+from functools import lru_cache
+from logging import getLogger
+
+if sys.version_info >= (3, 8):
+    from typing import final
+else:
+    try:
+        from typing_extensions import final
+    except ImportError:
+
+        def final(cls):  # type: ignore[misc,no-untyped-def]
+            return cls
+
+
+from .constant import (
+    COMMON_CJK_CHARACTERS,
+    COMMON_SAFE_ASCII_CHARACTERS,
+    TRACE,
+    UNICODE_SECONDARY_RANGE_KEYWORD,
+    _ACCENTUATED,
+    _ARABIC,
+    _ARABIC_ISOLATED_FORM,
+    _CJK,
+    _HANGUL,
+    _HIRAGANA,
+    _KATAKANA,
+    _LATIN,
+    _THAI,
+)
+from .utils import (
+    _character_flags,
+    is_emoticon,
+    is_punctuation,
+    is_separator,
+    is_symbol,
+    remove_accent,
+    unicode_range,
+)
+
+# Combined bitmask for CJK/Hangul/Katakana/Hiragana/Thai glyph detection.
+_GLYPH_MASK: int = _CJK | _HANGUL | _KATAKANA | _HIRAGANA | _THAI
+
+
+@final
+class CharInfo:
+    """Pre-computed character properties shared across all detectors.
+
+    Instantiated once and reused via :meth:`update` on every character
+    in the hot loop so that redundant calls to str methods
+    (``isalpha``, ``isupper``, …) and cached utility functions
+    (``_character_flags``, ``is_punctuation``, …) are avoided when
+    several plugins need the same information.
+    """
+
+    __slots__ = (
+        "character",
+        "printable",
+        "alpha",
+        "upper",
+        "lower",
+        "space",
+        "digit",
+        "is_ascii",
+        "case_variable",
+        "flags",
+        "accentuated",
+        "latin",
+        "is_cjk",
+        "is_arabic",
+        "is_glyph",
+        "punct",
+        "sym",
+    )
+
+    def __init__(self) -> None:
+        self.character: str = ""
+        self.printable: bool = False
+        self.alpha: bool = False
+        self.upper: bool = False
+        self.lower: bool = False
+        self.space: bool = False
+        self.digit: bool = False
+        self.is_ascii: bool = False
+        self.case_variable: bool = False
+        self.flags: int = 0
+        self.accentuated: bool = False
+        self.latin: bool = False
+        self.is_cjk: bool = False
+        self.is_arabic: bool = False
+        self.is_glyph: bool = False
+        self.punct: bool = False
+        self.sym: bool = False
+
+    def update(self, character: str) -> None:
+        """Update all properties for *character* (called once per character)."""
+        self.character = character
+
+        # ASCII fast-path: for characters with ord < 128, we can skip
+        # _character_flags() entirely and derive most properties from ord.
+        o: int = ord(character)
+        if o < 128:
+            self.is_ascii = True
+            self.accentuated = False
+            self.is_cjk = False
+            self.is_arabic = False
+            self.is_glyph = False
+            # ASCII alpha: a-z (97-122) or A-Z (65-90)
+            if 65 <= o <= 90:
+                # Uppercase ASCII letter
+                self.alpha = True
+                self.upper = True
+                self.lower = False
+                self.space = False
+                self.digit = False
+                self.printable = True
+                self.case_variable = True
+                self.flags = _LATIN
+                self.latin = True
+                self.punct = False
+                self.sym = False
+            elif 97 <= o <= 122:
+                # Lowercase ASCII letter
+                self.alpha = True
+                self.upper = False
+                self.lower = True
+                self.space = False
+                self.digit = False
+                self.printable = True
+                self.case_variable = True
+                self.flags = _LATIN
+                self.latin = True
+                self.punct = False
+                self.sym = False
+            elif 48 <= o <= 57:
+                # ASCII digit 0-9
+                self.alpha = False
+                self.upper = False
+                self.lower = False
+                self.space = False
+                self.digit = True
+                self.printable = True
+                self.case_variable = False
+                self.flags = 0
+                self.latin = False
+                self.punct = False
+                self.sym = False
+            elif o == 32 or (9 <= o <= 13):
+                # Space, tab, newline, etc.
+                self.alpha = False
+                self.upper = False
+                self.lower = False
+                self.space = True
+                self.digit = False
+                self.printable = o == 32
+                self.case_variable = False
+                self.flags = 0
+                self.latin = False
+                self.punct = False
+                self.sym = False
+            else:
+                # Other ASCII (punctuation, symbols, control chars)
+                self.printable = character.isprintable()
+                self.alpha = False
+                self.upper = False
+                self.lower = False
+                self.space = False
+                self.digit = False
+                self.case_variable = False
+                self.flags = 0
+                self.latin = False
+                self.punct = is_punctuation(character) if self.printable else False
+                self.sym = is_symbol(character) if self.printable else False
+        else:
+            # Non-ASCII path
+            self.is_ascii = False
+            self.printable = character.isprintable()
+            self.alpha = character.isalpha()
+            self.upper = character.isupper()
+            self.lower = character.islower()
+            self.space = character.isspace()
+            self.digit = character.isdigit()
+            self.case_variable = self.lower != self.upper
+
+            # Flag-based classification (single unicodedata.name() call, lru-cached)
+            flags: int
+            if self.alpha:
+                flags = _character_flags(character)
+            else:
+                flags = 0
+            self.flags = flags
+            self.accentuated = bool(flags & _ACCENTUATED)
+            self.latin = bool(flags & _LATIN)
+            self.is_cjk = bool(flags & _CJK)
+            self.is_arabic = bool(flags & _ARABIC)
+            self.is_glyph = bool(flags & _GLYPH_MASK)
+
+            # Eagerly compute punct and sym (avoids property dispatch overhead
+            # on 300K+ accesses in the hot loop).
+            self.punct = is_punctuation(character) if self.printable else False
+            self.sym = is_symbol(character) if self.printable else False
+
+
+class MessDetectorPlugin:
+    """
+    Base abstract class used for mess detection plugins.
+    All detectors MUST extend and implement given methods.
+    """
+
+    __slots__ = ()
+
+    def feed_info(self, character: str, info: CharInfo) -> None:
+        """
+        The main routine to be executed upon character.
+        Insert the logic in witch the text would be considered chaotic.
+        """
+        raise NotImplementedError  # Defensive:
+
+    def reset(self) -> None:  # Defensive:
+        """
+        Permit to reset the plugin to the initial state.
+        """
+        raise NotImplementedError
+
+    @property
+    def ratio(self) -> float:
+        """
+        Compute the chaos ratio based on what your feed() has seen.
+        Must NOT be lower than 0.; No restriction gt 0.
+        """
+        raise NotImplementedError  # Defensive:
+
+
+@final
+class TooManySymbolOrPunctuationPlugin(MessDetectorPlugin):
+    __slots__ = (
+        "_punctuation_count",
+        "_symbol_count",
+        "_character_count",
+        "_last_printable_char",
+        "_frenzy_symbol_in_word",
+    )
+
+    def __init__(self) -> None:
+        self._punctuation_count: int = 0
+        self._symbol_count: int = 0
+        self._character_count: int = 0
+
+        self._last_printable_char: str | None = None
+        self._frenzy_symbol_in_word: bool = False
+
+    def feed_info(self, character: str, info: CharInfo) -> None:
+        """Optimized feed using pre-computed character info."""
+        self._character_count += 1
+
+        if (
+            character != self._last_printable_char
+            and character not in COMMON_SAFE_ASCII_CHARACTERS
+        ):
+            if info.punct:
+                self._punctuation_count += 1
+            elif not info.digit and info.sym and not is_emoticon(character):
+                self._symbol_count += 2
+
+        self._last_printable_char = character
+
+    def reset(self) -> None:  # Abstract
+        self._punctuation_count = 0
+        self._character_count = 0
+        self._symbol_count = 0
+
+    @property
+    def ratio(self) -> float:
+        if self._character_count == 0:
+            return 0.0
+
+        ratio_of_punctuation: float = (
+            self._punctuation_count + self._symbol_count
+        ) / self._character_count
+
+        return ratio_of_punctuation if ratio_of_punctuation >= 0.3 else 0.0
+
+
+@final
+class TooManyAccentuatedPlugin(MessDetectorPlugin):
+    __slots__ = ("_character_count", "_accentuated_count")
+
+    def __init__(self) -> None:
+        self._character_count: int = 0
+        self._accentuated_count: int = 0
+
+    def feed_info(self, character: str, info: CharInfo) -> None:
+        """Optimized feed using pre-computed character info."""
+        self._character_count += 1
+
+        if info.accentuated:
+            self._accentuated_count += 1
+
+    def reset(self) -> None:  # Abstract
+        self._character_count = 0
+        self._accentuated_count = 0
+
+    @property
+    def ratio(self) -> float:
+        if self._character_count < 8:
+            return 0.0
+
+        ratio_of_accentuation: float = self._accentuated_count / self._character_count
+        return ratio_of_accentuation if ratio_of_accentuation >= 0.35 else 0.0
+
+
+@final
+class UnprintablePlugin(MessDetectorPlugin):
+    __slots__ = ("_unprintable_count", "_character_count")
+
+    def __init__(self) -> None:
+        self._unprintable_count: int = 0
+        self._character_count: int = 0
+
+    def feed_info(self, character: str, info: CharInfo) -> None:
+        """Optimized feed using pre-computed character info."""
+        if (
+            not info.space
+            and not info.printable
