unification: dossiers → CANON v1 (104/142 changed: H1 [dates] incl. open-ended for living, Homonym warnings unified, RU-name sections → canonical field, Round logs → Notes, Dossier status dropped)

tools/migrate_dossiers.py

@@ -0,0 +1,91 @@
+#!/usr/bin/env python3
+"""Normalize author dossiers to CANON v1 (lossless, mechanical).
+Usage: python3 tools/migrate_dossiers.py [--dry-run]"""
+import re, sys, glob, os
+
+def migrate(p, dry):
+    c = open(p, encoding='utf-8').read()
+    o = c
+    # A1: H1 dates (19xx-19yy) / (19xx) → [..]
+    c = re.sub(r'^(# .+?) \(((?:19|18)\d{2})(?:[-–](?:19|20)\d{2})?\)\s*$',
+               r'\1 [\2\3]'.replace('\\3', lambda m: ''), c, count=1, flags=re.M) if False else c
+    m = re.match(r'^# (.+?) \(((?:19|18)\d{2}(?:[-–](?:19|20)\d{2})?)\)\s*$', c, re.M)
+    if m:
+        c = c.replace(f'# {m.group(1)} ({m.group(2)})', f'# {m.group(1)} [{m.group(2)}]', 1)
+    # open-ended (LIVING authors): (19xx-) → [19xx-] — never close an open range
+    m = re.match(r'^# (.+?) \(((?:19|18)\d{2})-\)\s*$', c, re.M)
+    if m:
+        c = c.replace(f'# {m.group(1)} ({m.group(2)}-)', f'# {m.group(1)} [{m.group(2)}-]', 1)
+    # A2: homonym heading variants
+    c = re.sub(r'^## Homonym warning \([0-9-]+\)\s*$', '## Homonym warnings', c, flags=re.M)
+    c = re.sub(r'^## Homonyms / notes\s*$', '## Homonym warnings', c, flags=re.M)
+    # A3: ## RU name (RSL query) → bold field
+    c = re.sub(r'^## RU name \(RSL query\)\s*\n+(.+?)\n',
+               r'**RU name (canonical):** \1\n', c, flags=re.M)
+    c = re.sub(r'^## RU name \([^)]*\)\s*\n+(.+?)\n',
+               r'**RU name (canonical):** \1\n', c, flags=re.M)
+    # A4: RU editions / RU works sections → Verified RU works
+    for m in re.finditer(r'^## (RU editions|RU works)[^\n]*\s*$', c, re.M):
+        pass
+    def merge_ru(m):
+        title = m.group(0)
+        if '## Verified RU works' in c:
+            return f'### {title[3:] if title.startswith("## ") else title}'
+        return '## Verified RU works'
+    c = re.sub(r'^## RU editions[^\n]*$', merge_ru, c, flags=re.M)
+    c = re.sub(r'^## RU works[^\n]*$', merge_ru, c, flags=re.M)
+    # A5: Round sections → under Notes (or Notes itself)
+    for m in re.finditer(r'^## (Round [^\n]+)\s*$', c, re.M):
+        title = m.group(1)
+        if '## Notes' in c:
+            c = c.replace(f'## {title}\n', f'### {title}\n', 1)
+        else:
+            c = c.replace(f'## {title}\n', f'## Notes\n\n*{title}*\n', 1)
+    # A6: drop Dossier status line
+    c = re.sub(r'\*\*Dossier status:\*\* [^\n]*\n', '', c)
+    # A7: EN identity → Field / identity
+    c = re.sub(r'^\*\*EN identity:\*\*', '**Field / identity:**', c, flags=re.M)
+    c = re.sub(r'^\*\*Identity:\*\*', '**Field / identity:**', c, flags=re.M)
+    c = re.sub(r'^\*\*Identity \(verified[^)]*\):\*\*', '**Field / identity:**', c, flags=re.M)
+    # collapse 3+ blank lines
+    c = re.sub(r'\n{3,}', '\n\n', c)
+    if c != o and not dry:
+        open(p, 'w', encoding='utf-8').write(c)
+    return c != o
+
+def validate(p):
+    c = open(p, encoding='utf-8').read()
+    iss = []
+    if not c.startswith('# '):
+        iss.append('no H1')
+    if c.count('## Verified RU works') > 1:
+        iss.append(f"Verified RU works x{c.count('## Verified RU works')}")
+    if c.count('## Notes') > 1:
+        iss.append(f"Notes x{c.count('## Notes')}")
+    if c.count('## Homonym warnings') > 1:
+        iss.append('Homonym warnings dup')
+    if c.count('## List items (ISAP Zurich)') > 1:
+        iss.append('List items dup')
+    import re as _re
+    for t in (r'^## RU name \(RSL query\)', r'^\*\*Dossier status:\*\*', r'^\*\*EN identity:\*\*',
+              r'^## Round ', r'^## RU editions', r'^## RU works'):
+        if _re.search(t, c, _re.M):
+            iss.append(f'leftover: {t}')
+    return iss
+
+def main():
+    dry = '--dry-run' in sys.argv
+    n = 0
+    for p in sorted(glob.glob('authors/*.md')):
+        if migrate(p, dry):
+            n += 1
+    print(f'changed: {n} / {len(glob.glob("authors/*.md"))}')
+    bad = 0
+    for p in sorted(glob.glob('authors/*.md')):
+        for i in validate(p):
+            bad += 1
+            print(f'{p}: {i}')
+    print(f'validation issues: {bad}')
+
+if __name__ == '__main__':
+    main()