Dataset Viewer
The dataset viewer is not available for this split.
Cannot load the dataset split (in streaming mode) to extract the first rows.
Error code: StreamingRowsError
Exception: CastError
Message: Couldn't cast
rows: int64
excluded_by_reason: struct<challenge_page (HTTP 200 "One moment, please..." auto-reload interstitial): int64, fetched af (... 240 chars omitted)
child 0, challenge_page (HTTP 200 "One moment, please..." auto-reload interstitial): int64
child 1, fetched after the first challenge page (collector missed the challenge); discarded unread, not in pa (... 2 chars omitted): int64
child 2, http_error_000: int64
child 3, named minor (named_minor_screen v2, rule A): int64
child 4, sealed/in-camera/anonymised-for-protection (ZA/MW terms): int64
exclusions_ids_only: list<item: struct<item_id: string, reason: string, status: string>>
child 0, item: struct<item_id: string, reason: string, status: string>
child 0, item_id: string
child 1, reason: string
child 2, status: string
redactions_by_part: struct<cases-00000.parquet: struct<email_named: int64, email_role_kept: int64, national_id: int64, p (... 20 chars omitted)
child 0, cases-00000.parquet: struct<email_named: int64, email_role_kept: int64, national_id: int64, payment_card: int64>
child 0, email_named: int64
child 1, email_role_kept: int64
child 2, national_id: int64
child 3, payment_card: int64
redactions_total: struct<email_named: int64, email_role_kept: int64, national_id: int64, payment_card: int64>
child 0, email_named: int64
child 1, email_role_kept: int64
child 2, national_id: int64
child 3, payment_card: int64
empty_by_column: struct<cases-00000.parque
...
ail_role_kept: int64, national_id: int64, payment_card: int64>
child 0, email_named: int64
child 1, email_role_kept: int64
child 2, national_id: int64
child 3, payment_card: int64
child 9, results: int64
child 10, robots_fetch_utc: timestamp[s]
child 11, rows: int64
child 12, rows_by_court: struct<Industrial Relations Court of Malawi: int64>
child 0, Industrial Relations Court of Malawi: int64
child 13, screen_version: string
screen_version: string
parser_version: string
files: struct<scrapers/README.md: string, scrapers/build_pack.py: string, scrapers/collect.py: string, scra (... 531 chars omitted)
child 0, scrapers/README.md: string
child 1, scrapers/build_pack.py: string
child 2, scrapers/collect.py: string
child 3, scrapers/extract_text.py: string
child 4, scrapers/list_jud.py: string
child 5, scrapers/list_sca.py: string
child 6, scrapers/minors_text_screen.py: string
child 7, scrapers/named_minor_screen.py: string
child 8, scrapers/polite_fetch.py: string
child 9, scrapers/requirements.txt: string
child 10, scrapers/rescreen.py: string
child 11, scrapers/robots_guard.py: string
child 12, scrapers/run_build_gated.sh: string
child 13, scrapers/sealed_screen.py: string
child 14, scrapers/secret_screen.py: string
child 15, scrapers/sexual_exploitation_screen.py: string
child 16, scrapers/youth_text_screen.py: string
child 17, scrapers/za_mw_screens.py: string
to
{'parser_version': Value('string'), 'screen_version': Value('string'), 'files': {'scrapers/README.md': Value('string'), 'scrapers/build_pack.py': Value('string'), 'scrapers/collect.py': Value('string'), 'scrapers/extract_text.py': Value('string'), 'scrapers/list_jud.py': Value('string'), 'scrapers/list_sca.py': Value('string'), 'scrapers/minors_text_screen.py': Value('string'), 'scrapers/named_minor_screen.py': Value('string'), 'scrapers/polite_fetch.py': Value('string'), 'scrapers/requirements.txt': Value('string'), 'scrapers/rescreen.py': Value('string'), 'scrapers/robots_guard.py': Value('string'), 'scrapers/run_build_gated.sh': Value('string'), 'scrapers/sealed_screen.py': Value('string'), 'scrapers/secret_screen.py': Value('string'), 'scrapers/sexual_exploitation_screen.py': Value('string'), 'scrapers/youth_text_screen.py': Value('string'), 'scrapers/za_mw_screens.py': Value('string')}}
because column names don't match
Traceback: Traceback (most recent call last):
File "/src/services/worker/src/worker/utils.py", line 149, in get_rows_or_raise
return get_rows(
dataset=dataset,
...<4 lines>...
