text stringlengths 185 73.3k | repo stringlengths 7 100 | path stringlengths 4 146 | language stringclasses 7
values | hash stringlengths 16 16 | score float64 7 8.5 | stars int64 0 237k |
|---|---|---|---|---|---|---|
"""图片 OCR 脱敏单元测试(v0.4,可选特性)。
核心逻辑(mask_blocks / render_text / redact_image)是纯函数,不依赖
pytesseract / Pillow 即可测试;真机 tesseract 相关测试按可用性跳过。
"""
import os
import subprocess
import sys
import unittest
sys.path.insert(0, os.path.dirname(os.path.dirname(os.path.abspath(__file__))))
from llm_sanitizer import ocr # noqa: E40... | JunyuZhan/llm-sanitizer | tests/test_ocr.py | .py | f3a3626aa4635af7 | 7.5 | 0 |
#!/usr/bin/env python3
# SPDX-FileCopyrightText: 2026 mokume-metal
# SPDX-License-Identifier: MIT
"""版を決め、リリースノートを組む。
リリースは**リポジトリのファイルを 1 つも変えない**。まとめた変更履歴を main へ
入れるには PR が要り、Actions の既定のトークンで作った PR は他のワークフローを
起こさないので必須チェックが走らない — その迂回はリリースのたびに壊れる場所を
増やす。置き場は GitHub の Releases で足り、ファイルに写すと二重管理になる
(ADR-0001 原則 9)。
... | mokume-metal/mokume | scripts/release.py | .py | ced7aa0c777df643 | 7 | 0 |
#!/usr/bin/env python3
# SPDX-FileCopyrightText: 2026 mokume-metal
# SPDX-License-Identifier: MIT
"""ルールセット定義ファイルの検査と、実設定との照合 (#98)。
正本はリポジトリの `.github/rulesets/*.json` で、GitHub 側の状態はその写しに
すぎない (ADR-0006)。ここは二つの問いだけに答える:
shape <定義ディレクトリ> 定義が API に投げられる形か (token 不要)
diff <定義ディレクトリ> <実設定ディレクトリ>... | mokume-metal/mokume | scripts/rulesets_lib.py | .py | cb5f146e36e2dd5a | 7 | 0 |
#!/usr/bin/env python3
# SPDX-FileCopyrightText: 2026 mokume-metal
# SPDX-License-Identifier: MIT
"""scripts/release.py の検査。
守りたいのは 3 つ:
1. 版の上げ幅が履歴どおりに決まる (1.0 未満では破壊的変更も minor)
2. 中身の無い版を出さない (断片が 1 つも増えていなければ中断)
3. 書いた断片がノートから黙って消えない (知らない分類も落とさない)
git は一時リポジトリを本物で回す — 「前回のタグ以降に追加されたか」は履歴の話で、
そこを模造すると検査が確かめた... | mokume-metal/mokume | scripts/tests/release_test.py | .py | 72d616fa3cd70063 | 7.5 | 0 |
#!/usr/bin/env python3
# SPDX-FileCopyrightText: 2026 mokume-metal
# SPDX-License-Identifier: MIT
"""ルールセットのドリフト起票の検査 (#99)。
固定したいのは六つ:
1. ドリフトを検出したら Issue が立ち、本文に差分そのものが載る
(run のログを開かないと何がずれたか分からない起票は、無いのとあまり変わらない)
2. 同じ Issue が open なら二重に立てない — 日次で回る検査なので、放置された
ドリフトが毎日 1 本ずつ Issue を産まない
3. 重複判定はタイトル... | mokume-metal/mokume | scripts/tests/ruleset_drift_test.py | .py | 675e3334d01b2714 | 7.5 | 0 |
#!/usr/bin/env python3
# SPDX-FileCopyrightText: 2026 mokume-metal
# SPDX-License-Identifier: MIT
"""ルールセットの定義ファイルと照合・適用の検査 (#98)。
固定したいのは三つ:
1. 定義ファイルの形の検査が、GET の応答をそのまま置いた事故 (id 等の混入) と
bypass_actors の欠落を落とす — ADR-0003 決定 1 は書かれていて初めて検査できる
2. 実設定に bypass_actors が無い (= その認証では読めない) とき、黙って一致と
言わない — 既定は赤... | mokume-metal/mokume | scripts/tests/rulesets_test.py | .py | acf738ab8642e0f5 | 7.5 | 0 |
#!/usr/bin/env python3
"""Strict conformance adapter for OSIRIS v3 spiral compute.
The frozen parent OSIRIS v2/S𓂸ḥ/2 technical_forward keeps the full verified SAT
assignment inside its engine but intentionally exports only assignment_sha256 in
the public projection. The spiral contract requires a reusable SAT witness... | Hawkar-usls/Janus-Cosmos | workspace/osiris_spiral_runtime_strict.py | .py | f989df1681b8c0e7 | 7 | 0 |
#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""Start the loopback-only ViralX Connector and open the hosted pairing page."""
from __future__ import annotations
import argparse
import json
import socket
import threading
import time
import webbrowser
from urllib.error import URLError
from urllib.parse import parse_q... | chongchonghaoman/ViralX | connector.py | .py | 529bc2b4d160f97b | 7.39 | 5 |
#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""将分析结果导出为 Markdown 到 Obsidian"""
import sys
sys.stdout.reconfigure(encoding='utf-8')
import json
from pathlib import Path
from datetime import datetime
def generate_markdown(analysis_file: Path, keyword: str) -> str:
"""生成详细的 Markdown 报告"""
with open(analysis_fi... | chongchonghaoman/ViralX | export_to_obsidian.py | .py | bbc0b36e202d9917 | 7.39 | 5 |
#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""完整分析流程 + 导出到 Obsidian"""
import sys
sys.stdout.reconfigure(encoding='utf-8')
import json
from pathlib import Path
from datetime import datetime
from tiktok_viral_analyzer import TikTokViralAnalyzer
from ai_analyzer import AIAnalyzer
from report_generator import generate... | chongchonghaoman/ViralX | full_analysis.py | .py | 1efd6276e27dc23a | 7.39 | 5 |
#!/usr/bin/env python3
"""
Ask-the-user bridge — the Discord surface for a multiple-choice question.
