diff --git a/.gitea/workflows/check.yml b/.gitea/workflows/check.yml index e5d4118..1d8ae31 100644 --- a/.gitea/workflows/check.yml +++ b/.gitea/workflows/check.yml @@ -46,6 +46,26 @@ jobs: subprocess.run([python,'-B','-m','pip','install','--no-deps',str(wheel)],check=True) outside = root / 'outside-checkout' outside.mkdir() + mm_fixture = json.loads(gzip.decompress((src/'tests/fixtures/mm2026-source.json.gz').read_bytes())) + mm_snapshots = root/'mm2026-snapshots' + for name, body in mm_fixture['files'].items(): + path = mm_snapshots/name + path.parent.mkdir(parents=True,exist_ok=True) + path.write_bytes(body.encode('utf-8')) + subprocess.run([python,'-I','-B','-m','paper_atlas','collect','mm2026-program','--snapshots',str(mm_snapshots),'--out',str(root/'installed-mm2026')],cwd=outside,check=True) + mm_report = json.loads((root/'installed-mm2026/collection-report.json').read_text()) + assert (mm_report['records'],mm_report['program_observations'],mm_report['extra_presentation_observations']) == (2143,4326,2183) + assert len(mm_report['files']) == 28 and mm_report['missing_abstract_count'] == 216 + assert mm_report['session_count'] == 157 and not mm_report['database_written'] and mm_report['network_requests'] == 0 and mm_report['approved_for_import'] == 0 + mm_rows = {r['paper_id']:r for f in mm_report['files'] for r in map(json.loads,pathlib.Path(f['file']).read_text().splitlines())} + import hashlib + assert set(mm_rows) == {g['paper_id'] for g in mm_fixture['independent_inventory']['inventory']} + for gold in mm_fixture['independent_inventory']['inventory']: + row = mm_rows[gold['paper_id']] + assert (row['title'],row['authors'],row['affiliations'],row['track']) == (gold['title'],gold['authors'],gold['affiliations'],gold['track_literal']) + assert hashlib.sha256(row['abstract'].encode()).hexdigest() == gold['abstract_sha256'] + assert [c['source_pointer'] for c in row['contexts']] == gold['locations'] + assert not row['approved_for_import'] and not row['final_decision_verified'] and 'emails' not in row subprocess.run([str(venv/'bin/paper-atlas'),'--help'],cwd=outside,check=True) subprocess.run([python,'-I','-B','-m','paper_atlas','collect','cvf','--help'],cwd=outside,check=True) subprocess.run([python,'-I','-B','-m','paper_atlas','collect','rss2026-proceedings','--help'],cwd=outside,check=True) diff --git a/README.md b/README.md index 6831df8..914f413 100644 --- a/README.md +++ b/README.md @@ -140,3 +140,5 @@ ECCV 2026 的[暂定节目采集器](docs/eccv2026-program.md)合并重复节目 `collect icdm2026-program` 离线核验 324 条 ICDM2026 官方线下日程,按明确的研究/应用标签输出 241/83 条;展示长短、缺失摘要及未核实出版身份分别保留,不自动入库。见 [ICDM2026 使用与恢复说明](docs/icdm2026-program.md)。 `collect recsys2026-program` 核验九个 track 的 263 项官方录用贡献观察(260 个不同 DOI),保留摘要、原始作者和重复 DOI 的来源版本。见 [RecSys 2026 使用与恢复说明](docs/recsys2026-program.md)。 + +`collect mm2026-program` 从已冻结官方快照恢复全部 28 个来源 track 的 2,143 个节目 ID,保留 4,326 次展示的来源位置与场次上下文,单列 216 个摘要缺失项;输出供贡献类型、出版及作品身份核查使用,不是可直接导入的论文/决定队列。见 [MM 2026 使用、来源和恢复说明](docs/mm2026-source-status.md)。 diff --git a/docs/mm2026-source-status.md b/docs/mm2026-source-status.md index c66dd76..0d44474 100644 --- a/docs/mm2026-source-status.md +++ b/docs/mm2026-source-status.md @@ -73,4 +73,20 @@ 服务器恢复包:`/data/paper-index/validation/mm2026-source-discovery-and-inventory-recovery-v1.tar.gz`;本地备份:`/Users/zzyang/Documents/Codex/2026-11-06/paper-index/` 同名文件。先安全解包并核验回执,再做规范化,不能重新抓取 403 入口作为恢复方式。 -本批尚未形成生产导入队列。下一步需要可复用采集器保留每个节目 ID 的全部展示上下文、明确贡献类型与来源边界,单列缺失字段,再做作品/出版身份及既往投稿审查。已消费的其他会议导入队列不可重放;最终全库去重和完整审计仍为独立必做阶段。 +## 可复用采集命令 + +在已配置的服务器环境中运行,`--snapshots` 是包含 `mm2026-program-and-public-probes-v1/` 的父目录;恢复时先安全解包上述已核验恢复包。命令只读快照,不请求网络、不连接数据库,不重新访问403入口,也不下载PDF: + +```sh +paper-atlas collect mm2026-program \ + --snapshots /data/paper-index/collected \ + --out /data/paper-index/collected/mm2026-native-v1 +``` + +输出目录必须不存在;已存在、失败或部分输出均不能覆盖重跑,应先查看报告,另选新的目录。`collection-report.json` 仅在全量验证成功后生成;失败保留 `failed.json`,不冒充完成。 + +采集器核验固定回执、正文哈希、URL、抓取时间、完整逐轨数量和字段缺失盘点。28个逐轨 JSONL 保留原始标题、作者、机构、逐字摘要以及同一节目ID的全部展示位置、时间、poster/oral号和展示kind;157个场次及子场次/非论文活动另存 `sessions.jsonl`。外层空track不覆盖 `paper.track`,不同核心版本、未知字段或未处理嵌套需要重新审查,不能静默丢弃或取第一条。公开联系邮箱数组只留在原始证据中,不进入规范记录。`evidence/` 保留原始节目JSON、完整回执及四个既有403响应。 + +这些文件的格式为 `mm_program_observation_v1_not_ingestion_record`,所有项 `approved_for_import=false`。节目ID仅为 `mm_program:program_id` 来源定位符,track标签使用 `mm_program:track`;不建立科学作品、原始投稿ID或最终决定,不推断 DOI/arXiv/OpenReview 标识、论文类型、拒稿或撤稿。行业报告、艺术、workshop及其他来源贡献均保留,尚需逐轨核对贡献类型与正式出版依据。 + +本批尚未形成生产导入队列。后续做贡献类型、作品/出版身份及既往投稿审查,再生成经批准的增量队列。已消费的其他会议导入队列不可重放;最终全库去重和完整审计仍为独立必做阶段。 diff --git a/paper_atlas/cli.py b/paper_atlas/cli.py index 3cf457e..71fb7e4 100644 --- a/paper_atlas/cli.py +++ b/paper_atlas/cli.py @@ -32,6 +32,7 @@ COMMANDS = { ('collect','icaps2026-program'): ('paper_atlas.collectors.icaps2026', {}, '核验 ICAPS 2026 官方录用名单并分开既有论文展示'), ('collect','rss2026-proceedings'): ('paper_atlas.collectors.rss2026', {}, '核验 RSS 2026 官方论文集与独立录用表'), ('collect','recsys2026-program'): ('paper_atlas.collectors.recsys2026', {}, '核验 RecSys 2026 公开录用贡献并保留 DOI 来源版本'), + ('collect','mm2026-program'): ('paper_atlas.collectors.mm2026', {}, '恢复 MM 2026 全部来源 track 与重复展示上下文,待贡献和身份核查'), ('collect','cvf'): ('paper_atlas.collectors.vision', {'source':'cvf'}, 'CVPR / ICCV 官方开放目录'), ('collect','ecva'): ('paper_atlas.collectors.vision', {'source':'ecva'}, 'ECCV 官方开放目录'), ('collect','aamas-proceedings'): ('paper_atlas.collectors.aamas', {}, '按官方目录区分 AAMAS 发表形式和摘要'), diff --git a/paper_atlas/collectors/mm2026.py b/paper_atlas/collectors/mm2026.py new file mode 100644 index 0000000..ddf1ecc --- /dev/null +++ b/paper_atlas/collectors/mm2026.py @@ -0,0 +1,202 @@ +"""Restore frozen ACM MM 2026 program observations, not publication decisions.""" +import argparse +from collections import Counter +from datetime import datetime +import hashlib +import json +from pathlib import Path +import re + +SOURCE = 'mm_program' +SNAPSHOT = 'mm2026-program-and-public-probes-v1' +URL = 'https://2026.acmmm.org/site/program-data.json' +BODY_SHA = '08e2976ecba2d0d535000b0935f41c85076ec674e518a1df2c5843d9bde4d083' +MANIFEST_SHA = 'e43ac6cdfc86bac61305553a044dd6148800b5aa41a5d1bf06c486e90d0ed7b8' +TRACK_COUNTS = {'MRAC':35,'SUMAC':9,'NeuroMM':10,'IMPACT':8,'RichMediaGAI':10, + 'MMSports':16,'MUWS':5,'EReL-MIR':7,'GreenMM':10,'DFF':12,'Main Track':1703, + 'Grand Challenge':97,'Doctoral Symposium':4,'Datasets':50,'Interactive Art':23, + 'Reproducibility':1,'Demo and Video':27,'Open Source':14,'Industry Talk':7, + 'BNI':25,'QoEVMA':4,'HUMA':14,'WIPAI':8,'AMI':11,'Mulsemedia':9, + 'UAVM':11,'MCHM':7,'JAV-CG':6} +CORE = ('paper_id','title','authors','affiliations','abstract','track') +PRESENTATION = ('kind','start_time','end_time','poster_id','oral_id') + + +def digest(raw): + return hashlib.sha256(raw).hexdigest() + + +def safe_read(path, limit): + if