column_names=column_names,
)
File "/src/libs/libcommon/src/libcommon/utils.py", line 272, in decorator
return func(*args, **kwargs)
File "/src/services/worker/src/worker/utils.py", line 129, in get_rows
rows_plus_one = list(itertools.islice(safe_iter(ds, dataset=dataset), rows_max_number + 1))
File "/src/services/worker/src/worker/utils.py", line 544, in safe_iter
yield from ds.decode(False) if ds.features else ds
File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 3052, in __iter__
for key, example in ex_iterable:
^^^^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 2559, in __iter__
for key, pa_table in self._iter_arrow():
~~~~~~~~~~~~~~~~^^
File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 2584, in _iter_arrow
for key, pa_table in self.ex_iterable._iter_arrow():
~~~~~~~~~~~~~~~~~~~~~~~~~~~~^^
File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 558, in _iter_arrow
for key, pa_table in iterator:
^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 440, in _iter_arrow
for key, pa_table in self.generate_tables_fn(**gen_kwags):
~~~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/json/json.py", line 361, in _generate_tables
self._cast_table(pa_table, json_field_paths=json_field_paths),
~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/json/json.py", line 132, in _cast_table
pa_table = table_cast(pa_table, self.info.features.arrow_schema)
File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2381, in table_cast
return cast_table_to_schema(table, schema)
File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2309, in cast_table_to_schema
raise CastError(
...<3 lines>...
)
datasets.table.CastError: Couldn't cast
rows: int64
excluded_by_reason: struct<challenge_page (HTTP 200 "One moment, please..." auto-reload interstitial): int64, fetched af (... 240 chars omitted)
child 0, challenge_page (HTTP 200 "One moment, please..." auto-reload interstitial): int64
child 1, fetched after the first challenge page (collector missed the challenge); discarded unread, not in pa (... 2 chars omitted): int64
child 2, http_error_000: int64
child 3, named minor (named_minor_screen v2, rule A): int64
child 4, sealed/in-camera/anonymised-for-protection (ZA/MW terms): int64
exclusions_ids_only: list<item: struct<item_id: string, reason: string, status: string>>
child 0, item: struct<item_id: string, reason: string, status: string>
child 0, item_id: string
child 1, reason: string
child 2, status: string
redactions_by_part: struct<cases-00000.parquet: struct<email_named: int64, email_role_kept: int64, national_id: int64, p (... 20 chars omitted)
child 0, cases-00000.parquet: struct<email_named: int64, email_role_kept: int64, national_id: int64, payment_card: int64>
child 0, email_named: int64
child 1, email_role_kept: int64
child 2, national_id: int64
child 3, payment_card: int64
redactions_total: struct<email_named: int64, email_role_kept: int64, national_id: int64, payment_card: int64>
child 0, email_named: int64
child 1, email_role_kept: int64
child 2, national_id: int64
child 3, payment_card: int64
empty_by_column: struct<cases-00000.parque
...
