Commit Graph

15 Commits

Author SHA1 Message Date
zhihaovme50 f937203c9e fix(deps): stop pointing users at markitdown[all] in hints and docs
Six places still told users to `pip install 'markitdown[all]'`: the
"markitdown library not available" error hints in arxiv_tool, core_tool and
semantic_scholar_tool, plus the fulltext access guide and the markitdown
example.

That is the extra the previous commit removed, and for the same reasons: it
pins youtube-transcript-api~=1.0.0 (`Requires-Python <3.14`), and since
markitdown 0.1.6 it also requires a prerelease-only Azure dependency, so a
default resolver silently backtracks to markitdown 0.1.5. A user following
the hint on Python 3.14 lands on a MarkItDown whose YouTube converter is
registered but cannot fetch transcripts.

All six now name the same converter extras the project declares. No test
asserts on these strings.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01A2xx4ELgtuMxiDema4LaEh
2026-09-01 20:26:44 -05:00
Shanghua Gao b4d9354b8d Merge main into codex/sufianta-batch-review, reconciling two independent weekly-healthcheck reliability designs
Both branches solved the same problem (a lost CI runner discarding the
whole weekly tool sweep) independently: main shipped sharded matrix jobs
with per-shard log parsing (scripts/summarize_health_sweep.py, already in
production), while this branch restores an unsharded checkpoint/resume
system (--json-output/--resume, atomic JSON checkpointing).

Resolution:
- .github/workflows/weekly-tool-healthcheck.yml and the run_all_patterns()
  call site: took main's sharded design, since it's already shipped and
  replacing it would reintroduce the exact failure mode it was built to
  fix. The checkpoint/resume machinery itself stays -- it's still useful
  standalone for a developer resuming an interrupted local sweep.
- Fixed a real interaction bug this merge would otherwise have introduced:
  checkpoint_complete compared `results` against the full `expected_patterns`
  universe, which is never true for a single shard's `selected` subset, so
  every successful sharded CI run would have incorrectly exited 1. Now
  compared against `selected` -- identical to `expected_patterns` when
  --shard isn't used, correctly scoped when it is. Verified with a manual
  shard simulation.
- Removed test_weekly_workflow_consumes_the_checkpoint_contract: it
  asserted the workflow YAML wires up --json-output and executes an inline
  job-summary heredoc, both part of the design main's sharding replaced.
  Its premise no longer holds under the shipped design.
2026-08-25 07:51:17 -04:00
Shanghua 59545fe7f2 Make PyMuPDF an explicit opt-in backend 2026-08-12 08:04:32 -04:00
Shanghua 868bbed53f Fix reload and health-check validation gaps 2026-08-11 01:36:29 -04:00
Shanghua 39c28ac00a Trim pseudocode and redundant registry refactor 2026-08-10 21:23:54 -04:00
Shanghua 27548bcf87 Remove non-functional formatting churn 2026-08-10 17:30:56 -04:00
Shanghua dabbb65192 Fix MCP registration examples semantically 2026-08-10 16:02:44 -04:00
Shanghua 27396a6fb4 Finish documentation artifact cleanup 2026-08-10 11:55:17 -04:00
SufianTA 0b2a523b4d Validate documentation examples recursively 2026-08-03 17:14:32 -07:00
Shanghua 1abee3938c clean docs 2026-02-23 12:38:53 -05:00
Shanghua fbb6b284de update docs 2026-02-23 11:58:57 -05:00
Shanghua a003d3b158 update 2026-02-22 18:49:50 -05:00
Shanghua 519eeb8a80 update 2026-02-22 18:35:06 -05:00
Shanghua 5bdfa66221 update 2026-02-22 17:59:54 -05:00
Shanghua 7f6a4b3706 update docs 2026-02-11 00:58:17 -05:00