SJP.PL

AGENTS.md

@@ -1,74 +0,0 @@
-# AGENTS.md
-
-Guidance for agents and contributors adding a source and cutting a release.
-Read this before opening a PR — several recent contributions added data but
-skipped the doc/version bump, forcing a maintainer cleanup release.
-
-## How the docs are actually maintained
-
-`src/make_docs.py` is a **scaffolding/assembly helper, not a round-trip source
-of truth.** It regenerates the README source table + release totals, `LICENSE`,
-a `CHANGELOG` stub, and *template* datasheets — but it does **not** reproduce the
-committed docs on its own:
-
-- It iterates `SOURCES` in `src/sources.py`; any source in `data/` but missing
-  from `SOURCES` is silently dropped from the table and totals.
-- It stamps every regenerated datasheet's "Added" with the single global `ADDED`
-  constant, overwriting real per-source add dates.
-- It cannot reproduce hand-written README narrative (audit notes, policy notes,
-  the phrase-frequency section).
-
-So datasheets and the README narrative are **hand/contributor-maintained**.
-Never run `make_docs.py` and commit the result blind — `git diff` first and
-restore anything it dropped or restamped.
-
-## Add a source
-
-1. Produce the artifact under `data/<key>/`:
-   - `<key>.parquet` — built by `src/build_dynaword.py` (LFS-tracked automatically).
-   - `<key>.stats.json` — doc/token/char counts (drives all totals).
-   - `<key>.md` — datasheet (see below).
-2. Add the ingestion script: `src/fetch_<key>.py` or `src/clean_<key>.py`, so the
-   source is reproducible. `build_source` (build_dynaword.py) expects a
-   `<file_key|speakleash_key>.jsonl.zst` intermediate.
-3. Register the source in `src/sources.py` (`SOURCES`) — required, or make_docs
-   never sees it. Copy an existing entry's shape: `pretty`, `license`,
-   `license_spdx`, `traceable`, `upstream`, `domain`, `created`, `is_ocr`, and the
-   `speakleash_key`/`file_key`. If the datasheet is hand-authored (rich provenance
-   or a fixed add date), add `"custom_datasheet": True` so make_docs leaves it
-   alone.
-4. Add a contract test: `src/test_<key>_contract.py` (canonical schema, non-empty
-   text, positive token counts, uniform source/license, stats-file consistency).
-   Run `python3 -m pytest src/` before committing.
-
-## Cut the release (the part that gets skipped)
-
-1. In `src/make_docs.py`: bump `VERSION` and `RELEASE_DATE`. Add a row for the new
-   version to the version table and a row for yourself to the Contributors table
-   (both are literal rows in the template).
-2. Regenerate the phrase-frequency report and charts (registry-independent — it
-   globs `data/*/*.parquet`):
-   ```bash
-   python3 src/pattern_frequency_report.py
-   ```
-   Writes `artifacts/pattern_frequency_hf_snippet.md` + 10 PNGs. Splice the tables
-   and chart embeds into the README "Results" section by hand.
-3. Update `README.md` by hand: header totals, version table, source table row,
-   Contributors row. The document/token totals must equal the sum of the
-   per-source `stats.json` files.
-4. Prepend the release notes to `CHANGELOG.md` under a new `## vX.Y.Z (date)`
-   heading. History is append-only — never rewrite earlier releases.
-
-## Do not commit
-
-`*.log` (fetch/build logs), `.DS_Store`, `src/__pycache__/`. Only `logs/` is in
-`.gitignore` — root-level logs and OS cruft are not, so check `git status`
-before staging. `*.parquet` is LFS-tracked; commit the pointer, not the blob.
-
-## Known drift / cleanup opportunities
-
-- `biblioteka_nauki`, `europeana`, `parlamint_pl` are in `SOURCES` but have no
-  `data/` shard (make_docs skips them with `! no stats`). Intentional placeholders
-  or stale — confirm before relying on `build_dynaword.py --all`.
-- The global `ADDED` datasheet stamp is a footgun: give datasheets a per-source
-  add date (or `custom_datasheet`) before making make_docs regenerate them.

CHANGELOG.md

@@ -1,44 +1,5 @@
 # Changelog
 
-## v0.2.3 (2026-07-26)
-
-- Added the community contribution `european_hplt_v3_pl` by Arkadiusz Słota
-  (`Maggio33`): HPLT v3.0 Polish WDS quality bins 10+9, 110,590 documents and
-  512,383,778 cl100k-proxy tokens under CC0-1.0 packaging.
-- Added the reviewed `global_voices` contribution by Dawid Majewski
-  (`dawidmajewski`): 2,040 documents / 3,715,020 tokens with CC-BY-3.0 and
-  per-document author attribution metadata.
-- Added `nkjp1m` from contributor `1am`: 18,381 documents / 2,601,201 tokens
-  from the manually annotated 1-million-word NKJP subcorpus under the upstream
-  version-unspecified CC-BY statement.
-- Revalidated the contributed parquet against the canonical schema, non-empty
-  text, positive token counts, uniform source/license metadata, and stats-file
-  consistency.
-- Updated the complete 16-source release totals from per-source parquet stats:
-  2,710,974 documents and 6,881,277,362 cl100k-proxy tokens.
-- Recomputed the full-corpus phrase-frequency report and charts. Updated the
-  Arrow reporting path to use 64-bit string offsets for large row groups.
-- Fixed `src/test_hplt_v3_contract.py` to resolve its fixture relative to the
-  repository instead of a contributor-local Windows path.
-- Added public contributor credit for Arkadiusz Słota (`Maggio33`) and Paweł
-  Puzio (`ppuzio`, author of the `govpl` contribution released in v0.2.2), plus
-  Dawid Majewski (`dawidmajewski`) for Global Voices Polish.
-- HPLT caveat: CC0 applies to the HPLT packaging; underlying web documents can
-  carry independent rights. Downstream near-dedup and web-content review remain
-  recommended before training.
-
-## v0.2.2 (2026-07-19)
-
-- Added `govpl` — gov.pl Polish government press releases (CC-BY-SA-4.0),
-  88,190 documents / 80.7M tokens, fetched directly from gov.pl across 133
-  ministry/agency subsites. New scripts: `src/discover_govpl.py` (discovers each
-  subsite's news-section path into the committed `src/govpl_subsites.json`
-  manifest) and `src/fetch_govpl.py` (pure-stdlib crawl of that manifest).
-- Stable totals after adding govpl: 13 sources, 2,579,963 docs, 6.36B tokens
-  (tiktoken proxy).
-- Article date range 2005-09-26 → 2026-07-19. Build dropped 24 non-Polish and
-  deduped 1,926 cross-posted articles (exact sha1).
-
 ## v0.3.1-preview (2026-07-10)
 
 - Added a separate 810-document Europeana direct-ingestion preview parquet at

README.md

@@ -19,89 +19,29 @@ A continuously developed, **openly-licensed**, human-text Polish corpus — a Po
 edition in the [Dynaword](https://huggingface.co/datasets/danish-foundation-models/danish-dynaword)
 family (Enevoldsen et al., [arXiv:2508.02271](https://arxiv.org/abs/2508.02271)).
 
-> **v0.2.3 stable** · 2,710,974 documents · **6.88B tokens**
-> (6,881,277,362 tiktoken cl100k-proxy tokens) · 16 sources
-> Updated: **2026-07-26**
+> **v0.2.1 stable** · 2,491,773 documents · **6.28B tokens**
+> (tiktoken proxy; canonical Llama-3 count at release) · 12 sources
+> Updated: **2026-07-02**
 
-> **v0.3-dev experimental track** · quality/diversity workflow, source-gate
-> validation and candidate-data audits. This is development work, not a released
-> corpus version, and it does not replace the v0.2.3 stable parquets.
+> **v0.3.0-preview validation release** · live source-gate validation for
+> SpeakLeash shards and direct Europeana candidate data, plus the quality/diversity
+> remix workflow. This is candidate-only metadata and audit work: it does not
+> replace the v0.2.1 stable parquets.
 
-> **Europeana validation artifact (2026-07-10)** · first reproducible
-> direct-ingestion shard:
+> **v0.3.1-preview** · first reproducible Europeana direct-ingestion shard:
 > 1,000 API records → 810 quality-gated parquet documents, each preserving
 > per-record rights and creator metadata. Stored separately under
-> the legacy path `previews/v0.3.1/europeana.parquet`; it is not a release of the
-> full dataset and does not alter stable totals.
+> `previews/v0.3.1/europeana.parquet`; the paged Search API run took 59.4 s and
+> does not alter v0.2.1 stable totals.
 
-## Releases
-
-### Stable releases
+## Versions
 
 | version | status | documents | tokens | notes |
 |---|---|---:|---:|---|
-| `v0.2.3` | stable release | 2,710,974 | 6.88B | Adds community-contributed HPLT v3 Polish WDS 10+9, Global Voices Polish and the NKJP manually annotated 1M-word subcorpus. |
-| `v0.2.2` | previous stable | 2,579,963 | 6.36B | Added community-contributed `govpl`: 88,190 docs / 80.7M tokens fetched directly across 133 ministry/agency subsites. |
-| `v0.2.1` | previous stable | 2,491,773 | 6.28B | 12-source stable corpus with `license` and `author` metadata columns; adds `1000_novels`. |
+| `v0.2.1` | stable release | 2,491,773 | 6.28B | 12-source stable corpus with `license` and `author` metadata columns; adds `1000_novels`. |
 | `v0.2.0` | previous stable | 2,490,773 | 6.22B | Provenance-first corpus from 11 open/official sources. |
-
-![Stable dataset size by version](artifacts/dataset_size_by_version.png)
-
-### Development and validation artifacts
-
-| name | type | scope | notes |
-|---|---|---|---|
-| `v0.3-dev` | experimental development track | workflow and candidate audits | Quality remix, legal-source caps, source QA, deduplication and direct-upstream ingestion. Not a corpus release. |
-| `Europeana validation artifact 2026-07-10` | reproducible sample | 810 documents / 140,042 tokens | Stored at the legacy path `previews/v0.3.1/europeana.parquet`; preserves per-record rights and creator metadata. Not a corpus release. |
-
-### Version details
-
-#### v0.2.3 — current stable
-
-Community expansion release. Adds `european_hplt_v3_pl` from
-[PR #5](https://huggingface.co/datasets/SlayerLab/polish-dynaword/discussions/5)
-and `global_voices` from
-[PR #7](https://huggingface.co/datasets/SlayerLab/polish-dynaword/discussions/7),
-plus `nkjp1m` from
-[PR #8](https://huggingface.co/datasets/SlayerLab/polish-dynaword/discussions/8).
-The complete 16-source release contains **2,710,974 documents /
-6,881,277,362 cl100k-proxy tokens**, summed from the released per-source parquet
-statistics.
-
-#### v0.2.2 — previous stable
-
-Added `govpl` from
-[PR #9](https://huggingface.co/datasets/SlayerLab/polish-dynaword/discussions/9):
-88,190 Polish government press releases collected directly from 133 gov.pl
-ministry and agency subsites. Result: 13 sources, 2,579,963 documents and
-6.36B proxy tokens.
-
-#### v0.2.1 — previous stable
-
-Introduced the canonical eight-column release schema:
-`id, text, source, added, created, token_count, license, author`. Added the
-CC-BY-4.0 `1000_novels` source from
-[PR #1](https://huggingface.co/datasets/SlayerLab/polish-dynaword/discussions/1)
-and recomputed all release statistics from parquet files.
-
-#### v0.2.0 — initial stable corpus
-
-The provenance-first baseline: 11 reviewed open or official sources,
-2,490,773 documents and 6.22B proxy tokens. This version established the
-per-source datasheet, traceable-license and exclusion-policy workflow.
-
-#### v0.3-dev — experimental quality workflow
-
-Candidate-only work on a better-balanced training mixture: legal-source caps,
-temperature sampling, source QA, deduplication and direct-upstream ingestion.
-It is a workflow and validation track, not a replacement for the stable corpus.
-
-#### Europeana validation artifact — 2026-07-10
-
-A separate reproducible 810-document Europeana sample preserving per-record
-rights and creator metadata. It validates the direct-ingestion contract without
-changing stable-release totals. `v0.3.1-preview` remains only as the legacy
-storage-path label.
+| `v0.3.0-preview` | validation release, candidate-only | candidate-only | TBD | Live source gates validated; Biblioteka Nauki, Europeana, and HPLT/Common Corpus expansion remain subject to per-source QA, dedup, and legal review. |
+| `v0.3.1-preview` | reproducible Europeana shard | 810 | 140,042 | 1,000 direct API records sampled; 810 passed short/language/OCR gates, with per-record rights and author metadata. |
 
 ## What this dataset contributes
 The raw texts come from existing open corpora (redistributed via SpeakLeash and,
@@ -126,18 +66,10 @@ as the redistributing aggregator; this release does not claim ownership of them
 
 ## Contributors
 
-| contributor | contribution | release / PR |
-|---|---|---|
-| [Kacper Wikieł](https://huggingface.co/kacperwikiel) | Project maintainer; corpus curation, source and license review, release engineering, documentation, validation and reproducible build workflow. | all releases |
-| [Bart Kobyliński](https://huggingface.co/bartoszkobylinski1) | Added `1000_novels`; expanded Biblioteka Nauki and Europeana ingestion with per-document license and author metadata. | `v0.2.1`, [PR #1](https://huggingface.co/datasets/SlayerLab/polish-dynaword/discussions/1) |
-| [Paweł Puzio](https://huggingface.co/ppuzio) | Built `govpl`: subsite discovery, direct ingestion pipeline, dataset artifact, contract tests and documentation. | `v0.2.2`, [PR #9](https://huggingface.co/datasets/SlayerLab/polish-dynaword/discussions/9) |
-| [Arkadiusz Słota](https://huggingface.co/Maggio33) | Built `european_hplt_v3_pl`: HPLT v3 WDS 10+9 cleaning pipeline, 110,590-document artifact, validation and documentation. | `v0.2.3`, [PR #5](https://huggingface.co/datasets/SlayerLab/polish-dynaword/discussions/5) |
-| [Dawid Majewski](https://huggingface.co/dawidmajewski) | Added the reviewed Global Voices Polish corpus with per-document author attribution and documentation. | `v0.2.3`, [PR #7](https://huggingface.co/datasets/SlayerLab/polish-dynaword/discussions/7) |
-| [1am](https://huggingface.co/1am) | Added `nkjp1m`: the manually annotated 1-million-word NKJP subcorpus, direct fetch/build pipeline, source documentation and release artifact. | `v0.2.3`, [PR #8](https://huggingface.co/datasets/SlayerLab/polish-dynaword/discussions/8) |
-
-Contributions are credited when they add a verifiable dataset artifact, pipeline,
-validation, documentation or release work. A merged discussion with no resulting
-files is not listed as a data contribution.
+- **Kacper Wikieł** — corpus curation, release engineering, documentation, and
+  reproducible build workflow.
+- **Bart Kobyliński** — source expansion work for Biblioteka Nauki and Europeana,
+  including per-document license/author metadata collection and rebuild planning.
 
 ## Guiding principles
 1. **Open & traceable licensing** — every source is *openly licensed* with a documented
@@ -153,19 +85,15 @@ files is not listed as a data contribution.
 | [parliamentary](data/parliamentary/parliamentary.md) | Polish Parliamentary Corpus (Sejm/Senat) | `public-domain (official documents)` | 324,622 | 1,646.8M |
 | [wikisource](data/wikisource/wikisource.md) | Polish Wikisource | `CC-BY-SA-3.0` | 632,005 | 801.9M |
 | [wikipedia](data/wikipedia/wikipedia.md) | Polish Wikipedia | `CC-BY-SA-3.0` | 1,171,897 | 707.2M |
-| [european_hplt_v3_pl](data/european_hplt_v3_pl/european_hplt_v3_pl.md) | HPLT v3.0 Polish — top WDS quality bins 10+9 | `CC0-1.0` | 110,590 | 512.4M |
 | [dziennik_ustaw](data/dziennik_ustaw/dziennik_ustaw.md) | Dziennik Ustaw + Monitor Polski (Polish primary legislation) | `public-domain (official documents)` | 35,442 | 486.1M |
 | [wolne_lektury](data/wolne_lektury/wolne_lektury.md) | Wolne Lektury (school readings) | `CC-BY-SA-4.0 / Wolna Sztuka 1.3` | 6,141 | 103.0M |
-| [govpl](data/govpl/govpl.md) | gov.pl — Polish government press releases | `CC-BY-SA-4.0` | 88,190 | 80.7M |
 | [1000_novels](data/1000_novels/1000_novels.md) | 1000 Novels Corpus (CLARIN-PL) | `CC-BY-4.0` | 1,000 | 60.5M |
 | [wikiquote](data/wikiquote/wikiquote.md) | Polish Wikiquote (quotations) | `CC-BY-SA-3.0` | 30,363 | 31.9M |
 | [eltec_pol](data/eltec_pol/eltec_pol.md) | ELTeC-pol (European Literary Text Collection, Polish) | `CC-BY-4.0` | 100 | 21.5M |
 | [wikivoyage](data/wikivoyage/wikivoyage.md) | Polish Wikivoyage (travel guides) | `CC-BY-SA-3.0` | 13,645 | 17.1M |
 | [wikibooks](data/wikibooks/wikibooks.md) | Polish Wikibooks (open textbooks) | `CC-BY-SA-3.0` | 9,112 | 15.6M |
 | [wikinews](data/wikinews/wikinews.md) | Polish Wikinews | `CC-BY-2.5` | 24,386 | 12.1M |
-| [global_voices](data/global_voices/global_voices.md) | Global Voices Polish | `CC-BY-3.0` | 2,040 | 3.7M |
-| [nkjp1m](data/nkjp1m/nkjp1m.md) | NKJP manually annotated 1-million-word subcorpus | `CC-BY` | 18,381 | 2.6M |
-| **total** | | | **2,710,974** | **6,881.3M** |
+| **total** | | | **2,491,773** | **6,281.9M** |
 
 ## Method
 Only **human-authored** text — no synthetic, machine-translated, or auto-transcribed
@@ -216,10 +144,9 @@ Current v0.3 source-ingestion status:
 - Europeana release policy: split conservatively at pre-1929 records for
   US-sensitive downstream reuse, and keep later/unknown records separately
   labeled or held until legal review.
-- `european_hplt_v3_pl`: WDS bins 10+9 are included in v0.2.3 after contract
-  validation. HPLT packaging is CC0, while underlying crawled web documents can
-  carry independent rights; downstream near-dedup and web-content review remain
-  recommended before training.
+- `ashtok897/european-hplt-v1`: candidate workflow exists, but web-crawl
+  provenance, dedup, QA, and final mix weighting are still pending before stable
+  inclusion.
 
 Current review artifacts:
 
@@ -287,19 +214,19 @@ python3 src/make_docs.py
 
 Raw counts and token-normalized shares are regenerated from the current parquet files with `src/pattern_frequency_report.py`.
 