+            and character != "\x1a"
+            and character != "\ufeff"
+        ):
+            self._unprintable_count += 1
+        self._character_count += 1
+
+    def reset(self) -> None:  # Abstract
+        self._unprintable_count = 0
+
+    @property
+    def ratio(self) -> float:
+        if self._character_count == 0:  # Defensive:
+            return 0.0
+
+        return (self._unprintable_count * 8) / self._character_count
+
+
+@final
+class SuspiciousDuplicateAccentPlugin(MessDetectorPlugin):
+    __slots__ = (
+        "_successive_count",
+        "_character_count",
+        "_last_latin_character",
+        "_last_was_accentuated",
+    )
+
+    def __init__(self) -> None:
+        self._successive_count: int = 0
+        self._character_count: int = 0
+
+        self._last_latin_character: str | None = None
+        self._last_was_accentuated: bool = False
+
+    def feed_info(self, character: str, info: CharInfo) -> None:
+        """Optimized feed using pre-computed character info."""
+        self._character_count += 1
+        if (
+            self._last_latin_character is not None
+            and info.accentuated
+            and self._last_was_accentuated
+        ):
+            if info.upper and self._last_latin_character.isupper():
+                self._successive_count += 1
+            if remove_accent(character) == remove_accent(self._last_latin_character):
+                self._successive_count += 1
+        self._last_latin_character = character
+        self._last_was_accentuated = info.accentuated
+
+    def reset(self) -> None:  # Abstract
+        self._successive_count = 0
+        self._character_count = 0
+        self._last_latin_character = None
+        self._last_was_accentuated = False
+
+    @property
+    def ratio(self) -> float:
+        if self._character_count == 0:
+            return 0.0
+
+        return (self._successive_count * 2) / self._character_count
+
+
+@final
+class SuspiciousRange(MessDetectorPlugin):
+    __slots__ = (
+        "_suspicious_successive_range_count",
+        "_character_count",
+        "_last_printable_seen",
+        "_last_printable_range",
+    )
+
+    def __init__(self) -> None:
+        self._suspicious_successive_range_count: int = 0
+        self._character_count: int = 0
+        self._last_printable_seen: str | None = None
+        self._last_printable_range: str | None = None
+
+    def feed_info(self, character: str, info: CharInfo) -> None:
+        """Optimized feed using pre-computed character info."""
+        self._character_count += 1
+
+        if info.space or info.punct or character in COMMON_SAFE_ASCII_CHARACTERS:
+            self._last_printable_seen = None
+            self._last_printable_range = None
+            return
+
+        if self._last_printable_seen is None:
+            self._last_printable_seen = character
+            self._last_printable_range = unicode_range(character)
+            return
+
+        unicode_range_a: str | None = self._last_printable_range
+        unicode_range_b: str | None = unicode_range(character)
+
+        if is_suspiciously_successive_range(unicode_range_a, unicode_range_b):
+            self._suspicious_successive_range_count += 1
+
+        self._last_printable_seen = character
+        self._last_printable_range = unicode_range_b
+
+    def reset(self) -> None:  # Abstract
+        self._character_count = 0
+        self._suspicious_successive_range_count = 0
+        self._last_printable_seen = None
+        self._last_printable_range = None
+
+    @property
+    def ratio(self) -> float:
+        if self._character_count <= 13:
+            return 0.0
+
+        ratio_of_suspicious_range_usage: float = (
+            self._suspicious_successive_range_count * 2
+        ) / self._character_count
+
+        return ratio_of_suspicious_range_usage
+
+
+@final
+class SuperWeirdWordPlugin(MessDetectorPlugin):
+    __slots__ = (
+        "_word_count",
+        "_bad_word_count",
+        "_foreign_long_count",
+        "_is_current_word_bad",
+        "_foreign_long_watch",
+        "_character_count",
+        "_bad_character_count",
+        "_buffer_length",
+        "_buffer_last_char",
+        "_buffer_last_char_accentuated",
+        "_buffer_accent_count",
+        "_buffer_glyph_count",
+        "_buffer_upper_count",
+    )
+
+    def __init__(self) -> None:
+        self._word_count: int = 0
+        self._bad_word_count: int = 0
+        self._foreign_long_count: int = 0
+
+        self._is_current_word_bad: bool = False
+        self._foreign_long_watch: bool = False
+
+        self._character_count: int = 0
+        self._bad_character_count: int = 0
+
+        self._buffer_length: int = 0
+        self._buffer_last_char: str | None = None
+        self._buffer_last_char_accentuated: bool = False
+        self._buffer_accent_count: int = 0
+        self._buffer_glyph_count: int = 0
+        self._buffer_upper_count: int = 0
+
+    def feed_info(self, character: str, info: CharInfo) -> None:
+        """Optimized feed using pre-computed character info."""
+        if info.alpha:
+            self._buffer_length += 1
+            self._buffer_last_char = character
+
+            if info.upper:
+                self._buffer_upper_count += 1
+
+            self._buffer_last_char_accentuated = info.accentuated
+
+            if info.accentuated:
+                self._buffer_accent_count += 1
+            if (
+                not self._foreign_long_watch
+                and (not info.latin or info.accentuated)
+                and not info.is_glyph
+            ):
+                self._foreign_long_watch = True
+            if info.is_glyph:
+                self._buffer_glyph_count += 1
+            return
+        if not self._buffer_length:
+            return
+        if info.space or info.punct or is_separator(character):
+            self._word_count += 1
+            buffer_length: int = self._buffer_length
+
+            self._character_count += buffer_length
+
+            if buffer_length >= 4:
+                if self._buffer_accent_count / buffer_length >= 0.5:
+                    self._is_current_word_bad = True
+                elif (
+                    self._buffer_last_char_accentuated
+                    and self._buffer_last_char.isupper()  # type: ignore[union-attr]
+                    and self._buffer_upper_count != buffer_length
+                ):
+                    self._foreign_long_count += 1
+                    self._is_current_word_bad = True
+                elif self._buffer_glyph_count == 1:
+                    self._is_current_word_bad = True
+                    self._foreign_long_count += 1
+            if buffer_length >= 24 and self._foreign_long_watch:
+                probable_camel_cased: bool = (
+                    self._buffer_upper_count > 0
+                    and self._buffer_upper_count / buffer_length <= 0.3
+                )
+
+                if not probable_camel_cased:
+                    self._foreign_long_count += 1
+                    self._is_current_word_bad = True
+
+            if self._is_current_word_bad:
+                self._bad_word_count += 1
+                self._bad_character_count += buffer_length
+                self._is_current_word_bad = False
+
+            self._foreign_long_watch = False
+            self._buffer_length = 0
+            self._buffer_last_char = None
+            self._buffer_last_char_accentuated = False
+            self._buffer_accent_count = 0
+            self._buffer_glyph_count = 0
+            self._buffer_upper_count = 0
+        elif (
+            character not in {"<", ">", "-", "=", "~", "|", "_"}
+            and not info.digit
+            and info.sym
+        ):
+            self._is_current_word_bad = True
+            self._buffer_length += 1
+            self._buffer_last_char = character
+            self._buffer_last_char_accentuated = False
+
+    def reset(self) -> None:  # Abstract
+        self._buffer_length = 0
+        self._buffer_last_char = None
+        self._buffer_last_char_accentuated = False
+        self._is_current_word_bad = False
+        self._foreign_long_watch = False
+        self._bad_word_count = 0
+        self._word_count = 0
+        self._character_count = 0
+        self._bad_character_count = 0
+        self._foreign_long_count = 0
+        self._buffer_accent_count = 0
+        self._buffer_glyph_count = 0
+        self._buffer_upper_count = 0
+
+    @property
+    def ratio(self) -> float:
+        if self._word_count <= 10 and self._foreign_long_count == 0:
+            return 0.0
+
+        return self._bad_character_count / self._character_count
+
+
+@final
+class CjkUncommonPlugin(MessDetectorPlugin):
+    """
+    Detect messy CJK text that probably means nothing.
+    """
+
+    __slots__ = ("_character_count", "_uncommon_count")
+
+    def __init__(self) -> None:
+        self._character_count: int = 0
+        self._uncommon_count: int = 0
+
+    def feed_info(self, character: str, info: CharInfo) -> None:
+        """Optimized feed using pre-computed character info."""
+        self._character_count += 1
+
+        if character not in COMMON_CJK_CHARACTERS:
+            self._uncommon_count += 1
+
+    def reset(self) -> None:  # Abstract
+        self._character_count = 0
+        self._uncommon_count = 0
+
+    @property
+    def ratio(self) -> float:
+        if self._character_count < 8:
+            return 0.0
+
+        uncommon_form_usage: float = self._uncommon_count / self._character_count
+
+        # we can be pretty sure it's garbage when uncommon characters are widely
+        # used. otherwise it could just be traditional chinese for example.
+        return uncommon_form_usage / 10 if uncommon_form_usage > 0.5 else 0.0
+
+
+@final
+class ArchaicUpperLowerPlugin(MessDetectorPlugin):
+    __slots__ = (
+        "_buf",
+        "_character_count_since_last_sep",
+        "_successive_upper_lower_count",
+        "_successive_upper_lower_count_final",
+        "_character_count",
+        "_last_alpha_seen",
+        "_last_alpha_seen_upper",
+        "_last_alpha_seen_lower",
+        "_current_ascii_only",
+    )
+
+    def __init__(self) -> None:
+        self._buf: bool = False
+
+        self._character_count_since_last_sep: int = 0
+
+        self._successive_upper_lower_count: int = 0