Why this exists at all
----------------------
Claude Code's built-in `AskUserQuestion` tool does not exist over this
transport. The agents here are spawned as `claude -p --input-format
stream-json`, and in that mode th... | mcarmody/karakos-package | bin/ask_handler.py | .py | d1f80b573f7bea69 | 7 | 0 |
#!/usr/bin/env python3
"""
Message Capture — Persist Discord messages to daily JSONL logs.
Reads from the agent server's message queue and writes structured JSONL
entries to data/messages/messages-YYYY-MM-DD.jsonl. Called by the relay
after processing each message batch.
Usage:
capture.py --message '{"channel":"g... | mcarmody/karakos-package | bin/capture.py | .py | d4b72d70aad5200a | 7 | 0 |
#!/usr/bin/env python3
"""
Health Monitor — Checks component health and alerts on staleness
"""
import json
import logging
import os
import subprocess
import sys
import time
from pathlib import Path
from datetime import datetime, timedelta, timezone
from logging.handlers import RotatingFileHandler
WORKSPACE_ROOT = Pa... | mcarmody/karakos-package | bin/health-monitor.py | .py | 2b1f6abe8c54e712 | 7 | 0 |
#!/usr/bin/env python3
"""
Memory Maintenance — Episodic consolidation and embedding generation.
Processes recent messages from JSONL files:
1. Reads previous day's messages from JSONL
2. Scores importance (using Claude Haiku for cheap importance scoring)
3. Creates episodes in SQLite episodes table
4. Decays existing... | mcarmody/karakos-package | bin/memory-maintenance.py | .py | 6c1e2c40cd11cd98 | 7 | 0 |
#!/usr/bin/env python3
"""
Data Purging — Clean up old logs and messages.
Purges:
1. Message JSONL files older than MESSAGE_RETENTION_DAYS (default 90)
2. Tool audit records older than TOOL_AUDIT_RETENTION_DAYS (default 30)
3. Session summary archives (keeps last 30 per agent)
4. Agent stream logs older than STREAM_LO... | mcarmody/karakos-package | bin/purge-data.py | .py | 61733edf74b1763f | 7 | 0 |
#!/usr/bin/env python3
"""Register Karakos's Discord application ("slash") commands.
Guild-scoped registration (not global): guild commands appear immediately,
global ones take up to an hour to propagate.
Run automatically by `bin/entrypoint.sh` on every container start, so a
fresh install has working `/` commands wi... | mcarmody/karakos-package | bin/register-discord-commands.py | .py | 423ae53da0e4d90e | 7 | 0 |
#!/usr/bin/env python3
"""
Python-based Scheduler — Replaces cron inside Docker
Runs scheduled tasks with full environment variable access.
Health heartbeat confirms liveness.
"""
import schedule
import subprocess
import sys
import os
import json
import time
import logging
from pathlib import Path
from datetime impor... | mcarmody/karakos-package | bin/scheduler.py | .py | 48f30b084be9020d | 7 | 0 |
#!/usr/bin/env python3
"""
Session Summarizer — Generates session summaries for agent context preservation
Reads recent agent stream logs, calls Claude to generate a summary, validates
required headers, and outputs to checkpoint file for next session re-injection.
"""
import argparse
import json
import os
import sys
... | mcarmody/karakos-package | bin/summarize-session.py | .py | c0a9fab4847d056c | 7 | 0 |
#!/usr/bin/env python3
"""
wedge-check.py — catch an agent that is alive and stuck.
health-monitor.py checks whether components are *running*. This checks
whether a running agent is *moving*, which is a different failure and the one
that actually strands a user: the claude subprocess is up, the agent server's
event lo... | mcarmody/karakos-package | bin/wedge-check.py | .py | 2c2e9a217d7c9676 | 7 | 0 |
#!/usr/bin/env python3
"""
Protected Paths Checker — Verifies staged git files against protection rules.
Called by pre-commit hook. Prints blocked paths and exits non-zero if any
Tier 1 protected files are staged for commit.
Usage:
check-protected-paths.py --staged
"""
import json
import os
import subprocess
imp... | mcarmody/karakos-package | system/check-protected-paths.py | .py | fbc8675a8772621d | 7 | 0 |
#!/usr/bin/env python3
"""
stop-deferred-work.py — Stop hook: catch a promised-but-never-done turn.
An agent can end its turn with "I'll check that shortly" or "On it." and
nothing re-invokes it — no background self, no next beat. The user is left
holding a commitment with nothing behind it. This hook runs a regex cat... | mcarmody/karakos-package | system/hooks/stop-deferred-work.py | .py | 4e24e959877ccb0e | 7 | 0 |
"""
Shared pytest fixtures for Karakos test suite.