-- Total token count (tiktoken proxy): **6,881,277,362**
+- Total token count (tiktoken proxy): **6,281,911,234**
 
 | Pattern | Count | Share of all tokens |
 |---|---:|---:|
-| `w roku` | 460,249 | 0.0067% |
-| `klasyfikacji` | 134,093 | 0.0019% |
-| `ustawa` | 598,176 | 0.0087% |
-| `artykuł` | 2,096,271 | 0.0305% |
-| `parlament` | 1,216,808 | 0.0177% |
-| `rozporządzenie` | 1,497,273 | 0.0218% |
-| `w pobliżu` | 89,774 | 0.0013% |
-| `mieszkańców` | 288,911 | 0.0042% |
-| `Dz.U.` | 944,565 | 0.0137% |
+| `w roku` | 435,541 | 0.0069% |
+| `klasyfikacji` | 129,977 | 0.0021% |
+| `ustawa` | 587,459 | 0.0094% |
+| `artykuł` | 2,036,359 | 0.0324% |
+| `parlament` | 1,201,641 | 0.0191% |
+| `rozporządzenie` | 1,490,464 | 0.0237% |
+| `w pobliżu` | 78,823 | 0.0013% |
+| `mieszkańców` | 241,200 | 0.0038% |
+| `Dz.U.` | 939,966 | 0.0150% |
 
 ### Per-source shares
 
@@ -341,42 +268,6 @@ Raw counts and token-normalized shares are regenerated from the current parquet
 | `eurlex` | `w pobliżu` | 6,088 | 0.00026% |
 | `eurlex` | `mieszkańców` | 9,441 | 0.00040% |
 | `eurlex` | `Dz.U.` | 915,707 | 0.03851% |
-| `european_hplt_v3_pl` | `w roku` | 21,067 | 0.00411% |
-| `european_hplt_v3_pl` | `klasyfikacji` | 3,576 | 0.00070% |
-| `european_hplt_v3_pl` | `ustawa` | 5,754 | 0.00112% |
-| `european_hplt_v3_pl` | `artykuł` | 56,295 | 0.01099% |
-| `european_hplt_v3_pl` | `parlament` | 10,618 | 0.00207% |
-| `european_hplt_v3_pl` | `rozporządzenie` | 2,661 | 0.00052% |
-| `european_hplt_v3_pl` | `w pobliżu` | 9,877 | 0.00193% |
-| `european_hplt_v3_pl` | `mieszkańców` | 33,939 | 0.00662% |
-| `european_hplt_v3_pl` | `Dz.U.` | 1,482 | 0.00029% |
-| `global_voices` | `w roku` | 160 | 0.00431% |
-| `global_voices` | `klasyfikacji` | 5 | 0.00013% |
-| `global_voices` | `ustawa` | 157 | 0.00423% |
-| `global_voices` | `artykuł` | 1,047 | 0.02818% |
-| `global_voices` | `parlament` | 366 | 0.00985% |
-| `global_voices` | `rozporządzenie` | 26 | 0.00070% |
-| `global_voices` | `w pobliżu` | 105 | 0.00283% |
-| `global_voices` | `mieszkańców` | 508 | 0.01367% |
-| `global_voices` | `Dz.U.` | 1 | 0.00003% |
-| `govpl` | `w roku` | 3,372 | 0.00418% |
-| `govpl` | `klasyfikacji` | 526 | 0.00065% |
-| `govpl` | `ustawa` | 4,727 | 0.00586% |
-| `govpl` | `artykuł` | 2,417 | 0.00300% |
-| `govpl` | `parlament` | 3,967 | 0.00492% |
-| `govpl` | `rozporządzenie` | 4,102 | 0.00509% |
-| `govpl` | `w pobliżu` | 892 | 0.00111% |
-| `govpl` | `mieszkańców` | 13,105 | 0.01625% |
-| `govpl` | `Dz.U.` | 3,114 | 0.00386% |
-| `nkjp1m` | `w roku` | 109 | 0.00419% |
-| `nkjp1m` | `klasyfikacji` | 9 | 0.00035% |
-| `nkjp1m` | `ustawa` | 79 | 0.00304% |
-| `nkjp1m` | `artykuł` | 153 | 0.00588% |
-| `nkjp1m` | `parlament` | 216 | 0.00830% |
-| `nkjp1m` | `rozporządzenie` | 20 | 0.00077% |
-| `nkjp1m` | `w pobliżu` | 77 | 0.00296% |
-| `nkjp1m` | `mieszkańców` | 159 | 0.00611% |
-| `nkjp1m` | `Dz.U.` | 2 | 0.00008% |
 | `parliamentary` | `w roku` | 198,192 | 0.01203% |
 | `parliamentary` | `klasyfikacji` | 12,637 | 0.00077% |
 | `parliamentary` | `ustawa` | 459,179 | 0.02788% |

artifacts/pattern_frequency_hf_snippet.md

@@ -1,18 +1,18 @@
 ## Phrase frequency in corpus (token-normalized)
 
-- Total token count (tiktoken proxy): **6,881,277,362**
+- Total token count (tiktoken proxy): **6,281,911,234**
 
 | Pattern | Count | Share of all tokens |
 |---|---:|---:|
-| `w roku` | 460,249 | 0.0067% |
-| `klasyfikacji` | 134,093 | 0.0019% |
-| `ustawa` | 598,176 | 0.0087% |
-| `artykuł` | 2,096,271 | 0.0305% |
-| `parlament` | 1,216,808 | 0.0177% |
-| `rozporządzenie` | 1,497,273 | 0.0218% |
-| `w pobliżu` | 89,774 | 0.0013% |
-| `mieszkańców` | 288,911 | 0.0042% |
-| `Dz.U.` | 944,565 | 0.0137% |
+| `w roku` | 435,541 | 0.0069% |
+| `klasyfikacji` | 129,977 | 0.0021% |
+| `ustawa` | 587,459 | 0.0094% |
+| `artykuł` | 2,036,359 | 0.0324% |
+| `parlament` | 1,201,641 | 0.0191% |
+| `rozporządzenie` | 1,490,464 | 0.0237% |
+| `w pobliżu` | 78,823 | 0.0013% |
+| `mieszkańców` | 241,200 | 0.0038% |
+| `Dz.U.` | 939,966 | 0.0150% |
 
 ### Per-source shares
 
@@ -54,42 +54,6 @@
 | `eurlex` | `w pobliżu` | 6,088 | 0.00026% |
 | `eurlex` | `mieszkańców` | 9,441 | 0.00040% |
 | `eurlex` | `Dz.U.` | 915,707 | 0.03851% |
-| `european_hplt_v3_pl` | `w roku` | 21,067 | 0.00411% |
-| `european_hplt_v3_pl` | `klasyfikacji` | 3,576 | 0.00070% |
-| `european_hplt_v3_pl` | `ustawa` | 5,754 | 0.00112% |
-| `european_hplt_v3_pl` | `artykuł` | 56,295 | 0.01099% |
-| `european_hplt_v3_pl` | `parlament` | 10,618 | 0.00207% |
-| `european_hplt_v3_pl` | `rozporządzenie` | 2,661 | 0.00052% |
-| `european_hplt_v3_pl` | `w pobliżu` | 9,877 | 0.00193% |
-| `european_hplt_v3_pl` | `mieszkańców` | 33,939 | 0.00662% |
-| `european_hplt_v3_pl` | `Dz.U.` | 1,482 | 0.00029% |
-| `global_voices` | `w roku` | 160 | 0.00431% |
-| `global_voices` | `klasyfikacji` | 5 | 0.00013% |
-| `global_voices` | `ustawa` | 157 | 0.00423% |
-| `global_voices` | `artykuł` | 1,047 | 0.02818% |
-| `global_voices` | `parlament` | 366 | 0.00985% |
-| `global_voices` | `rozporządzenie` | 26 | 0.00070% |
-| `global_voices` | `w pobliżu` | 105 | 0.00283% |
-| `global_voices` | `mieszkańców` | 508 | 0.01367% |
-| `global_voices` | `Dz.U.` | 1 | 0.00003% |
-| `govpl` | `w roku` | 3,372 | 0.00418% |
-| `govpl` | `klasyfikacji` | 526 | 0.00065% |
-| `govpl` | `ustawa` | 4,727 | 0.00586% |
-| `govpl` | `artykuł` | 2,417 | 0.00300% |
-| `govpl` | `parlament` | 3,967 | 0.00492% |
-| `govpl` | `rozporządzenie` | 4,102 | 0.00509% |
-| `govpl` | `w pobliżu` | 892 | 0.00111% |
-| `govpl` | `mieszkańców` | 13,105 | 0.01625% |
-| `govpl` | `Dz.U.` | 3,114 | 0.00386% |
-| `nkjp1m` | `w roku` | 109 | 0.00419% |
-| `nkjp1m` | `klasyfikacji` | 9 | 0.00035% |
-| `nkjp1m` | `ustawa` | 79 | 0.00304% |
-| `nkjp1m` | `artykuł` | 153 | 0.00588% |
-| `nkjp1m` | `parlament` | 216 | 0.00830% |
-| `nkjp1m` | `rozporządzenie` | 20 | 0.00077% |
-| `nkjp1m` | `w pobliżu` | 77 | 0.00296% |
-| `nkjp1m` | `mieszkańców` | 159 | 0.00611% |
-| `nkjp1m` | `Dz.U.` | 2 | 0.00008% |
 | `parliamentary` | `w roku` | 198,192 | 0.01203% |
 | `parliamentary` | `klasyfikacji` | 12,637 | 0.00077% |
 | `parliamentary` | `ustawa` | 459,179 | 0.02788% |

data/biblioteka_nauki/biblioteka_nauki.md

@@ -1,60 +0,0 @@
-# biblioteka_nauki
-
-Biblioteka Nauki — Polish open-access scholarly corpus (scientific articles, books, chapters)
-
-## Dataset description
-- **Source (upstream):** https://bibliotekanauki.pl (operated by ICM, University of Warsaw — Centrum Otwartej Nauki)
-- **Domain:** academic (scientific articles, books, chapters)
-- **Language:** Polish (pl)
-- **License:** per-document — CC BY 4.0 28,335 · CC BY-SA 4.0 13,733 · Public Domain 3
-- **Created (range):** per-document publication year, from each item's bibliotekanauki.pl metadata
-- **Added:** 2026-06-27
-
-## Licensing — traceable basis
-The SpeakLeash blob carries no per-document license field, and ~53% of the upstream collection is
-all-rights-reserved, so a per-item gate was mandatory. Each item's license was recovered from the
-bibliotekanauki.pl API (`bibliotekanauki.pl/api/{type}/<id>` → `license.iconId`). Only openly-licensed
-items are kept; all-rights-reserved and unknown items are gated out. Recovery yielded 43,576 clean ids.
-
-| license | documents |
-|---|---:|
-| CC BY 4.0 | 28,335 |
-| CC BY-SA 4.0 | 13,733 |
-| Public Domain | 3 |
-
-The per-document license is carried in the parquet `license` column.
-
-## Attribution
-Per-document attribution (author · year · title · publisher) was re-fetched 1:1 from
-bibliotekanauki.pl (43,576 / 43,576 ok) and is carried in the parquet `attribution` column,
-satisfying the CC BY / CC BY-SA attribution requirement that applies to every document here.
-
-## Provenance
-Pulled from SpeakLeash's public redistribution (`speakleash-ds-pub`, key `biblioteka_nauki_pl_corpus`)
-of the upstream source above. SpeakLeash credited as intermediate aggregator; upstream license and
-attribution recovered per document and preserved.
-
-## Statistics
-| documents | characters | tokens (tiktoken proxy) |
-|---:|---:|---:|
-| 42,071 | 1,748,333,935 | 673,488,831 |
-
-## Filters applied (build_dynaword.py)
-Minimal, per Dynaword guidelines (heavy filtering left to downstream use):
-- drop documents < 200 chars: **0**
-- drop non-Polish (diacritic ratio): **925**
-- exact cross-source dedup (sha1): **180**
-- OCR alpha-ratio < 0.70 (OCR source): **400**
-- read 43,576 → kept 42,071
-
-Token counts are a fast tiktoken (cl100k) proxy (~1% off Llama-3); the canonical
-Llama-3 count is computed at release.
-
-## Limitations
-- **Mixed license.** Roughly one third of the documents are CC BY-SA 4.0 (share-alike); the exact
-  per-document license is in the `license` column. Downstream redistribution must respect share-alike
-  for those documents.
-- **OCR / extraction noise.** Text is extracted from publisher PDFs; some layout and hyphenation
-  artefacts remain after the alpha-ratio filter.
-- **Per-document dates.** `created` is the publication year from upstream metadata; finer-grained
-  dates are not available.

data/biblioteka_nauki/biblioteka_nauki.parquet

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:cc93cdd5ca3e6576a1da0aa294df1308e879b8516ac30911289e430c69a07b26
-size 1076638904

data/biblioteka_nauki/biblioteka_nauki.stats.json

@@ -1,12 +0,0 @@
-{
-  "read": 43576,
-  "kept": 42071,
-  "drop_short": 0,
-  "drop_lang": 925,
-  "drop_dup": 180,
-  "drop_ocr": 400,
-  "chars": 1748333935,
-  "tokens": 673488831,
-  "secs": 484.1,
-  "license": "CC-BY-4.0 / CC-BY-SA-4.0"
-}

data/biblioteka_nauki/recover_attribution.py

@@ -1,125 +0,0 @@
-#!/usr/bin/env python3
-"""Targeted attribution pass for the CLEAN (CC-BY/CC-BY-SA/PD) Biblioteka Nauki ids ONLY.
-The first recovery pass stored the license correctly but parsed the wrong attribution fields.
-CC-BY/CC-BY-SA legally REQUIRE attribution, so re-fetch just the ~43.6k clean ids and extract:
-  authors = contributors[role=AUTHOR] firstName+lastName ; year = dates[PUBLISHED].date.year ;
-  title = mainTitle.text ; publisher = publishingCompanyInfo.name
-
-  licenses.jsonl -> attribution.jsonl  {type,id,license,authors,year,title,publisher}
-Resumable, threaded, polite. ~43.6k calls.
-"""
-import json, os, random, sys, threading, time
-from concurrent.futures import ThreadPoolExecutor, as_completed
-import requests
-
-HERE = os.path.dirname(os.path.abspath(__file__))
-LIC = os.path.join(HERE, "licenses.jsonl")
-OUT = os.path.join(HERE, "attribution.jsonl")
-PROG = os.path.join(HERE, "attribution_progress.txt")
-CLEAN = {"cc-by", "cc-by-sa", "cc0", "public-domain"}
-WORKERS = 8
-UA = ("Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 "
-      "(KHTML, like Gecko) Chrome/124 Safari/537.36")
-HEADERS = {"User-Agent": UA, "Accept": "application/json", "Referer": "https://bibliotekanauki.pl/"}
-_local = threading.local()
-_wlock = threading.Lock()
-
-
-def session():
-    s = getattr(_local, "s", None)
-    if s is None:
-        s = requests.Session()
-        s.headers.update(HEADERS)
-        s.mount("https://", requests.adapters.HTTPAdapter(pool_connections=4, pool_maxsize=4))
-        _local.s = s
-    return s
-
-
-def parse(d):
-    authors = []
-    for c in (d.get("contributors") or []):
-        if c.get("role") == "AUTHOR":
-            nm = f"{c.get('firstName','') or ''} {c.get('lastName','') or ''}".strip()
-            if nm:
-                authors.append(nm)
-    year = None
-    dates = d.get("dates") or []
-    for dt in dates:
-        if dt.get("type") == "PUBLISHED":
-            year = (dt.get("date") or {}).get("year")
-            break
-    if not year and dates:
-        year = (dates[0].get("date") or {}).get("year")
-    title = (d.get("mainTitle") or {}).get("text")
-    pub = (d.get("publishingCompanyInfo") or {}).get("name")
-    return authors[:20], (str(year) if year else None), title, pub
-
-
-def fetch(item):
-    typ, _id, lic = item
-    url = f"https://bibliotekanauki.pl/api/{typ}/{_id}"
-    for attempt in range(4):
-        try:
-            r = session().get(url, timeout=25)
-            if r.status_code == 200:
-                a, y, t, p = parse(r.json())
-                return {"type": typ, "id": _id, "license": lic,
-                        "authors": a, "year": y, "title": t, "publisher": p, "status": "ok"}
-            if r.status_code in (403, 404, 410):
-                return {"type": typ, "id": _id, "license": lic, "status": f"http{r.status_code}"}
-            time.sleep((2 ** attempt) + random.random())
-        except Exception as e:
-            if attempt == 3:
-                return {"type": typ, "id": _id, "license": lic, "status": f"err:{type(e).__name__}"}
-            time.sleep((2 ** attempt) + random.random())
-    return {"type": typ, "id": _id, "license": lic, "status": "err:retries"}
-
-
-def main():
-    best = {}
-    for l in open(LIC, encoding="utf-8"):
-        try:
-            r = json.loads(l)
-        except Exception:
-            continue
-        best[(r["type"], r["id"])] = r
-    clean = [(t, i, r["iconId"]) for (t, i), r in best.items()
-             if r.get("status") == "ok" and r.get("iconId") in CLEAN]
-    done = set()
-    if os.path.exists(OUT):
-        for l in open(OUT, encoding="utf-8"):
-            try:
-                r = json.loads(l)
-                done.add((r["type"], r["id"]))
-            except Exception:
-                pass
-    todo = [c for c in clean if (c[0], c[1]) not in done]
-    print(f"clean={len(clean)} done={len(done)} todo={len(todo)} workers={WORKERS}", flush=True)
-
-    out = open(OUT, "a", encoding="utf-8")
-    n = len(done)
-    t0 = time.time()
-    counts = {}
-    with ThreadPoolExecutor(max_workers=WORKERS) as ex:
-        futs = [ex.submit(fetch, c) for c in todo]
-        for fut in as_completed(futs):
-            res = fut.result()
-            with _wlock:
-                out.write(json.dumps(res, ensure_ascii=False) + "\n")
-            n += 1
-            counts[res.get("status")] = counts.get(res.get("status"), 0) + 1
-            time.sleep(random.uniform(0.02, 0.10))
-            if n % 500 == 0:
-                out.flush()
-                rate = (n - len(done)) / max(time.time() - t0, 1)
-                eta = (len(clean) - n) / max(rate, 0.1) / 60
-                msg = f"{n}/{len(clean)} {rate:.1f}/s ETA {eta:.0f}m " + \
-                      " ".join(f"{k}={v}" for k, v in sorted(counts.items()))
-                print(msg, flush=True)
-                open(PROG, "w").write(msg + "\n")
-    out.flush(); out.close()
-    print(f"DONE attribution: {n}/{len(clean)} counts={counts}", flush=True)
-
-
-if __name__ == "__main__":
-    main()

