endomorphosis/ipfs_uzbekistan_district_laws

Uzbekistan district laws: lex.uz local-authority decisions (pilot)

Research snapshot of Yunusobod district (Tashkent city) hokim and Kengash decisions, plus district/city acts listed under the viloyat hokim filter, from Qonunchilik ma'lumotlari milliy bazasi (lex.uz), run by the "Adolat" National Legal Information Centre under the Ministry of Justice of Uzbekistan.

Not legal advice. The official source prevails over this corpus. This is a PILOT batch (a small sample of issuing bodies). It is not complete coverage.

Field	Value
Laws / instruments, cumulative (all batches through batch2)	614
Rows (articles.parquet), cumulative (all batches through batch2)	1107
Snapshot date	2026-10-24
Source	lex.uz, local search (/uz/search/loc), fbody 20316 Yunusobod tumani hokimi, 20322 Xalq deputatlari Yunusobod tumani Kengashi; plus Toshkent tumani / Nurafshon / Ohangaron city acts listed under fbody 10012
Laws / instruments (first batch)	15
Rows (articles.parquet, first batch)	29 (body 15, appendix 14, article 0)
Language	uz (verbatim; Latin/Cyrillic as published)
Known gaps	42 (see manifests/known_gaps.json)

Cumulative note: this dataset now holds several batches. The table above gives cumulative totals first; the remaining rows describe the first batch. Each later batch has its own section below.

Rows

• laws.text is the verbatim act text as published in HTML on lex.uz (all non-empty document elements, joined by newlines).
• Each articles row is a verbatim, in-order exact substring of its law text: the main body block, then each appendix block. These decisions are numbered-paragraph instruments without "modda" articles, so they are not split further. body_status is text, or repealed_placeholder for repeal stubs (there are none in this pilot).
• Acts whose text lex.uz publishes only as PDF or image, or with an empty body (header and signature only), are excluded and listed in known_gaps.

Licence and source terms

lex.uz "Axborotni taqdim etish shartlari" (verbatim, uz): Ushbu axborot resursi mulk huquqi obyekti bo‘lib, O‘zbekiston Respublikasining Fuqarolik kodeksi, “Mualliflik huquqi va turdosh huquqlar to‘g‘risida”gi Qonuni va intellektual mulk to‘g‘risidagi boshqa normativ huquqiy hujjatlar bilan himoyalanadi. O‘zbekiston Respublikasining “Axborotlashtirish to‘g‘risida”gi Qonuniga muvofiq ushbu axborot resursning egalik huquqi Markazga tegishli. Ushbu axborot resursidan olgan axborotni faqat shaxsiy maqsadlarda foydalanishingiz mumkin. Daromad olish maqsadida maʼlumotlardan nusxa olish, tarqatish, sotish, berish resurs egasining yozma roziligisiz taqiqlanadi. Boshqa resurslarda axborotdan foydalanish (normativ huquqiy hujjatlarni istisno qilgan xolda) mazkur axborot resursiga havola qilgan xolda yo‘l qo‘yiladi.

(Unofficial translation) This information resource is an object of property protected by the Civil Code, the Law on Copyright and Related Rights and other IP acts. Under the Law on Informatization, ownership of the resource belongs to the Centre. Information obtained may be used only for personal purposes. Copying, distributing, selling or transferring data for profit without the owner's written consent is prohibited. Use of the information on other resources (with the exception of normative legal acts) is permitted with a reference to this resource.

The "Foydalanish shartlari" page (/agreement) is a privacy policy and has no reuse clause. Official documents of state bodies are not protected under Uzbek copyright law, and the terms themselves except normative legal acts from the reuse restriction. The full record is in manifests/terms.json.

Provenance

Collected with an honest user agent, at 8 s per-host pacing, after a robots.txt check (404, so nothing is disallowed), under fleet light gates. See scrapers/README.md and manifests/batch1_code.json (code pin endomorphosis/ipfs_datasets_py@74621da9dcd9b0200486e8b006f8f68f793c552c).