"""
import importlib.util
import json
import os
import sqlite3
import sys
import tempfile
from pathlib import Path
import pytest
PACKAGE_ROOT = Path(__file__).parent.parent
def import_script(name: str, file_path: Path = None):
"""Import a Python script by name... | mcarmody/karakos-package | tests/conftest.py | .py | 05bb987d42ad3933 | 7.5 | 0 |
# A simple torch style logger
# (C) Wei YANG 2017
from __future__ import absolute_import
import matplotlib.pyplot as plt
import os
import sys
import numpy as np
__all__ = ['Logger', 'LoggerMonitor', 'savefig']
def savefig(fname, dpi=None):
dpi = 150 if dpi == None else dpi
plt.savefig(fname, dpi=dpi)
def... | yoonsangPark0422/Semi-supervised-learning | baseline compare/ABC-main/ABCcode0920/utils/logger.py | .py | 55b5e593428afa6d | 7 | 0 |
'''Some helper functions for PyTorch, including:
- get_mean_and_std: calculate the mean and std value of dataset.
- msr_init: net parameter initialization.
- progress_bar: progress bar mimic xlua.progress.
'''
import errno
import os
import sys
import time
import math
import torch.nn as nn
import torch.nn.i... | yoonsangPark0422/Semi-supervised-learning | baseline compare/ABC-main/ABCcode0920/utils/misc.py | .py | 014c138f0ea7c49d | 7 | 0 |
# Copyright 2019 Google LLC
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# https://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, ... | yoonsangPark0422/Semi-supervised-learning | baseline compare/crest-master/third_party/augment_ops.py | .py | d38cd3442f152a6d | 7 | 0 |
# Copyright 2019 Google LLC
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# https://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, ... | yoonsangPark0422/Semi-supervised-learning | baseline compare/crest-master/third_party/rand_augment.py | .py | 6ab1d5737ab07ed9 | 7 | 0 |
# Copyright 2021 Google LLC
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# https://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, ... | yoonsangPark0422/Semi-supervised-learning | baseline compare/crest-master/train_and_eval_loop.py | .py | b3932a66f3e18456 | 7 | 0 |
import json
import os
import torch
import torch.nn.functional as F
from yacs.config import CfgNode
from lib.dataset.utils import get_class_counts
from lib.engine import BaseTrainer
from lib.models.losses import Accuracy
from lib.utils import Meters, get_last_n_median
from lib.utils.dist_logger import DistributionLogg... | yoonsangPark0422/Semi-supervised-learning | baseline compare/daso-main/lib/algorithm/base_algorithm.py | .py | 27883d15b9b99002 | 7 | 0 |
import math
import os
import numpy as np
import torch
from scipy import optimize
from lib.dataset.utils import get_data_config, get_imb_num
def get_target_dist(cfg, to_prob=False, device=None):
data_cfg = get_data_config(cfg)
num_l_head = data_cfg.NUM_LABELED_HEAD
num_ul_head = data_cfg.NUM_UNLABELED_H... | yoonsangPark0422/Semi-supervised-learning | baseline compare/daso-main/lib/algorithm/darp_reproduce.py | .py | bbe1a03f62e84273 | 7 | 0 |
"""Process-wide registry of the running :class:`AgentRuntime`.
z4j has two independent install paths that can fire in the same
Python process when a Django app uses Celery:
1. ``z4j_django.apps.Z4JDjangoConfig.ready()`` - runs when Django
loads ``INSTALLED_APPS`` (which happens for *both* the web
process and th... | z4jdev/z4j-bare | src/z4j_bare/_process_singleton.py | .py | 8e8c66b44ca2bd7e | 7 | 0 |
"""Out-of-band agent control: pidfile + SIGHUP-driven reconnect.
The agent normally lives inside another host process (Django,
Flask, FastAPI, a Celery worker, a bare Python script). When an
operator wants to force the agent to drop its current connection
and reconnect immediately - skipping the supervisor's exponenti... | z4jdev/z4j-bare | src/z4j_bare/control.py | .py | 6c1cdf22abfc6db4 | 7 | 0 |
"""Framework-agnostic declarative scheduler reconciler (1.2.2+).
Shared by every framework adapter (z4j-django, z4j-flask, z4j-fastapi).
Each adapter's own ``declarative.py`` is a thin shim that:
1. Reads ``Z4J_SCHEDULES`` (and optional ``CELERY_BEAT_SCHEDULE``)
from the framework's config surface (Django settings... | z4jdev/z4j-bare | src/z4j_bare/declarative.py | .py | 13da1c56b6337f0f | 7 | 0 |
"""Reusable agent diagnostics for ``z4j-doctor`` style commands.
Each framework adapter (z4j-django, z4j-flask, z4j-fastapi) ships a
``z4j-doctor`` entry point that wraps the same probes here. Probes
return structured :class:`ProbeResult` records, never raise: a doctor
command's job is to *report* a failure, not bubbl... | z4jdev/z4j-bare | src/z4j_bare/diagnostics.py | .py | d7ab36c31fe35032 | 7 | 0 |
"""Public ``install_agent()`` entry point for bare Python projects.
This is the function bare-Python projects call from their startup
module (e.g. a Celery app's ``celery.py``, an RQ worker bootstrap,
a Dramatiq broker init module, or any other engine entry point) to
wire up z4j. Framework adapters (Django, Flask, Fas... | z4jdev/z4j-bare | src/z4j_bare/install.py | .py | 08247598e607f60a | 7 | 0 |
"""Best-effort detection of "am I running under a process supervisor?"
Used by the ``restart_worker`` self-exit path to refuse a restart
when no supervisor can be detected - otherwise the agent would
silently kill the worker process with nothing to respawn it.
Detection signals (any one → orchestrated):
- ``Z4J_ORCH... | z4jdev/z4j-bare | src/z4j_bare/orchestrator_detect.py | .py | 5465d7632eaf59ac | 7 | 0 |
"""Exception-safety wrappers.
These helpers isolate z4j failures at synchronous host-call boundaries that
opt into them. Some engine and framework integrations implement an equivalent
local boundary instead, so this module is not a universal wrapper around every
public entry point. ``KeyboardInterrupt`` and ``SystemEx... | z4jdev/z4j-bare | src/z4j_bare/safety.py | .py | 127e378eb73edcf8 | 7 | 0 |
"""Resolve writable directories for the agent's on-disk buffer.