any(p.is_symlink() for p in (path, *path.parents)) or not path.is_file() or path.stat().st_size > limit: + raise ValueError('Unsafe snapshot path or size') + return path.read_bytes() + + +def load_snapshot(root): + directory = Path(root)/SNAPSHOT + raw = safe_read(directory/'manifest.json', 128*1024) + if digest(raw) != MANIFEST_SHA: + raise ValueError('Reviewed manifest hash changed') + manifest = json.loads(raw) + if manifest.get('database_written') is not False or len(manifest['entries']) != 5: + raise ValueError('Unreviewed manifest scope') + evidence = {'manifest.json':raw} + for receipt in manifest['entries']: + name = receipt['file'] + if Path(name).name != name or name in evidence: + raise ValueError('Unsafe or duplicate receipt name') + body = safe_read(directory/name, 12*1024**2) + if (receipt['bytes'] != len(body) or receipt['sha256'] != digest(body) + or receipt['url'] != receipt['final_url'] + or not receipt['content_type'].startswith('application/json') + or datetime.fromisoformat(receipt['at']).utcoffset() is None): + raise ValueError('Snapshot receipt/body disagree') + if name == 'program-data.json': + if receipt['url'] != URL or receipt['http_status'] != 200 or digest(body) != BODY_SHA: + raise ValueError('Wrong program source') + elif receipt['http_status'] != 403: + raise ValueError('Unreviewed public-access result') + evidence[name] = body + if 'program-data.json' not in evidence: + raise ValueError('Missing program source') + return evidence + + +def without_emails(value): + """Keep public session structure; contact arrays remain in raw evidence only.""" + if isinstance(value, dict): + return {k:without_emails(v) for k,v in value.items() if k != 'emails'} + if isinstance(value, list): + return [without_emails(v) for v in value] + return value + + +def program_locations(node, pointer=''): + """Independent structural guard against silently skipping a new nesting form.""" + if isinstance(node, dict): + if 'paper_id' in node: + yield pointer + for key,value in node.items(): + escaped = key.replace('~','~0').replace('/','~1') + yield from program_locations(value, pointer+'/'+escaped) + elif isinstance(node, list): + for i,value in enumerate(node): + yield from program_locations(value, pointer+'/'+str(i)) + + +def parse_program(raw): + data = json.loads(raw) + if (data.get('generated_at') != '2026-11-14T18:36:26.860677+00:00' + or data.get('days') != [1,2,3,4,5] or len(data.get('rooms',[])) != 20 + or len(data.get('entries',[])) != 157): + raise ValueError('Unreviewed program edition or census') + records = {}; sessions = []; seen = [] + + def observe(paper, pointer, session_pointer, subsession_pointer=None): + if not set(CORE) <= paper.keys() or set(paper)-set(CORE)-set(PRESENTATION)-{'emails'}: + raise ValueError('Unreviewed program fields') + for key in ('paper_id','title','abstract','track'): + if not isinstance(paper[key],str): + raise ValueError('Wrong program field type') + for key in ('authors','affiliations'): + if not isinstance(paper[key],list) or any(not isinstance(v,str) for v in paper[key]): + raise ValueError('Wrong author or affiliation list') + if not paper['paper_id'].strip() or not paper['title'].strip() or paper['track'] not in TRACK_COUNTS: + raise ValueError('Missing identity/title or unreviewed source track') + core = {k:paper[k] for k in CORE} + pid = paper['paper_id']; seen.append(pointer) + if