data/biblioteka_nauki/recover_licenses.py

@@ -1,171 +0,0 @@
-#!/usr/bin/env python3
-"""Pass 2: recover the per-item license for every Biblioteka Nauki doc via the public API.
-  index.jsonl  ->  GET bibliotekanauki.pl/api/{type}/{id}  ->  licenses.jsonl
-
-Resumable (skips ids already in licenses.jsonl), polite (modest concurrency + jitter + backoff),
-keep-alive sessions, and a circuit-breaker if the host starts blocking us.
-
-Output line: {type,id, iconId, cc(bool), restricted(bool), authors, year, source, status}
-  status: "ok" | "http404" | "http403" | "err:<Type>"
-"""
-import json, os, random, sys, threading, time
-from concurrent.futures import ThreadPoolExecutor, as_completed
-import requests
-
-HERE = os.path.dirname(os.path.abspath(__file__))
-INDEX = os.path.join(HERE, "index.jsonl")
-OUT = os.path.join(HERE, "licenses.jsonl")
-PROG = os.path.join(HERE, "recover_progress.txt")
-
-WORKERS = 8
-TIMEOUT = 25
-MAX_RETRY = 4
-UA = ("Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 "
-      "(KHTML, like Gecko) Chrome/124 Safari/537.36")
-HEADERS = {"User-Agent": UA, "Accept": "application/json", "Referer": "https://bibliotekanauki.pl/"}
-
-_local = threading.local()
-_write_lock = threading.Lock()
-_recent = []            # rolling window of recent statuses for the circuit breaker
-_recent_lock = threading.Lock()
-_stop = threading.Event()
-
-
-def session():
-    s = getattr(_local, "s", None)
-    if s is None:
-        s = requests.Session()
-        s.headers.update(HEADERS)
-        a = requests.adapters.HTTPAdapter(pool_connections=4, pool_maxsize=4)
-        s.mount("https://", a)
-        _local.s = s
-    return s
-
-
-def load_done():
-    done = set()
-    if os.path.exists(OUT):
-        with open(OUT, encoding="utf-8") as f:
-            for line in f:
-                try:
-                    r = json.loads(line)
-                    done.add((r["type"], r["id"]))
-                except Exception:
-                    pass
-    return done
-
-
-def fetch(item):
-    if _stop.is_set():
-        return None
-    typ, _id = item["type"], item["id"]
-    url = f"https://bibliotekanauki.pl/api/{typ}/{_id}"
-    for attempt in range(MAX_RETRY):
-        try:
-            r = session().get(url, timeout=TIMEOUT)
-            if r.status_code == 200:
-                d = r.json()
-                lic = d.get("license") or {}
-                authors = []
-                for c in (d.get("contributors") or []):
-                    nm = (c.get("name") or "").strip()
-                    if nm:
-                        authors.append(nm)
-                year = None
-                for dt in (d.get("dates") or []):
-                    y = str(dt.get("year") or dt.get("value") or "")[:4]
-                    if y.isdigit():
-                        year = y
-                        break
-                insts = d.get("publicationInstitutions") or d.get("journal") or None
-                if isinstance(insts, list):
-                    insts = "; ".join(x.get("name", "") if isinstance(x, dict) else str(x) for x in insts) or None
-                return {
-                    "type": typ, "id": _id,
-                    "iconId": lic.get("iconId"),
-                    "cc": bool(lic.get("creativeCommons")),
-                    "restricted": bool(d.get("restricted")),
-                    "authors": authors[:12],
-                    "year": year,
-                    "source": insts,
-                    "status": "ok",
-                }
-            if r.status_code in (403,):
-                return {"type": typ, "id": _id, "status": "http403"}
-            if r.status_code in (404, 410):
-                return {"type": typ, "id": _id, "status": "http404"}
-            # 429 / 5xx -> backoff & retry
-            time.sleep((2 ** attempt) + random.random())
-        except Exception as e:
-            if attempt == MAX_RETRY - 1:
-                return {"type": typ, "id": _id, "status": f"err:{type(e).__name__}"}
-            time.sleep((2 ** attempt) + random.random())
-    return {"type": typ, "id": _id, "status": "err:retries"}
-
-
-def note(status):
-    with _recent_lock:
-        _recent.append(status)
-        if len(_recent) > 200:
-            _recent.pop(0)
-        # circuit breaker: >60% 403 over a full window => host is blocking us
-        if len(_recent) == 200 and sum(1 for s in _recent if s == "http403") > 120:
-            _stop.set()
-
-
-def main():
-    if not os.path.exists(INDEX):
-        sys.exit(f"missing {INDEX} — run build_index.py first")
-    items = []
-    with open(INDEX, encoding="utf-8") as f:
-        for line in f:
-            try:
-                r = json.loads(line)
-                items.append({"type": r["type"], "id": r["id"]})
-            except Exception:
-                pass
-    done = load_done()
-    todo = [it for it in items if (it["type"], it["id"]) not in done]
-    total = len(items)
-    print(f"index={total}  done={len(done)}  todo={len(todo)}  workers={WORKERS}", flush=True)
-
-    n = len(done)
-    t0 = time.time()
-    out = open(OUT, "a", encoding="utf-8")
-    counts = {}
-    try:
-        with ThreadPoolExecutor(max_workers=WORKERS) as ex:
-            futs = [ex.submit(fetch, it) for it in todo]
-            for fut in as_completed(futs):
-                res = fut.result()
-                if res is None:
-                    continue
-                with _write_lock:
-                    out.write(json.dumps(res, ensure_ascii=False) + "\n")
-                n += 1
-                st = res.get("status", "?")
-                counts[st] = counts.get(st, 0) + 1
-                note(st)
-                # politeness jitter (per completed task; with 8 workers ~ a few hundred req/min)
-                time.sleep(random.uniform(0.02, 0.12))
-                if n % 500 == 0:
-                    out.flush()
-                    rate = (n - len(done)) / max(time.time() - t0, 1)
-                    eta = (total - n) / max(rate, 0.1) / 60
-                    msg = (f"{n}/{total}  {rate:.1f}/s  ETA {eta:.0f}m  "
-                           + " ".join(f"{k}={v}" for k, v in sorted(counts.items())))
-                    print(msg, flush=True)
-                    with open(PROG, "w") as p:
-                        p.write(msg + "\n")
-                if _stop.is_set():
-                    print("!! circuit breaker tripped (sustained 403) — stopping. Re-run later to resume.", flush=True)
-                    break
-    finally:
-        out.flush()
-        out.close()
-    print(f"FINISHED pass: wrote {n}/{total}. counts={counts}", flush=True)
-    print("Next: build_parquet.py to assemble the CC-BY/SA subset.", flush=True)
-
-
-if __name__ == "__main__":
-    main()

data/european_hplt_v3_pl/european_hplt_v3_pl.md

@@ -1,57 +0,0 @@
-# european_hplt_v3_pl
-
-Filtered Polish subset of **HPLT v3.0** (`pol_Latn`), top quality bins (WDS = 10 and 9), merged into one file.
-
-## Dataset description
-- **Source (upstream):** https://hplt-project.org/datasets/v3.0 (`pol_Latn`, shards `10_1` [WDS=10] + `9_1` prefix [WDS=9])
-- **Domain:** web (modern natural Polish — informational, narrative, opinion, how-to, discussion)
-- **Language:** Polish (pl)
-- **License:** `CC0-1.0` (packaging; see licensing note)
-- **Created (range):** web crawls 2012-2024 (Internet Archive + Common Crawl)
-- **Added:** 2026-07-14
-
-## Licensing - traceable basis
-HPLT licenses the **packaging** of these text data under Creative Commons CC0 1.0 ("no rights reserved"),
-per the HPLT v3.0 Terms of Use (hplt-project.org/datasets/v3.0) and the HF card `HPLT/HPLT3.0`.
-HPLT does not own the underlying web text; a notice-and-takedown policy applies and downstream users
-remain responsible for compliance (EU CDSM Directive 2019/790, GDPR). This is a web-crawl source: it is
-offered at parity with the already-listed `european-hplt-v1` (CC0) basis and still needs maintainer
-web/legal review (`accept_after_web_review`), not auto-accept.
-
-## Provenance
-Fetched directly from the HPLT distribution (`data.hplt-project.org/three/sorted/pol_Latn/{10_1,9_1}.jsonl.zst`),
-NOT via SpeakLeash. HPLT v3 is globally deduplicated upstream and sorted into Web-Docs-Scorer (WDS) quality
-bins; only the two top bins are used. Per-document URL, WDS score, web-register, MT-probability and PII spans
-are available in the raw HPLT record and were used for filtering (not retained in the released canonical schema).
-
-## Statistics
-| documents | characters | tokens (tiktoken cl100k proxy) |
-|---:|---:|---:|
-| 110,590 | 1,426,290,944 | 512,383,778 |
-
-Composition (single merged parquet): bin-10 (WDS=10) 55,590 docs / 261,468,813 tok;
-bin-9 (WDS=9) 55,000 docs / 250,914,965 tok. See `european_hplt_v3_pl.stats.json` for the per-bin + drop breakdown.
-
-## Per-document license metadata
-| license | documents |
-|---|---:|
-| `CC0-1.0` | 110,590 |
-
-## Filters applied (src/clean_hplt_v3.py; thresholds inherited from src/filter_european_hplt.py)
-Per-shard keep: bin-10 55,590/195,874 (28.4%), bin-9 55,000/152,380 (36.1%). Drop reasons:
-`register` (not ID/OP/HI/IN/NA), `machine_translated` (MT>=0.20), `domain`
-(wikipedia/wikisource/wikimedia + file-hosts excluded), `adult_spam`, `legalish` (drops legal/parliamentary
-phrasing - serves the anti-legalese goal), `boilerplate`, `too_long` (>120k chars = book-length copyright/OCR
-risk), `mojibake` (>=3 CJK/replacement chars - upstream HPLT UTF-8->GBK mis-decode), `filehost` (PDF/file-share
-headers), `domain_cap` (>250 docs/domain). MT-probability of kept docs: mean ~0.08. Mojibake residual: 0.
-
-## Status & caveats (web-crawl - NOT "release-clean")
-- **Copyright is not fully solved by filtering.** The length cap removes obvious book-length works, but
-  in-copyright text recurs in web-crawl at ALL lengths (news reprints, blog reposts, excerpts). Inherent to
-  web sources; the reason HPLT ships CC0-on-packaging + notice-and-takedown.
-- **Near-dup NOT removed** (Dynaword leaves heavy filtering downstream). ~11.8% of an HPLT sample's 8-grams
-  overlap Polish Wikipedia; wikipedia/wikisource domain exclusion mitigates the largest sources, but
-  cross-source near/containment dedup vs the full corpus remains a downstream step.
-- Surface-metric parity with accepted sources: mojibake 0%, boilerplate ~0.2%, diacritic-ratio ~0.066,
-  TTR@500 ~0.64 (within the accepted range, ~= parliamentary). ~9% commercial/SEO-adjacent pages (web-inherent).
-- This is 2 top bins merged. The full `pol_Latn` corpus is ~270B raw tokens across WDS bins 10->5.

data/european_hplt_v3_pl/european_hplt_v3_pl.parquet

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:6295b3280cef2b99e9bf511a1c48be9c94cd219455d2e9666ebf17363771b714
-size 610242503

data/european_hplt_v3_pl/european_hplt_v3_pl.stats.json

@@ -1,41 +0,0 @@
-{
-  "read": 348254,
-  "kept": 110590,
-  "drop_short": 0,
-  "drop_lang": 2,
-  "drop_dup": 0,
-  "drop_ocr": 0,
-  "chars": 1426290944,
-  "tokens": 512383778,
-  "licenses": {
-    "CC0-1.0": 110590
-  },
-  "authors_with_value": 0,
-  "license": "CC0-1.0",
-  "stats_recomputed_from_parquet": true,
-  "drop_detail": {
-    "machine_translated": 46477,
-    "adult_spam": 18682,
-    "register": 128956,
-    "legalish": 15881,
-    "domain": 23234,
-    "boilerplate": 3183,
-    "mojibake": 241,
-    "too_long": 216,
-    "filehost": 7,
-    "lang_prob": 2,
-    "domain_cap": 785
-  },
-  "shards": {
-    "bin10_WDS10": {
-      "kept": 55590,
-      "tokens": 261468813
-    },
-    "bin9_WDS9": {
-      "kept": 55000,
-      "tokens": 250914965
-    }
-  },
-  "mt_prob_mean_kept": 0.08,
-  "source_repo": "HPLT/HPLT3.0 pol_Latn WDS bins 10+9 (data.hplt-project.org)"
-}

data/global_voices/NOTICE.md

@@ -1,11 +0,0 @@
-# NOTICE — Global Voices source
-
-The exported article text is attributed to Global Voices and the credited
-original authors and translators under Creative Commons Attribution 3.0.
-
-Source: https://pl.globalvoices.org/
-License: https://creativecommons.org/licenses/by/3.0/
-
-Per-record contributor attribution is stored in the `author` column. The accompanying `global_voices.attribution.jsonl` sidecar additionally preserves each released article's title, canonical URL, license URL and modification notice. Twelve historical records had no machine-readable contributor field; full-text review recovered Didem Tali and Khalid Ibrahim for two of them. Ten records without an identifiable contributor use the explicitly organizational attribution `Global Voices`; this must not be interpreted as a known individual author.
-
-Binary media files are not included. Articles may nevertheless contain captions, URLs, social-media posts and quotations from third-party works; rights in those materials may remain with their respective authors. Article-level CC BY 3.0 is not asserted independently for photographs, videos, embeds or separately quoted works.

data/global_voices/global_voices.md

@@ -1,65 +0,0 @@
-# global_voices
-
-Polish-language civic-media and public-interest journalism from [Global Voices](https://pl.globalvoices.org/), produced by its international community of writers and volunteer translators.
-
-## Dataset description
-
-- **Source (upstream):** https://pl.globalvoices.org/
-- **Domain:** civic media, news and public-interest journalism
-- **Language:** Polish (pl)
-- **License:** `CC-BY-3.0`
-- **Created (range):** 2008-11-04, 2023-05-24
-- **Added:** 2026-07-15
-
-Global Voices describes its Lingua translations as human-made work performed by a volunteer translator community. The dataset primarily represents edited, translated journalism rather than conversational Polish or text necessarily authored originally in Polish.
-
-## Statistics
-
-| documents | characters | tokens (`cl100k_base` proxy) |
-|---:|---:|---:|
-| 2,040 | 10,073,628 | 3,715,020 |
-
-The Parquet schema matches DynaWord:
-
-`id, text, source, added, created, token_count, license, author`
-
-Published files:
-
-- `global_voices.parquet` — cleaned article text and DynaWord metadata;
-- `global_voices.stats.json` — release statistics;
-- `global_voices.attribution.jsonl` — title, canonical article URL, contributor attribution, license URL and modification notice for every document;
-- `NOTICE.md` — licensing and attribution notice.
-
-## Provenance
-
-Articles were acquired directly from the Polish Global Voices website through its official publication interfaces, rather than from a third-party corpus. Each released ID is linked to the corresponding canonical article URL in `global_voices.attribution.jsonl`.
-
-The source release contained 2,044 records. Four records below the minimum quality and length requirements were excluded, leaving 2,040 documents. The final source contains no exact duplicate documents.
-
-## Cleaning
-
-The release received a reviewed cleaning pass. Redundant source-language passages were removed where the corresponding Polish translation was retained, together with extraction artifacts and obsolete interface text. Intentional quotations and source-language material required for meaning or documentary context were preserved.
-
-The cleaning process did not use machine translation or automatic speech transcription. It should not be interpreted as making every document strictly monolingual.
-
-## Licensing — traceable basis and attribution
-
-CC BY 3.0 was verified from the article-level license link published with each retained Global Voices article. Each released record carries `CC-BY-3.0` metadata. Contributor roles are preserved in the `author` column where available, distinguishing original authors from Polish translators.
-
-Because the fixed DynaWord schema does not contain article titles or URLs, the attribution sidecar maps every released `id` to its title, canonical URL, contributor attribution, license URL and modification notice. It contains one unique URL-bearing record for each of the 2,040 Parquet documents.
-
-Recommended attribution:
-
-> Global Voices; the original author(s) and Polish translator(s) listed in the attribution sidecar; canonical article URL; CC BY 3.0. Article text was extracted and cleaned; redundant source-language passages were removed where a Polish translation was retained.
-
-Binary photographs, videos and embedded media are not redistributed. Articles may contain quotations, captions or social-media text whose separate rights remain with their respective authors.
-
-## Limitations and recommended use
-
-- The dataset represents edited online journalism and human translation, not the full distribution of contemporary Polish.
-- Translationese and source-specific editorial conventions may be present.
-- Some intentional source-language quotations remain inside otherwise Polish articles.
-- Coverage ends in May 2023.
-- Cross-source and near-duplicate removal against other DynaWord sources remains part of the complete corpus build.
-
-Global Voices is best used as a complementary source of contemporary civic-media and journalistic Polish, rather than as a standalone representation of colloquial or natively authored Polish.

data/global_voices/global_voices.parquet

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:1ca02fdb6bad32b20757ddcd9b9f072056ff35b68c487143138cc7912b3a45c3
-size 5253559

data/global_voices/global_voices.stats.json

@@ -1,15 +0,0 @@
-{
-  "read": 2044,
-  "kept": 2040,
-  "drop_short": 4,
-  "drop_lang": 0,
-  "drop_dup": 0,
-  "drop_ocr": 0,
-  "chars": 10073628,
-  "tokens": 3715020,
-  "licenses": {
-    "CC-BY-3.0": 2040
-  },
-  "authors_with_value": 2040,
-  "stats_recomputed_from_parquet": true
-}

data/govpl/govpl.md

@@ -1,41 +0,0 @@
-# govpl
-
-gov.pl — Polish government press releases
-
-## Dataset description
-- **Source (upstream):** https://www.gov.pl/
-- **Domain:** government
-- **Language:** Polish (pl)
-- **License:** `CC-BY-SA-4.0`
-- **Created (range):** 2005-09-26, 2026-07-19
-- **Added:** 2026-07-19
-
-## Licensing — traceable basis
-All textual content on gov.pl is published under CC-BY-SA 4.0 (site footer: "Treści tekstowe publikowane w serwisie … są udostępniane na licencji CC BY-SA 4.0"). Authored and hosted by the Polish government, which holds the right to license it — same open family as Polish Wikipedia.
-
-## Provenance
-Ministry/agency press releases fetched directly from gov.pl (editor-content per article, HTML stripped) by src/fetch_govpl.py. Subsites + their non-uniform news-section paths are enumerated by src/discover_govpl.py into the committed manifest src/govpl_subsites.json (gov.pl has no sitemap/API). Each record's meta.url resolves on gov.pl.
-
-## Statistics
-| documents | characters | tokens (tiktoken proxy) |
-|---:|---:|---:|
-| 88,190 | 218,099,000 | 80,666,129 |
-
-
-## Per-document license metadata
-| license | documents |
-|---|---:|
-| `CC-BY-SA-4.0` | 88,190 |
-
-Author metadata present for **0** documents. Empty values mean the upstream record did not expose a machine-readable author field.
-
-## Filters applied (build_dynaword.py)
-Minimal, per Dynaword guidelines (heavy filtering left to downstream use):
-- drop documents < 200 chars: **0**
-- drop non-Polish (diacritic ratio): **24**
-- exact cross-source dedup (sha1): **1,926**
-- OCR alpha-ratio < 0.70 (OCR sources only): **0**
-- read 90,140 → kept 88,190
-
-Token counts are a fast tiktoken (cl100k) proxy (~1% off Llama-3); the canonical
-Llama-3 count is computed at release.

data/govpl/govpl.parquet

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:9c0310fcdf45804e6547bd197cba55f58f12df618cd3d8b92b8e9077a3720b9f
-size 82940575

data/govpl/govpl.stats.json

@@ -1,16 +0,0 @@
-{
-  "read": 90140,
-  "kept": 88190,
-  "drop_short": 0,
-  "drop_lang": 24,
-  "drop_dup": 1926,
-  "drop_ocr": 0,
-  "chars": 218099000,
-  "tokens": 80666129,
-  "licenses": {
-    "CC-BY-SA-4.0": 88190
-  },
-  "authors_with_value": 0,
-  "secs": 2.9,
-  "license": "CC-BY-SA-4.0"
-}

data/nkjp1m/nkjp1m.md

@@ -1,42 +0,0 @@
-# nkjp1m
-
-The manually annotated 1-million word subcorpus of the National Corpus of Polish
-
-## Dataset description
-- **Source (upstream):** https://clip.ipipan.waw.pl/NationalCorpusOfPolish
-- **Domain:** mixed (balanced reference corpus)
-- **Language:** Polish (pl)
-- **License:** `CC-BY`
-- **Created (range):** 1957-2008 (recorded for 6,581 of 18,381 documents; per-document in the created column)
-- **Added:** 2026-06-15
-
-## Licensing — traceable basis
-The NKJP download page (clip.ipipan.waw.pl) states the manually annotated 1-million-word subcorpus (Podkorpus Milionowy NKJP 1.2) is available under a CC-BY licence; attribution to the NKJP Consortium (IPI PAN) preserved. The page states 'CC-BY' with no version, normalized to CC-BY-4.0 for SPDX. The CC-BY licence covers the corpus text; NKJP's separate tools are GPL and are not included here.
-
-## Provenance
-Downloaded the IPI PAN tarball (NKJP-PodkorpusMilionowy-1.2.tar.gz) directly by src/fetch_njkp.py, one row per TEI <div> passage (its <ab> paragraphs joined by newlines, ellipsis markers stripped). The same minimal gates as build_dynaword.py are applied per passage (drop < 200 chars, drop non-Polish by diacritic ratio, exact sha1 dedup; the OCR gate is inapplicable). Per-document 'created' is the sample's TEI publication date (<date type="published">) where the header records one. Normalized into the DynaWord schema; cl100k token counts via tiktoken (encode_ordinary), and the stats are recomputed from the written parquet.
-
-## Statistics
-| documents | characters | tokens (tiktoken proxy) |
-| --------: | ---------: | ----------------------: |
-|    18,381 |  6,824,905 |               2,601,201 |
-
-
-## Per-document license metadata
-| license | documents |
-| ------- | --------: |
-| `CC-BY` |    18,381 |
-
-Author metadata present for **18,381** documents. Empty values mean the upstream record did not expose a machine-readable author field.
-
-Statistics were recomputed directly from the released parquet file.
-
-## Filters applied (build_dynaword.py)
-Minimal, per Dynaword guidelines (heavy filtering left to downstream use):
-- drop documents < 200 chars: **92**
-- drop non-Polish (diacritic ratio): **2**
-- exact cross-source dedup (sha1): **9**
-- OCR alpha-ratio < 0.70 (OCR sources only): **0**
-- read 18,484 → kept 18,381
-
-Token counts are a fast tiktoken (cl100k) proxy.