Notes on status, gaps, schema and terms

• laws.law_status is in_force_per_source_flag_unverified: lex.uz shows an in-force flag on each document, and this corpus has not independently checked whether each act is still in force.
• Acts that lex.uz publishes only as PDF, or with an empty body (header and signature only), were not fetched as PDF and were not OCR'd. Their PDFs were not downloaded, and the acts are listed in manifests/known_gaps.json.
• articles.body_status is a column added beyond the Malta/Taiwan schema (endomorphosis/ipfs_malta_laws@404a1774ff3b276b5a92ae998c2592537614cf84). Its values are text and repealed_placeholder.
• Per the user's ruling of 2026-10-13, restrictive source terms do not block the inclusion of official statute text. The lex.uz terms are quoted verbatim above and in manifests/terms.json.
• The secret scan (manifests/scan.json) ran on the staged files, not as a pre-commit hook, because nothing has been committed.

Batch 2 (district, scale): appended on parent commit 64695584342f

• Files: data/batch2/laws.parquet (599 laws) and data/batch2/articles.parquet (1078 rows: body 601, appendix 477, article 0; repealed_placeholder 0).
• Issuing bodies (lex.uz local search, Uzbek Latin filter): see manifests/batch2_bounds.json. Already-published laws are not re-included (0 skipped).
• Known gaps (manifests/batch2_known_gaps.json): 671 PDF-only acts and 1246 empty-body acts (header and signature only) excluded, plus 501 heading-only or PDF-only blocks dropped from included acts. No PDFs were downloaded or OCR'd.
• 1,244 acts listed under these filters are not in this batch because their issuing-body badge is not district-level (manifests/batch2_deferred_breakdown.json). 1,229 of them come from the 22 provincial-body lists and are provincial-tier acts already handled in ipfs_uzbekistan_provincial_laws: 441 are published there (420 in batch 2, 21 in the pilot) and 788 are listed there as known gaps (357 PDF-only, 431 empty-body). The other 15 come from this batch's district-body lists and are not published in either dataset: 3 Toshkent city hokim acts (capital city, provincial tier), 2 viloyat hokim acts, 4 acts with truncated badges, and 6 Bekobod city acts whose badge spells "shaxar", which the tier rule does not match. These 15 are left for a later batch.
• Fetch log: manifests/batch2_fetch_log.jsonl records every request of this batch's session (robots.txt, list GETs, pager POSTs, act pages) with t_start (request sent) and t_end (response received); per-host pacing of 8 s is measured from t_end. Counts: 2667 requests: act_page error: 1, act_page 200: 2492, list_GET 200: 109, pager_POST 200: 63, robots 404: 2.
• Code: new scrapers/ files collect_lexuz_scale.py, pack_lexuz_batch.py, verify_pack_batch.py, stage_append_lexuz.py, fetch_lexuz_v2.py, collect_lexuz_v3.py, stage_append_lexuz_v3.py, README_batch2.md; published scrapers/ files are unchanged. Record: manifests/batch2_code.json.

Correction: lex.uz fetch counts reported for provincial batch 2

The provincial batch 2 hand-off said 1,389 fetch responses with one 404 on /agreement. That was wrong in two ways. The 1,389/404 figures came from the whole cumulative work log (1390 lines: pilot sessions plus the provincial scale session), not from batch 2, and the single 404 was on /uz/agreement (a wrong path tried in the pilot); /agreement itself returned 200. The provincial batch 2 scale session made 1,266 requests (21 list GETs, 52 pager POSTs, 1,193 act pages), all HTTP 200. manifests/batch2_fetch_log.jsonl in ipfs_uzbekistan_provincial_laws has 1,268 entries, all HTTP 200: the act-page responses for all 1,268 indexed provincial acts, 75 of which were fetched in the pilot sessions; list and pager requests were left out. robots.txt (404) was checked at each session start but not logged in the old log format. From district batch 2 on, every request is logged with t_start and t_end.