+        self._successive_upper_lower_count_final: int = 0
+
+        self._character_count: int = 0
+
+        self._last_alpha_seen: str | None = None
+        self._last_alpha_seen_upper: bool = False
+        self._last_alpha_seen_lower: bool = False
+        self._current_ascii_only: bool = True
+
+    def feed_info(self, character: str, info: CharInfo) -> None:
+        """Optimized feed using pre-computed character info."""
+        is_concerned: bool = info.alpha and info.case_variable
+        chunk_sep: bool = not is_concerned
+
+        if chunk_sep and self._character_count_since_last_sep > 0:
+            if (
+                self._character_count_since_last_sep <= 64
+                and not info.digit
+                and not self._current_ascii_only
+            ):
+                self._successive_upper_lower_count_final += (
+                    self._successive_upper_lower_count
+                )
+
+            self._successive_upper_lower_count = 0
+            self._character_count_since_last_sep = 0
+            self._last_alpha_seen = None
+            self._buf = False
+            self._character_count += 1
+            self._current_ascii_only = True
+
+            return
+
+        if self._current_ascii_only and not info.is_ascii:
+            self._current_ascii_only = False
+
+        if self._last_alpha_seen is not None:
+            if (info.upper and self._last_alpha_seen_lower) or (
+                info.lower and self._last_alpha_seen_upper
+            ):
+                if self._buf:
+                    self._successive_upper_lower_count += 2
+                    self._buf = False
+                else:
+                    self._buf = True
+            else:
+                self._buf = False
+
+        self._character_count += 1
+        self._character_count_since_last_sep += 1
+        self._last_alpha_seen = character
+        self._last_alpha_seen_upper = info.upper
+        self._last_alpha_seen_lower = info.lower
+
+    def reset(self) -> None:  # Abstract
+        self._character_count = 0
+        self._character_count_since_last_sep = 0
+        self._successive_upper_lower_count = 0
+        self._successive_upper_lower_count_final = 0
+        self._last_alpha_seen = None
+        self._last_alpha_seen_upper = False
+        self._last_alpha_seen_lower = False
+        self._buf = False
+        self._current_ascii_only = True
+
+    @property
+    def ratio(self) -> float:
+        if self._character_count == 0:  # Defensive:
+            return 0.0
+
+        return self._successive_upper_lower_count_final / self._character_count
+
+
+@final
+class ArabicIsolatedFormPlugin(MessDetectorPlugin):
+    __slots__ = ("_character_count", "_isolated_form_count")
+
+    def __init__(self) -> None:
+        self._character_count: int = 0
+        self._isolated_form_count: int = 0
+
+    def reset(self) -> None:  # Abstract
+        self._character_count = 0
+        self._isolated_form_count = 0
+
+    def feed_info(self, character: str, info: CharInfo) -> None:
+        """Optimized feed using pre-computed character info."""
+        self._character_count += 1
+
+        if info.flags & _ARABIC_ISOLATED_FORM:
+            self._isolated_form_count += 1
+
+    @property
+    def ratio(self) -> float:
+        if self._character_count < 8:
+            return 0.0
+
+        isolated_form_usage: float = self._isolated_form_count / self._character_count
+
+        return isolated_form_usage
+
+
+@lru_cache(maxsize=1024)
+def is_suspiciously_successive_range(
+    unicode_range_a: str | None, unicode_range_b: str | None
+) -> bool:
+    """
+    Determine if two Unicode range seen next to each other can be considered as suspicious.
+    """
+    if unicode_range_a is None or unicode_range_b is None:
+        return True
+
+    if unicode_range_a == unicode_range_b:
+        return False
+
+    if "Latin" in unicode_range_a and "Latin" in unicode_range_b:
+        return False
+
+    if "Emoticons" in unicode_range_a or "Emoticons" in unicode_range_b:
+        return False
+
+    # Latin characters can be accompanied with a combining diacritical mark
+    # eg. Vietnamese.
+    if ("Latin" in unicode_range_a or "Latin" in unicode_range_b) and (
+        "Combining" in unicode_range_a or "Combining" in unicode_range_b
+    ):
+        return False
+
+    keywords_range_a, keywords_range_b = (
+        unicode_range_a.split(" "),
+        unicode_range_b.split(" "),
+    )
+
+    for el in keywords_range_a:
+        if el in UNICODE_SECONDARY_RANGE_KEYWORD:
+            continue
+        if el in keywords_range_b:
+            return False
+
+    # Japanese Exception
+    range_a_jp_chars, range_b_jp_chars = (
+        unicode_range_a
+        in (
+            "Hiragana",
+            "Katakana",
+        ),
+        unicode_range_b in ("Hiragana", "Katakana"),
+    )
+    if (range_a_jp_chars or range_b_jp_chars) and (
+        "CJK" in unicode_range_a or "CJK" in unicode_range_b
+    ):
+        return False
+    if range_a_jp_chars and range_b_jp_chars:
+        return False
+
+    if "Hangul" in unicode_range_a or "Hangul" in unicode_range_b:
+        if "CJK" in unicode_range_a or "CJK" in unicode_range_b:
+            return False
+        if unicode_range_a == "Basic Latin" or unicode_range_b == "Basic Latin":
+            return False
+
+    # Chinese/Japanese use dedicated range for punctuation and/or separators.
+    if ("CJK" in unicode_range_a or "CJK" in unicode_range_b) or (
+        unicode_range_a in ["Katakana", "Hiragana"]
+        and unicode_range_b in ["Katakana", "Hiragana"]
+    ):
+        if "Punctuation" in unicode_range_a or "Punctuation" in unicode_range_b:
+            return False
+        if "Forms" in unicode_range_a or "Forms" in unicode_range_b:
+            return False
+        if unicode_range_a == "Basic Latin" or unicode_range_b == "Basic Latin":
+            return False
+
+    return True
+
+
+@lru_cache(maxsize=2048)
+def mess_ratio(
+    decoded_sequence: str, maximum_threshold: float = 0.2, debug: bool = False
+) -> float:
+    """
+    Compute a mess ratio given a decoded bytes sequence. The maximum threshold does stop the computation earlier.
+    """
+
+    seq_len: int = len(decoded_sequence)
+
+    if seq_len < 511:
+        step: int = 32
+    elif seq_len < 1024:
+        step = 64
+    else:
+        step = 128
+
+    # Create each detector as a named local variable (unrolled from the generic loop).
+    # This eliminates per-character iteration over the detector list and
+    # per-character eligible() virtual dispatch, while keeping every plugin class
+    # intact and fully readable.
+    d_sp: TooManySymbolOrPunctuationPlugin = TooManySymbolOrPunctuationPlugin()
+    d_ta: TooManyAccentuatedPlugin = TooManyAccentuatedPlugin()
+    d_up: UnprintablePlugin = UnprintablePlugin()
+    d_sda: SuspiciousDuplicateAccentPlugin = SuspiciousDuplicateAccentPlugin()
+    d_sr: SuspiciousRange = SuspiciousRange()
+    d_sw: SuperWeirdWordPlugin = SuperWeirdWordPlugin()
+    d_cu: CjkUncommonPlugin = CjkUncommonPlugin()
+    d_au: ArchaicUpperLowerPlugin = ArchaicUpperLowerPlugin()
+    d_ai: ArabicIsolatedFormPlugin = ArabicIsolatedFormPlugin()
+
+    # Local references for feed_info methods called in the hot loop.
+    d_sp_feed = d_sp.feed_info
+    d_ta_feed = d_ta.feed_info
+    d_up_feed = d_up.feed_info
+    d_sda_feed = d_sda.feed_info
+    d_sr_feed = d_sr.feed_info
+    d_sw_feed = d_sw.feed_info
+    d_cu_feed = d_cu.feed_info
+    d_au_feed = d_au.feed_info
+    d_ai_feed = d_ai.feed_info
+
+    # Single reusable CharInfo object (avoids per-character allocation).
+    info: CharInfo = CharInfo()
+    info_update = info.update
+
+    mean_mess_ratio: float
+
+    for block_start in range(0, seq_len, step):
+        for character in decoded_sequence[block_start : block_start + step]:
+            # Pre-compute all character properties once (shared across all plugins).
+            info_update(character)
+
+            # Detectors with eligible() == always True
+            d_up_feed(character, info)
+            d_sw_feed(character, info)
+            d_au_feed(character, info)
+
+            # Detectors with eligible() == isprintable
+            if info.printable:
+                d_sp_feed(character, info)
+                d_sr_feed(character, info)
+
+            # Detectors with eligible() == isalpha
+            if info.alpha:
+                d_ta_feed(character, info)
+                # SuspiciousDuplicateAccent: isalpha() and is_latin()
+                if info.latin:
+                    d_sda_feed(character, info)
+                # CjkUncommon: is_cjk()
+                if info.is_cjk:
+                    d_cu_feed(character, info)
+                # ArabicIsolatedForm: is_arabic()
+                if info.is_arabic:
+                    d_ai_feed(character, info)
+
+        mean_mess_ratio = (
+            d_sp.ratio
+            + d_ta.ratio
+            + d_up.ratio
+            + d_sda.ratio
+            + d_sr.ratio
+            + d_sw.ratio
+            + d_cu.ratio
+            + d_au.ratio
+            + d_ai.ratio
+        )
+
+        if mean_mess_ratio >= maximum_threshold:
+            break
+    else:
+        # Flush last word buffer in SuperWeirdWordPlugin via trailing newline.
+        info_update("\n")
+        d_sw_feed("\n", info)
+        d_au_feed("\n", info)
+        d_up_feed("\n", info)
+
+        mean_mess_ratio = (
+            d_sp.ratio
+            + d_ta.ratio
+            + d_up.ratio
+            + d_sda.ratio
+            + d_sr.ratio
+            + d_sw.ratio
+            + d_cu.ratio
+            + d_au.ratio
+            + d_ai.ratio
+        )
+
+    if debug:  # Defensive:
+        logger = getLogger("charset_normalizer")
+
+        logger.log(
+            TRACE,
+            "Mess-detector extended-analysis start. "
+            f"intermediary_mean_mess_ratio_calc={step} mean_mess_ratio={mean_mess_ratio} "
+            f"maximum_threshold={maximum_threshold}",
+        )
+
+        if seq_len > 16:
+            logger.log(TRACE, f"Starting with: {decoded_sequence[:16]}")
+            logger.log(TRACE, f"Ending with: {decoded_sequence[-16::]}")
+
+        for dt in [d_sp, d_ta, d_up, d_sda, d_sr, d_sw, d_cu, d_au, d_ai]:
+            logger.log(TRACE, f"{dt.__class__}: {dt.ratio}")
+
+    return round(mean_mess_ratio, 3)