Service deployments often run the agent process under a low-privilege
user (``www-data``, ``nobody``, systemd ``DynamicUser=yes``) whose
``$HOME`` resolves to a directory the process cannot write to -
``/var/www``, ``/nonexistent``, or a transient ``/run/... | z4jdev/z4j-bare | src/z4j_bare/storage.py | .py | 8b52746a4eb87c54 | 7 | 0 |
"""Transport layer - how the agent talks to the brain.
Two implementations, selected by configuration:
- :class:`z4j_bare.transport.websocket.WebSocketTransport` - the
primary transport. A long-lived bi-directional WebSocket to the
brain at ``/ws/agent``. Preferred in production.
- :class:`z4j_bare.transport.long... | z4jdev/z4j-bare | src/z4j_bare/transport/__init__.py | .py | 5820359c1683fe30 | 7 | 0 |
"""Exactly one place in z4j may swallow a ``BaseException``.
``safe_boundary`` is the firewall between z4j and the host application: a bug in
our signal handler or middleware must not crash the user's worker. That is worth
having, and it is worth having exactly once. Every other handler has to re-raise
or forward, bec... | z4jdev/z4j-bare | tests/unit/test_base_exception_firewall.py | .py | 71cd7afe59861c5d | 7.5 | 0 |
"""Tests for the process-wide ``AgentRuntime`` singleton.
The singleton prevents duplicate WebSocket sessions when two
independent install paths fire in the same Python process - the
concrete case being a Django+Celery worker where both
``Z4JDjangoConfig.ready()`` and ``celery.signals.worker_init``
try to install an a... | z4jdev/z4j-bare | tests/unit/test_process_singleton.py | .py | b2cf348e3aa5b5a8 | 7.5 | 0 |
"""Unit tests for ``z4j_bare.safety``.
The safety wrappers are the single most important firewall between
z4j code and the host application. Every test here must pass 100%
of the time.
"""
from __future__ import annotations
import logging
import pytest
from z4j_bare.safety import safe_boundary, safe_call
class Te... | z4jdev/z4j-bare | tests/unit/test_safety.py | .py | 868f6334f7ad8ce6 | 7.5 | 0 |
"""``except*`` runs every matching arm, so arm order cannot decide the class.
The supervisor picked a failure class by assigning inside each ``except*`` arm,
with a comment explaining that the AgentIncompatibleError arm came first
deliberately. That reasoning holds for ``except`` and not for ``except*``: PEP
654 runs ... | z4jdev/z4j-bare | tests/unit/test_supervisor_classifies_by_rank.py | .py | 6e2076dbeb0df15d | 7.5 | 0 |
"""The supervisor must count a failure under the class that classified it.
A disconnect is logged in tiers: the first failure of a streak is loud and
carries the remedy, later ones drop to DEBUG, and every tenth re-surfaces as an
"still trying" INFO. The tier is chosen from the failure's position in its
streak, so the... | z4jdev/z4j-bare | tests/unit/test_supervisor_failure_streaks.py | .py | 698f85c1e8a1bc42 | 7.5 | 0 |
"""Deterministic repro: silent event_batch loss across a WS reconnect.
Safety property under test (the delivery contract from runtime.py's
own comments): an ``event_batch`` buffer entry may only be deleted
after ITS frame id was acked by the brain in the CURRENT session,
or after the entry was re-sent and acked later.... | z4jdev/z4j-bare | tests/unit/test_ws_ack_loss_repro.py | .py | 7f93aac1e111ade1 | 7.5 | 0 |
"""Shared backend contracts (design doc 17.1).
These are the data shapes that cross the MemoryBackend seam. ``ExecutionRecord``
is the canonical one defined in ``tehm/canonical/capture.py`` (single source of
truth); the rest are backend-neutral.
Provenance rule for every candidate/proposal: ``source`` is always one o... | Zhang-D-Y/Typed-Executable-Hardware-Memory | memory/contracts.py | .py | b2d13e3cf8e4d15b | 7 | 0 |
"""Resolve the canonical TEHM freeze without host-specific paths."""
from __future__ import annotations
import json
import os
from pathlib import Path
POINTER_PATH = Path(__file__).resolve().parent / "canonical_freeze_pointer_v1.json"
def load_pointer(path: Path | None = None) -> dict:
path = Path(path or POIN... | Zhang-D-Y/Typed-Executable-Hardware-Memory | memory/evaluation/freeze_pointer.py | .py | ff0fafa4ed9f6b1c | 7 | 0 |
"""MemoryBackend factory (design doc 17.2-17.4).
Backend selection is locked at process start via ``R2G_MEMORY_BACKEND``
(``none`` | ``legacy`` | ``tehm``, default ``legacy``). Switching mid-run is
forbidden; an invalid name is a hard error, never a silent fallback (H12).
"""
from __future__ import annotations
import... | Zhang-D-Y/Typed-Executable-Hardware-Memory | memory/factory.py | .py | 4a911cdd4a774405 | 7 | 0 |
"""MemoryBackend interface (design doc 17.1).
The single seam every runtime memory touchpoint goes through. The three
implementations (none / legacy / tehm) are mutually isolated: TEHM never reads
legacy authority and legacy never reads ``tehm.sqlite`` (H5/H8).
"""
from __future__ import annotations
from typing impor... | Zhang-D-Y/Typed-Executable-Hardware-Memory | memory/interface.py | .py | 686afb568e3d94ac | 7 | 0 |
"""``none`` backend: no-memory baseline (design doc 17.2).