data/nkjp1m/nkjp1m.parquet

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:75f6a0d37c344960baa70e1d582f9908bc60f8bfee75bba97a68cecadbb745c7
-size 3718378

data/nkjp1m/nkjp1m.stats.json

@@ -1,18 +0,0 @@
-{
-  "read": 18484,
-  "kept": 18381,
-  "drop_short": 92,
-  "drop_lang": 2,
-  "drop_dup": 9,
-  "drop_ocr": 0,
-  "chars": 6824905,
-  "tokens": 2601201,
-  "licenses": {
-    "CC-BY": 18381
-  },
-  "authors_with_value": 18381,
-  "documents_with_created": 6581,
-  "created_range": "1957-2008",
-  "license": "CC-BY",
-  "stats_recomputed_from_parquet": true
-}

pattern_frequency_report.md

@@ -1,18 +1,18 @@
 ## Pattern frequency on corpus
 
-- total tokens (tiktoken proxy): `6,881,277,362`
+- total tokens (tiktoken proxy): `6,281,911,234`
 
 | pattern | count | share of all tokens |
 |---|---:|---:|
-| `w roku` | 460,249 | 0.0067% |
-| `klasyfikacji` | 134,093 | 0.0019% |
-| `ustawa` | 598,176 | 0.0087% |
-| `artykuł` | 2,096,271 | 0.0305% |
-| `parlament` | 1,216,808 | 0.0177% |
-| `rozporządzenie` | 1,497,273 | 0.0218% |
-| `w pobliżu` | 89,774 | 0.0013% |
-| `mieszkańców` | 288,911 | 0.0042% |
-| `Dz.U.` | 944,565 | 0.0137% |
+| `w roku` | 435,541 | 0.0069% |
+| `klasyfikacji` | 129,977 | 0.0021% |
+| `ustawa` | 587,459 | 0.0094% |
+| `artykuł` | 2,036,359 | 0.0324% |
+| `parlament` | 1,201,641 | 0.0191% |
+| `rozporządzenie` | 1,490,464 | 0.0237% |
+| `w pobliżu` | 78,823 | 0.0013% |
+| `mieszkańców` | 241,200 | 0.0038% |
+| `Dz.U.` | 939,966 | 0.0150% |
 
 | source | pattern | count | per-token share |
 |---|---|---:|---:|
@@ -52,42 +52,6 @@
 | eurlex | `w pobliżu` | 6,088 | 0.00026% |
 | eurlex | `mieszkańców` | 9,441 | 0.00040% |
 | eurlex | `Dz.U.` | 915,707 | 0.03851% |
-| european_hplt_v3_pl | `w roku` | 21,067 | 0.00411% |
-| european_hplt_v3_pl | `klasyfikacji` | 3,576 | 0.00070% |
-| european_hplt_v3_pl | `ustawa` | 5,754 | 0.00112% |
-| european_hplt_v3_pl | `artykuł` | 56,295 | 0.01099% |
-| european_hplt_v3_pl | `parlament` | 10,618 | 0.00207% |
-| european_hplt_v3_pl | `rozporządzenie` | 2,661 | 0.00052% |
-| european_hplt_v3_pl | `w pobliżu` | 9,877 | 0.00193% |
-| european_hplt_v3_pl | `mieszkańców` | 33,939 | 0.00662% |
-| european_hplt_v3_pl | `Dz.U.` | 1,482 | 0.00029% |
-| global_voices | `w roku` | 160 | 0.00431% |
-| global_voices | `klasyfikacji` | 5 | 0.00013% |
-| global_voices | `ustawa` | 157 | 0.00423% |
-| global_voices | `artykuł` | 1,047 | 0.02818% |
-| global_voices | `parlament` | 366 | 0.00985% |
-| global_voices | `rozporządzenie` | 26 | 0.00070% |
-| global_voices | `w pobliżu` | 105 | 0.00283% |
-| global_voices | `mieszkańców` | 508 | 0.01367% |
-| global_voices | `Dz.U.` | 1 | 0.00003% |
-| govpl | `w roku` | 3,372 | 0.00418% |
-| govpl | `klasyfikacji` | 526 | 0.00065% |
-| govpl | `ustawa` | 4,727 | 0.00586% |
-| govpl | `artykuł` | 2,417 | 0.00300% |
-| govpl | `parlament` | 3,967 | 0.00492% |
-| govpl | `rozporządzenie` | 4,102 | 0.00509% |
-| govpl | `w pobliżu` | 892 | 0.00111% |
-| govpl | `mieszkańców` | 13,105 | 0.01625% |
-| govpl | `Dz.U.` | 3,114 | 0.00386% |
-| nkjp1m | `w roku` | 109 | 0.00419% |
-| nkjp1m | `klasyfikacji` | 9 | 0.00035% |
-| nkjp1m | `ustawa` | 79 | 0.00304% |
-| nkjp1m | `artykuł` | 153 | 0.00588% |
-| nkjp1m | `parlament` | 216 | 0.00830% |
-| nkjp1m | `rozporządzenie` | 20 | 0.00077% |
-| nkjp1m | `w pobliżu` | 77 | 0.00296% |
-| nkjp1m | `mieszkańców` | 159 | 0.00611% |
-| nkjp1m | `Dz.U.` | 2 | 0.00008% |
 | parliamentary | `w roku` | 198,192 | 0.01203% |
 | parliamentary | `klasyfikacji` | 12,637 | 0.00077% |
 | parliamentary | `ustawa` | 459,179 | 0.02788% |

src/clean_hplt_v3.py

@@ -1,202 +0,0 @@
-#!/usr/bin/env python3
-"""clean_hplt_v3.py — filtr Polish HPLT v3 bin-10 (WDS=10) -> kanoniczna schema DynaWord.
-
-Adaptacja src/filter_european_hplt.py maintainera do SUROWEGO formatu HPLT v3 (jsonl.zst,
-pobrany z data.hplt-project.org). Te same progi jakosci; mapowanie pol v3 -> pol v1.
-Output: kanoniczna 8-kolumnowa schema (id/text/source/added/created/token_count/license/author)
-+ stats.json (format DynaWord + drop-reasons + register/MT/domain) + karta .md.
-
-Pola v3: text, lang[list], prob[list], u(url), doc_scores[list], web-register{dict}, c(mime), id.
-"""
-from __future__ import annotations
-import argparse, json, os, re, sys, time
-from pathlib import Path
-from urllib.parse import urlparse
-import zstandard as zstd
-import pyarrow as pa, pyarrow.parquet as pq
-import tiktoken
-
-ADDED = "2026-07-14"
-SOURCE = "european_hplt_v3_pl"
-LICENSE = "CC0-1.0"
-
-BOILERPLATE_RE = re.compile(
-    r"(cookies?|privacy policy|terms of service|all rights reserved|"
-    r"polityka prywatno\u015bci|regulamin|wszelkie prawa zastrze\u017cone|"
-    r"zaloguj|rejestracja|newsletter)", re.I)
-ADULT_SPAM_RE = re.compile(
-    r"(porn|sex|escort|casino|viagra|cialis|bukmacher|hazard|"
-    r"porno|seks|erotycz|kasyno|randki)", re.I)
-LEGALISH_RE = re.compile(
-    r"\b(ustawa|rozporz\u0105dzenie|dz\.u\.|sejm|senat|parlament|eur-lex|"
-    r"trybuna\u0142|s\u0105d|wyrok|kodeks|komisja europejska)\b", re.I)
-EXCLUDED_DOMAIN_PARTS = {".bip.", "bip.", "docplayer.pl", "mapa-kodow-pocztowych.pl",
-                         "slideplayer.pl", "wikipedia.org", "wikisource.org", "wikibooks.org",
-                         "wikiquote.org", "wikivoyage.org", "chomikuj.pl", "scribd.com",
-                         "pdfcoffee.com", "dokumen.pub", "docer.pl", "ebookpoint", "wolnelektury.pl"}
-# mojibake: polskie znaki UTF-8 zdekodowane jako CJK/inne (upstream HPLT extraction bug)
-MOJIBAKE_RE = re.compile(r"[\u2e00-\u9fff\uff00-\uffef\ufffd]")
-# naglowki serwisow wymiany plikow / PDF-ripow (ryzyko piractwa)
-PDFHOST_RE = re.compile(r"(\d+\s*Pages?\s*[\u2022\u00b7]|Uploaded at|\d+\s*Words?\s*[\u2022\u00b7])", re.I)
-MAX_CHARS = 120000  # ~40k tok: powyzej to ksiazki (copyright/OCR-risk), nie strony web
-ALLOWED_REGISTERS = {"ID", "OP", "HI", "IN", "NA"}
-
-CANON = pa.schema([("id", pa.string()), ("text", pa.string()), ("source", pa.string()),
-                   ("added", pa.string()), ("created", pa.string()),
-                   ("token_count", pa.int64()), ("license", pa.string()), ("author", pa.string())])
-
-ENC = tiktoken.get_encoding("cl100k_base")
-
-def domain(url):
-    if not url:
-        return ""
-    h = urlparse(url).netloc.lower()
-    return h[4:] if h.startswith("www.") else h
-
-def top_register(reg):
-    if not isinstance(reg, dict) or not reg:
-        return "", 0.0
-    k, v = max(reg.items(), key=lambda kv: kv[1] or 0.0)
-    return str(k), float(v or 0.0)
-
-def keep(o, min_chars, min_words, min_lang_prob, max_mt, dom_counts, max_per_domain):
-    lang = o.get("lang")
-    if not (isinstance(lang, list) and lang and lang[0] == "pol_Latn"):
-        return None, "language"
-    prob = o.get("prob")
-    lp = float(prob[0]) if isinstance(prob, list) and prob else 0.0
-    if lp < min_lang_prob:
-        return None, "lang_prob"
-    host = domain(o.get("u"))
-    if any(p in host for p in EXCLUDED_DOMAIN_PARTS):
-        return None, "domain"
-    if max_per_domain and dom_counts.get(host, 0) >= max_per_domain:
-        return None, "domain_cap"
-    reg = o.get("web-register") or {}
-    mt = float(reg.get("MT", 0.0)) if isinstance(reg, dict) else 0.0
-    if mt >= max_mt:
-        return None, "machine_translated"
-    rtop, _ = top_register(reg)
-    if ALLOWED_REGISTERS and rtop not in ALLOWED_REGISTERS:
-        return None, "register"
-    text = (o.get("text") or "").strip()
-    if len(text) < min_chars:
-        return None, "short"
-    if len(text.split()) < min_words:
-        return None, "few_words"
-    if len(text) > MAX_CHARS:
-        return None, "too_long"
-    if PDFHOST_RE.search(text[:200]):
-        return None, "filehost"
-    if len(MOJIBAKE_RE.findall(text[:3000])) >= 3:
-        return None, "mojibake"
-    s = text[:5000]
-    if len(BOILERPLATE_RE.findall(s)) >= 3:
-        return None, "boilerplate"
-    if ADULT_SPAM_RE.search(s):
-        return None, "adult_spam"
-    if LEGALISH_RE.search(s):
-        return None, "legalish"
-    return text, "kept"
-
-def open_input(path):
-    """Local file or URL (stream) -> binary readable context manager."""
-    if str(path).startswith("http"):
-        import urllib.request
-        return urllib.request.urlopen(urllib.request.Request(path, headers={"User-Agent": "slayer-dsbench/1.0"}), timeout=120)
-    return open(path, "rb")
-
-def main():
-    ap = argparse.ArgumentParser()
-    ap.add_argument("--in", dest="inp", default=r"C:\ProjektyPublic\datasets\hplt_v3_pl\10_1.jsonl.zst")
-    ap.add_argument("--out-dir", default=r"C:\ProjektyPublic\datasets\hplt_v3_pl\data\european_hplt_v3_pl")
-    ap.add_argument("--source", default="european_hplt_v3_pl", help="wartosc kolumny source")
-    ap.add_argument("--out-name", default="", help="basename plikow (domyslnie = source); id prefix")
-    ap.add_argument("--bin-label", default="WDS=10")
-    ap.add_argument("--min-chars", type=int, default=400)
-    ap.add_argument("--min-words", type=int, default=80)
-    ap.add_argument("--min-lang-prob", type=float, default=0.80)
-    ap.add_argument("--max-mt-prob", type=float, default=0.20)
-    ap.add_argument("--max-per-domain", type=int, default=250)
-    ap.add_argument("--max-records", type=int, default=0)
-    a = ap.parse_args()
-    src = a.source; out = a.out_name or a.source
-    outd = Path(a.out_dir); outd.mkdir(parents=True, exist_ok=True)
-    t0 = time.time()
-
-    drops = {}; regs = {}; doms = {}; dom_counts = {}
-    ids, texts, tokens = [], [], []
-    read = kept = chars = toks = 0
-    mt_sum = 0.0
-    writer = pq.ParquetWriter(outd / f"{out}.parquet", CANON, compression="zstd")
-
-    def flush():
-        nonlocal ids, texts, tokens
-        if not ids:
-            return
-        n = len(ids)
-        writer.write_table(pa.table({
-            "id": ids, "text": texts, "source": [src]*n, "added": [ADDED]*n,
-            "created": [""]*n, "token_count": tokens, "license": [LICENSE]*n, "author": [""]*n},
-            schema=CANON))
-        ids, texts, tokens = [], [], []
-
-    with open_input(a.inp) as fh:
-        reader = zstd.ZstdDecompressor().stream_reader(fh)
-        buf = b""
-        while True:
-            chunk = reader.read(1 << 20)
-            if not chunk:
-                break
-            buf += chunk
-            while b"\n" in buf:
-                line, buf = buf.split(b"\n", 1)
-                if not line.strip():
-                    continue
-                try:
-                    o = json.loads(line)
-                except Exception:
-                    continue
-                read += 1
-                text, reason = keep(o, a.min_chars, a.min_words, a.min_lang_prob,
-                                    a.max_mt_prob, dom_counts, a.max_per_domain)
-                if text is None:
-                    drops[reason] = drops.get(reason, 0) + 1
-                    continue
-                host = domain(o.get("u")); dom_counts[host] = dom_counts.get(host, 0) + 1
-                if len(doms) < 500:
-                    doms[host] = doms.get(host, 0) + 1
-                reg = o.get("web-register") or {}
-                rtop, _ = top_register(reg); regs[rtop] = regs.get(rtop, 0) + 1
-                mt_sum += float(reg.get("MT", 0.0)) if isinstance(reg, dict) else 0.0
-                tk = len(ENC.encode(text, disallowed_special=()))
-                ids.append(f"{out}_{kept}"); texts.append(text); tokens.append(tk)
-                kept += 1; chars += len(text); toks += tk
-                if len(ids) >= 1000:
-                    flush()
-                if read % 20000 == 0:
-                    print(f"  read={read:,} kept={kept:,} tok={toks:,} ({time.time()-t0:.0f}s)", flush=True)
-                if a.max_records and kept >= a.max_records:
-                    buf = b""; break
-            else:
-                continue
-            break
-    flush(); writer.close()
-
-    stats = {"read": read, "kept": kept, "drop_short": drops.get("short", 0),
-             "drop_lang": drops.get("language", 0) + drops.get("lang_prob", 0),
-             "drop_dup": 0, "drop_ocr": 0, "chars": chars, "tokens": toks,
-             "licenses": {LICENSE: kept}, "authors_with_value": 0, "license": LICENSE,
-             "stats_recomputed_from_parquet": True,
-             "drop_detail": drops, "registers": regs,
-             "mt_prob_mean_kept": round(mt_sum / max(1, kept), 3),
-             "domains_top_sample": dict(sorted(doms.items(), key=lambda x: -x[1])[:30]),
-             "secs": round(time.time()-t0, 1),
-             "source_repo": f"HPLT/HPLT3.0 pol_Latn {a.bin_label} via {a.inp}"}
-    (outd / f"{out}.stats.json").write_text(json.dumps(stats, ensure_ascii=False, indent=2)+"\n", encoding="utf-8")
-    print("=== STATS ===")
-    print(json.dumps({k: v for k, v in stats.items() if k not in ("domains_top_sample",)}, ensure_ascii=False, indent=2))
-    print(f"parquet={outd/(out+'.parquet')} ({os.path.getsize(outd/(out+'.parquet')):,}B)")
-
-if __name__ == "__main__":
-    main()