ail_role_kept: int64, national_id: int64, payment_card: int64>
child 0, email_named: int64
child 1, email_role_kept: int64
child 2, national_id: int64
child 3, payment_card: int64
child 9, results: int64
child 10, robots_fetch_utc: timestamp[s]
child 11, rows: int64
child 12, rows_by_court: struct<Industrial Relations Court of Malawi: int64>
child 0, Industrial Relations Court of Malawi: int64
child 13, screen_version: string
screen_version: string
parser_version: string
files: struct<scrapers/README.md: string, scrapers/build_pack.py: string, scrapers/collect.py: string, scra (... 531 chars omitted)
child 0, scrapers/README.md: string
child 1, scrapers/build_pack.py: string
child 2, scrapers/collect.py: string
child 3, scrapers/extract_text.py: string
child 4, scrapers/list_jud.py: string
child 5, scrapers/list_sca.py: string
child 6, scrapers/minors_text_screen.py: string
child 7, scrapers/named_minor_screen.py: string
child 8, scrapers/polite_fetch.py: string
child 9, scrapers/requirements.txt: string
child 10, scrapers/rescreen.py: string
child 11, scrapers/robots_guard.py: string
child 12, scrapers/run_build_gated.sh: string
child 13, scrapers/sealed_screen.py: string
child 14, scrapers/secret_screen.py: string
child 15, scrapers/sexual_exploitation_screen.py: string
child 16, scrapers/youth_text_screen.py: string
child 17, scrapers/za_mw_screens.py: string
to
{'parser_version': Value('string'), 'screen_version': Value('string'), 'files': {'scrapers/README.md': Value('string'), 'scrapers/build_pack.py': Value('string'), 'scrapers/collect.py': Value('string'), 'scrapers/extract_text.py': Value('string'), 'scrapers/list_jud.py': Value('string'), 'scrapers/list_sca.py': Value('string'), 'scrapers/minors_text_screen.py': Value('string'), 'scrapers/named_minor_screen.py': Value('string'), 'scrapers/polite_fetch.py': Value('string'), 'scrapers/requirements.txt': Value('string'), 'scrapers/rescreen.py': Value('string'), 'scrapers/robots_guard.py': Value('string'), 'scrapers/run_build_gated.sh': Value('string'), 'scrapers/sealed_screen.py': Value('string'), 'scrapers/secret_screen.py': Value('string'), 'scrapers/sexual_exploitation_screen.py': Value('string'), 'scrapers/youth_text_screen.py': Value('string'), 'scrapers/za_mw_screens.py': Value('string')}}
because column names don't matchNeed help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.
Malawi caselaw pilot p1
Warning: licence status
No source site grants an explicit licence. Copyright footers, quoted verbatim:
- judiciary.mw (fetched 2026-10-30 11:12 PT): "© 2026 Malawi Judiciary" (rendered; the year is written by a script). The site disclaimer has no reuse restriction; it covers warranty and linking only.
Malawi, Copyright Act 26 of 2016 s28(1)(a): protection does not extend to "written laws and decisions of courts and administrative bodies and to official translations thereof". UNVERIFIED against a primary text: WIPO Lex record 17267 confirms the Act (assented 2016-09-01, published 2016-09-02, commenced 2017-03-13), but its PDF host (wipolex-res.wipo.int) returned 403 for robots.txt and www.wipo.int disallows /edocs/lexdocs/, so the PDF was not fetched; the wording matches secondary copies only.
These readings are not legal advice. Every row carries licence = no-licence-stated-statutory-exclusion-claimed.
Contents
10 judgments. Text extracted with pdftotext (PDF) or python-docx (DOCX); no OCR. Scanned PDFs excluded as no_text_layer.
Exclusions (whole-decision, ids and reasons only; no names or titles stored)
- challenge_page (HTTP 200 "One moment, please..." auto-reload interstitial): 2
- fetched after the first challenge page (collector missed the challenge); discarded unread, not in pack: 5
- http_error_000: 1
- named minor (named_minor_screen v2, rule A): 1
- sealed/in-camera/anonymised-for-protection (ZA/MW terms): 1
Screens cover minors/youth, family and children's matters, sexual offences, sealed/anonymised matters, initials-only parties and PII. PII screen (za-mw-1.2): Luhn-valid ZA ID numbers and payment cards are redacted; ALL email addresses are screened: a named individual's address is replaced with '[redacted email]', role/office mailboxes (info@, registrar@, admin@, office-type) are kept. Minor rule: excluded on doubt where an identifiable child has health, abuse, care or injury facts. ID lists: manifests/stats.json.