.venv/Lib/site-packages/charset_normalizer/models.py

@@ -0,0 +1,369 @@
+from __future__ import annotations
+
+from encodings.aliases import aliases
+from json import dumps
+from re import sub
+from typing import Any, Iterator, List, Tuple
+
+from .constant import RE_POSSIBLE_ENCODING_INDICATION, TOO_BIG_SEQUENCE
+from .utils import iana_name, is_multi_byte_encoding, unicode_range
+
+
+class CharsetMatch:
+    def __init__(
+        self,
+        payload: bytes | bytearray,
+        guessed_encoding: str,
+        mean_mess_ratio: float,
+        has_sig_or_bom: bool,
+        languages: CoherenceMatches,
+        decoded_payload: str | None = None,
+        preemptive_declaration: str | None = None,
+    ):
+        self._payload: bytes | bytearray = payload
+
+        self._encoding: str = guessed_encoding
+        self._mean_mess_ratio: float = mean_mess_ratio
+        self._languages: CoherenceMatches = languages
+        self._has_sig_or_bom: bool = has_sig_or_bom
+        self._unicode_ranges: list[str] | None = None
+
+        self._leaves: list[CharsetMatch] = []
+        self._mean_coherence_ratio: float = 0.0
+
+        self._output_payload: bytes | None = None
+        self._output_encoding: str | None = None
+
+        self._string: str | None = decoded_payload
+
+        self._preemptive_declaration: str | None = preemptive_declaration
+
+    def __eq__(self, other: object) -> bool:
+        if not isinstance(other, CharsetMatch):
+            if isinstance(other, str):
+                return iana_name(other) == self.encoding
+            return False
+        return self.encoding == other.encoding and self.fingerprint == other.fingerprint
+
+    def __lt__(self, other: object) -> bool:
+        """
+        Implemented to make sorted available upon CharsetMatches items.
+        """
+        if not isinstance(other, CharsetMatch):
+            raise ValueError
+
+        chaos_difference: float = abs(self.chaos - other.chaos)
+        coherence_difference: float = abs(self.coherence - other.coherence)
+
+        # Below 0.5% difference --> Use Coherence
+        if chaos_difference < 0.005 and coherence_difference > 0.02:
+            return self.coherence > other.coherence
+        elif chaos_difference < 0.005 and coherence_difference <= 0.02:
+            # When having a difficult decision, use the result that decoded as many multi-byte as possible.
+            # preserve RAM usage!
+            if len(self._payload) >= TOO_BIG_SEQUENCE:
+                return self.chaos < other.chaos
+            return self.multi_byte_usage > other.multi_byte_usage
+
+        return self.chaos < other.chaos
+
+    @property
+    def multi_byte_usage(self) -> float:
+        return 1.0 - (len(str(self)) / len(self.raw))
+
+    def __str__(self) -> str:
+        # Lazy Str Loading
+        if self._string is None:
+            self._string = str(self._payload, self._encoding, "strict")
+            # UTF-7 BOM is encoded in modified Base64 whose byte boundary
+            # can overlap with the next character, so raw-byte stripping
+            # is unreliable. Strip the decoded BOM character instead.
+            if (
+                self._has_sig_or_bom
+                and self._encoding == "utf_7"
+                and self._string
+                and self._string[0] == "\ufeff"
+            ):
+                self._string = self._string[1:]
+        return self._string
+
+    def __repr__(self) -> str:
+        return f"<CharsetMatch '{self.encoding}' fp({self.fingerprint})>"
+
+    def add_submatch(self, other: CharsetMatch) -> None:
+        if not isinstance(other, CharsetMatch) or other == self:
+            raise ValueError(
+                "Unable to add instance <{}> as a submatch of a CharsetMatch".format(
+                    other.__class__
+                )
+            )
+
+        other._string = None  # Unload RAM usage; dirty trick.
+        self._leaves.append(other)
+
+    @property
+    def encoding(self) -> str:
+        return self._encoding
+
+    @property
+    def encoding_aliases(self) -> list[str]:
+        """
+        Encoding name are known by many name, using this could help when searching for IBM855 when it's listed as CP855.
+        """
+        also_known_as: list[str] = []
+        for u, p in aliases.items():
+            if self.encoding == u:
+                also_known_as.append(p)
+            elif self.encoding == p:
+                also_known_as.append(u)
+        return also_known_as
+
+    @property
+    def bom(self) -> bool:
+        return self._has_sig_or_bom
+
+    @property
+    def byte_order_mark(self) -> bool:
+        return self._has_sig_or_bom
+
+    @property
+    def languages(self) -> list[str]:
+        """
+        Return the complete list of possible languages found in decoded sequence.
+        Usually not really useful. Returned list may be empty even if 'language' property return something != 'Unknown'.
+        """
+        return [e[0] for e in self._languages]
+
+    @property
+    def language(self) -> str:
+        """
+        Most probable language found in decoded sequence. If none were detected or inferred, the property will return
+        "Unknown".
+        """
+        if not self._languages:
+            # Trying to infer the language based on the given encoding
+            # Its either English or we should not pronounce ourselves in certain cases.
+            if "ascii" in self.could_be_from_charset:
+                return "English"
+
+            # doing it there to avoid circular import
+            from charset_normalizer.cd import encoding_languages, mb_encoding_languages
+
+            languages = (
+                mb_encoding_languages(self.encoding)
+                if is_multi_byte_encoding(self.encoding)
+                else encoding_languages(self.encoding)
+            )
+
+            if len(languages) == 0 or "Latin Based" in languages:
+                return "Unknown"
+
+            return languages[0]
+
+        return self._languages[0][0]
+
+    @property
+    def chaos(self) -> float:
+        return self._mean_mess_ratio
+
+    @property
+    def coherence(self) -> float:
+        if not self._languages:
+            return 0.0
+        return self._languages[0][1]
+
+    @property
+    def percent_chaos(self) -> float:
+        return round(self.chaos * 100, ndigits=3)
+
+    @property
+    def percent_coherence(self) -> float:
+        return round(self.coherence * 100, ndigits=3)
+
+    @property
+    def raw(self) -> bytes | bytearray:
+        """
+        Original untouched bytes.
+        """
+        return self._payload
+
+    @property
+    def submatch(self) -> list[CharsetMatch]:
+        return self._leaves
+
+    @property
+    def has_submatch(self) -> bool:
+        return len(self._leaves) > 0
+
+    @property
+    def alphabets(self) -> list[str]:
+        if self._unicode_ranges is not None:
+            return self._unicode_ranges
+        # list detected ranges
+        detected_ranges: list[str | None] = [unicode_range(char) for char in str(self)]
+        # filter and sort
+        self._unicode_ranges = sorted(list({r for r in detected_ranges if r}))
+        return self._unicode_ranges
+
+    @property
+    def could_be_from_charset(self) -> list[str]:
+        """
+        The complete list of encoding that output the exact SAME str result and therefore could be the originating
+        encoding.
+        This list does include the encoding available in property 'encoding'.
+        """
+        return [self._encoding] + [m.encoding for m in self._leaves]
+
+    def output(self, encoding: str = "utf_8") -> bytes:
+        """
+        Method to get re-encoded bytes payload using given target encoding. Default to UTF-8.
+        Any errors will be simply ignored by the encoder NOT replaced.
+        """
+        if self._output_encoding is None or self._output_encoding != encoding:
+            self._output_encoding = encoding
+            decoded_string = str(self)
+            if (
+                self._preemptive_declaration is not None
+                and self._preemptive_declaration.lower()
+                not in ["utf-8", "utf8", "utf_8"]
+            ):
+                patched_header = sub(
+                    RE_POSSIBLE_ENCODING_INDICATION,
+                    lambda m: m.string[m.span()[0] : m.span()[1]].replace(
+                        m.groups()[0],
+                        iana_name(self._output_encoding).replace("_", "-"),  # type: ignore[arg-type]
+                    ),
+                    decoded_string[:8192],
+                    count=1,
+                )
+
+                decoded_string = patched_header + decoded_string[8192:]
+
+            self._output_payload = decoded_string.encode(encoding, "replace")
+
+        return self._output_payload  # type: ignore
+
+    @property
+    def fingerprint(self) -> int:
+        """
+        Retrieve a hash fingerprint of the decoded payload, used for deduplication.
+        """
+        return hash(str(self))
+
+
+class CharsetMatches:
+    """
+    Container with every CharsetMatch items ordered by default from most probable to the less one.
+    Act like a list(iterable) but does not implements all related methods.
+    """
+
+    def __init__(self, results: list[CharsetMatch] | None = None):
+        self._results: list[CharsetMatch] = sorted(results) if results else []
+
+    def __iter__(self) -> Iterator[CharsetMatch]:
+        yield from self._results
+
+    def __getitem__(self, item: int | str) -> CharsetMatch:
+        """
+        Retrieve a single item either by its position or encoding name (alias may be used here).
+        Raise KeyError upon invalid index or encoding not present in results.
+        """
+        if isinstance(item, int):
+            return self._results[item]
+        if isinstance(item, str):
+            item = iana_name(item, False)
+            for result in self._results:
+                if item in result.could_be_from_charset:
+                    return result
+        raise KeyError
+
+    def __len__(self) -> int:
+        return len(self._results)
+
+    def __bool__(self) -> bool:
+        return len(self._results) > 0
+
+    def append(self, item: CharsetMatch) -> None:
+        """
+        Insert a single match. Will be inserted accordingly to preserve sort.
+        Can be inserted as a submatch.
+        """
+        if not isinstance(item, CharsetMatch):
+            raise ValueError(
+                "Cannot append instance '{}' to CharsetMatches".format(
+                    str(item.__class__)
+                )
+            )
+        # We should disable the submatch factoring when the input file is too heavy (conserve RAM usage)
+        if len(item.raw) < TOO_BIG_SEQUENCE:
+            for match in self._results:
+                if match.fingerprint == item.fingerprint and match.chaos == item.chaos:
+                    match.add_submatch(item)
+                    return
+        self._results.append(item)
+        self._results = sorted(self._results)
+
+    def best(self) -> CharsetMatch | None:
+        """
+        Simply return the first match. Strict equivalent to matches[0].
+        """
+        if not self._results:
+            return None
+        return self._results[0]
+
+    def first(self) -> CharsetMatch | None:
+        """
+        Redundant method, call the method best(). Kept for BC reasons.
+        """
+        return self.best()
+
+
+CoherenceMatch = Tuple[str, float]
+CoherenceMatches = List[CoherenceMatch]
+
+
+class CliDetectionResult:
+    def __init__(
+        self,
+        path: str,
+        encoding: str | None,
+        encoding_aliases: list[str],
+        alternative_encodings: list[str],
+        language: str,
+        alphabets: list[str],
+        has_sig_or_bom: bool,
+        chaos: float,
+        coherence: float,
+        unicode_path: str | None,
+        is_preferred: bool,
+    ):
+        self.path: str = path
+        self.unicode_path: str | None = unicode_path
+        self.encoding: str | None = encoding
+        self.encoding_aliases: list[str] = encoding_aliases
+        self.alternative_encodings: list[str] = alternative_encodings
+        self.language: str = language
+        self.alphabets: list[str] = alphabets
+        self.has_sig_or_bom: bool = has_sig_or_bom
+        self.chaos: float = chaos
+        self.coherence: float = coherence
+        self.is_preferred: bool = is_preferred
+
+    @property
+    def __dict__(self) -> dict[str, Any]:  # type: ignore
+        return {
+            "path": self.path,
+            "encoding": self.encoding,
+            "encoding_aliases": self.encoding_aliases,
+            "alternative_encodings": self.alternative_encodings,
+            "language": self.language,
+            "alphabets": self.alphabets,
+            "has_sig_or_bom": self.has_sig_or_bom,
+            "chaos": self.chaos,
+            "coherence": self.coherence,
+            "unicode_path": self.unicode_path,
+            "is_preferred": self.is_preferred,
+        }
+
+    def to_json(self) -> str:
+        return dumps(self.__dict__, ensure_ascii=True, indent=4)