src/discover_govpl.py

@@ -1,194 +0,0 @@
-#!/usr/bin/env python3
-"""Discover gov.pl subsites + their news-section paths and write a committed
-manifest (govpl_subsites.json) that fetch_govpl.py consumes.
-
-gov.pl offers no sitemap or API for enumeration, and news-section slugs are
-non-uniform: ministries use `/aktualnosci`, but voivodeship/agency subsites use
-`/wiadomosci2`, `/aktualnosci10`, `/wszystkie-aktualnosci`, etc. (gov.pl appends
-numeric suffixes to disambiguate repeated slugs). So discovery is a separate,
-auditable step — freeze the JS-rendered subsite list, then probe each landing
-page for its deepest-paginating news section. fetch_govpl.py then reads the
-manifest with plain stdlib (no browser), keeping the fetch reproducible.
-
-SEED_SLUGS was captured from the rendered catalog at
-https://www.gov.pl/web/gov/katalog-jednostek on 2026-07-18 (a client-side list,
-NOT fetchable over plain HTTP). To refresh, open that page in a browser console
-and run:
-    [...new Set([...document.querySelectorAll('a[href*="/web/"]')]
-      .map(a => (a.getAttribute('href').match(/\\/web\\/([a-z0-9-]+)/i)||[])[1])
-      .filter(Boolean))].filter(s => s !== 'gov')
-then paste the array into SEED_SLUGS below.
-
-All gov.pl text is CC-BY-SA 4.0 (site footer), government-authored/hosted.
-
-Usage:
-  python3 src/discover_govpl.py --out src/govpl_subsites.json
-"""
-from __future__ import annotations
-import argparse, http.client, json, re, time
-from datetime import date
-from html import unescape
-from urllib.request import urlopen, Request
-from urllib.error import URLError, HTTPError
-
-BASE = "https://www.gov.pl"
-UA = {"User-Agent": "polish-dynaword/0.1 (+research; openly-licensed corpus)",
-      "Accept": "text/html"}
-
-# Frozen JS-rendered catalog (see module docstring for refresh recipe).
-SEED_SLUGS = [
-    "premier", "aktywa-panstwowe", "cyfryzacja", "edukacja", "energia",
-    "finanse", "fundusze-regiony", "infrastruktura", "klimat", "kultura",
-    "nauka", "obrona-narodowa", "rodzina", "rolnictwo", "rozwoj-technologia",
-    "sport", "sprawiedliwosc", "mswia", "dyplomacja", "zdrowie", "zermswia",
-    "arimr", "chemikalia", "rpp", "gdos", "gif", "piorin", "gios", "gis",
-    "gitd", "gugik", "krus", "paa", "rcb", "udsc", "prokuratoria", "urpl",
-    "uzp", "uw-kujawsko-pomorski", "uw-lubelski", "uw-lubuski", "uw-lodzki",
-    "uw-mazowiecki", "uw-opolski", "uw-podkarpacki", "uw-podlaski",
-    "uw-warminsko-mazurski", "uw-zachodniopomorski", "gddkia-bialystok",
-    "gddkia-bydgoszcz", "gddkia-gdansk", "gddkia-katowice", "gddkia-kielce",
-    "gddkia-krakow", "gddkia-lublin", "gddkia-lodz", "gddkia-olsztyn",
-    "gddkia-opole", "gddkia-poznan", "gddkia-rzeszow", "gddkia-szczecin",
-    "gddkia-warszawa", "gddkia-wroclaw", "gddkia-zielona-gora", "wsse-bialystok",
-    "wsse-bydgoszcz", "wsse-gdansk", "wsse-gorzowwlkp", "wsse-katowice",
-    "wsse-kielce", "wsse-krakow", "wsse-lublin", "wsse-lodz", "wsse-olsztyn",
-    "wsse-opole", "wsse-poznan", "wsse-rzeszow", "wsse-szczecin",
-    "wsse-warszawa", "wsse-wroclaw", "prokuratura-krajowa", "po-bialystok",
-    "po-bielsko-biala", "po-bydgoszcz", "po-elblag", "po-gdansk", "po-gliwice",
-    "po-gorzow-wielkopolski", "po-katowice", "po-kielce", "po-krakow",
-    "po-legnica", "po-lublin", "po-lodz", "po-lomza", "po-nowy-sacz",
-    "po-olsztyn", "po-opole", "po-ostroleka", "po-ostrow-wielkopolski",
-    "po-piotrkow-trybunalski", "po-plock", "po-poznan", "po-rzeszow",
-    "po-siedlce", "po-sieradz", "po-slupsk", "po-sosnowiec", "po-swidnica",
-    "po-tarnobrzeg", "po-tarnow", "po-torun", "po-warszawa", "po-warszawa-praga",
-    "po-wloclawek", "po-wroclaw", "po-zielona-gora", "pr-gdansk", "pr-katowice",
-    "pr-krakow", "pr-lublin", "pr-lodz", "pr-poznan", "pr-rzeszow",
-    "pr-szczecin", "pr-warszawa", "pr-wroclaw", "kgpsp", "kwpsp-gdansk",
-    "kwpsp-opole", "kwpsp-poznan", "cskmswia", "spzoz-mswia-katowice",
-    "spzoz-mswia-lublin", "spzoz-mswia-opole", "spzoz-mswia-szczecin",
-    "spzoz-mswia-zielona-gora", "spzoz-mswia-glucholazy", "krrit", "kowr",
-    "kzn", "nck", "wody-polskie",
-]
-
-# A landing-page sub-page counts as a news section if its slug starts with one
-# of these (gov.pl appends -suffix / digits, e.g. aktualnosci10, wiadomosci2).
-_NEWS_PREFIX = re.compile(r"^(wszystkie-aktualnosci|aktualnosci|wiadomosci|"
-                          r"serwis-prasowy|komunikaty)", re.I)
-# On a listing page each article preview is an <span class="date"> ... </span>
-# (inside a class="event" block) followed by <div class="title"><a href=...>.
-# Binding the two in one match keeps date/title/url aligned per article.
-_ARTICLE = re.compile(
-    r'<span class="date">\s*([\d.]+)\s*</span>.*?'
-    r'<div class="title">\s*<a href="([^"]+)"[^>]*>\s*(.*?)\s*</a>',
-    re.I | re.S)
-_PAGE = re.compile(r'[?&]page=(\d+)')
-
-
-def http_get_url(url: str, tries: int = 4) -> tuple[str | None, str]:
-    """GET, returning (text, final_url_after_redirects). text is None on 404 or
-    exhausted retries. Needed because gov.pl news aliases (e.g. /aktualnosci)
-    302-redirect to a canonical feed (/wydarzenia) and DROP the ?page= param —
-    so pagination must run against the resolved canonical URL, not the alias."""
-    for i in range(tries):
-        try:
-            with urlopen(Request(url, headers=UA), timeout=60) as r:
-                return r.read().decode("utf-8", "replace"), r.geturl()
-        # OSError covers URLError/TimeoutError/socket/ssl/ConnectionReset;
-        # HTTPException covers IncompleteRead (truncated body) — all transient, retry.
-        except (HTTPError, OSError, http.client.HTTPException) as e:
-            if isinstance(e, HTTPError) and e.code == 404:
-                return None, url
-            time.sleep(1.5 * (i + 1))
-    return None, url
-
-
-def http_get(url: str, tries: int = 4) -> str | None:
-    """GET text, retrying transient errors; None on 404 or exhausted retries."""
-    return http_get_url(url, tries)[0]
-
-
-def listing_articles(html: str, slug: str) -> list[dict]:
-    """{url,title,date} for real articles on a listing page of subsite `slug`."""
-    out = []
-    prefix = f"/web/{slug}/"
-    for date_str, href, title in _ARTICLE.findall(html):
-        if href.startswith(prefix):
-            out.append({
-                "url": href,
-                "title": unescape(re.sub(r"\s+", " ", title)).strip(),
-                "date": date_str,
-            })
-    return out
-
-
-def max_page(html: str) -> int:
-    nums = [int(n) for n in _PAGE.findall(html)]
-    return max(nums) if nums else 1
-
-
-def news_candidates(landing_html: str, slug: str) -> list[str]:
-    """Distinct same-subsite sub-slugs that look like a news section."""
-    seen = {}
-    for sub in re.findall(rf'href="/web/{re.escape(slug)}/([a-z0-9-]+)"', landing_html):
-        if _NEWS_PREFIX.match(sub):
-            seen[sub] = True
-    return list(seen)
-
-
-def probe_subsite(slug: str) -> dict | None:
-    """Resolve a subsite's best news section -> {slug, news_path, pages} or None."""
-    landing = http_get(f"{BASE}/web/{slug}")
-    if not landing:
-        return None
-    best = None
-    for sub in news_candidates(landing, slug):
-        html = http_get(f"{BASE}/web/{slug}/{sub}")
-        if not html:
-            continue
-        n_art = len(listing_articles(html, slug))
-        if n_art == 0:
-            continue
-        pages = max_page(html)
-        # Prefer the deepest-paginating feed; tie-break the shorter (more general) slug.
-        key = (pages, -len(sub))
-        if best is None or key > best[0]:
-            best = (key, sub, pages)
-        time.sleep(0.2)
-    if best is None:
-        return None
-    return {"slug": slug, "news_path": best[1], "pages": best[2]}
-
-
-def main():
-    ap = argparse.ArgumentParser()
-    ap.add_argument("--out", default="src/govpl_subsites.json")
-    ap.add_argument("--slugs", nargs="*", help="override SEED_SLUGS (for testing)")
-    args = ap.parse_args()
-
-    slugs = args.slugs or SEED_SLUGS
-    manifest, skipped = [], []
-    for i, slug in enumerate(slugs, 1):
-        rec = probe_subsite(slug)
-        if rec:
-            manifest.append(rec)
-            print(f"[{i}/{len(slugs)}] {slug} -> {rec['news_path']} ({rec['pages']}p)", flush=True)
-        else:
-            skipped.append(slug)
-            print(f"[{i}/{len(slugs)}] {slug} -> NO NEWS SECTION (skipped)", flush=True)
-
-    total_pages = sum(r["pages"] for r in manifest)
-    out = {
-        "generated": date.today().isoformat(),
-        "source": f"{BASE}/web/gov/katalog-jednostek",
-        "subsites": manifest,
-        "skipped": skipped,  # logged, not silently dropped
-        "total_pages": total_pages,
-    }
-    with open(args.out, "w", encoding="utf-8") as f:
-        json.dump(out, f, ensure_ascii=False, indent=2)
-    print(f"\nwrote {args.out}: {len(manifest)} subsites, {total_pages} listing pages, "
-          f"{len(skipped)} skipped")
-
-
-if __name__ == "__main__":
-    main()

src/fetch_govpl.py

@@ -1,183 +0,0 @@
-#!/usr/bin/env python3
-"""Fetch gov.pl ministry/agency press releases into a SpeakLeash-style
-.jsonl.zst, so build_dynaword.py consumes it unchanged (native source, not via
-SpeakLeash). All textual content on gov.pl is published under CC-BY-SA 4.0 (site
-footer: "Treści tekstowe publikowane w serwisie … CC BY-SA 4.0"), authored and
-hosted by the Polish government — so the redistributor genuinely holds the right
-to license it.
-
-Enumeration is driven by the committed manifest src/govpl_subsites.json (built
-by discover_govpl.py): {slug, news_path, pages} per subsite. This keeps the
-fetch pure-stdlib and reproducible — no headless browser at fetch time.
-
-Resumable: appends to <out>/govpl.jsonl, skips article URLs already present,
-then compresses to govpl.jsonl.zst at the end of a run. Per-request failures are
-retried then logged and skipped (never silently truncating the crawl).
-
-Usage:
-  python3 src/discover_govpl.py --out src/govpl_subsites.json   # once / to refresh
-  python3 src/fetch_govpl.py --out ~/speakleash --manifest src/govpl_subsites.json
-"""
-from __future__ import annotations
-import argparse, json, re, subprocess, sys, threading, time
-from concurrent.futures import ThreadPoolExecutor
-from html import unescape
-from pathlib import Path
-
-sys.path.insert(0, str(Path(__file__).resolve().parent))
-from discover_govpl import BASE, http_get, http_get_url, listing_articles  # shared, DRY
-
-KEY = "govpl"  # build_dynaword reads <KEY>.jsonl.zst
-MIN_CHARS = 200  # same floor as build_dynaword; skip stubs early
-
-_SCRIPT = re.compile(r"(?is)<(script|style)[^>]*>.*?</\1>")
-_BLOCK_END = re.compile(r"(?i)</(p|div|h[1-6]|li|tr|table|ul|ol|blockquote)\s*>|<br\s*/?>")
-_TAG = re.compile(r"<[^>]+>")
-# Body ends where the editor-content region gives way to gallery/attachments/tags.
-_END_MARKER = re.compile(
-    r'<[^>]+class="[^"]*(?:attachments|art-tags|tags|social|share|gallery|files)[^"]*"'
-    r'|data-group="gallery"|</article', re.I)
-_EDITOR = re.compile(r'<div class="editor-content"[^>]*>', re.I)
-
-
-def html_to_text(html: str) -> str:
-    html = _SCRIPT.sub("", html)
-    html = _BLOCK_END.sub("\n", html)
-    html = _TAG.sub("", html)
-    text = unescape(html)
-    text = re.sub(r"[ \t]+", " ", text)
-    text = re.sub(r" *\n *", "\n", text)
-    text = re.sub(r"\n{3,}", "\n\n", text)
-    return text.strip()
-
-
-def article_body(html: str) -> str:
-    """Text of the article body. gov.pl pages carry the class twice; pick the
-    richest editor-content region and cut at the trailing gallery/attachments.
-    ponytail: regex slice, not a DOM parse — end-markers drop gallery captions;
-    upgrade to an HTML parser only if boilerplate leakage shows up in stats."""
-    best = ""
-    for m in _EDITOR.finditer(html):
-        start = m.end()
-        end = _END_MARKER.search(html, start)
-        chunk = html[start:end.start()] if end else html[start:]
-        text = html_to_text(chunk)
-        if len(text) > len(best):
-            best = text
-    return best
-
-
-def normalize(art: dict, html: str, slug: str) -> dict | None:
-    """Listing entry {url,title,date} + article HTML -> record, or None if thin."""
-    body = article_body(html)
-    if len(body) < MIN_CHARS:
-        return None
-    title = art.get("title") or ""
-    text = f"{title}\n\n{body}" if title else body
-    return {
-        "text": text,
-        "meta": {
-            "url": BASE + art["url"],
-            "subsite": slug,
-            "title": title,
-            "date": art.get("date") or "",
-        },
-    }
-
-
-def iter_listing(slug: str, news_path: str, pages: int):
-    """Yield {url,title,date} article entries across a subsite's listing pages.
-
-    Resolves the news section's canonical URL first: gov.pl aliases (e.g.
-    /aktualnosci) redirect to a canonical feed (/wydarzenia) and drop ?page=,
-    so paginating the alias returns page 1 forever. We paginate the resolved
-    base with ?page=N&size=10 (the exact form gov.pl's own pager emits)."""
-    first_html, canonical = http_get_url(f"{BASE}/web/{slug}/{news_path}")
-    if first_html is None:
-        print(f"  WARN news section failed, skipped: {slug}/{news_path}",
-              file=sys.stderr, flush=True)
-        return
-    base = canonical.split("?", 1)[0]  # canonical path, no query
-    for page in range(1, pages + 1):
-        html = http_get(f"{base}?page={page}&size=10")
-        if html is None:
-            print(f"  WARN listing failed, skipped: {base}?page={page}",
-                  file=sys.stderr, flush=True)
-            continue
-        yield from listing_articles(html, slug)
-
-
-def fetch_subsite(si: int, n_sub: int, sub: dict, done: set, lock: threading.Lock,
-                  fo, st: dict) -> None:
-    """Crawl one subsite. Runs in a worker thread: the slow article GET stays
-    outside the lock so workers actually overlap; only shared state (done set,
-    file, counters) is lock-guarded."""
-    slug, news_path, pages = sub["slug"], sub["news_path"], sub["pages"]
-    print(f"[{si}/{n_sub}] {slug}/{news_path} ({pages}p)", flush=True)
-    try:
-        for art in iter_listing(slug, news_path, pages):
-            url = BASE + art["url"]
-            with lock:
-                if url in done:
-                    continue
-                done.add(url)
-                st["seen"] += 1
-                seen = st["seen"]
-            html = http_get(url)  # network — deliberately outside the lock
-            if html is None:
-                with lock: st["skipped"] += 1
-                print(f"  WARN article failed, skipped: {url}", file=sys.stderr, flush=True)
-                continue
-            rec = normalize(art, html, slug)
-            line = json.dumps(rec, ensure_ascii=False) + "\n" if rec else None
-            with lock:
-                if line:
-                    fo.write(line); st["kept"] += 1
-                if seen % 500 == 0:
-                    print(f"  seen {st['seen']:,} | kept {st['kept']:,} | skipped {st['skipped']:,} "
-                          f"| {st['seen']/(time.time()-st['t0']):.1f}/s", flush=True)
-    except Exception as e:
-        # A subsite that dies must not abort the whole crawl (ex.map would
-        # propagate and kill every worker). Log, skip the rest, keep going.
-        print(f"  WARN subsite crashed, skipped rest: {slug}: {e!r}",
-              file=sys.stderr, flush=True)
-
-
-def main():
-    ap = argparse.ArgumentParser()
-    ap.add_argument("--out", default="~/speakleash")
-    ap.add_argument("--manifest", default="src/govpl_subsites.json")
-    ap.add_argument("--workers", type=int, default=2,
-                    help="concurrent subsite crawlers (be polite to gov.pl)")
-    args = ap.parse_args()
-
-    manifest = json.loads(Path(args.manifest).read_text(encoding="utf-8"))
-    subsites = manifest["subsites"]
-
-    out_dir = Path(args.out).expanduser(); out_dir.mkdir(parents=True, exist_ok=True)
-    jsonl = out_dir / f"{KEY}.jsonl"
-    done = set()
-    if jsonl.exists():
-        for ln in jsonl.open(encoding="utf-8"):
-            try: done.add(json.loads(ln)["meta"]["url"])
-            except Exception: pass
-        print(f"resume: {len(done):,} already fetched")
-
-    st = {"seen": 0, "kept": 0, "skipped": 0, "t0": time.time()}
-    lock = threading.Lock()
-    with jsonl.open("a", encoding="utf-8") as fo:
-        with ThreadPoolExecutor(max_workers=args.workers) as ex:
-            # list() drains the map so any worker exception surfaces here.
-            list(ex.map(
-                lambda item: fetch_subsite(item[0], len(subsites), item[1],
-                                           done, lock, fo, st),
-                enumerate(subsites, 1)))
-
-    print(f"fetched {st['kept']:,} articles ({st['skipped']:,} skipped); compressing...", flush=True)
-    subprocess.run(["zstd", "-19", "-f", "--rm", str(jsonl), "-o",
-                    str(out_dir / f"{KEY}.jsonl.zst")], check=True)
-    print(f"wrote {out_dir / (KEY + '.jsonl.zst')} in {round(time.time()-st['t0'])}s")
-
-
-if __name__ == "__main__":
-    main()