PII redactions
- data/cases-00000.parquet: named-individual emails redacted 0, role/office mailboxes kept 0, national IDs 0, payment cards 0
- pack total: named-individual emails redacted 0, role/office mailboxes kept 0, national IDs 0, payment cards 0
Packaged scrapers/robots_guard.py vs the fleet copy
The packaged robots_guard.py is NOT the current fleet copy (sha256 fa8823164631..., 1277 lines). It derives from an older 1028-line copy (sha256 6a372966c1c3..., the one the fetches actually ran with) and therefore LACKS fleet lines 373-599: the TLS / per-host CA-bundle code for the robots.txt fetch (ROBOTS_CA_BUNDLES, set_ca_bundle, ROBOTS_GUARD_CA_BUNDLES env, cross-host redirect refusal, _default_fetch) plus smaller related edits (fleet 601-651, 1002-1005). It is used here only for robots_guard.evaluate() on robots.txt text fetched separately with curl, so the missing fetch code was not exercised. One further change for packaging: the default ROOT was changed from an absolute box path to the relative 'legal-corpora' (the fetches ran with the absolute default; evaluate() does not use ROOT).
polite_fetch.py: version that ran vs version packaged
Three versions of polite_fetch.py were involved (reconstructed from fetch logs; no backups of the earlier versions exist):
- Version A (no transport retry): made fetch-log lines 1 (11:12:23 PT), 3 (11:13:07 PT) and 4 (11:13:28 PT). These lines have no transport_retry field. It also made all ZA fetches.
- Version B (edited mid-run, between 11:13:28 and 11:13:38 PT): added one automatic retry after 15 s when curl gets no HTTP response (code 000), never on 403/429/challenge. The first automatic retry is at line 5 (11:13:38 PT); retries also appear at lines 6, 7, 10, 14 and 23. Version B ran with PACE_S=8 and did NOT detect 'one moment, please' or 'window.location.reload()', which is why the interstitial below was missed.
- Version C (edited 11:23 PT, after the run): added those two challenge strings. This is the packaged copy. v2 fixed a code comment that said 10 s; 15 s is what ran.
Empty columns
- cases-00000.parquet: case_number (10 rows empty), neutral_citation (10 rows empty)
Collection incidents
Transport failures (HTTP 000)
The host repeatedly dropped TLS connections. Five fetch-log lines record no HTTP response (curl code 000): lines 1, 3, 6, 7 and 10. Reconstructed from fetch logs:
- Line 1 (11:12:23 PT, listing /judgements): version A, no automatic retry. Line 2 (11:12:46 PT, HTTP 200) is a separate manual re-request, logged by hand (no memory/disk gate fields; note "single transport-error retry"). It was not an automatic retry.
- Line 3 (11:13:07 PT, /judiciary/disclaimer): version A, no automatic retry. Line 4 (11:13:28 PT, HTTP 200) is a separate manual re-request made through version A. It was not an automatic retry.
- Line 5 (listing page 1): first attempt failed, automatic retry succeeded (HTTP 200).
- Lines 6 and 7 (listing /judgements?page=2): two separate requests, each with a failed automatic retry. Listing pages 2-4 were abandoned, so only pages 0-1 (20 entries) were used.
- Line 10 (one judgment PDF): failed, and its automatic retry also failed. It was not fetched. This is the only one that became an exclusion: mw:a398e70fb17a, reason http_error_000.
- Lines 14 and 23 (judgment PDFs): first attempt failed, automatic retry succeeded (HTTP 200, kept as normal fetches). Lines 1, 3, 6 and 7 were listing or disclaimer pages, not judgments, and lines 1 and 3 were later re-requested successfully, so none of them is an exclusion.
Challenge page
After fetch 12 the host served an HTTP 200 "One moment, please..." interstitial that reloads itself, for 2 fetches (mw:a4dda2f56438, mw:7f3a4194e43a). The collector missed it and made 5 more fetches (mw:10622bd40082, mw:2ae9ff75cb58, mw:eece9d526549, mw:732e1a3b5976, mw:52326f7c3c98). Those 5 were discarded unread and are not in the pack. The host was then stopped and has not been contacted since. The detector has since been fixed.
Takedown
To request removal of a judgment, open an issue or discussion on the dataset repository with the case_id or source_url; the row will be removed in the next revision.
Checksums
SHA256SUMS lists 24 files (the pack has 25 files including SHA256SUMS itself, which it does not list).
- Downloads last month
- 83