Records execution evidence (as a receipt) but never retrieves historical
experience. Used as the M0 / no-memory experimental arm.
"""
from __future__ import annotations
from contracts import (
ActivationProposal,
ActivationResult,
BuildReport,
... | Zhang-D-Y/Typed-Executable-Hardware-Memory | memory/none_backend.py | .py | dde98280deb7ea4d | 7 | 0 |
#!/usr/bin/env python3
"""Build the second held-out evaluation snapshot after the training expansion.
The physical training database is the post-WP2 ``memory/tehm.sqlite``. The
RTL training evidence is copied from the first freeze because the current
working database is physical-only; only RTL rows/rules are imported... | Zhang-D-Y/Typed-Executable-Hardware-Memory | memory/scripts/build_second_heldout_freeze.py | .py | 5d7b41cf028546bd | 7 | 0 |
#!/usr/bin/env python3
"""E2E tests for CLI flags using kimi model."""
from __future__ import annotations
import json
import os
import subprocess
import sys
from pathlib import Path
# Colors for output
GREEN = "\033[92m"
RED = "\033[91m"
RESET = "\033[0m"
BOLD = "\033[1m"
def _env() -> dict[str, str]:
"""Retur... | AkshatTilak/codeless | scripts/test_cli_flags.py | .py | 3e95f52f7e17bcca | 7.74 | 2 |
#!/usr/bin/env python3
"""E2E tests for Harness features: retry, skills, parallel tools, path permissions.
Uses kimi model for real API calls.
"""
from __future__ import annotations
import asyncio
import os
import subprocess
import sys
from pathlib import Path
sys.path.insert(0, str(Path(__file__).resolve().parents... | AkshatTilak/codeless | scripts/test_harness_features.py | .py | 372f47348eddd4de | 7.74 | 2 |
"""Git Checkpoints and Rollback Engine for Codeless & ABB workspaces."""
from __future__ import annotations
import datetime
import json
import shutil
import subprocess
import uuid
from dataclasses import asdict, dataclass
from pathlib import Path
from codeless.abb.shadow import get_project_storage_dir, resolve_abb_w... | AkshatTilak/codeless | src/codeless/abb/checkpoints.py | .py | dd68e1ec3aee75e9 | 7.24 | 2 |
"""ABB Lifecycle Hook Bridge for intercepting tool execution."""
from __future__ import annotations
from pathlib import Path
from typing import Any
from codeless.abb.hooks.dag_guard import check_dag_dependencies
from codeless.abb.hooks.frontmatter import parse_frontmatter, validate_task_frontmatter
from codeless.abb... | AkshatTilak/codeless | src/codeless/abb/hooks/bridge.py | .py | 4a79c93e6c92d660 | 7.24 | 2 |
"""DAG Dependency Gating hook for ABB task transitions."""
from __future__ import annotations
from pathlib import Path
from typing import Any
from codeless.abb.hooks.frontmatter import parse_frontmatter
def index_tasks(tasks_dir: Path) -> dict[str, tuple[Path, dict[str, Any]]]:
"""
Index all task markdown ... | AkshatTilak/codeless | src/codeless/abb/hooks/dag_guard.py | .py | 9fc45caab55c1232 | 7.24 | 2 |
"""YAML frontmatter parser and strict schema validator for ABB task files."""
from __future__ import annotations
import re
from pathlib import Path
from typing import Any
import yaml
SEMVER_REGEX = re.compile(r"^\d+\.\d+\.\d+(-[a-zA-Z0-9.]+)?$")
VALID_STATUSES = frozenset(
{
"not_started",
"pend... | AkshatTilak/codeless | src/codeless/abb/hooks/frontmatter.py | .py | e28e790916e71867 | 7.24 | 2 |
"""4-Mode Permission Controller and Persona Composition Engine for Codeless."""
from __future__ import annotations
from enum import Enum
from pathlib import Path
from codeless.abb.shadow import resolve_abb_workspace
from codeless.abb.virtualization import find_project_root, is_abb_path
from codeless.permissions.mode... | AkshatTilak/codeless | src/codeless/abb/permissions.py | .py | 69f6fe31bcadde70 | 7.24 | 2 |
"""Shadow workspace bootstrapper, location manager, and path resolver for Codeless."""
from __future__ import annotations
import datetime
import hashlib
import json
import os
import shutil
from pathlib import Path
from typing import Any, Literal
AbbLocation = Literal["shadow", "local"]
def get_project_hash(project... | AkshatTilak/codeless | src/codeless/abb/shadow.py | .py | 64038b307b712b85 | 7.24 | 2 |
"""Two-Track Verification Runner and Manifest Engine for Codeless."""
from __future__ import annotations
import asyncio
import datetime
import subprocess
import time
from dataclasses import dataclass, field
from pathlib import Path
from typing import Any
from codeless.abb.hooks.frontmatter import parse_frontmatter
f... | AkshatTilak/codeless | src/codeless/abb/verification.py | .py | ae0b06355673ee3d | 7.24 | 2 |
"""Shadow path virtualization engine for Codeless.
Transparently maps Agent Buildable Base (ABB) domains (tasks, references,
workflows, design, features, skills, etc.) to the project's active shadow
or local workspace while mapping codebase source and tests to the repo root.