src/fetch_njkp.py

@@ -1,342 +0,0 @@
-#!/usr/bin/env python3
-"""Fetch the NKJP 1-million-word subcorpus (Podkorpus Milionowy NKJP 1.2) and
-build a DynaWord-style parquet shard from the TEI source.
-
-Downloads + extracts the IPI PAN tarball under /tmp, then walks every sample
-folder and emits one parquet row per <div> in text.xml. Each <div> is a single
-contiguous excerpt from one source document (its <ab> paragraphs joined by
-newlines, ellipses stripped); across <div>s the excerpts are unrelated, so they
-are kept as separate rows rather than merged.
-
-The same minimal gates as src/build_dynaword.py are applied per passage (drop
-< 200 chars, drop non-Polish by diacritic ratio, exact sha1 dedup; the OCR gate
-is inapplicable - NKJP1M is not OCR), so the shard matches what the DynaWord
-build would keep. Per-row `created` is the sample's TEI publication date
-(<date type="published">) when the header records one. Token counts use
-tiktoken cl100k (encode_ordinary), same as build_dynaword.py, and the emitted
-stats are recomputed from the written parquet.
-
-Text extraction + the stats report are adapted from tmp/count_tokens.py.
-
-Usage:
-  python3 src/fetch_njkp.py --out . --workers 8
-"""
-from __future__ import annotations
-
-import argparse
-import hashlib
-import json
-import re
-import ssl
-import sys
-import tarfile
-import time
-from concurrent.futures import ProcessPoolExecutor
-from datetime import date
-from pathlib import Path
-from urllib.error import URLError
-from urllib.request import urlopen, Request
-from xml.etree import ElementTree as ET
-
-import pyarrow as pa
-import pyarrow.parquet as pq
-import tiktoken
-
-URL = ("https://clip.ipipan.waw.pl/NationalCorpusOfPolish"
-       "?action=AttachFile&do=get&target=NKJP-PodkorpusMilionowy-1.2.tar.gz")
-ROOT_NAME = "NKJP-PodkorpusMilionowy-1.2"
-UA = {"User-Agent": "polish-dynaword/0.1 (+research; openly-licensed corpus)"}
-
-KEY = "nkjp1m"          # source name / parquet stem, consumed by build_dynaword
-LICENSE = "CC-BY"       # stated on the NKJP download page (clip.ipipan.waw.pl)
-AUTHOR = "NKJP"         # compiled/distributed by the NKJP Consortium (IPI PAN)
-
-# Gates - identical thresholds to src/build_dynaword.py so the shard matches
-# what the DynaWord build itself would keep.
-MIN_CHARS = 200
-MIN_POLISH_RATIO = 0.005
-POLISH_RE = re.compile(r"[ąćęłńóśźżĄĆĘŁŃÓŚŹŻ]")
-ALPHA_RE = re.compile(r"[^\W\d_]", re.UNICODE)
-
-NS = "{http://www.tei-c.org/ns/1.0}"
-DIV_TAG, AB_TAG, DATE_TAG = f"{NS}div", f"{NS}ab", f"{NS}date"
-_ELLIPSIS = re.compile(r"…|\.{3,}")  # ellipsis markers stripped from text
-
-# Same schema as src/build_dynaword.py so the shard drops straight into DynaWord.
-SCHEMA = pa.schema([
-    ("id", pa.string()), ("text", pa.string()), ("source", pa.string()),
-    ("added", pa.string()), ("created", pa.string()), ("token_count", pa.int64()),
-    ("license", pa.string()), ("author", pa.string()),
-])
-
-_ENC = None  # per-worker tiktoken encoder
-
-
-def download(url: str, dst: Path) -> Path:
-    if dst.exists() and dst.stat().st_size:
-        print(f"  cached {dst}", flush=True)
-        return dst
-    print(f"  downloading {url}", flush=True)
-    dst.parent.mkdir(parents=True, exist_ok=True)
-
-    def stream(ctx):
-        with urlopen(Request(url, headers=UA), timeout=120, context=ctx) as r, \
-                dst.open("wb") as f:
-            while chunk := r.read(1 << 20):
-                f.write(chunk)
-
-    try:
-        stream(None)
-    except URLError as e:
-        # clip.ipipan.waw.pl ships an incomplete cert chain Python rejects (curl
-        # accepts it); retry unverified for this known source only.
-        if not isinstance(e.reason, ssl.SSLError):
-            raise
-        print("  ! TLS verify failed; retrying unverified", file=sys.stderr, flush=True)
-        stream(ssl._create_unverified_context())
-    return dst
-
-
-def extract_archive(archive: Path, dest: Path) -> Path:
-    # The tarball has no top folder, so extract into a dedicated dir to scope
-    # the text.xml scan.
-    root = dest / ROOT_NAME
-    if root.is_dir():
-        print(f"  already extracted {root}", flush=True)
-        return root
-    print(f"  extracting {archive} -> {root}", flush=True)
-    root.mkdir(parents=True)
-    with tarfile.open(archive, "r:gz") as tar:
-        tar.extractall(root, filter="data")  # py3.12+ safe extraction
-    return root
-
-
-def _clean(text: str) -> str:
-    """Strip ellipsis markers (… and "...") and tidy the whitespace they leave."""
-    return re.sub(r"[ \t]{2,}", " ", _ELLIPSIS.sub(" ", text)).strip()
-
-
-def _polish_ratio(text: str) -> float:
-    """Fraction of letters that are Polish-specific diacritics (build_dynaword)."""
-    letters = ALPHA_RE.findall(text)
-    return len(POLISH_RE.findall(text)) / len(letters) if letters else 0.0
-
-
-def _norm_date(raw: str) -> str:
-    """Normalize a TEI @when value to an ISO date or bare year, else "".
-
-    Keeps full YYYY-MM-DD, keeps bare YYYY, and salvages a leading 4-digit year
-    from anything else (e.g. "YYYY-MM", a stray trailing space). Values whose
-    year is implausible as a publication year (NKJP1M has a few malformed ones,
-    such as "200") are treated as missing.
-    """
-    raw = (raw or "").strip()
-    if re.fullmatch(r"\d{4}-\d{2}-\d{2}", raw):
-        return raw if 1500 <= int(raw[:4]) <= 2014 else ""
-    m = re.match(r"(\d{4})", raw)
-    return m.group(1) if m and 1500 <= int(m.group(1)) <= 2014 else ""
-
-
-def _published_date(header_path: Path) -> str:
-    """Publication date from a sample's TEI header (<date type='published'>)."""
-    try:
-        tree = ET.parse(header_path)
-    except Exception:  # noqa: BLE001 - missing/unreadable header -> no date
-        return ""
-    for el in tree.iter(DATE_TAG):
-        if el.get("type") == "published":
-            d = _norm_date(el.get("when") or (el.text or ""))
-            if d:
-                return d
-    return ""
-
-
-def extract_passages(xml_path: str) -> list[str]:
-    """One passage per TEI <div> (its <ab> paragraphs joined by newlines).
-
-    <ab> blocks inside a <div> are adjacent source paragraphs (real coherence);
-    different <div>s are unrelated sampled excerpts, so each becomes its own row.
-    Adapted from tmp/count_tokens.py, which instead merged every <ab> per file.
-    """
-    passages, current = [], []
-    for _, elem in ET.iterparse(xml_path, events=("end",)):
-        if elem.tag == AB_TAG:
-            # itertext() also captures any nested inline markup text.
-            text = _clean("".join(elem.itertext()))
-            if text:
-                current.append(text)
-            elem.clear()
-        elif elem.tag == DIV_TAG:
-            if current:
-                passages.append("\n".join(current))
-                current = []
-            elem.clear()
-    if current:  # <ab> outside any <div> (not expected) - keep as one passage
-        passages.append("\n".join(current))
-    return passages
-
-
-def process_file(xml_path: str) -> tuple[list[dict], list[int]]:
-    """Sample file -> (kept row dicts per surviving <div>, [read, short, lang]).
-
-    Gates run per passage; dedup is deferred to main() so it can span files.
-    """
-    global _ENC
-    if _ENC is None:
-        _ENC = tiktoken.get_encoding("cl100k_base")
-    folder_dir = Path(xml_path).parent
-    folder = folder_dir.name
-    created = _published_date(folder_dir / "header.xml")
-    try:
-        passages = extract_passages(xml_path)
-    except Exception:  # noqa: BLE001 - skip unreadable samples, don't crash
-        return [], [0, 0, 0]
-
-    read = short = lang = 0
-    kept = []
-    for i, text in enumerate(passages):
-        read += 1
-        text = text.strip()
-        if len(text) < MIN_CHARS:
-            short += 1
-            continue
-        if _polish_ratio(text) < MIN_POLISH_RATIO:
-            lang += 1
-            continue
-        kept.append({
-            "id": f"{KEY}_{folder}_{i}",   # passage index kept stable across gating
-            "text": text,
-            "created": created,
-            "tokens": len(_ENC.encode_ordinary(text)),
-            "chars": len(text),
-            "sha1": hashlib.sha1(text.encode("utf-8")).digest(),
-        })
-    return kept, [read, short, lang]
-
-
-def build_stats(parquet_path: Path, gate: dict) -> dict:
-    """Recompute the sidecar stats directly from the written parquet.
-
-    Counts (kept/chars/tokens/licenses/authors/dates) come from the bytes on
-    disk; the drop_* tallies are build-time artifacts carried over from `gate`.
-    A cross-check asserts read == kept + drops so the two can't silently drift.
-    """
-    t = pq.read_table(parquet_path)
-    d = t.to_pydict()
-    n = t.num_rows
-    dated = [c for c in d["created"] if c]
-    years = sorted({int(c[:4]) for c in dated})
-    read = gate["read"]
-    drops = gate["drop_short"] + gate["drop_lang"] + gate["drop_dup"] + gate["drop_ocr"]
-    assert n == read - drops, f"gate arithmetic: kept {n} != read {read} - drops {drops}"
-    return {
-        "read": read,
-        "kept": n,
-        "drop_short": gate["drop_short"],
-        "drop_lang": gate["drop_lang"],
-        "drop_dup": gate["drop_dup"],
-        "drop_ocr": gate["drop_ocr"],
-        "chars": sum(len(x) for x in d["text"]),
-        "tokens": sum(d["token_count"]),
-        "licenses": {LICENSE: n},
-        "authors_with_value": sum(1 for a in d["author"] if a),
-        "documents_with_created": len(dated),
-        "created_range": f"{years[0]}-{years[-1]}" if years else "",
-        "license": LICENSE,
-        "stats_recomputed_from_parquet": True,
-    }
-
-
-def report(rows: list[dict], stats: dict) -> None:
-    """Aggregate cl100k token stats. (adapted from count_tokens.py)"""
-    toks = sorted(r["tokens"] for r in rows)
-    total_tok, total_chars = sum(toks), sum(r["chars"] for r in rows)
-
-    def pct(p: float) -> int:
-        return toks[max(0, min(len(toks) - 1, round(p / 100 * (len(toks) - 1))))]
-
-    print("\n" + "=" * 60)
-    print("NKJP cl100k token statistics")
-    print("=" * 60)
-    print(f"Passages kept:      {len(rows):,}")
-    print(f"Total tokens:       {total_tok:,}")
-    print(f"Total characters:   {total_chars:,}")
-    if toks:
-        print(f"Mean tokens/pass.:  {total_tok / len(toks):,.1f}")
-        print(f"Median / p90 / max: {pct(50):,} / {pct(90):,} / {max(toks):,}")
-        if total_tok:
-            print(f"Chars per token:    {total_chars / total_tok:.2f}")
-    if stats.get("created_range"):
-        print(f"Created range:      {stats['created_range']} "
-              f"({stats['documents_with_created']:,}/{len(rows):,} dated)")
-
-
-def main() -> int:
-    ap = argparse.ArgumentParser(description=__doc__)
-    ap.add_argument("--out", default=".",
-                    help="DynaWord root; shard -> <out>/data/nkjp1m/nkjp1m.parquet")
-    ap.add_argument("--tmp", default="/tmp", help="Download + extraction dir")
-    ap.add_argument("--workers", type=int, default=None, help="Process pool size")
-    ap.add_argument("--added", default=date.today().isoformat(),
-                    help="Value for the 'added' column (default: today)")
-    args = ap.parse_args()
-
-    t0 = time.time()
-    tmp = Path(args.tmp).expanduser()
-    root = extract_archive(download(URL, tmp / f"{ROOT_NAME}.tar.gz"), tmp)
-
-    files = sorted(str(p) for p in root.rglob("text.xml"))
-    if not files:
-        print(f"error: no text.xml under {root}", file=sys.stderr)
-        return 1
-    print(f"Found {len(files)} text.xml files under {root}", flush=True)
-
-    read = short = lang = 0
-    collected = []  # kept row dicts, in sorted file + passage order (deterministic)
-    with ProcessPoolExecutor(max_workers=args.workers) as pool:
-        for i, (kept, st3) in enumerate(pool.map(process_file, files, chunksize=16), 1):
-            read += st3[0]; short += st3[1]; lang += st3[2]
-            collected.extend(kept)
-            if i % 1000 == 0 or i == len(files):
-                print(f"  processed {i}/{len(files)} files, {len(collected):,} kept passages",
-                      file=sys.stderr)
-
-    # Exact dedup (sha1), first-wins over the deterministic order above.
-    seen, rows, drop_dup = set(), [], 0
-    for r in collected:
-        if r["sha1"] in seen:
-            drop_dup += 1
-            continue
-        seen.add(r["sha1"])
-        rows.append(r)
-
-    n = len(rows)
-    out = Path(args.out).expanduser().resolve() / "data" / KEY / f"{KEY}.parquet"
-    out.parent.mkdir(parents=True, exist_ok=True)
-    pq.write_table(pa.table({
-        "id": [r["id"] for r in rows],
-        "text": [r["text"] for r in rows],
-        "source": [KEY] * n,
-        "added": [args.added] * n,
-        "created": [r["created"] for r in rows],
-        "token_count": [r["tokens"] for r in rows],
-        "license": [LICENSE] * n,
-        "author": [AUTHOR] * n,
-    }, schema=SCHEMA), out, compression="zstd")
-
-    # DynaWord-style sidecar stats, recomputed from the parquet just written.
-    gate = {"read": read, "drop_short": short, "drop_lang": lang,
-            "drop_dup": drop_dup, "drop_ocr": 0}
-    stats = build_stats(out, gate)
-    out.with_name(f"{KEY}.stats.json").write_text(
-        json.dumps(stats, ensure_ascii=False, indent=2) + "\n")
-
-    report(rows, stats)
-    print(f"\nWrote {n:,} passages from {len(files):,} files "
-          f"(read {read:,}, -short {short:,} -lang {lang:,} -dup {drop_dup:,}) "
-          f"-> {out} in {round(time.time() - t0)}s", flush=True)
-    return 0
-
-
-if __name__ == "__main__":
-    raise SystemExit(main())