pid not in records: + missing = [k for k in ('authors','abstract') if not paper[k]] + records[pid] = {**core,'source':SOURCE,'year':2026,'venue':'MM', + 'source_locator':{'namespace':'mm_program:program_id','value':pid}, + 'observed_status':'listed_in_official_program', + 'publication_identity_verified':False,'final_decision_verified':False, + 'approved_for_import':False,'missing_fields':missing, + 'tags':[{'namespace':'mm_program:track','value':paper['track']}], + 'contexts':[]} + elif {k:records[pid][k] for k in CORE} != core: + raise ValueError('Conflicting core versions for one program ID; review required') + records[pid]['contexts'].append({'source_pointer':pointer,'session_pointer':session_pointer, + 'subsession_pointer':subsession_pointer, + 'presentation':{k:paper[k] for k in PRESENTATION if k in paper}}) + + for i,entry in enumerate(data['entries']): + pointer = '/entries/'+str(i) + if entry['track'] != '': + raise ValueError('Unreviewed outer track; do not override paper.track') + session = without_emails({k:v for k,v in entry.items() if k not in ('papers','subsessions')}) + session.update(source_pointer=pointer,subsessions=[]) + for j,paper in enumerate(entry.get('papers') or []): + observe(paper,pointer+'/papers/'+str(j),pointer) + for j,sub in enumerate(entry.get('subsessions') or []): + sp = pointer+'/subsessions/'+str(j) + saved = without_emails({k:v for k,v in sub.items() if k != 'items'}) + saved.update(source_pointer=sp,nonpaper_items=[]) + for k,item in enumerate(sub.get('items') or []): + ip = sp+'/items/'+str(k) + if 'paper_id' in item: + observe(item,ip,pointer,sp) + else: + saved['nonpaper_items'].append({'source_pointer':ip,'fields':without_emails(item)}) + session['subsessions'].append(saved) + sessions.append(session) + if set(seen) != set(program_locations(data)) or len(set(seen)) != len(seen): + raise ValueError('Unhandled or duplicated program location') + items = list(records.values()) + if (len(seen) != 4326 or len(items) != 2143 + or Counter(r['track'] for r in items) != TRACK_COUNTS + or sum(not r['abstract'] for r in items) != 216 + or {r['paper_id'] for r in items if not r['authors']} != {'mfp4971','mfp7801'}): + raise ValueError('Program count or missing-field census changed') + return items,sessions + + +def collect(out, *, snapshots): + out = Path(out) + if out.exists() or out.is_symlink(): + raise ValueError('Output exists; use a fresh directory') + if any(p.is_symlink() for p in out.parents): + raise ValueError('Unsafe output parent') + out.mkdir(parents=True) + try: + evidence = load_snapshot(snapshots) + rows,sessions = parse_program(evidence['program-data.json']) + (out/'evidence').mkdir() + for name,body in evidence.items(): + (out/'evidence'/name).write_bytes(body) + outputs = [] + for track,count in TRACK_COUNTS.items(): + slug = re.sub('[^a-z0-9]+','_',track.lower()).strip('_') + path = out/('mm-program-2026-'+slug+'.jsonl') + raw = ''.join(json.dumps(r,ensure_ascii=False)+'\n' for r in rows if r['track']==track).encode() + path.write_bytes(raw) + outputs.append({'file':str(path),'file_sha256':digest(raw),'count':count, + 'track_literal':track,'format':'mm_program_observation_v1_not_ingestion_record', + 'status':'pending_contribution_and_identity_review','approved_for_import':0}) + session_raw = ''.join(json.dumps(r,ensure_ascii=False)+'\n' for r in