Merge code for conf27/inductive-nlg-for-shared-tasks/ - #1

Merged

mam10eks merged 39 commits into

mainfrom

inductive-nlg-for-shared-tasks

Aug 18, 2026

Commits on Aug 16, 2026

Scaffold project layout per AGENT.md

pa27cufo
and
claude
committed
30da7cd
Add Stage 1: fetch and cache CEUR-WS + DBLP volume pages

pa27cufo
and
claude
committed
05e9494
Add Stage 2: parse volume index pages into sections and papers

pa27cufo
and
claude
committed
068ad8a
Merge pull request #2 from pierre-achkar/stage-1-fetch-volumes
pierre-achkar
authored
dbc6547
Merge pull request #5 from pierre-achkar/stage-2-parse-sections
pierre-achkar
authored
2570296
Add Stage 3: group papers into shared-task entries

pa27cufo
and
claude
committed
74c4253
Merge pull request #7 from pierre-achkar/stage-3-group-tasks
pierre-achkar
authored
b733759
Fix three title-matching bugs found during full needs_review audit

pa27cufo
and
claude
committed
a90a6f7
Merge pull request #9 from pierre-achkar/stage-3-fix-title-matching-bugs
pierre-achkar
authored
36a4342
Add Stage 4: extract claimed team/run counts from overview PDFs

pa27cufo
and
claude
committed
26bf5af
Merge pull request #11 from pierre-achkar/stage-4-extract-counts
pierre-achkar
authored
1e18064
Add Stage 5: resolve participant code links

pa27cufo
and
claude
committed
3763b44
Add final benchmark assembly: shared_tasks.jsonl / .csv / report.md

pa27cufo
and
claude
committed
2c91add
Merge pull request #14 from pierre-achkar/stage-5-find-code
pierre-achkar
authored
9d61049
Merge pull request #15 from pierre-achkar/build-final-corpus
pierre-achkar
authored
7ba52dc
Fix three real bugs found in full pipeline audit

pa27cufo
and
claude
committed
6876e71
Merge pull request #17 from pierre-achkar/full-audit-fixes
pierre-achkar
authored
0967614
Repair grouping: detect organizer task papers, implement is_umbrella/team_name

pa27cufo
and
claude
committed
53e3dca
Repair code_urls: exclude dependencies, expose link status and evidence

pa27cufo
and
claude
committed
3284405
Merge pull request #20 from pierre-achkar/repair-grouping-and-schema
pierre-achkar
authored
6d0a2c0
Add Stage 7: parse full text of corpus PDFs to markdown for participants

pa27cufo
and
claude
committed
3e146bf
data
pa27cufo
committed
d00c9ed
Align parsed full text with shared_tasks.jsonl / .csv

pa27cufo
and
claude
committed
ea9e33c
Stop tracking raw fetched sources in git

pa27cufo
and
claude
committed
f22b70e
Extract figures and tables from each PDF alongside the full text

pa27cufo
and
claude
committed
efc16a7
Render table images from the PDF instead of emitting CSV

pa27cufo
and
claude
committed
bfb1e11
Derive table images from page geometry, not from markdown pairing

pa27cufo
and
claude
committed
a5b8035
Merge pull request #22 from pierre-achkar/stage-7-fulltext
pierre-achkar
authored
1eb07a1
Add CORPUS.md: build method, layout, and usage guide

pa27cufo
and
claude
committed
54cca55
Rename CORPUS.md to README.md

pa27cufo
and
claude
committed
6c2a3dc
Remove files the project does not use

pa27cufo
and
claude
committed
a1d1fb4
Add requirements.txt; stop tracking .python-version

pa27cufo
and
claude
committed
e570e0e
Strip requirements.txt to bare pins

pa27cufo
and
claude
committed
6c47679

Commits on Aug 17, 2026

mf
mam10eks
committed
b8dc36a
mf
mam10eks
committed
0e270bc
Merge main into inductive-nlg-for-shared-tasks

mam10eks
and
Copilot
committed
2980526
some more prototyping
mam10eks
committed
7ff7aa0
mf
mam10eks
committed
dcff4a5
add prototypical example for task 1
mam10eks
committed
cb2cacd