.venv/Lib/site-packages/charset_normalizer/utils.py

@@ -0,0 +1,422 @@
+from __future__ import annotations
+
+import importlib
+import logging
+import unicodedata
+from bisect import bisect_right
+from codecs import IncrementalDecoder
+from encodings.aliases import aliases
+from functools import lru_cache
+from re import findall
+from typing import Generator
+
+from _multibytecodec import (  # type: ignore[import-not-found,import]
+    MultibyteIncrementalDecoder,
+)
+
+from .constant import (
+    ENCODING_MARKS,
+    IANA_SUPPORTED_SIMILAR,
+    RE_POSSIBLE_ENCODING_INDICATION,
+    UNICODE_RANGES_COMBINED,
+    UNICODE_SECONDARY_RANGE_KEYWORD,
+    UTF8_MAXIMAL_ALLOCATION,
+    COMMON_CJK_CHARACTERS,
+    _LATIN,
+    _CJK,
+    _HANGUL,
+    _KATAKANA,
+    _HIRAGANA,
+    _THAI,
+    _ARABIC,
+    _ARABIC_ISOLATED_FORM,
+    _ACCENT_KEYWORDS,
+    _ACCENTUATED,
+)
+
+
+@lru_cache(maxsize=UTF8_MAXIMAL_ALLOCATION)
+def _character_flags(character: str) -> int:
+    """Compute all name-based classification flags with a single unicodedata.name() call."""
+    try:
+        desc: str = unicodedata.name(character)
+    except ValueError:
+        return 0
+
+    flags: int = 0
+
+    if "LATIN" in desc:
+        flags |= _LATIN
+    if "CJK" in desc:
+        flags |= _CJK
+    if "HANGUL" in desc:
+        flags |= _HANGUL
+    if "KATAKANA" in desc:
+        flags |= _KATAKANA
+    if "HIRAGANA" in desc:
+        flags |= _HIRAGANA
+    if "THAI" in desc:
+        flags |= _THAI
+    if "ARABIC" in desc:
+        flags |= _ARABIC
+        if "ISOLATED FORM" in desc:
+            flags |= _ARABIC_ISOLATED_FORM
+
+    for kw in _ACCENT_KEYWORDS:
+        if kw in desc:
+            flags |= _ACCENTUATED
+            break
+
+    return flags
+
+
+@lru_cache(maxsize=UTF8_MAXIMAL_ALLOCATION)
+def is_accentuated(character: str) -> bool:
+    return bool(_character_flags(character) & _ACCENTUATED)
+
+
+@lru_cache(maxsize=UTF8_MAXIMAL_ALLOCATION)
+def remove_accent(character: str) -> str:
+    decomposed: str = unicodedata.decomposition(character)
+    if not decomposed:
+        return character
+
+    codes: list[str] = decomposed.split(" ")
+
+    return chr(int(codes[0], 16))
+
+
+# Pre-built sorted lookup table for O(log n) binary search in unicode_range().
+# Each entry is (range_start, range_end_exclusive, range_name).
+_UNICODE_RANGES_SORTED: list[tuple[int, int, str]] = sorted(
+    (ord_range.start, ord_range.stop, name)
+    for name, ord_range in UNICODE_RANGES_COMBINED.items()
+)
+_UNICODE_RANGE_STARTS: list[int] = [e[0] for e in _UNICODE_RANGES_SORTED]
+
+
+@lru_cache(maxsize=UTF8_MAXIMAL_ALLOCATION)
+def unicode_range(character: str) -> str | None:
+    """
+    Retrieve the Unicode range official name from a single character.
+    """
+    character_ord: int = ord(character)
+
+    # Binary search: find the rightmost range whose start <= character_ord
+    idx = bisect_right(_UNICODE_RANGE_STARTS, character_ord) - 1
+    if idx >= 0:
+        start, stop, name = _UNICODE_RANGES_SORTED[idx]
+        if character_ord < stop:
+            return name
+
+    return None
+
+
+@lru_cache(maxsize=UTF8_MAXIMAL_ALLOCATION)
+def is_latin(character: str) -> bool:
+    return bool(_character_flags(character) & _LATIN)
+
+
+@lru_cache(maxsize=UTF8_MAXIMAL_ALLOCATION)
+def is_punctuation(character: str) -> bool:
+    character_category: str = unicodedata.category(character)
+
+    if "P" in character_category:
+        return True
+
+    character_range: str | None = unicode_range(character)
+
+    if character_range is None:
+        return False
+
+    return "Punctuation" in character_range
+
+
+@lru_cache(maxsize=UTF8_MAXIMAL_ALLOCATION)
+def is_symbol(character: str) -> bool:
+    character_category: str = unicodedata.category(character)
+
+    if "S" in character_category or "N" in character_category:
+        return True
+
+    character_range: str | None = unicode_range(character)
+
+    if character_range is None:
+        return False
+
+    return "Forms" in character_range and character_category != "Lo"
+
+
+@lru_cache(maxsize=UTF8_MAXIMAL_ALLOCATION)
+def is_emoticon(character: str) -> bool:
+    character_range: str | None = unicode_range(character)
+
+    if character_range is None:
+        return False
+
+    return "Emoticons" in character_range or "Pictographs" in character_range
+
+
+@lru_cache(maxsize=UTF8_MAXIMAL_ALLOCATION)
+def is_separator(character: str) -> bool:
+    if character.isspace() or character in {"｜", "+", "<", ">"}:
+        return True
+
+    character_category: str = unicodedata.category(character)
+
+    return "Z" in character_category or character_category in {"Po", "Pd", "Pc"}
+
+
+@lru_cache(maxsize=UTF8_MAXIMAL_ALLOCATION)
+def is_case_variable(character: str) -> bool:
+    return character.islower() != character.isupper()
+
+
+@lru_cache(maxsize=UTF8_MAXIMAL_ALLOCATION)
+def is_cjk(character: str) -> bool:
+    return bool(_character_flags(character) & _CJK)
+
+
+@lru_cache(maxsize=UTF8_MAXIMAL_ALLOCATION)
+def is_hiragana(character: str) -> bool:
+    return bool(_character_flags(character) & _HIRAGANA)
+
+
+@lru_cache(maxsize=UTF8_MAXIMAL_ALLOCATION)
+def is_katakana(character: str) -> bool:
+    return bool(_character_flags(character) & _KATAKANA)
+
+
+@lru_cache(maxsize=UTF8_MAXIMAL_ALLOCATION)
+def is_hangul(character: str) -> bool:
+    return bool(_character_flags(character) & _HANGUL)
+
+
+@lru_cache(maxsize=UTF8_MAXIMAL_ALLOCATION)
+def is_thai(character: str) -> bool:
+    return bool(_character_flags(character) & _THAI)
+
+
+@lru_cache(maxsize=UTF8_MAXIMAL_ALLOCATION)
+def is_arabic(character: str) -> bool:
+    return bool(_character_flags(character) & _ARABIC)
+
+
+@lru_cache(maxsize=UTF8_MAXIMAL_ALLOCATION)
+def is_arabic_isolated_form(character: str) -> bool:
+    return bool(_character_flags(character) & _ARABIC_ISOLATED_FORM)
+
+
+@lru_cache(maxsize=UTF8_MAXIMAL_ALLOCATION)
+def is_cjk_uncommon(character: str) -> bool:
+    return character not in COMMON_CJK_CHARACTERS
+
+
+@lru_cache(maxsize=len(UNICODE_RANGES_COMBINED))
+def is_unicode_range_secondary(range_name: str) -> bool:
+    return any(keyword in range_name for keyword in UNICODE_SECONDARY_RANGE_KEYWORD)
+
+
+@lru_cache(maxsize=UTF8_MAXIMAL_ALLOCATION)
+def is_unprintable(character: str) -> bool:
+    return (
+        character.isspace() is False  # includes \n \t \r \v
+        and character.isprintable() is False
+        and character != "\x1a"  # Why? Its the ASCII substitute character.
+        and character != "\ufeff"  # bug discovered in Python,
+        # Zero Width No-Break Space located in 	Arabic Presentation Forms-B, Unicode 1.1 not acknowledged as space.
+    )
+
+
+def any_specified_encoding(
+    sequence: bytes | bytearray, search_zone: int = 8192
+) -> str | None:
+    """
+    Extract using ASCII-only decoder any specified encoding in the first n-bytes.
+    """
+    if not isinstance(sequence, (bytes, bytearray)):
+        raise TypeError
+
+    seq_len: int = len(sequence)
+
+    results: list[str] = findall(
+        RE_POSSIBLE_ENCODING_INDICATION,
+        sequence[: min(seq_len, search_zone)].decode("ascii", errors="ignore"),
+    )
+
+    if len(results) == 0:
+        return None
+
+    for specified_encoding in results:
+        specified_encoding = specified_encoding.lower().replace("-", "_")
+
+        encoding_alias: str
+        encoding_iana: str
+
+        for encoding_alias, encoding_iana in aliases.items():
+            if encoding_alias == specified_encoding:
+                return encoding_iana
+            if encoding_iana == specified_encoding:
+                return encoding_iana
+
+    return None
+
+
+@lru_cache(maxsize=128)
+def is_multi_byte_encoding(name: str) -> bool:
+    """
+    Verify is a specific encoding is a multi byte one based on it IANA name
+    """
+    return name in {
+        "utf_8",
+        "utf_8_sig",
+        "utf_16",
+        "utf_16_be",
+        "utf_16_le",
+        "utf_32",
+        "utf_32_le",
+        "utf_32_be",
+        "utf_7",
+    } or issubclass(
+        importlib.import_module(f"encodings.{name}").IncrementalDecoder,
+        MultibyteIncrementalDecoder,
+    )
+
+
+def identify_sig_or_bom(sequence: bytes | bytearray) -> tuple[str | None, bytes]:
+    """
+    Identify and extract SIG/BOM in given sequence.
+    """
+
+    for iana_encoding in ENCODING_MARKS:
+        marks: bytes | list[bytes] = ENCODING_MARKS[iana_encoding]
+
+        if isinstance(marks, bytes):
+            marks = [marks]
+
+        for mark in marks:
+            if sequence.startswith(mark):
+                return iana_encoding, mark
+
+    return None, b""
+
+
+def should_strip_sig_or_bom(iana_encoding: str) -> bool:
+    return iana_encoding not in {"utf_16", "utf_32"}
+
+
+def iana_name(cp_name: str, strict: bool = True) -> str:
+    """Returns the Python normalized encoding name (Not the IANA official name)."""
+    cp_name = cp_name.lower().replace("-", "_")
+
+    encoding_alias: str
+    encoding_iana: str
+
+    for encoding_alias, encoding_iana in aliases.items():
+        if cp_name in [encoding_alias, encoding_iana]:
+            return encoding_iana
+
+    if strict:
+        raise ValueError(f"Unable to retrieve IANA for '{cp_name}'")
+
+    return cp_name
+
+
+def cp_similarity(iana_name_a: str, iana_name_b: str) -> float:
+    if is_multi_byte_encoding(iana_name_a) or is_multi_byte_encoding(iana_name_b):
+        return 0.0
+
+    decoder_a = importlib.import_module(f"encodings.{iana_name_a}").IncrementalDecoder
+    decoder_b = importlib.import_module(f"encodings.{iana_name_b}").IncrementalDecoder
+
+    id_a: IncrementalDecoder = decoder_a(errors="ignore")
+    id_b: IncrementalDecoder = decoder_b(errors="ignore")
+
+    character_match_count: int = 0
+
+    for i in range(256):
+        to_be_decoded: bytes = bytes([i])
+        if id_a.decode(to_be_decoded) == id_b.decode(to_be_decoded):
+            character_match_count += 1
+
+    return character_match_count / 256
+
+
+def is_cp_similar(iana_name_a: str, iana_name_b: str) -> bool:
+    """
+    Determine if two code page are at least 80% similar. IANA_SUPPORTED_SIMILAR dict was generated using
+    the function cp_similarity.
+    """
+    return (
+        iana_name_a in IANA_SUPPORTED_SIMILAR
+        and iana_name_b in IANA_SUPPORTED_SIMILAR[iana_name_a]
+    )
+
+
+def set_logging_handler(
+    name: str = "charset_normalizer",
+    level: int = logging.INFO,
+    format_string: str = "%(asctime)s | %(levelname)s | %(message)s",
+) -> None:
+    logger = logging.getLogger(name)
+    logger.setLevel(level)
+
+    handler = logging.StreamHandler()
+    handler.setFormatter(logging.Formatter(format_string))
+    logger.addHandler(handler)
+
+
+def cut_sequence_chunks(
+    sequences: bytes | bytearray,
+    encoding_iana: str,
+    offsets: range,
+    chunk_size: int,
+    bom_or_sig_available: bool,
+    strip_sig_or_bom: bool,
+    sig_payload: bytes,
+    is_multi_byte_decoder: bool,
+    decoded_payload: str | None = None,
+) -> Generator[str, None, None]:
+    if decoded_payload and is_multi_byte_decoder is False:
+        for i in offsets:
+            chunk = decoded_payload[i : i + chunk_size]
+            if not chunk:
+                break
+            yield chunk
+    else:
+        for i in offsets:
+            chunk_end = i + chunk_size
+            if chunk_end > len(sequences) + 8:
+                continue
+
+            cut_sequence = sequences[i : i + chunk_size]
+
+            if bom_or_sig_available and strip_sig_or_bom is False:
+                cut_sequence = sig_payload + cut_sequence
+
+            chunk = cut_sequence.decode(
+                encoding_iana,
+                errors="ignore" if is_multi_byte_decoder else "strict",
+            )
+
+            # multi-byte bad cutting detector and adjustment
+            # not the cleanest way to perform that fix but clever enough for now.
+            if is_multi_byte_decoder and i > 0:
+                chunk_partial_size_chk: int = min(chunk_size, 16)
+
+                if (
+                    decoded_payload
+                    and chunk[:chunk_partial_size_chk] not in decoded_payload
+                ):
+                    for j in range(i, i - 4, -1):
+                        cut_sequence = sequences[j:chunk_end]
+
+                        if bom_or_sig_available and strip_sig_or_bom is False:
+                            cut_sequence = sig_payload + cut_sequence
+
+                        chunk = cut_sequence.decode(encoding_iana, errors="ignore")
+
+                        if chunk[:chunk_partial_size_chk] in decoded_payload:
+                            break
+
+            yield chunk