src/govpl_subsites.json

@@ -1,684 +0,0 @@
-{
-  "generated": "2026-07-18",
-  "source": "https://www.gov.pl/web/gov/katalog-jednostek",
-  "subsites": [
-    {
-      "slug": "premier",
-      "news_path": "aktualnosci",
-      "pages": 294
-    },
-    {
-      "slug": "aktywa-panstwowe",
-      "news_path": "wiadomosci",
-      "pages": 87
-    },
-    {
-      "slug": "cyfryzacja",
-      "news_path": "wiadomosci",
-      "pages": 353
-    },
-    {
-      "slug": "edukacja",
-      "news_path": "wiadomosci",
-      "pages": 104
-    },
-    {
-      "slug": "energia",
-      "news_path": "wiadomosci",
-      "pages": 36
-    },
-    {
-      "slug": "finanse",
-      "news_path": "wiadomosci",
-      "pages": 266
-    },
-    {
-      "slug": "fundusze-regiony",
-      "news_path": "aktualnosci1",
-      "pages": 378
-    },
-    {
-      "slug": "infrastruktura",
-      "news_path": "aktualnosci",
-      "pages": 325
-    },
-    {
-      "slug": "klimat",
-      "news_path": "wiadomosci",
-      "pages": 289
-    },
-    {
-      "slug": "kultura",
-      "news_path": "aktualnosci",
-      "pages": 481
-    },
-    {
-      "slug": "nauka",
-      "news_path": "wiadomosci",
-      "pages": 78
-    },
-    {
-      "slug": "obrona-narodowa",
-      "news_path": "aktualnosci5",
-      "pages": 186
-    },
-    {
-      "slug": "rodzina",
-      "news_path": "wiadomosci-polityka-rodzinna",
-      "pages": 111
-    },
-    {
-      "slug": "rolnictwo",
-      "news_path": "wiadomosci",
-      "pages": 5
-    },
-    {
-      "slug": "rozwoj-technologia",
-      "news_path": "wiadomosci",
-      "pages": 52
-    },
-    {
-      "slug": "sport",
-      "news_path": "wiadomosci",
-      "pages": 219
-    },
-    {
-      "slug": "sprawiedliwosc",
-      "news_path": "aktualnosci-ogloszenia-i-informacje",
-      "pages": 285
-    },
-    {
-      "slug": "mswia",
-      "news_path": "aktualnosci",
-      "pages": 216
-    },
-    {
-      "slug": "dyplomacja",
-      "news_path": "aktualnosci",
-      "pages": 302
-    },
-    {
-      "slug": "zdrowie",
-      "news_path": "wiadomosci",
-      "pages": 145
-    },
-    {
-      "slug": "zermswia",
-      "news_path": "aktualnosci",
-      "pages": 12
-    },
-    {
-      "slug": "arimr",
-      "news_path": "aktualnosci10",
-      "pages": 6
-    },
-    {
-      "slug": "rpp",
-      "news_path": "aktualnosci",
-      "pages": 93
-    },
-    {
-      "slug": "gdos",
-      "news_path": "aktualnosci",
-      "pages": 57
-    },
-    {
-      "slug": "gif",
-      "news_path": "wiadomosci",
-      "pages": 37
-    },
-    {
-      "slug": "gios",
-      "news_path": "wiadomosci",
-      "pages": 53
-    },
-    {
-      "slug": "gis",
-      "news_path": "wiadomosci",
-      "pages": 49
-    },
-    {
-      "slug": "gitd",
-      "news_path": "komunikaty-btm2",
-      "pages": 27
-    },
-    {
-      "slug": "gugik",
-      "news_path": "wiadomosci",
-      "pages": 183
-    },
-    {
-      "slug": "krus",
-      "news_path": "aktualnosci",
-      "pages": 18
-    },
-    {
-      "slug": "paa",
-      "news_path": "aktualnosci",
-      "pages": 49
-    },
-    {
-      "slug": "rcb",
-      "news_path": "komunikaty",
-      "pages": 90
-    },
-    {
-      "slug": "udsc",
-      "news_path": "aktualnosci",
-      "pages": 23
-    },
-    {
-      "slug": "prokuratoria",
-      "news_path": "aktualnosci",
-      "pages": 9
-    },
-    {
-      "slug": "urpl",
-      "news_path": "aktualnosci",
-      "pages": 14
-    },
-    {
-      "slug": "uzp",
-      "news_path": "aktualnosci",
-      "pages": 18
-    },
-    {
-      "slug": "uw-kujawsko-pomorski",
-      "news_path": "aktualnosci",
-      "pages": 216
-    },
-    {
-      "slug": "uw-lubelski",
-      "news_path": "wiadomosci",
-      "pages": 95
-    },
-    {
-      "slug": "uw-lubuski",
-      "news_path": "wiadomosci",
-      "pages": 138
-    },
-    {
-      "slug": "uw-lodzki",
-      "news_path": "wiadomosci",
-      "pages": 161
-    },
-    {
-      "slug": "uw-mazowiecki",
-      "news_path": "wiadomosci2",
-      "pages": 132
-    },
-    {
-      "slug": "uw-opolski",
-      "news_path": "aktualnosci",
-      "pages": 138
-    },
-    {
-      "slug": "uw-podkarpacki",
-      "news_path": "wiadomosci",
-      "pages": 149
-    },
-    {
-      "slug": "uw-podlaski",
-      "news_path": "wiadomosci",
-      "pages": 310
-    },
-    {
-      "slug": "uw-warminsko-mazurski",
-      "news_path": "aktualnosci",
-      "pages": 377
-    },
-    {
-      "slug": "uw-zachodniopomorski",
-      "news_path": "wiadomosci",
-      "pages": 4
-    },
-    {
-      "slug": "gddkia-bialystok",
-      "news_path": "aktualnosci",
-      "pages": 2
-    },
-    {
-      "slug": "gddkia-bydgoszcz",
-      "news_path": "aktualnosci",
-      "pages": 3
-    },
-    {
-      "slug": "gddkia-gdansk",
-      "news_path": "aktualnosci",
-      "pages": 5
-    },
-    {
-      "slug": "gddkia-katowice",
-      "news_path": "aktualnosci",
-      "pages": 5
-    },
-    {
-      "slug": "gddkia-kielce",
-      "news_path": "aktualnosci",
-      "pages": 6
-    },
-    {
-      "slug": "gddkia-krakow",
-      "news_path": "aktualnosci",
-      "pages": 7
-    },
-    {
-      "slug": "gddkia-lublin",
-      "news_path": "aktualnosci",
-      "pages": 5
-    },
-    {
-      "slug": "gddkia-lodz",
-      "news_path": "aktualnosci",
-      "pages": 4
-    },
-    {
-      "slug": "gddkia-olsztyn",
-      "news_path": "aktualnosci",
-      "pages": 1
-    },
-    {
-      "slug": "gddkia-opole",
-      "news_path": "aktualnosci",
-      "pages": 5
-    },
-    {
-      "slug": "gddkia-poznan",
-      "news_path": "aktualnosci",
-      "pages": 5
-    },
-    {
-      "slug": "gddkia-rzeszow",
-      "news_path": "aktualnosci",
-      "pages": 6
-    },
-    {
-      "slug": "gddkia-szczecin",
-      "news_path": "aktualnosci",
-      "pages": 3
-    },
-    {
-      "slug": "gddkia-warszawa",
-      "news_path": "aktualnosci",
-      "pages": 4
-    },
-    {
-      "slug": "gddkia-wroclaw",
-      "news_path": "aktualnosci",
-      "pages": 7
-    },
-    {
-      "slug": "gddkia-zielona-gora",
-      "news_path": "aktualnosci",
-      "pages": 2
-    },
-    {
-      "slug": "wsse-bialystok",
-      "news_path": "wiadomosci",
-      "pages": 20
-    },
-    {
-      "slug": "wsse-bydgoszcz",
-      "news_path": "aktualnosci",
-      "pages": 71
-    },
-    {
-      "slug": "wsse-gdansk",
-      "news_path": "wiadomosci",
-      "pages": 48
-    },
-    {
-      "slug": "wsse-katowice",
-      "news_path": "wiadomosci",
-      "pages": 35
-    },
-    {
-      "slug": "wsse-kielce",
-      "news_path": "wiadomosci",
-      "pages": 28
-    },
-    {
-      "slug": "wsse-lublin",
-      "news_path": "komunikaty-i-meldunki",
-      "pages": 53
-    },
-    {
-      "slug": "wsse-lodz",
-      "news_path": "komunikaty-o-wodzie-do-spozycia2",
-      "pages": 104
-    },
-    {
-      "slug": "wsse-olsztyn",
-      "news_path": "komunikaty-i-meldunki",
-      "pages": 52
-    },
-    {
-      "slug": "wsse-opole",
-      "news_path": "wiadomosci",
-      "pages": 17
-    },
-    {
-      "slug": "wsse-poznan",
-      "news_path": "wiadomosci",
-      "pages": 31
-    },
-    {
-      "slug": "wsse-rzeszow",
-      "news_path": "wiadomosci",
-      "pages": 24
-    },
-    {
-      "slug": "wsse-szczecin",
-      "news_path": "wiadomosci",
-      "pages": 44
-    },
-    {
-      "slug": "wsse-warszawa",
-      "news_path": "wiadomosci",
-      "pages": 36
-    },
-    {
-      "slug": "wsse-wroclaw",
-      "news_path": "komunikaty-i-meldunki",
-      "pages": 25
-    },
-    {
-      "slug": "prokuratura-krajowa",
-      "news_path": "aktualnosci",
-      "pages": 541
-    },
-    {
-      "slug": "po-bialystok",
-      "news_path": "aktualnosci",
-      "pages": 25
-    },
-    {
-      "slug": "po-bielsko-biala",
-      "news_path": "komunikaty",
-      "pages": 1
-    },
-    {
-      "slug": "po-bydgoszcz",
-      "news_path": "aktualnosci1",
-      "pages": 17
-    },
-    {
-      "slug": "po-elblag",
-      "news_path": "komunikaty",
-      "pages": 8
-    },
-    {
-      "slug": "po-gdansk",
-      "news_path": "aktualnosci",
-      "pages": 1
-    },
-    {
-      "slug": "po-gliwice",
-      "news_path": "komunikaty",
-      "pages": 32
-    },
-    {
-      "slug": "po-gorzow-wielkopolski",
-      "news_path": "aktualnosci1",
-      "pages": 17
-    },
-    {
-      "slug": "po-katowice",
-      "news_path": "komunikaty-rzecznika-prasowego",
-      "pages": 8
-    },
-    {
-      "slug": "po-kielce",
-      "news_path": "aktualnosci1",
-      "pages": 25
-    },
-    {
-      "slug": "po-krakow",
-      "news_path": "komunikaty-rzecznika2",
-      "pages": 11
-    },
-    {
-      "slug": "po-legnica",
-      "news_path": "aktualnosci1",
-      "pages": 36
-    },
-    {
-      "slug": "po-lublin",
-      "news_path": "aktualnosci",
-      "pages": 9
-    },
-    {
-      "slug": "po-lodz",
-      "news_path": "aktualnosci1",
-      "pages": 26
-    },
-    {
-      "slug": "po-lomza",
-      "news_path": "komunikaty-rzecznika-prasowego---archiwalne",
-      "pages": 13
-    },
-    {
-      "slug": "po-nowy-sacz",
-      "news_path": "aktualnosci",
-      "pages": 25
-    },
-    {
-      "slug": "po-olsztyn",
-      "news_path": "aktualnosci1",
-      "pages": 16
-    },
-    {
-      "slug": "po-opole",
-      "news_path": "aktualnosci",
-      "pages": 21
-    },
-    {
-      "slug": "po-ostroleka",
-      "news_path": "komunikaty-prasowe",
-      "pages": 6
-    },
-    {
-      "slug": "po-ostrow-wielkopolski",
-      "news_path": "aktualnosci1",
-      "pages": 18
-    },
-    {
-      "slug": "po-piotrkow-trybunalski",
-      "news_path": "aktualnosci",
-      "pages": 17
-    },
-    {
-      "slug": "po-plock",
-      "news_path": "aktualnosci",
-      "pages": 13
-    },
-    {
-      "slug": "po-poznan",
-      "news_path": "aktualnosci1",
-      "pages": 51
-    },
-    {
-      "slug": "po-rzeszow",
-      "news_path": "komunikaty-prasowe",
-      "pages": 38
-    },
-    {
-      "slug": "po-siedlce",
-      "news_path": "aktualnosci1",
-      "pages": 15
-    },
-    {
-      "slug": "po-sieradz",
-      "news_path": "aktualnosci",
-      "pages": 15
-    },
-    {
-      "slug": "po-sosnowiec",
-      "news_path": "aktualnosci",
-      "pages": 28
-    },
-    {
-      "slug": "po-swidnica",
-      "news_path": "aktualnosci",
-      "pages": 23
-    },
-    {
-      "slug": "po-tarnow",
-      "news_path": "aktualnosci1",
-      "pages": 14
-    },
-    {
-      "slug": "po-torun",
-      "news_path": "aktualnosci1",
-      "pages": 19
-    },
-    {
-      "slug": "po-warszawa",
-      "news_path": "komunikaty-rzecznika-prasowego",
-      "pages": 28
-    },
-    {
-      "slug": "po-warszawa-praga",
-      "news_path": "aktualnosci",
-      "pages": 31
-    },
-    {
-      "slug": "po-wloclawek",
-      "news_path": "aktualnosci1",
-      "pages": 18
-    },
-    {
-      "slug": "po-wroclaw",
-      "news_path": "aktualnosci",
-      "pages": 23
-    },
-    {
-      "slug": "po-zielona-gora",
-      "news_path": "aktualnosci1",
-      "pages": 35
-    },
-    {
-      "slug": "pr-gdansk",
-      "news_path": "aktualnosci",
-      "pages": 23
-    },
-    {
-      "slug": "pr-katowice",
-      "news_path": "aktualnosci-prokuratury-regionalnej",
-      "pages": 14
-    },
-    {
-      "slug": "pr-krakow",
-      "news_path": "komunikaty",
-      "pages": 6
-    },
-    {
-      "slug": "pr-lublin",
-      "news_path": "aktualnosci1",
-      "pages": 8
-    },
-    {
-      "slug": "pr-lodz",
-      "news_path": "aktualnosci",
-      "pages": 15
-    },
-    {
-      "slug": "pr-poznan",
-      "news_path": "wiadomosci",
-      "pages": 17
-    },
-    {
-      "slug": "pr-rzeszow",
-      "news_path": "aktualnosci",
-      "pages": 11
-    },
-    {
-      "slug": "pr-szczecin",
-      "news_path": "aktualnosci",
-      "pages": 13
-    },
-    {
-      "slug": "pr-warszawa",
-      "news_path": "aktualnosci1",
-      "pages": 19
-    },
-    {
-      "slug": "pr-wroclaw",
-      "news_path": "aktualnosci1",
-      "pages": 32
-    },
-    {
-      "slug": "kgpsp",
-      "news_path": "aktualnosci",
-      "pages": 130
-    },
-    {
-      "slug": "kwpsp-gdansk",
-      "news_path": "aktualnosci",
-      "pages": 103
-    },
-    {
-      "slug": "kwpsp-opole",
-      "news_path": "aktualnosci",
-      "pages": 6
-    },
-    {
-      "slug": "kwpsp-poznan",
-      "news_path": "wiadomosci2021",
-      "pages": 24
-    },
-    {
-      "slug": "spzoz-mswia-opole",
-      "news_path": "wiadomosci",
-      "pages": 3
-    },
-    {
-      "slug": "spzoz-mswia-szczecin",
-      "news_path": "aktualnosci",
-      "pages": 62
-    },
-    {
-      "slug": "spzoz-mswia-zielona-gora",
-      "news_path": "aktualnosci",
-      "pages": 2
-    },
-    {
-      "slug": "krrit",
-      "news_path": "aktualnosci",
-      "pages": 61
-    },
-    {
-      "slug": "kowr",
-      "news_path": "aktualnosci",
-      "pages": 5
-    },
-    {
-      "slug": "kzn",
-      "news_path": "aktualnosci",
-      "pages": 24
-    },
-    {
-      "slug": "nck",
-      "news_path": "aktualnosci",
-      "pages": 55
-    },
-    {
-      "slug": "wody-polskie",
-      "news_path": "aktualnosci",
-      "pages": 62
-    }
-  ],
-  "skipped": [
-    "chemikalia",
-    "piorin",
-    "wsse-gorzowwlkp",
-    "wsse-krakow",
-    "po-slupsk",
-    "po-tarnobrzeg",
-    "cskmswia",
-    "spzoz-mswia-katowice",
-    "spzoz-mswia-lublin",
-    "spzoz-mswia-glucholazy"
-  ],
-  "total_pages": 9247
-}

src/make_docs.py

@@ -13,8 +13,7 @@ sys.path.insert(0, str(Path(__file__).resolve().parent))
 from sources import SOURCES, EXCLUDED, ADDED
 
 ROOT = Path(__file__).resolve().parent.parent
-VERSION = "0.2.3"
-RELEASE_DATE = "2026-07-26"
+VERSION = "0.2.1"
 CONTACT = "k.wikiel@gmail.com"  # notice-and-takedown / data-removal requests
 
 # Dupochron: documented good-faith provenance + no-warranty + takedown + PII.
@@ -120,10 +119,7 @@ def main():
         st = load_stats(name)
         if not st:
             print(f"  ! no stats for {name}"); continue
-        # Sources with a hand-authored datasheet (rich provenance, per-source add
-        # date) opt out of template regeneration to avoid clobbering it.
-        if not cfg.get("custom_datasheet"):
-            (ROOT / "data" / name / f"{name}.md").write_text(datasheet(name, cfg, st))
+        (ROOT / "data" / name / f"{name}.md").write_text(datasheet(name, cfg, st))
         rows.append((name, cfg, st))
         tot_doc += st["kept"]; tot_tok += st["tokens"]; tot_chr += st["chars"]
     rows.sort(key=lambda r: -r[2]["tokens"])
@@ -170,81 +166,21 @@ family (Enevoldsen et al., [arXiv:2508.02271](https://arxiv.org/abs/2508.02271))
 
 > **v{VERSION} stable** · {tot_doc:,} documents · **{tot_tok/1e9:.2f}B tokens**
 > (tiktoken proxy; canonical Llama-3 count at release) · {len(rows)} sources
-> Updated: **{RELEASE_DATE}**
+> Updated: **{ADDED}**
 
-> **v0.3-dev experimental track** · quality/diversity workflow, source-gate
-> validation and candidate-data audits. This is development work, not a released
-> corpus version, and it does not replace the v{VERSION} stable parquets.
+> **v0.3.0-preview in progress** · quality/diversity remix workflow, legal-style
+> downweighting, and filtered contemporary Polish web candidates. Biblioteka
+> Nauki and Europeana are prepared as source-ingestion PR targets with
+> per-document `license` and `author` metadata, but they are not part of this
+> stable parquet release yet.
 
-> **Europeana validation artifact (2026-07-10)** · a separate reproducible
-> 810-document direct-ingestion sample stored under the legacy path
-> `previews/v0.3.1/europeana.parquet`. It is not a full dataset release.
-
-## Releases
-
-### Stable releases
+## Versions
 
 | version | status | documents | tokens | notes |
 |---|---|---:|---:|---|
-| `v0.2.3` | stable release | {tot_doc:,} | {tot_tok/1e9:.2f}B | Adds community-contributed `european_hplt_v3_pl`, `global_voices` and `nkjp1m`. |
-| `v0.2.2` | previous stable | 2,579,963 | 6.36B | Added community-contributed `govpl`: 88,190 docs / 80.7M tokens. |
-| `v0.2.1` | previous stable | 2,491,773 | 6.28B | 12-source corpus with `license` and `author` metadata columns; added `1000_novels`. |
+| `v0.2.1` | stable release | {tot_doc:,} | {tot_tok/1e9:.2f}B | 12-source stable corpus with `license` and `author` metadata columns; adds `1000_novels`. |
 | `v0.2.0` | previous stable | 2,490,773 | 6.22B | Provenance-first corpus from 11 open/official sources. |
-
-![Stable dataset size by version](artifacts/dataset_size_by_version.png)
-
-### Development and validation artifacts
-
-| name | type | scope | notes |
-|---|---|---|---|
-| `v0.3-dev` | experimental development track | workflow and candidate audits | Quality remix, legal-source caps, source QA, deduplication and direct-upstream ingestion. Not a corpus release. |
-| `Europeana validation artifact 2026-07-10` | reproducible sample | 810 documents / 140,042 tokens | Stored at the legacy path `previews/v0.3.1/europeana.parquet`; preserves per-record rights and creator metadata. Not a corpus release. |
-
-### Version details
-
-#### v0.2.3 — current stable
-
-Community expansion release. Adds `european_hplt_v3_pl` from
-[PR #5](https://huggingface.co/datasets/SlayerLab/polish-dynaword/discussions/5)
-and `global_voices` from
-[PR #7](https://huggingface.co/datasets/SlayerLab/polish-dynaword/discussions/7),
-plus `nkjp1m` from
-[PR #8](https://huggingface.co/datasets/SlayerLab/polish-dynaword/discussions/8).
-The complete 16-source release contains **2,710,974 documents /
-6,881,277,362 cl100k-proxy tokens**, summed from the released per-source parquet
-statistics.
-
-#### v0.2.2 — previous stable
-
-Added `govpl` from
-[PR #9](https://huggingface.co/datasets/SlayerLab/polish-dynaword/discussions/9):
-88,190 Polish government press releases collected directly from 133 gov.pl
-ministry and agency subsites.
-
-#### v0.2.1 — previous stable
-
-Introduced the canonical eight-column release schema:
-`id, text, source, added, created, token_count, license, author`. Added
-`1000_novels` from
-[PR #1](https://huggingface.co/datasets/SlayerLab/polish-dynaword/discussions/1)
-and recomputed all release statistics from parquet files.
-
-#### v0.2.0 — initial stable corpus
-
-The provenance-first baseline: 11 reviewed open or official sources,
-2,490,773 documents and 6.22B proxy tokens.
-
-#### v0.3-dev — experimental quality workflow
-
-Candidate-only work on a better-balanced training mixture: legal-source caps,
-temperature sampling, source QA, deduplication and direct-upstream ingestion.
-It is not a replacement for the stable corpus.
-
-#### Europeana validation artifact — 2026-07-10
-
-A separate reproducible 810-document Europeana sample preserving per-record
-rights and creator metadata. It does not change stable-release totals.
-`v0.3.1-preview` remains only as the legacy storage-path label.
+| `v0.3.0-preview` | workflow + candidate data in progress | candidate-only | TBD | Biblioteka Nauki, Europeana, and HPLT/Common Corpus style expansion pending per-source QA, dedup, and legal review. |
 
 ## What this dataset contributes
 The raw texts come from existing open corpora (redistributed via SpeakLeash and,
@@ -269,18 +205,10 @@ as the redistributing aggregator; this release does not claim ownership of them
 
 ## Contributors
 
-| contributor | contribution | release / PR |
-|---|---|---|
-| [Kacper Wikieł](https://huggingface.co/kacperwikiel) | Project maintainer; corpus curation, source and license review, release engineering, documentation, validation and reproducible build workflow. | all releases |
-| [Bart Kobyliński](https://huggingface.co/bartoszkobylinski1) | Added `1000_novels`; expanded Biblioteka Nauki and Europeana ingestion with per-document license and author metadata. | `v0.2.1`, [PR #1](https://huggingface.co/datasets/SlayerLab/polish-dynaword/discussions/1) |
-| [Paweł Puzio](https://huggingface.co/ppuzio) | Built `govpl`: subsite discovery, direct ingestion pipeline, dataset artifact, contract tests and documentation. | `v0.2.2`, [PR #9](https://huggingface.co/datasets/SlayerLab/polish-dynaword/discussions/9) |
-| [Arkadiusz Słota](https://huggingface.co/Maggio33) | Built `european_hplt_v3_pl`: HPLT v3 WDS 10+9 cleaning pipeline, 110,590-document artifact, validation and documentation. | `v0.2.3`, [PR #5](https://huggingface.co/datasets/SlayerLab/polish-dynaword/discussions/5) |
-| [Dawid Majewski](https://huggingface.co/dawidmajewski) | Added the reviewed Global Voices Polish corpus with per-document author attribution and documentation. | `v0.2.3`, [PR #7](https://huggingface.co/datasets/SlayerLab/polish-dynaword/discussions/7) |
-| [1am](https://huggingface.co/1am) | Added `nkjp1m`: the manually annotated 1-million-word NKJP subcorpus, direct fetch/build pipeline, source documentation and release artifact. | `v0.2.3`, [PR #8](https://huggingface.co/datasets/SlayerLab/polish-dynaword/discussions/8) |
-
-Contributions are credited when they add a verifiable dataset artifact, pipeline,
-validation, documentation or release work. A merged discussion with no resulting
-files is not listed as a data contribution.
+- **Kacper Wikieł** — corpus curation, release engineering, documentation, and
+  reproducible build workflow.
+- **Bart Kobyliński** — source expansion work for Biblioteka Nauki and Europeana,
+  including per-document license/author metadata collection and rebuild planning.
 
 ## Guiding principles
 1. **Open & traceable licensing** — every source is *openly licensed* with a documented
@@ -333,10 +261,9 @@ Current v0.3 source-ingestion status:
 - Europeana release policy: split conservatively at pre-1929 records for
   US-sensitive downstream reuse, and keep later/unknown records separately
   labeled or held until legal review.
-- `european_hplt_v3_pl`: WDS bins 10+9 are included in v{VERSION} after contract
-  validation. HPLT packaging is CC0, while underlying crawled web documents can
-  carry independent rights; downstream near-dedup and web-content review remain
-  recommended before training.
+- `ashtok897/european-hplt-v1`: candidate workflow exists, but web-crawl
+  provenance, dedup, QA, and final mix weighting are still pending before stable
+  inclusion.
 