"""
from __future__ import annotations
fr... | AkshatTilak/codeless | src/codeless/abb/virtualization.py | .py | 94a38b72ca99e187 | 7.24 | 2 |
"""Anthropic API client wrapper with retry logic."""
from __future__ import annotations
import asyncio
import json
import logging
import uuid
from dataclasses import dataclass, field
from typing import Any, AsyncIterator, Callable, Protocol
from anthropic import APIError, APIStatusError, AsyncAnthropic
from codeles... | AkshatTilak/codeless | src/codeless/api/client.py | .py | bba6cf80444682d6 | 7.24 | 2 |
# Copyright 2026 Alibaba Cloud (Qwen3-ASR)
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed t... | Karin-java/qwen3-aligner-refactor | qwen3_aligner/srt_utils.py | .py | 62928ac8c6a234ec | 7 | 0 |
# Copyright 2026 Alibaba Cloud (Qwen3-ASR)
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed t... | Karin-java/qwen3-aligner-refactor | qwen3_aligner/text_utils.py | .py | bf797f17747e336e | 7 | 0 |
"""Semantic actions and the action engine.
An *action* is a named unit of work against a device (inventory, run a set of
commands / on-demand commands, collect, ...), implemented as a class and
registered in ``ActionRegistry``. ``run_action`` resolves a vendor's command
catalog and executes the action, saving outputs ... | liliangnike/fncollect | src/fncollect/actions.py | .py | 07df1f305922c09b | 7 | 0 |
"""ONT cutthrough.
An ONT's management plane is not directly reachable: it is reached *through*
the OLT. Before any ONT session can open, the OLT side must be provisioned
(prepare a debug/vlan path, push the workstation client IP, register the
ONT serial, etc.) and often restored afterwards.
This module models that a... | liliangnike/fncollect | src/fncollect/cutthrough.py | .py | 95682915294b5810 | 7 | 0 |
"""Per-device single-task lock (cross-platform).
For a Fixed Network device, only one log-collect task may run at a time.
Each fncollect process is separate, so we use an advisory OS file lock keyed
by the device identifier (e.g. its IP). Another process targeting the same
device fails fast with ``DeviceLockedError``.... | liliangnike/fncollect | src/fncollect/device_lock.py | .py | fada41defc040007 | 7 | 0 |
"""Hardware-type auto-discovery.
Mirrors the ngalexx approach of dispatching to a concrete hardware class by
import path, so that adding a new hardware board is a single-file change with
no core edits. Given a vendor package import path and a hardware type string,
this resolves and imports the matching class.
Convent... | liliangnike/fncollect | src/fncollect/discovery.py | .py | 229483b2c90e5227 | 7 | 0 |
"""Real interactive SSH session.
Nokia ISAM-style devices do not accept a plain ``exec`` request: they need
an interactive PTY shell, prompt detection, paging (``--more--``) handling
and support for the context-based CLI navigation. This module provides a
session that meets those needs.
It also works around the SSH `... | liliangnike/fncollect | src/fncollect/net.py | .py | 84be8e100efe811c | 7 | 0 |
"""ONT devices, specialised by management SoC/chipset.
An ONT's management plane runs on vendor-specific silicon (MediaTek, Broadcom,
Realtek, ...). Different chipsets speak different CLI dialects, so each is a
distinct device class. ``ont_device_for`` dispatches an ONT to the class for
its chipset (falling back to a ... | liliangnike/fncollect | src/fncollect/ont.py | .py | c9827e92897be96b | 7 | 0 |
"""Reusable value processors.
Turns raw device output into exact values. Each processor is a *named*
transform referenced from YAML (``extract`` / ``probe.mappings``), so "get
value" steps stay declarative while the parsing logic lives once in code.
Supported processors:
* ``regex`` -- capture one/many groups via a... | liliangnike/fncollect | src/fncollect/processors.py | .py | 6fffc3ad1503390d | 7 | 0 |
"""Progress bars for stages/steps/procedures/command executions.
Built on ``tqdm`` and styled to look like a Linux package-installation
progress bar: a fixed square-bracket bar with a percentage and a description
of the current unit (stage -> procedure -> step -> command).
Nested levels are rendered as stacked tqdm b... | liliangnike/fncollect | src/fncollect/progress.py | .py | 2eeb8c9591771d5a | 7 | 0 |
"""Abstract contracts for fncollect.
A *vendor* is a collection of device types that share a command dialect,
login behaviour and transport options. A *device* is a concrete box the tool
collects from, composed of a *session* (how to talk) and a *role* / hardware
type (what it is, e.g. OLT/NT/LT/ONT). An *action* is o... | liliangnike/fncollect | src/fncollect/vendor.py | .py | 6aeef0a1647220bd | 7 | 0 |
"""Nokia ISAM (7360/7362) vendor pack.
Uses the real interactive SSH session with context-based CLI navigation.
Bare-bones but functional; extend the command catalog in
``config/vendors/isam/vendor.yml`` and add hardware models under
``vendors/isam/hardware/``.
"""
from __future__ import annotations
from fncollect.n... | liliangnike/fncollect | src/fncollect/vendors/isam/__init__.py | .py | cbcc2e655ebfc256 | 7 | 0 |
"""Tests for the DCP-driven ONT cutthrough provider."""
import pytest
from fncollect.config import LoggingConfig, RunConfig
from fncollect.cutthrough import DcpCutthroughProvider, OntTarget
from fncollect.dcp import DcpDefinition, DcpStep, parse_dcp
from fncollect.logging_setup import build_logger
from fncollect.ses... | liliangnike/fncollect | tests/test_dcp_cutthrough.py | .py | 556997fab7471491 | 7.5 | 0 |
"""Tests for the per-device single-task lock."""
import pytest
from fncollect.device_lock import DeviceLock, DeviceLockedError
def test_lock_prevents_concurrent(tmp_path):
a = DeviceLock(tmp_path, "10.10.10.10")
b = DeviceLock(tmp_path, "10.10.10.10")
with a, pytest.raises(DeviceLockedError):
b.... | liliangnike/fncollect | tests/test_device_lock.py | .py | 7d2b3d15081d20ac | 7.5 | 0 |
"""Tests for the non-programmer UX: auto-regex + procedure catalog."""
from fncollect.processors import auto_regex, extract_values
from fncollect.vendors.isam import IsamVendor
def test_auto_regex_extracts_version():
line = " isam-feature-group : R6.6.02g"
pattern = auto_regex(line, "6.6.02")
assert p... | liliangnike/fncollect | tests/test_procedures.py | .py | 0e50735ea15fd79c | 7.5 | 0 |
"""Self-contained HTML5 player for the fill-front frame sequence.