.venv/Lib/site-packages/charset_normalizer/version.py

@@ -0,0 +1,8 @@
+"""
+Expose version
+"""
+
+from __future__ import annotations
+
+__version__ = "3.4.7"
+VERSION = __version__.split(".")

.venv/Lib/site-packages/feedparser-6.0.12.dist-info/INSTALLER

@@ -0,0 +1 @@
+pip

.venv/Lib/site-packages/feedparser-6.0.12.dist-info/METADATA

@@ -0,0 +1,102 @@
+Metadata-Version: 2.4
+Name: feedparser
+Version: 6.0.12
+Summary: Universal feed parser, handles RSS 0.9x, RSS 1.0, RSS 2.0, CDF, Atom 0.3, and Atom 1.0 feeds
+Home-page: https://github.com/kurtmckee/feedparser
+Download-URL: https://pypi.python.org/pypi/feedparser
+Author: Kurt McKee
+Author-email: contactme@kurtmckee.org
+License: BSD-2-Clause
+Keywords: atom,cdf,feed,parser,rdf,rss
+Platform: POSIX
+Platform: Windows
+Classifier: Development Status :: 5 - Production/Stable
+Classifier: Intended Audience :: Developers
+Classifier: License :: OSI Approved :: BSD License
+Classifier: Operating System :: OS Independent
+Classifier: Programming Language :: Python
+Classifier: Programming Language :: Python :: 3.6
+Classifier: Programming Language :: Python :: 3.7
+Classifier: Programming Language :: Python :: 3.8
+Classifier: Programming Language :: Python :: 3.9
+Classifier: Topic :: Software Development :: Libraries :: Python Modules
+Classifier: Topic :: Text Processing :: Markup :: XML
+Requires-Python: >=3.6
+Description-Content-Type: text/x-rst
+License-File: LICENSE
+Requires-Dist: sgmllib3k
+Dynamic: author
+Dynamic: author-email
+Dynamic: classifier
+Dynamic: description
+Dynamic: description-content-type
+Dynamic: download-url
+Dynamic: home-page
+Dynamic: keywords
+Dynamic: license
+Dynamic: license-file
+Dynamic: platform
+Dynamic: requires-dist
+Dynamic: requires-python
+Dynamic: summary
+
+..
+    This file is part of feedparser.
+    Copyright 2010-2025 Kurt McKee <contactme@kurtmckee.org>
+    Copyright 2002-2008 Mark Pilgrim
+    Released under the BSD 2-clause license.
+
+
+feedparser
+##########
+
+Parse Atom and RSS feeds in Python.
+
+----
+
+
+Installation
+============
+
+feedparser can be installed by running pip:
+
+..  code-block:: console
+
+    $ pip install feedparser
+
+
+Documentation
+=============
+
+The feedparser documentation is available on the web at:
+
+    https://feedparser.readthedocs.io/en/latest/
+
+It is also included in its source format, ReST, in the ``docs/`` directory.
+To build the documentation you'll need the Sphinx package, which is available at:
+
+    https://www.sphinx-doc.org/
+
+You can then build HTML pages using a command similar to:
+
+..  code-block:: console
+
+    $ sphinx-build -b html docs/ fpdocs
+
+This will produce HTML documentation in the ``fpdocs/`` directory.
+
+
+Testing
+=======
+
+Feedparser has an extensive test suite, powered by tox. To run it, type this:
+
+..  code-block:: console
+
+    $ python -m venv venv
+    $ source venv/bin/activate  # or "venv\bin\activate.ps1" on Windows
+    (venv) $ pip install -r requirements-dev.txt
+    (venv) $ tox
+
+This will spawn an HTTP server that will listen on port 8097. The tests will
+fail if that port is in use.