sessions).encode() + (out/'sessions.jsonl').write_bytes(session_raw) + report = {'status':'collected_pending_contribution_and_identity_review','source':SOURCE, + 'records':2143,'program_observations':4326,'extra_presentation_observations':2183, + 'track_counts':TRACK_COUNTS,'files':outputs,'session_count':157, + 'sessions_file_sha256':digest(session_raw),'missing_abstract_count':216, + 'missing_author_program_ids':['mfp4971','mfp7801'], + 'source_body_sha256':BODY_SHA,'source_manifest_sha256':MANIFEST_SHA, + 'database_written':False,'pdfs_downloaded':False,'network_requests':0,'approved_for_import':0, + 'public_negative_access':{'anonymous_notes_probes':4,'http_statuses':[403]*4, + 'negative_decisions_inferred':0,'scope':'Observed endpoints only; not proof of zero public negative decisions.'}, + 'scope':'Program IDs are source locators, not scientific-work or original submission IDs. All28source tracks retained; presentation times/kinds/locators and session context kept separately. No publication status, paper type, DOI, arxiv/OpenReview ID, rejection or withdrawal inferred. Blank fields are not negative decisions. Contribution/publication/work/history review and final whole-database audit remain.'} + part = out/'collection-report.json.part' + part.write_text(json.dumps(report,ensure_ascii=False,indent=2)+'\n') + part.replace(out/'collection-report.json') + return report + except Exception as error: + (out/'failed.json').write_text(json.dumps({'status':'failed_not_complete','error_type':type(error).__name__,'database_written':False})+'\n') + raise + + +def main(argv=None, *, prog=None): + parser = argparse.ArgumentParser(prog=prog,description=__doc__) + parser.add_argument('--snapshots',required=True) + parser.add_argument('--out',required=True) + args = parser.parse_args(argv) + print(json.dumps(collect(args.out,snapshots=args.snapshots),ensure_ascii=False)) diff --git a/tests/fixtures/mm2026-source.json.gz b/tests/fixtures/mm2026-source.json.gz new file mode 100644 index 0000000..a855eae Binary files /dev/null and b/tests/fixtures/mm2026-source.json.gz differ diff --git a/tests/test_mm2026.py b/tests/test_mm2026.py new file mode 100644 index 0000000..2eab367 --- /dev/null +++ b/tests/test_mm2026.py @@ -0,0 +1,146 @@ +import copy +import gzip +import hashlib +import json +from pathlib import Path +import tempfile +import unittest + +from paper_atlas.collectors import mm2026 as c + + +class MM2026Test(unittest.TestCase): + @classmethod + def setUpClass(cls): + cls.fixture = json.loads(gzip.decompress((Path(__file__).parent/'fixtures/mm2026-source.json.gz').read_bytes())) + cls.files = {name:body.encode('utf-8') for name,body in cls.fixture['files'].items()} + cls.raw = cls.files[c.SNAPSHOT+'/program-data.json'] + cls.data = json.loads(cls.raw) + cls.gold = cls.fixture['independent_inventory'] + + def materialize(self, root): + for name,body in self.files.items(): + path = root/name; path.parent.mkdir(parents=True,exist_ok=True); path.write_bytes(body) + + def parse_changed(self, data): + return c.parse_program(json.dumps(data,ensure_ascii=False).encode()) + + def test_all2143_ids_and4326_locations_against_independent_inventory(self): + rows,sessions = c.parse_program(self.raw) + self.assertEqual(len(rows),2143); self.assertEqual(len(sessions),157) + self.assertEqual(sum(len(r['contexts']) for r in rows),4326) + actual = {r['paper_id']:r for r in