``export_frames`` already writes one PNG per frame, so the scrubber costs no
extra rendering: the PNGs are inlined as data URIs and driven by a small
vanilla-JS player. Everything (play/pause, seek, speed) happens in the
browser, so dragging the scrubbe... | shostako/mold-flow-fangate | build/lib/core/fill_player.py | .py | e3b3dc6efa939651 | 7 | 0 |
"""Cavity geometry definition.
The simulation domain is a 2D structured grid where each cell is either
inside the cavity (mask=True) or outside (mask=False). Each in-cavity
cell carries a thickness h [mm] (gap between mold halves). Gates are
point-like Dirichlet boundaries at tau=0.
This module provides:
- Geometry: ... | shostako/mold-flow-fangate | build/lib/core/geometry.py | .py | 4055eb99495268fa | 7 | 0 |
"""Material database and Cross-WLF viscosity model."""
from __future__ import annotations
import json
from collections.abc import Iterable
from dataclasses import dataclass
from pathlib import Path
import numpy as np
@dataclass(frozen=True)
class Material:
key: str
name: str
n: float
tau_star: floa... | shostako/mold-flow-fangate | build/lib/core/materials.py | .py | 21c6ab7d3cbef7f7 | 7 | 0 |
"""Record of the inputs that produced a run, for the results ZIP.
``metadata.json`` holds what the solver *computed* (volumes, tau, iteration
counts). It never held what the user *entered*, so a downloaded ZIP could not
be traced back to the settings that made it -- the geometry had to be
reverse-engineered from the i... | shostako/mold-flow-fangate | build/lib/core/settings_record.py | .py | 37b97c6af806319e | 7 | 0 |
"""Single source of truth for the application version.
``__version__`` is the release version (kept in sync with
``pyproject.toml`` and ``CHANGELOG.md``). ``build_label()`` additionally
reports the git commit the running code came from, which is what tells
you whether a deployed instance is actually up to date.
To cu... | shostako/mold-flow-fangate | build/lib/core/version.py | .py | 78aa1fc8eb704eaa | 7 | 0 |
"""Cavity geometry definition.
The simulation domain is a 2D structured grid where each cell is either
inside the cavity (mask=True) or outside (mask=False). Each in-cavity
cell carries a thickness h [mm] (gap between mold halves). Gates are
point-like Dirichlet boundaries at tau=0.
This module provides:
- Geometry: ... | shostako/mold-flow-fangate | core/geometry.py | .py | 4338f785e62ae4d1 | 7 | 0 |
"""Single source of truth for the application version.
``__version__`` is the release version (kept in sync with
``pyproject.toml`` and ``CHANGELOG.md``). ``build_label()`` additionally
reports the git commit the running code came from, which is what tells
you whether a deployed instance is actually up to date.
To cu... | shostako/mold-flow-fangate | core/version.py | .py | 09fab4305ea2d52c | 7 | 0 |
"""Color helpers shared by the thickness-colormap tests.
These live in one place on purpose. The luminance formula was originally
written out by hand in two test files, and *both copies were wrong the same
way* — reason enough to stop having two copies. See :func:`relative_luminance`.
"""
from __future__ import annot... | shostako/mold-flow-fangate | tests/colorimetry.py | .py | c4b28ca9bee5f251 | 7.5 | 0 |
"""AppTest wiring for the fan-gate geometry sidebar in ``app.py``."""
from __future__ import annotations
import dataclasses
from core import FanGatePlateConfig
from tests.ui_helpers import app as _app
from tests.ui_helpers import texts as _texts
def test_the_defaults_are_the_spec_mold_and_are_recorded_in_full():
... | shostako/mold-flow-fangate | tests/test_fan_gate_ui.py | .py | b4d3ecd73ed35663 | 7.5 | 0 |
"""Tests for the run-settings record bundled with the results ZIP.
The point of this record is that a downloaded ZIP can be reproduced. Before
it existed, ``metadata.json`` held only what the solver computed, so the
geometry had to be recovered by measuring the rendered images and searching
for a config whose volume a... | shostako/mold-flow-fangate | tests/test_settings_record.py | .py | c8b569ca308d3f33 | 7.5 | 0 |
"""Smoke tests: import-level sanity checks for the core package.
These tests do not solve the Hele-Shaw system; they only verify that
the package imports cleanly and the bundled material data is loadable.
"""
from __future__ import annotations
def test_core_imports() -> None:
from core import ( # noqa: F401
... | shostako/mold-flow-fangate | tests/test_smoke.py | .py | 9ba5455b021ec8b3 | 7.5 | 0 |
"""Analytical-solution verification: 1D strip, full-edge gate.
Setup
-----
Uniform strip with width Ny cells, length Nx cells, uniform thickness h,
uniform effective viscosity. Dirichlet τ=0 imposed on the entire left edge
(x=0), no-flux on the other three edges. Reduces to a strict 1D problem
in x.