.venv/Lib/site-packages/feedparser-6.0.12.dist-info/RECORD

@@ -0,0 +1,71 @@
+feedparser-6.0.12.dist-info/INSTALLER,sha256=zuuue4knoyJ-UwPPXg8fezS7VCrXJQrAP7zeNuwvFQg,4
+feedparser-6.0.12.dist-info/METADATA,sha256=sjwQUsly3f8iVU6mfuPJ07WcYfklLdtp1BQQMNV0CyU,2682
+feedparser-6.0.12.dist-info/RECORD,,
+feedparser-6.0.12.dist-info/REQUESTED,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
+feedparser-6.0.12.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
+feedparser-6.0.12.dist-info/licenses/LICENSE,sha256=n16SeD_ZvrKp3nas2jKo-W7oVSorVm45I_A38NdUIio,3146
+feedparser-6.0.12.dist-info/top_level.txt,sha256=V8OMyOFfOWI46em2E-SGpIywuPZEYtMPOcPua_gzvUk,11
+feedparser/__init__.py,sha256=I-6CoNSzaMv8ADF4iNAM6a7nrINO16f0KQpORWY1els,2177
+feedparser/__pycache__/__init__.cpython-314.pyc,,
+feedparser/__pycache__/api.cpython-314.pyc,,
+feedparser/__pycache__/encodings.cpython-314.pyc,,
+feedparser/__pycache__/exceptions.cpython-314.pyc,,
+feedparser/__pycache__/html.cpython-314.pyc,,
+feedparser/__pycache__/http.cpython-314.pyc,,
+feedparser/__pycache__/mixin.cpython-314.pyc,,
+feedparser/__pycache__/sanitizer.cpython-314.pyc,,
+feedparser/__pycache__/sgml.cpython-314.pyc,,
+feedparser/__pycache__/urls.cpython-314.pyc,,
+feedparser/__pycache__/util.cpython-314.pyc,,
+feedparser/api.py,sha256=APVBSiOxYRRmB6iLi5TY8VuQds6aj52SxcbR8XP3I3E,11379
+feedparser/datetimes/__init__.py,sha256=IiDI3Tq53l5GhiDk_gPkoCylm0-ayWHPFJC3-jsYQWQ,2786
+feedparser/datetimes/__pycache__/__init__.cpython-314.pyc,,
+feedparser/datetimes/__pycache__/asctime.cpython-314.pyc,,
+feedparser/datetimes/__pycache__/greek.cpython-314.pyc,,
+feedparser/datetimes/__pycache__/hungarian.cpython-314.pyc,,
+feedparser/datetimes/__pycache__/iso8601.cpython-314.pyc,,
+feedparser/datetimes/__pycache__/korean.cpython-314.pyc,,
+feedparser/datetimes/__pycache__/perforce.cpython-314.pyc,,
+feedparser/datetimes/__pycache__/rfc822.cpython-314.pyc,,
+feedparser/datetimes/__pycache__/w3dtf.cpython-314.pyc,,
+feedparser/datetimes/asctime.py,sha256=m_WvJum_ckvyw_ZNp6-Iz0W1920YfGECXhFtqbR34KA,2380
+feedparser/datetimes/greek.py,sha256=HIr5f2Ob9aJ-FEjtZKri1M0fiH9rfrpWbnNLbw3ovjc,4022
+feedparser/datetimes/hungarian.py,sha256=dQYEGiTcXnAEgxpy6BU8cCJzAwjtop2I9bpiDMYTG4c,2945
+feedparser/datetimes/iso8601.py,sha256=p1MfwSpMAoBGPJu1pAUQVULQNIMGuUmyLM65K3TSVko,5550
+feedparser/datetimes/korean.py,sha256=G7gLbJG1ULot_Cydb0_JQpgVuXex7lA3lv7lwcnNF8A,3354
+feedparser/datetimes/perforce.py,sha256=oFpA8cfGI7i-B2Ygx1Ow0OVuXDevDB6RZMHG9MnHneY,2213
+feedparser/datetimes/rfc822.py,sha256=wSrK28KbOYDIi72GUQDkr88TDMUH0ekwJ_S7CrZ54q0,5423
+feedparser/datetimes/w3dtf.py,sha256=prld3G3Z5YnNrE1nEV_4U6hApBSIToUNJ1N_l_CqKA4,4506
+feedparser/encodings.py,sha256=7Etu3zAuZpR3KiGWmquKlBZ-ir_TbPVq7aC6el93_ak,12558
+feedparser/exceptions.py,sha256=2jmZzsrJoIccUizZ3wGC4US0FzzZfn2HnRSHHxx4wts,1957
+feedparser/html.py,sha256=bEnb4YMFbuHlA-zQ788-OWl36Ro1K8UxEV7NAtQvWhQ,11282
+feedparser/http.py,sha256=5OBayWEn0JnWhywnArq6iKp1oei0lCJuty3p7h1BX8g,9844
+feedparser/mixin.py,sha256=0fCiJfDPrDY0OUnnhH3VY-m3Jv16F-Cn5P-3qFswZPw,32224
+feedparser/namespaces/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
+feedparser/namespaces/__pycache__/__init__.cpython-314.pyc,,
+feedparser/namespaces/__pycache__/_base.cpython-314.pyc,,
+feedparser/namespaces/__pycache__/admin.cpython-314.pyc,,
+feedparser/namespaces/__pycache__/cc.cpython-314.pyc,,
+feedparser/namespaces/__pycache__/dc.cpython-314.pyc,,
+feedparser/namespaces/__pycache__/georss.cpython-314.pyc,,
+feedparser/namespaces/__pycache__/itunes.cpython-314.pyc,,
+feedparser/namespaces/__pycache__/mediarss.cpython-314.pyc,,
+feedparser/namespaces/__pycache__/psc.cpython-314.pyc,,
+feedparser/namespaces/_base.py,sha256=-mEtrLkSYBSIUQcE3NegwyN9m8y3-EYQodbqdzgTEQY,17130
+feedparser/namespaces/admin.py,sha256=7QYxJEu0U-_AnI4pEfCyT_91NCLP6O7OmUJbqAxRKQQ,2317
+feedparser/namespaces/cc.py,sha256=_8Z6bVyMwD6dHbZ8jaTX2XMR1ARbmubArQbQdxbffuI,2866
+feedparser/namespaces/dc.py,sha256=9IrXjfKOiXrVW1lowSKnwwAPbH5h1wUC20i4Q5gDEcQ,4446
+feedparser/namespaces/georss.py,sha256=XUIGVB_OQcxZ08ttyXHAOXev1Gf6lddUI3Gb9_pR-Lk,11468
+feedparser/namespaces/itunes.py,sha256=0gREO92NR679WnXpP2K84H-E8dI9cFLuKQq5yW30oSw,4115
+feedparser/namespaces/mediarss.py,sha256=twS7pV0QcrOXAPXFjnHVpFNRsUxR8XJpxFehEzfzozg,5336
+feedparser/namespaces/psc.py,sha256=o_ito9jkMFp8tyX7j3mHGhaqh2AYBXMXhdG6pH3AvWo,2839
+feedparser/parsers/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
+feedparser/parsers/__pycache__/__init__.cpython-314.pyc,,
+feedparser/parsers/__pycache__/loose.cpython-314.pyc,,
+feedparser/parsers/__pycache__/strict.cpython-314.pyc,,
+feedparser/parsers/loose.py,sha256=ejZhqr5dbXm554tfkr-XPLh8fuIu9rTBO15RKOvfsHI,3452
+feedparser/parsers/strict.py,sha256=34iQHymDCQb89q9IKaPMjUe_Tl0iuBZT2IdKEFDmmOQ,5817
+feedparser/sanitizer.py,sha256=rZbj_hC2nCFk_WTighRNjzIWRiixaUO0mxXgvBPK5zs,23856
+feedparser/sgml.py,sha256=VgcjIARGQ3OCl8yeatDA9peaVb_XO13X20NICsmxNyU,3488
+feedparser/urls.py,sha256=4vQzNz2CY8MIdBsDmtvh7GSshOeqxcQSNU9GWelUx3Q,5490
+feedparser/util.py,sha256=iBMIYNIhMBbOSS9JZIr67LJtgLo4eLk0VifPx8E2HRY,6490

.venv/Lib/site-packages/feedparser-6.0.12.dist-info/WHEEL

@@ -0,0 +1,5 @@
+Wheel-Version: 1.0
+Generator: setuptools (80.9.0)
+Root-Is-Purelib: true
+Tag: py3-none-any
+

.venv/Lib/site-packages/feedparser-6.0.12.dist-info/licenses/LICENSE

@@ -0,0 +1,65 @@
+feedparser and its unit tests are released under the following license:
+
+----- begin license block -----
+
+Copyright (C) 2010-2025 Kurt McKee <contactme@kurtmckee.org>
+Copyright (C) 2002-2008 Mark Pilgrim
+All rights reserved.
+
+Redistribution and use in source and binary forms, with or without modification,
+are permitted provided that the following conditions are met:
+
+* Redistributions of source code must retain the above copyright notice,
+  this list of conditions and the following disclaimer.
+* Redistributions in binary form must reproduce the above copyright notice,
+  this list of conditions and the following disclaimer in the documentation
+  and/or other materials provided with the distribution.
+
+THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS 'AS IS'
+AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE
+LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
+CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
+SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
+INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
+CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
+ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
+POSSIBILITY OF SUCH DAMAGE.
+
+----- end license block -----
+
+
+
+
+
+The feedparser documentation (everything in the docs/ directory) is
+released under the following license:
+
+----- begin license block -----
+
+Copyright (C) 2010-2025 Kurt McKee <contactme@kurtmckee.org>
+Copyright (C) 2004-2008 Mark Pilgrim. All rights reserved.
+
+Redistribution and use in source (Sphinx ReST) and "compiled" forms (HTML, PDF,
+PostScript, RTF and so forth) with or without modification, are permitted
+provided that the following conditions are met:
+
+* Redistributions of source code (Sphinx ReST) must retain the above copyright
+  notice, this list of conditions and the following disclaimer.
+* Redistributions in compiled form (converted to HTML, PDF, PostScript, RTF and
+  other formats) must reproduce the above copyright notice, this list of
+  conditions and the following disclaimer in the documentation and/or other
+  materials provided with the distribution.
+
+THIS DOCUMENTATION IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS 'AS IS'
+AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE
+LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
+CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
+SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
+INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
+CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
+ARISING IN ANY WAY OUT OF THE USE OF THIS DOCUMENTATION, EVEN IF ADVISED OF THE
+POSSIBILITY OF SUCH DAMAGE.

.venv/Lib/site-packages/feedparser-6.0.12.dist-info/top_level.txt

@@ -0,0 +1 @@
+feedparser

.venv/Lib/site-packages/feedparser/__init__.py

@@ -0,0 +1,48 @@
+# Copyright 2010-2025 Kurt McKee <contactme@kurtmckee.org>
+# Copyright 2002-2008 Mark Pilgrim
+# All rights reserved.
+#
+# This file is part of feedparser.
+#
+# Redistribution and use in source and binary forms, with or without modification,
+# are permitted provided that the following conditions are met:
+#
+# * Redistributions of source code must retain the above copyright notice,
+#   this list of conditions and the following disclaimer.
+# * Redistributions in binary form must reproduce the above copyright notice,
+#   this list of conditions and the following disclaimer in the documentation
+#   and/or other materials provided with the distribution.
+#
+# THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS 'AS IS'
+# AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+# IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+# ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE
+# LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
+# CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
+# SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
+# INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
+# CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
+# ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
+# POSSIBILITY OF SUCH DAMAGE."""
+
+from .api import parse
+from .datetimes import registerDateHandler
+from .exceptions import *
+from .util import FeedParserDict
+
+__author__ = 'Kurt McKee <contactme@kurtmckee.org>'
+__license__ = 'BSD 2-clause'
+__version__ = '6.0.12'
+
+# HTTP "User-Agent" header to send to servers when downloading feeds.
+# If you are embedding feedparser in a larger application, you should
+# change this to your application name and URL.
+USER_AGENT = "feedparser/%s +https://github.com/kurtmckee/feedparser/" % __version__
+
+# If you want feedparser to automatically resolve all relative URIs, set this
+# to 1.
+RESOLVE_RELATIVE_URIS = 1
+
+# If you want feedparser to automatically sanitize all potentially unsafe
+# HTML content, set this to 1.
+SANITIZE_HTML = 1