rows} + self.assertEqual(set(actual),{g['paper_id'] for g in self.gold['inventory']}) + for gold in self.gold['inventory']: + r = actual[gold['paper_id']] + self.assertEqual((r['title'],r['authors'],r['track'],r['affiliations']), + (gold['title'],gold['authors'],gold['track_literal'],gold['affiliations'])) + self.assertEqual(hashlib.sha256(r['abstract'].encode()).hexdigest(),gold['abstract_sha256']) + self.assertEqual([x['source_pointer'] for x in r['contexts']],gold['locations']) + + def test_every_presentation_and_session_context_is_preserved(self): + rows,sessions = c.parse_program(self.raw) + for r in rows: + for context in r['contexts']: + node = self.data + for key in context['source_pointer'].split('/')[1:]: + node = node[int(key)] if isinstance(node,list) else node[key] + self.assertEqual(context['presentation'],{k:node[k] for k in c.PRESENTATION if k in node}) + self.assertEqual(r['paper_id'],node['paper_id']) + for i,session in enumerate(sessions): + original = self.data['entries'][i] + self.assertEqual(session['source_pointer'],'/entries/'+str(i)) + self.assertEqual(session['track'],'') + for key in original.keys()-{'papers','subsessions'}: + self.assertEqual(session[key],original[key]) + for j,sub in enumerate(original.get('subsessions') or []): + saved = session['subsessions'][j] + self.assertEqual(saved['title'],sub['title']) + expected = [x for x in sub.get('items') or [] if 'paper_id' not in x] + self.assertEqual([x['fields'] for x in saved['nonpaper_items']],expected) + + def test_all_tracks_retained_without_publication_or_negative_inference(self): + rows,_ = c.parse_program(self.raw) + self.assertEqual({r['track'] for r in rows},set(self.gold['unique_id_track_counts'])) + for r in rows: + self.assertFalse(r['approved_for_import']); self.assertFalse(r['final_decision_verified']) + self.assertFalse(r['publication_identity_verified']) + self.assertEqual(r['source_locator'],{'namespace':'mm_program:program_id','value':r['paper_id']}) + self.assertEqual(r['tags'],[{'namespace':'mm_program:track','value':r['track']}]) + self.assertEqual(r['observed_status'],'listed_in_official_program') + self.assertTrue({'status','identifiers','submission_identifiers','paper_type','emails'}.isdisjoint(r)) + missing = [r for r in rows if not r['authors']] + self.assertEqual({r['paper_id'] for r in missing},{'mfp4971','mfp7801'}) + self.assertTrue(all(r['missing_fields']==['authors','abstract'] for r in missing)) + self.assertEqual(sum(not r['abstract'] for r in rows),216) + + def test_collect_evidence_hashes_and_no_contact_arrays_in_normalized_output(self): + with tempfile.TemporaryDirectory() as tmp: + root=Path(tmp);self.materialize(root/'src') + report=c.collect(root/'out',snapshots=root/'src') + self.assertEqual((report['records'],report['program_observations'],report['extra_presentation_observations']),(2143,4326,2183)) + self.assertFalse(report['database_written']);self.assertFalse(report['pdfs_downloaded']) + self.assertEqual(report['network_requests'],0);self.assertEqual(report['approved_for_import'],0) + self.assertEqual(report['public_negative_access']['http_statuses'],[403]*4) + self.assertEqual(report['track_counts'],self.gold['unique_id_track_counts']) + for name,body in self.files.items(): + self.assertEqual((root/'out/evidence'/Path(name).name).read_bytes(),body) + n=0 + for item in report['files']: + raw=Path(item['file']).read_bytes();self.assertEqual(hashlib.sha256(raw).hexdigest(),item['file_sha256']) + self.assertNotIn(b'"emails"',raw);self.assertEqual(len(raw.splitlines()),item['count']);n+=item['count'] + self.assertEqual(item['format'],'mm_program_observation_v1_not_ingestion_record') + self.assertEqual(n,2143) + raw=(root/'out/sessions.jsonl').read_bytes() + self.assertEqual(hashlib.sha256(raw).hexdigest(),report['sessions_file_sha256']);self.assertNotIn(b'"emails"',raw) + with self.assertRaisesRegex(ValueError,'Output exists'):c.collect(root/'out',snapshots=root/'src') + + def test_conflicting_core_version_is_not_silently_merged(self): + data=copy.deepcopy(self.data) + first=data['entries'][0]['papers'][0] + data['entries'][0]['papers'].append({**first,'abstract':'conflicting abstract'}) + with self.assertRaisesRegex(ValueError,'Conflicting core'):self.parse_changed(data) + + def test_unknown_nesting_cannot_silently_lose_observations(self): + data=copy.deepcopy(self.data) + moved=data['entries'][0]['papers'].pop() + data['entries'][0]['unreviewed_bucket']=[moved] + with self.assertRaisesRegex(ValueError,'Unhandled'):self.parse_changed(data) + + def test_track_drift_unknown_fields_or_missing_title_require_review(self): + for kind in ('outer_track','paper_track','field','title','authors'): + data=copy.deepcopy(self.data);paper=data['entries'][0]['papers'][0] + if kind=='outer_track':data['entries'][0]['track']='Main Track' + elif kind=='paper_track':paper['track']='New Track' + elif kind=='field':paper['doi']='unreviewed' + elif kind=='title':paper['title']='' + else:paper['authors']='Unparsed author string' + with self.subTest(kind=kind),self.assertRaises(ValueError):self.parse_changed(data) + + def test_edition_or_census_change_requires_review(self): + for kind in ('edition','count'): + data=copy.deepcopy(self.data) + if kind=='edition':data['generated_at']='2025-11-14T18:36:26.860677+00:00' + else:data['entries'][0]['papers'].pop() + with self.subTest(kind=kind),self.assertRaises(ValueError):self.parse_changed(data) + + def test_corruption_missing_evidence_and_symlink_fail_without_completed_report(self): + for kind in ('manifest','program','probe','missing','symlink'): + with tempfile.TemporaryDirectory() as tmp: + root=Path(tmp);self.materialize(root/'src');directory=root/'src'/c.SNAPSHOT + name='manifest.json' if kind=='manifest' else 'openreview-rejected-probe.json' if kind=='probe' else 'program-data.json' + path=directory/name + if kind=='missing':path.unlink() + elif kind=='symlink': + original=directory/'original.json';path.rename(original);path.symlink_to(original) + else:path.write_bytes(path.read_bytes()+b' ') + with self.subTest(kind=kind),self.assertRaises((ValueError,FileNotFoundError)):c.collect(root/'out',snapshots=root/'src') + self.assertTrue((root/'out/failed.json').exists());self.assertFalse((root/'out/collection-report.json').exists()) + + def test_literal_abstract_math_and_nonpaper_session_events_retained(self): + rows,sessions=c.parse_program(self.raw) + self.assertTrue(any('\\' in r['abstract'] for r in rows)) + self.assertTrue(any('\n' in r['abstract'] for r in rows)) + self.assertTrue(any(s['type']=='coffee_break' for s in sessions)) + self.assertTrue(any(s['type']=='industry_talk' for s in sessions)) + self.assertEqual(c.without_emails({'emails':['contact'],'items':[{'emails':['contact'],'title':'talk'}]}),{'items':[{'title':'talk'}]})