 Current review artifacts:
 
@@ -371,19 +298,23 @@ python3 src/make_docs.py
 """
     (ROOT / "README.md").write_text(readme)
 
-    # CHANGELOG is append-only release history. Documentation regeneration must
-    # never erase earlier releases or hand-reviewed legal/release notes.
-    changelog_path = ROOT / "CHANGELOG.md"
-    changelog = changelog_path.read_text() if changelog_path.exists() else "# Changelog\n"
-    heading = f"## v{VERSION} ({RELEASE_DATE})"
-    if heading not in changelog:
-        entry = (
-            f"{heading}\n\n"
-            f"- Current release totals after parquet recount: {len(rows)} sources, "
-            f"{tot_doc:,} docs, {tot_tok:,} tokens (tiktoken cl100k proxy).\n\n"
-        )
-        changelog = changelog.replace("# Changelog\n", f"# Changelog\n\n{entry}", 1)
-        changelog_path.write_text(changelog)
+    (ROOT / "CHANGELOG.md").write_text(
+        f"# Changelog\n\n## v{VERSION} ({ADDED})\n\n"
+        f"- Released a stable metadata-schema update: "
+        f"`id, text, source, added, created, token_count, license, author`.\n"
+        f"- Current release totals after parquet recount: {len(rows)} sources, "
+        f"{tot_doc:,} docs, {tot_tok/1e9:.2f}B tokens (tiktoken proxy).\n"
+        f"- Added `1000_novels` as a stable CC-BY-4.0 literature source.\n"
+        f"- Added source registry entries and PR contract for future "
+        f"`biblioteka_nauki` and `europeana` ingestion with per-document "
+        f"license/author metadata. These sources are not included in v{VERSION} "
+        f"parquets yet.\n"
+        f"- Credited Bart Kobyliński for source expansion work on Biblioteka "
+        f"Nauki and Europeana metadata/rebuild preparation.\n"
+        f"- Kept raw SpeakLeash Europeana excluded unless a direct rebuild "
+        f"preserves per-record rights metadata.\n"
+        f"- Sources: {', '.join(n for n, _, _ in rows)}.\n"
+        f"- Excluded (see README): {', '.join(EXCLUDED)}.\n")
 
     (ROOT / "LICENSE").write_text(
         "Polish DynaWord is released under Creative Commons Attribution-ShareAlike 4.0\n"

src/pattern_frequency_report.py

@@ -45,10 +45,7 @@ def count_patterns_for_source(parquet_path: Path) -> dict[str, int]:
 
     for rg in range(pf.num_row_groups):
         table = pf.read_row_group(rg, columns=["text"])
-        # A single large source row group can exceed Arrow's 2 GiB utf8
-        # offset limit after lower-casing. Use 64-bit string offsets before
-        # applying any text kernels so community-contributed web shards scale.
-        text = pc.cast(table["text"], "large_string")
+        text = table["text"]
         text = pc.utf8_lower(text)
         text = pc.replace_substring_regex(text, pattern="\\r?\\n", replacement=" ")
 

src/plot_version_growth.py

@@ -1,66 +0,0 @@
-#!/usr/bin/env python3
-"""Plot stable Polish DynaWord corpus growth across released versions."""
-
-from pathlib import Path
-
-import matplotlib.pyplot as plt
-from matplotlib.ticker import FuncFormatter
-
-
-ROOT = Path(__file__).resolve().parent.parent
-OUTPUT = ROOT / "artifacts" / "dataset_size_by_version.png"
-
-VERSIONS = ["v0.2.0", "v0.2.1", "v0.2.2", "v0.2.3"]
-DOCUMENTS = [2_490_773, 2_491_773, 2_579_963, 2_710_974]
-TOKENS = [6_221_393_871, 6_281_911_234, 6_362_577_363, 6_881_277_362]
-
-
-def millions(value, _position):
-    return f"{value / 1_000_000:.1f}M"
-
-
-def billions(value, _position):
-    return f"{value / 1_000_000_000:.1f}B"
-
-
-def main():
-    plt.style.use("seaborn-v0_8-whitegrid")
-    fig, axes = plt.subplots(1, 2, figsize=(12, 4.8))
-
-    panels = [
-        (axes[0], DOCUMENTS, "Documents", millions, lambda v: f"{v / 1_000_000:.3f}M", "#2563eb"),
-        (axes[1], TOKENS, "Tokens (cl100k proxy)", billions, lambda v: f"{v / 1_000_000_000:.3f}B", "#0f766e"),
-    ]
-
-    for axis, values, ylabel, formatter, labeler, color in panels:
-        axis.plot(VERSIONS, values, marker="o", linewidth=2.5, color=color)
-        axis.fill_between(VERSIONS, values, alpha=0.12, color=color)
-        axis.set_ylabel(ylabel)
-        axis.yaxis.set_major_formatter(FuncFormatter(formatter))
-        axis.spines[["top", "right"]].set_visible(False)
-        axis.grid(axis="x", visible=False)
-        padding = (max(values) - min(values)) * 0.3 or max(values) * 0.02
-        axis.set_ylim(min(values) - padding, max(values) + padding)
-
-        for version, value in zip(VERSIONS, values):
-            axis.annotate(
-                labeler(value),
-                (version, value),
-                xytext=(0, 9),
-                textcoords="offset points",
-                ha="center",
-                fontsize=9,
-                fontweight="bold",
-                color=color,
-            )
-
-    fig.suptitle("Polish DynaWord — stable dataset growth by version", fontsize=15)
-    fig.tight_layout()
-    OUTPUT.parent.mkdir(parents=True, exist_ok=True)
-    fig.savefig(OUTPUT, dpi=180, bbox_inches="tight")
-    plt.close(fig)
-    print(OUTPUT)
-
-
-if __name__ == "__main__":
-    main()

src/sources.py

@@ -186,28 +186,6 @@ SOURCES = {
         "created": "1918-01-01, 2025-12-31",
         "is_ocr": False,
     },
-    "govpl": {
-        "file_key": "govpl",  # fetched directly from gov.pl, NOT via SpeakLeash
-        "pretty": "gov.pl — Polish government press releases",
-        "license": "CC-BY-SA-4.0",
-        "license_spdx": "CC-BY-SA-4.0",
-        "traceable": "All textual content on gov.pl is published under CC-BY-SA "
-                     "4.0 (site footer: \"Treści tekstowe publikowane w serwisie "
-                     "… są udostępniane na licencji CC BY-SA 4.0\"). Authored and "
-                     "hosted by the Polish government, which holds the right to "
-                     "license it — same open family as Polish Wikipedia.",
-        "upstream": "https://www.gov.pl/",
-        "provenance": "Ministry/agency press releases fetched directly from gov.pl "
-                      "(editor-content per article, HTML stripped) by "
-                      "src/fetch_govpl.py. Subsites + their non-uniform news-section "
-                      "paths are enumerated by src/discover_govpl.py into the "
-                      "committed manifest src/govpl_subsites.json (gov.pl has no "
-                      "sitemap/API). Each record's meta.url resolves on gov.pl.",
-        "domain": "government",
-        # actual min/max of article dates in the fetched corpus (90,124 dated).
-        "created": "2005-09-26, 2026-07-19",
-        "is_ocr": False,
-    },
     "biblioteka_nauki": {
         "file_key": "biblioteka_nauki_pl_corpus",
         "pretty": "Biblioteka Nauki",
@@ -251,81 +229,9 @@ SOURCES = {
                       "records should remain separately labeled or held until "
                       "legal review confirms their US status.",
     },
-    "nkjp1m": {
-        "file_key": "nkjp1m",
-        "pretty": "The manually annotated 1-million word subcorpus of the National Corpus of Polish",
-        "license": "CC-BY",
-        "license_spdx": "CC-BY-4.0",
-        "traceable": "The NKJP download page (clip.ipipan.waw.pl) states the manually annotated 1-million-word subcorpus (Podkorpus Milionowy NKJP 1.2) is available under a CC-BY licence; attribution to the NKJP Consortium (IPI PAN) preserved. The page states 'CC-BY' with no version, normalized to CC-BY-4.0 for SPDX. The CC-BY licence covers the corpus text; NKJP's separate tools are GPL and are not included here.",
-        "upstream": "https://clip.ipipan.waw.pl/NationalCorpusOfPolish",
-        "provenance": "Downloaded the IPI PAN tarball (NKJP-PodkorpusMilionowy-1.2.tar.gz) directly by src/fetch_njkp.py, one row per TEI <div> passage (its <ab> paragraphs joined by newlines, ellipsis markers stripped). The same minimal gates as build_dynaword.py are applied per passage (drop < 200 chars, drop non-Polish by diacritic ratio, exact sha1 dedup; the OCR gate is inapplicable). Per-document 'created' is the sample's TEI publication date (<date type=\"published\">) where the header records one. Normalized into the DynaWord schema; cl100k token counts via tiktoken (encode_ordinary), and the stats are recomputed from the written parquet.",
-        "domain": "mixed (balanced reference corpus)",
-        "created": "1957-2008 (recorded for 6,581 of 18,381 documents; per-document in the created column)",
-        "is_ocr": False,
-    },
-    "european_hplt_v3_pl": {
-        "file_key": "european_hplt_v3_pl",
-        "pretty": "HPLT v3.0 Polish (web, top WDS bins 10-9)",
-        "license": "CC0-1.0",
-        "license_spdx": "CC0-1.0",
-        "traceable": "HPLT v3.0 Terms of Use license the data packaging under CC0-1.0 "
-                     "(hplt-project.org/datasets/v3.0; HF card HPLT/HPLT3.0). HPLT does not "
-                     "own the underlying web text; notice-and-takedown applies and downstream "
-                     "legal/provenance review remains the user's responsibility.",
-        "upstream": "https://data.hplt-project.org/three/sorted/pol_Latn/",
-        "provenance": "Fetched directly from the HPLT v3 distribution (pol_Latn shards "
-                      "10_1 [WDS=10] and 9_1 prefix [WDS=9]) by src/clean_hplt_v3.py, NOT via "
-                      "SpeakLeash. Filtered: lang=pol_Latn + prob>=0.80, MT-prob<0.20, register "
-                      "in ID/OP/HI/IN/NA, min 400 chars / 80 words, boilerplate/adult/legalish "
-                      "drop, length cap 120k chars, CJK-mojibake drop, wikipedia/wikisource/"
-                      "wikimedia + file-host domains excluded. Token counts via tiktoken cl100k.",
-        "domain": "web",
-        "created": "2012-01-01, 2024-12-31",
-        "is_ocr": False,
-        "legal_note": "Web-crawl derived: accept_after_web_review. HPLT is globally "
-                      "deduplicated upstream and WDS quality-sorted; only top bins used. "
-                      "Copyright recurs at all lengths in web crawl, so per-source near/"
-                      "containment dedup vs existing sources is left to downstream per "
-                      "Dynaword guidelines. Parity with the accepted european-hplt-v1 CC0 basis.",
-    },
-    "1000_novels": {
-        "speakleash_key": "1000_novels_corpus_CLARIN-PL",
-        "pretty": "1000 Novels Corpus (CLARIN-PL)",
-        "license": "CC-BY-4.0",
-        "license_spdx": "CC-BY-4.0",
-        "traceable": "The 1000 Novels Corpus is released under Creative Commons "
-                     "Attribution 4.0 International by CLARIN-PL / Wrocław University "
-                     "of Technology, with attribution to the corpus authors and "
-                     "maintainers.",
-        "upstream": "https://clarin-pl.eu/dspace/handle/11321/312",
-        "domain": "literature",
-        "created": "unknown",
-        "is_ocr": False,
-        # Datasheet is contributor-committed (added 2026-07-02); skip regeneration
-        # so the global ADDED stamp does not overwrite the real add date.
-        "custom_datasheet": True,
-    },
-    "global_voices": {
-        # Data contributed directly (PR #7): no in-repo fetch/build script and a
-        # hand-authored datasheet. Registered for the source table + release totals;
-        # the datasheet stays contributor-owned via custom_datasheet.
-        "file_key": "global_voices",
-        "pretty": "Global Voices Polish",
-        "license": "CC-BY-3.0",
-        "license_spdx": "CC-BY-3.0",
-        "traceable": "CC BY 3.0 verified from the article-level license link published "
-                     "with each retained Global Voices article; each record carries "
-                     "CC-BY-3.0 metadata and per-document author/translator attribution "
-                     "in global_voices.attribution.jsonl.",
-        "upstream": "https://pl.globalvoices.org/",
-        "domain": "civic media, news and public-interest journalism",
-        "created": "2008-11-04, 2023-05-24",
-        "is_ocr": False,
-        "custom_datasheet": True,
-    },
 }
 
-ADDED = "2026-07-19"
+ADDED = "2026-06-15"
 EXCLUDED = {
     "open_subtitles_corpus": "Derivative of copyrighted film/TV dialogue; "
         "OpenSubtitles uploads largely unlicensed. Same copyright lesson as "

src/test_govpl_contract.py

@@ -1,84 +0,0 @@
-"""TDD contract for the direct gov.pl (press releases) ingestion path."""
-import sys
-import unittest
-from pathlib import Path
-
-sys.path.insert(0, str(Path(__file__).resolve().parent))
-
-from discover_govpl import listing_articles, news_candidates, max_page
-from fetch_govpl import article_body, normalize, BASE
-
-LISTING = """
-<div class="art-prev">
-  <div class="event"><span class="date">14.07.2026</span><span class="location">Paryż</span></div>
-  <div class="title"><a href="/web/premier/polska-gospodarzem-cwiczen">
-     Polska gospodarzem &#263;wicze&#324;</a></div>
-  <div class="intro">Lead...</div>
-</div>
-<div class="art-prev">
-  <div class="event"><span class="date">13.07.2026</span></div>
-  <div class="title"><a href="/web/premier/druga-wiadomosc">Druga wiadomo&#347;&#263;</a></div>
-</div>
-<!-- foreign link that must be ignored -->
-<div class="title"><a href="/web/gov/menu-link">nav</a></div>
-"""
-
-ARTICLE = """
-<article>
-  <h1>Tytu&#322;</h1>
-  <div class="editor-content"></div>
-  <div class="editor-content">
-    <p>Rada Ministr&oacute;w przyj&#281;&#322;a rozporz&#261;dzenie. """ + ("Tre&#347;&#263; komunikatu. " * 20) + """</p>
-    <div class="attachments"><a href="/x">Za&#322;&#261;cznik do pobrania</a></div>
-  </div>
-  <div class="gallery"><a href="/photo/1"><span class="sr-only">Poka&#380; zdj&#281;cie 1 z galerii</span></a></div>
-</article>
-"""
-
-
-class GovplContractTest(unittest.TestCase):
-    def test_listing_binds_date_title_url_and_filters_foreign(self):
-        arts = listing_articles(LISTING, "premier")
-        self.assertEqual(len(arts), 2)  # nav link (foreign slug) dropped
-        self.assertEqual(arts[0]["url"], "/web/premier/polska-gospodarzem-cwiczen")
-        self.assertEqual(arts[0]["date"], "14.07.2026")
-        self.assertIn("ćwiczeń", arts[0]["title"])  # entities decoded
-        self.assertEqual(arts[1]["date"], "13.07.2026")
-
-    def test_article_body_picks_richest_region_and_drops_gallery(self):
-        body = article_body(ARTICLE)
-        self.assertIn("Rada Ministrów", body)
-        self.assertNotIn("<", body)                       # tags stripped
-        self.assertNotIn("Pokaż zdjęcie", body)           # gallery caption cut
-        self.assertNotIn("Załącznik do pobrania", body)   # attachments cut
-
-    def test_normalize_schema_prepends_title_and_prefixes_url(self):
-        art = {"url": "/web/premier/x", "title": "Nagłówek", "date": "14.07.2026"}
-        row = normalize(art, ARTICLE, "premier")
-        self.assertEqual(set(row), {"text", "meta"})
-        self.assertEqual(set(row["meta"]), {"url", "subsite", "title", "date"})
-        self.assertTrue(row["text"].startswith("Nagłówek"))
-        self.assertEqual(row["meta"]["url"], BASE + "/web/premier/x")
-        self.assertEqual(row["meta"]["subsite"], "premier")
-
-    def test_thin_body_is_rejected(self):
-        thin = '<div class="editor-content"><p>Kr&#243;tko.</p></div>'
-        self.assertIsNone(normalize({"url": "/web/premier/x", "title": "T"}, thin, "premier"))
-        self.assertIsNone(normalize({"url": "/web/premier/x", "title": "T"}, "", "premier"))
-
-    def test_news_candidate_detection(self):
-        landing = ('<a href="/web/arimr/aktualnosci10">n</a>'
-                   '<a href="/web/arimr/wszystkie-aktualnosci">n</a>'
-                   '<a href="/web/arimr/serwis-prasowy-arimr2">n</a>'
-                   '<a href="/web/arimr/kontakt">x</a>'
-                   '<a href="/web/arimr/informacje-o-arimr">x</a>')
-        cands = set(news_candidates(landing, "arimr"))
-        self.assertEqual(cands, {"aktualnosci10", "wszystkie-aktualnosci", "serwis-prasowy-arimr2"})
-
-    def test_max_page(self):
-        self.assertEqual(max_page('<a href="?page=2">2</a><a href="?page=294">294</a>'), 294)
-        self.assertEqual(max_page("<html>no pager</html>"), 1)  # single-page listing
-
-
-if __name__ == "__main__":
-    unittest.main()

src/test_hplt_v3_contract.py

@@ -1,67 +0,0 @@
-#!/usr/bin/env python3
-"""Lokalny kontrakt-test artefaktu european_hplt_v3_pl (wzorzec: test_europeana_contract.py).
-
-Ćwiczy inwarianty bramki merge DynaWord, które da się sprawdzić bez repo/CI:
-canonical schema (typy+kolejność), licencja open+niepusta, brak pustego tekstu, token_count>0,
-stats.json spójny z parquetem. NIE zastępuje `dsbench audit`/CI po stronie repo (licencja finalnie
-podnoszona z src/sources.py — wpis MUSI być w PR).
-"""
-import json, sys
-from pathlib import Path
-import pyarrow as pa, pyarrow.parquet as pq
-import pyarrow.compute as pc
-
-ROOT = Path(__file__).resolve().parent.parent
-D = ROOT / "data" / "european_hplt_v3_pl"
-PARQUET = D / "european_hplt_v3_pl.parquet"
-STATS = D / "european_hplt_v3_pl.stats.json"
-ACCEPTED = {"cc0-1.0", "cc-by-3.0", "cc-by-4.0", "cc-by-sa-3.0", "cc-by-sa-4.0",
-            "public-domain", "mit", "apache-2.0", "bsd-3-clause"}
-CANON = ["id", "text", "source", "added", "created", "token_count", "license", "author"]
-CANON_TYPES = {"id": "string", "text": "string", "source": "string", "added": "string",
-               "created": "string", "token_count": "int64", "license": "string", "author": "string"}
-
-def main():
-    fails = []
-    pf = pq.ParquetFile(PARQUET)
-    sch = pf.schema_arrow
-    # 1. schema exact (order + types)
-    if sch.names != CANON:
-        fails.append(f"schema names != canonical: {sch.names}")
-    for name in CANON:
-        if name in sch.names and str(sch.field(name).type) != CANON_TYPES[name]:
-            fails.append(f"type {name}={sch.field(name).type} != {CANON_TYPES[name]}")
-    tbl = pf.read()
-    n = tbl.num_rows
-    # 2. licenses open + non-empty
-    lic = set(pc.unique(tbl["license"]).to_pylist())
-    bad = {l for l in lic if (l or "").lower() not in ACCEPTED}
-    if bad:
-        fails.append(f"non-accepted/empty licenses: {bad}")
-    # 3. no empty text
-    empty = pc.sum(pc.equal(pc.utf8_trim_whitespace(tbl["text"]), "")).as_py() or 0
-    if empty:
-        fails.append(f"empty text rows: {empty}")
-    # 4. token_count > 0
-    nonpos = pc.sum(pc.less_equal(tbl["token_count"], 0)).as_py() or 0
-    if nonpos:
-        fails.append(f"token_count<=0 rows: {nonpos}")
-    # 5. source uniform
-    srcs = set(pc.unique(tbl["source"]).to_pylist())
-    if srcs != {"european_hplt_v3_pl"}:
-        fails.append(f"source not uniform: {srcs}")
-    # 6. stats.json consistent
-    st = json.loads(STATS.read_text(encoding="utf-8"))
-    if st["kept"] != n:
-        fails.append(f"stats.kept={st['kept']} != parquet rows={n}")
-    tok_sum = pc.sum(tbl["token_count"]).as_py()
-    if st["tokens"] != tok_sum:
-        fails.append(f"stats.tokens={st['tokens']} != parquet sum={tok_sum}")
-
-    print(f"rows={n:,} tokens={tok_sum:,} licenses={lic} source={srcs}")
-    if fails:
-        print("CONTRACT FAIL:"); [print("  -", f) for f in fails]; sys.exit(1)
-    print("CONTRACT PASS (schema+license+text+token+stats invariants)")
-
-if __name__ == "__main__":
-    main()