.venv/Lib/site-packages/feedparser/api.py

@@ -0,0 +1,277 @@
+# The public API for feedparser
+# Copyright 2010-2025 Kurt McKee <contactme@kurtmckee.org>
+# Copyright 2002-2008 Mark Pilgrim
+# All rights reserved.
+#
+# This file is a part of feedparser.
+#
+# Redistribution and use in source and binary forms, with or without modification,
+# are permitted provided that the following conditions are met:
+#
+# * Redistributions of source code must retain the above copyright notice,
+#   this list of conditions and the following disclaimer.
+# * Redistributions in binary form must reproduce the above copyright notice,
+#   this list of conditions and the following disclaimer in the documentation
+#   and/or other materials provided with the distribution.
+#
+# THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS 'AS IS'
+# AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+# IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+# ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE
+# LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
+# CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
+# SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
+# INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
+# CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
+# ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
+# POSSIBILITY OF SUCH DAMAGE.
+
+import io
+import urllib.error
+import urllib.parse
+import xml.sax
+
+from .datetimes import registerDateHandler, _parse_date
+from .encodings import convert_to_utf8
+from .exceptions import *
+from .html import _BaseHTMLProcessor
+from . import http
+from . import mixin
+from .mixin import _FeedParserMixin
+from .parsers.loose import _LooseFeedParser
+from .parsers.strict import _StrictFeedParser
+from .sanitizer import replace_doctype
+from .sgml import *
+from .urls import convert_to_idn, make_safe_absolute_uri
+from .util import FeedParserDict
+
+
+# List of preferred XML parsers, by SAX driver name.  These will be tried first,
+# but if they're not installed, Python will keep searching through its own list
+# of pre-installed parsers until it finds one that supports everything we need.
+PREFERRED_XML_PARSERS = ["drv_libxml2"]
+
+_XML_AVAILABLE = True
+
+SUPPORTED_VERSIONS = {
+    '': 'unknown',
+    'rss090': 'RSS 0.90',
+    'rss091n': 'RSS 0.91 (Netscape)',
+    'rss091u': 'RSS 0.91 (Userland)',
+    'rss092': 'RSS 0.92',
+    'rss093': 'RSS 0.93',
+    'rss094': 'RSS 0.94',
+    'rss20': 'RSS 2.0',
+    'rss10': 'RSS 1.0',
+    'rss': 'RSS (unknown version)',
+    'atom01': 'Atom 0.1',
+    'atom02': 'Atom 0.2',
+    'atom03': 'Atom 0.3',
+    'atom10': 'Atom 1.0',
+    'atom': 'Atom (unknown version)',
+    'cdf': 'CDF',
+}
+
+
+def _open_resource(url_file_stream_or_string, etag, modified, agent, referrer, handlers, request_headers, result):
+    """URL, filename, or string --> stream
+
+    This function lets you define parsers that take any input source
+    (URL, pathname to local or network file, or actual data as a string)
+    and deal with it in a uniform manner.  Returned object is guaranteed
+    to have all the basic stdio read methods (read, readline, readlines).
+    Just .close() the object when you're done with it.
+
+    If the etag argument is supplied, it will be used as the value of an
+    If-None-Match request header.
+
+    If the modified argument is supplied, it can be a tuple of 9 integers
+    (as returned by gmtime() in the standard Python time module) or a date
+    string in any format supported by feedparser. Regardless, it MUST
+    be in GMT (Greenwich Mean Time). It will be reformatted into an
+    RFC 1123-compliant date and used as the value of an If-Modified-Since
+    request header.
+
+    If the agent argument is supplied, it will be used as the value of a
+    User-Agent request header.
+
+    If the referrer argument is supplied, it will be used as the value of a
+    Referer[sic] request header.
+
+    If handlers is supplied, it is a list of handlers used to build a
+    urllib2 opener.
+
+    if request_headers is supplied it is a dictionary of HTTP request headers
+    that will override the values generated by FeedParser.
+
+    :return: A bytes object.
+    """
+
+    if hasattr(url_file_stream_or_string, 'read'):
+        return url_file_stream_or_string.read()
+
+    if isinstance(url_file_stream_or_string, str) \
+       and urllib.parse.urlparse(url_file_stream_or_string)[0] in ('http', 'https', 'ftp', 'file', 'feed'):
+        return http.get(url_file_stream_or_string, etag, modified, agent, referrer, handlers, request_headers, result)
+
+    # try to open with native open function (if url_file_stream_or_string is a filename)
+    try:
+        with open(url_file_stream_or_string, 'rb') as f:
+            data = f.read()
+    except (IOError, UnicodeEncodeError, TypeError, ValueError):
+        # if url_file_stream_or_string is a str object that
+        # cannot be converted to the encoding returned by
+        # sys.getfilesystemencoding(), a UnicodeEncodeError
+        # will be thrown
+        # If url_file_stream_or_string is a string that contains NULL
+        # (such as an XML document encoded in UTF-32), TypeError will
+        # be thrown.
+        pass
+    else:
+        return data
+
+    # treat url_file_stream_or_string as string
+    if not isinstance(url_file_stream_or_string, bytes):
+        return url_file_stream_or_string.encode('utf-8')
+    return url_file_stream_or_string
+
+
+LooseFeedParser = type(
+    'LooseFeedParser',
+    (_LooseFeedParser, _FeedParserMixin, _BaseHTMLProcessor, object),
+    {},
+)
+
+StrictFeedParser = type(
+    'StrictFeedParser',
+    (_StrictFeedParser, _FeedParserMixin, xml.sax.handler.ContentHandler, object),
+    {},
+)
+
+
+def parse(url_file_stream_or_string, etag=None, modified=None, agent=None, referrer=None, handlers=None, request_headers=None, response_headers=None, resolve_relative_uris=None, sanitize_html=None):
+    """Parse a feed from a URL, file, stream, or string.
+
+    :param url_file_stream_or_string:
+        File-like object, URL, file path, or string. Both byte and text strings
+        are accepted. If necessary, encoding will be derived from the response
+        headers or automatically detected.
+
+        Note that strings may trigger network I/O or filesystem access
+        depending on the value. Wrap an untrusted string in
+        a :class:`io.StringIO` or :class:`io.BytesIO` to avoid this. Do not
+        pass untrusted strings to this function.
+
+        When a URL is not passed the feed location to use in relative URL
+        resolution should be passed in the ``Content-Location`` response header
+        (see ``response_headers`` below).
+
+    :param str etag: HTTP ``ETag`` request header.
+    :param modified: HTTP ``Last-Modified`` request header.
+    :type modified: :class:`str`, :class:`time.struct_time` 9-tuple, or
+        :class:`datetime.datetime`
+    :param str agent: HTTP ``User-Agent`` request header, which defaults to
+        the value of :data:`feedparser.USER_AGENT`.
+    :param referrer: HTTP ``Referer`` [sic] request header.
+    :param request_headers:
+        A mapping of HTTP header name to HTTP header value to add to the
+        request, overriding internally generated values.
+    :type request_headers: :class:`dict` mapping :class:`str` to :class:`str`
+    :param response_headers:
+        A mapping of HTTP header name to HTTP header value. Multiple values may
+        be joined with a comma. If a HTTP request was made, these headers
+        override any matching headers in the response. Otherwise this specifies
+        the entirety of the response headers.
+    :type response_headers: :class:`dict` mapping :class:`str` to :class:`str`
+
+    :param bool resolve_relative_uris:
+        Should feedparser attempt to resolve relative URIs absolute ones within
+        HTML content?  Defaults to the value of
+        :data:`feedparser.RESOLVE_RELATIVE_URIS`, which is ``True``.
+    :param bool sanitize_html:
+        Should feedparser skip HTML sanitization? Only disable this if you know
+        what you are doing!  Defaults to the value of
+        :data:`feedparser.SANITIZE_HTML`, which is ``True``.
+
+    :return: A :class:`FeedParserDict`.
+    """
+
+    if not agent or sanitize_html is None or resolve_relative_uris is None:
+        import feedparser
+    if not agent:
+        agent = feedparser.USER_AGENT
+    if sanitize_html is None:
+        sanitize_html = feedparser.SANITIZE_HTML
+    if resolve_relative_uris is None:
+        resolve_relative_uris = feedparser.RESOLVE_RELATIVE_URIS
+
+    result = FeedParserDict(
+        bozo=False,
+        entries=[],
+        feed=FeedParserDict(),
+        headers={},
+    )
+
+    try:
+        data = _open_resource(url_file_stream_or_string, etag, modified, agent, referrer, handlers, request_headers, result)
+    except urllib.error.URLError as error:
+        result.update({
+            'bozo': True,
+            'bozo_exception': error,
+        })
+        return result
+
+    if not data:
+        return result
+
+    # overwrite existing headers using response_headers
+    result['headers'].update(response_headers or {})
+
+    data = convert_to_utf8(result['headers'], data, result)
+    use_strict_parser = result['encoding'] and True or False
+
+    result['version'], data, entities = replace_doctype(data)
+
+    # Ensure that baseuri is an absolute URI using an acceptable URI scheme.
+    contentloc = result['headers'].get('content-location', '')
+    href = result.get('href', '')
+    baseuri = make_safe_absolute_uri(href, contentloc) or make_safe_absolute_uri(contentloc) or href
+
+    baselang = result['headers'].get('content-language', None)
+    if isinstance(baselang, bytes) and baselang is not None:
+        baselang = baselang.decode('utf-8', 'ignore')
+
+    if not _XML_AVAILABLE:
+        use_strict_parser = 0
+    if use_strict_parser:
+        # initialize the SAX parser
+        feedparser = StrictFeedParser(baseuri, baselang, 'utf-8')
+        feedparser.resolve_relative_uris = resolve_relative_uris
+        feedparser.sanitize_html = sanitize_html
+        saxparser = xml.sax.make_parser(PREFERRED_XML_PARSERS)
+        saxparser.setFeature(xml.sax.handler.feature_namespaces, 1)
+        try:
+            # disable downloading external doctype references, if possible
+            saxparser.setFeature(xml.sax.handler.feature_external_ges, 0)
+        except xml.sax.SAXNotSupportedException:
+            pass
+        saxparser.setContentHandler(feedparser)
+        saxparser.setErrorHandler(feedparser)
+        source = xml.sax.xmlreader.InputSource()
+        source.setByteStream(io.BytesIO(data))
+        try:
+            saxparser.parse(source)
+        except xml.sax.SAXException as e:
+            result['bozo'] = 1
+            result['bozo_exception'] = feedparser.exc or e
+            use_strict_parser = 0
+    if not use_strict_parser:
+        feedparser = LooseFeedParser(baseuri, baselang, 'utf-8', entities)
+        feedparser.resolve_relative_uris = resolve_relative_uris
+        feedparser.sanitize_html = sanitize_html
+        feedparser.feed(data.decode('utf-8', 'replace'))
+    result['feed'] = feedparser.feeddata
+    result['entries'] = feedparser.entries
+    result['version'] = result['version'] or feedparser.version
+    result['namespaces'] = feedparser.namespaces_in_use
+    return result