The discrete oper... | shostako/mold-flow-fangate | tests/test_solver_1d.py | .py | 4d08ef9ee5d94f86 | 7.5 | 0 |
"""AppTest wiring checks for the two-phase short-shot UI in ``app.py``."""
from __future__ import annotations
import io
import zipfile
from pathlib import Path
from streamlit.testing.v1 import AppTest
from tests.ui_helpers import app as _app
from tests.ui_helpers import texts as _texts
def test_the_defaults_are_t... | shostako/mold-flow-fangate | tests/test_two_phase_ui.py | .py | 9957329624984ed8 | 7.5 | 0 |
"""Smoke tests for the Plotly-based 3D visualizer.
These tests do not validate visual output (Plotly figure JSON is verbose
and brittle). They confirm:
- the 3D renderers run end-to-end on a small Hele-Shaw result,
- they return :class:`plotly.graph_objects.Figure` instances,
- the figure has the expected three-... | shostako/mold-flow-fangate | tests/test_visualizer_3d.py | .py | 2ebe415951bd6615 | 7.5 | 0 |
"""Smoke tests for the multilayer 2D renderers (PR-D)."""
from __future__ import annotations
import matplotlib.colors as mcolors
import matplotlib.pyplot as plt
import pytest
from core import (
MaterialDB,
MultilayerHeleShawSolver,
build_demo_geometry,
)
from core.visualizer import (
THICKNESS_CMAP,
... | shostako/mold-flow-fangate | tests/test_visualizer_layer.py | .py | c4ae2cf7cd9b185a | 7.5 | 0 |
"""Weld-line detector: confluence of flow directions, not a local maximum.
The old heuristic ("6 of 8 neighbours have a smaller tau") was a near-peak
test. A weld that keeps flowing -- the line behind a hole -- always has a
later downstream row, so at most 5 neighbours qualified and nothing was
drawn. The detector now... | shostako/mold-flow-fangate | tests/test_weld_detection.py | .py | 0348edb5ba8a443e | 7.5 | 0 |
"""Shared AppTest helpers for the ``app.py`` wiring tests."""
from __future__ import annotations
from pathlib import Path
from streamlit.testing.v1 import AppTest
APP = Path(__file__).resolve().parent.parent / "app.py"
def app(timeout: float = 240.0, *, fast: bool = True) -> AppTest:
"""Load the app; ``fast``... | shostako/mold-flow-fangate | tests/ui_helpers.py | .py | 6c64a4c216b17d37 | 7.5 | 0 |
"""Language packs: per-language data + hooks, registered centrally.
Every language's vocabulary lives in one pack — units, months, symbol words,
number-speaker hooks. Adding a language means adding one pack here and one
normalizer module; the pipeline reads packs through ``get_pack`` and never
branches on language cod... | Revolab-Sdn-Bhd/revo-norm | revo_norm/langpack.py | .py | 43d7b81f9967d726 | 7 | 0 |
import re
from revo_norm.currency_utils import CURRENCY_K_SUFFIX_PATTERN, expand_currency_k_suffix
from revo_norm.num2word_ms import to_cardinal as num2word
numbers_mapping_malay = {
"0": "kosong",
"1": "satu",
"2": "dua",
"3": "tiga",
"4": "empat",
"5": "lima",
"6": "enam",
"7": "tuju... | Revolab-Sdn-Bhd/revo-norm | revo_norm/normalizer_ms.py | .py | 873ef516822ef072 | 7 | 0 |
"""
Chinese number-to-words conversion for TTS.
Supports cardinal numbers from 0 to 10^16 (兆).
Handles special Chinese number conventions:
- 十 (10) standalone, 一十 (10 embedded in compound)
- 一百零一 (101)
- 一万零一 (10001) — 零 between groups with skipped digits
"""
_DIGITS = ["零", "一", "二", "三", "四", "五", "六", "七", "八", "九... | Revolab-Sdn-Bhd/revo-norm | revo_norm/num2word_zh.py | .py | 96432f27cc8d696a | 7 | 0 |
"""
TTS-specific utility functions for the ChatterBox backend.
This module contains features specific to TTS processing that should NOT
be in the general text normalization library (revo-norm).
Features:
- Sentence splitting and chunking for TTS
- Random comma insertion for pauses
- Sound word removal (laughter, appl... | Revolab-Sdn-Bhd/revo-norm | revo_norm/tts_utils.py | .py | e89e2f32f471bcc3 | 7 | 0 |
"""
Scrapes a headline from The Daily Pennsylvanian website and saves it to a
JSON file that tracks headlines over time.
"""
import os
import sys
import daily_event_monitor
import bs4
import requests
import loguru
def scrape_data_point():
"""
Scrapes the top article headline from The Daily Pennsylvanian O... | ang-won/daily-pennsylvanian-basic-scraper | script.py | .py | de021cfd9c0fa358 | 7 | 0 |
"""현행 법령 부칙에서 적용례·경과조치 문장만 고른다."""
from __future__ import annotations
import re
from utils.parsing import json_list, json_text
_TRANSITION_RE = re.compile(
r"적용례|경과조치|종전의\s*규정|시행\s*전에?|시행\s*당시|"
r"행위에\s*대하여|예에\s*따른다|불구하고.{0,20}적용"
)
def _flatten_addendum(value: object) -> list[str]:
if value is None:
... | hgkang17/law_info_search | llm/addendum.py | .py | abd1f5cbafb45d76 | 7 | 0 |
"""AI CLI 설치 여부 확인과 npm 전역 설치.
화면에서는 이 함수를 작업 스레드에서 부른다. npm 설치는 네트워크와 디스크를
쓰는 느린 작업이므로 GUI 스레드에서 직접 실행하면 안 된다.
"""
from __future__ import annotations
import json
import shutil
import subprocess
from collections.abc import Callable
from dataclasses import dataclass
from .desktop_mcp import clean_child_environment, ... | hgkang17/law_info_search | llm/ai_cli_setup.py | .py | 0b6d3cc12884c9a4 | 7 | 0 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.