ledger: extract stackchain magazine photo catalog into a table

2026-10-20 20:18:17 +00:00
1 parent f44a285454
commit c0dea95d77
@@ -0,0 +1,30 @@
+no,entry,type,text,words
+1,"img_67eedaf526e7.jpg","COVER","STACKCHAIN
+MAGAZINE
+ROUND 5
+INTERVIEW WITH MADEX
+BY ISBELL
+88K CHRONICLES: HOW ICO CONSPIRED
+TO SUPPRESS THE BITCOIN PRICE
+BY DENIZ
+REASONS TO BUY ALTCOINS
+BY SKYLER
+SELF-CUSTODY//BITCOIN ONLY
+MAGAZINE
+18+
+stackchain
+MAGAZINE
+MADE IN [unreadable]
+[unreadable graffiti tags on the owl artwork]",43
+2,"img_fcb2e6a8f066.jpg","PHOTO_OF_PERSON_HOLDING_MAGAZINE","B/H872.372
+[unreadable]O CHAIN
+[unreadable]INE
+[unreadable]OUND 5
+MACHINA
+SELF-CUSTODY//BIT[unreadable]",8
+3,"img_72f52468393f.jpg","(not transcribed)","—",0
+4,"img_d5ada9cee4e9.jpg","(not transcribed)","—",0
+5,"img_7333a3df0aac.jpg","(not transcribed)","—",0
+6,"img_05d51da1fd9d.jpg","(not transcribed)","—",0
+7,"img_125532ea2e01.jpg","(not transcribed)","—",0
+8,"img_1325775751e6.jpg","(not transcribed)","—",0
@@ -0,0 +1,58 @@
+[
+  {
+    "no": 1,
+    "entry": "img_67eedaf526e7.jpg",
+    "type": "COVER",
+    "text": "STACKCHAIN\nMAGAZINE\nROUND 5\nINTERVIEW WITH MADEX\nBY ISBELL\n88K CHRONICLES: HOW ICO CONSPIRED\nTO SUPPRESS THE BITCOIN PRICE\nBY DENIZ\nREASONS TO BUY ALTCOINS\nBY SKYLER\nSELF-CUSTODY//BITCOIN ONLY\nMAGAZINE\n18+\nstackchain\nMAGAZINE\nMADE IN [unreadable]\n[unreadable graffiti tags on the owl artwork]",
+    "words": 43
+  },
+  {
+    "no": 2,
+    "entry": "img_fcb2e6a8f066.jpg",
+    "type": "PHOTO_OF_PERSON_HOLDING_MAGAZINE",
+    "text": "B/H872.372\n[unreadable]O CHAIN\n[unreadable]INE\n[unreadable]OUND 5\nMACHINA\nSELF-CUSTODY//BIT[unreadable]",
+    "words": 8
+  },
+  {
+    "no": 3,
+    "entry": "img_72f52468393f.jpg",
+    "type": "(not transcribed)",
+    "text": null,
+    "words": 0
+  },
+  {
+    "no": 4,
+    "entry": "img_d5ada9cee4e9.jpg",
+    "type": "(not transcribed)",
+    "text": null,
+    "words": 0
+  },
+  {
+    "no": 5,
+    "entry": "img_7333a3df0aac.jpg",
+    "type": "(not transcribed)",
+    "text": null,
+    "words": 0
+  },
+  {
+    "no": 6,
+    "entry": "img_05d51da1fd9d.jpg",
+    "type": "(not transcribed)",
+    "text": null,
+    "words": 0
+  },
+  {
+    "no": 7,
+    "entry": "img_125532ea2e01.jpg",
+    "type": "(not transcribed)",
+    "text": null,
+    "words": 0
+  },
+  {
+    "no": 8,
+    "entry": "img_1325775751e6.jpg",
+    "type": "(not transcribed)",
+    "text": null,
+    "words": 0
+  }
+]
@@ -0,0 +1,15 @@
+# Stackchain Magazine photo catalog — extracted
+
+Source: `stackchain_catalog_raw.json` (8 entries).
+Transcription present: 2 of 8 entries; 6 entries carry no `text`.
+
+| No. | Entry | Type | Transcription | Words |
+|---|---|---|---|---|
+| 1 | `img_67eedaf526e7.jpg` | COVER | STACKCHAIN MAGAZINE ROUND 5 INTERVIEW WITH MADEX BY ISBELL 88K CHRONICLES: HOW ICO CONSPIRED TO SUPPRESS THE BITCOIN PRICE BY DENIZ REASONS… | 43 |
+| 2 | `img_fcb2e6a8f066.jpg` | PHOTO_OF_PERSON_HOLDING_MAGAZINE | B/H872.372 [unreadable]O CHAIN [unreadable]INE [unreadable]OUND 5 MACHINA SELF-CUSTODY//BIT[unreadable] | 8 |
+| 3 | `img_72f52468393f.jpg` | (not transcribed) | — | 0 |
+| 4 | `img_d5ada9cee4e9.jpg` | (not transcribed) | — | 0 |
+| 5 | `img_7333a3df0aac.jpg` | (not transcribed) | — | 0 |
+| 6 | `img_05d51da1fd9d.jpg` | (not transcribed) | — | 0 |
+| 7 | `img_125532ea2e01.jpg` | (not transcribed) | — | 0 |
+| 8 | `img_1325775751e6.jpg` | (not transcribed) | — | 0 |
@@ -0,0 +1,107 @@
+#!/usr/bin/env python3
+"""CHECK: re-derive the catalog from the raw JSON source and diff against the
+produced artifact (catalog.json + catalog.md). Exit 0 only if they agree.
+
+This is a real regression test, not a tautology: it re-reads the source file and
+re-derives the rows, then compares against the committed artifact bytes.
+"""
+import json
+import re
+import sys
+
+SRC = '/root/stackchain_catalog_raw.json'
+ACTUAL_JSON = '/root/.hermes/wizard-outputs/ledger/catalog.json'
+ACTUAL_MD = '/root/.hermes/wizard-outputs/ledger/catalog.md'
+
+TYPE_RE = re.compile(r'"type"\s*:\s*"(.*?)"')
+TEXT_RE = re.compile(r'"text"\s*:\s*"(.*?)"', re.DOTALL)
+DOUBLE_N = '\\\\n'
+SINGLE_N = '\\n'
+
+
+def reextract():
+    raw = json.load(open(SRC, encoding='utf-8'))
+    rows = []
+    for idx, (fname, v) in enumerate(raw.items(), 1):
+        m_type = TYPE_RE.search(v)
+        ftype = m_type.group(1) if m_type else None
+        m_text = TEXT_RE.search(v)
+        text = m_text.group(1) if m_text else None
+        if text is not None:
+            text = text.replace(DOUBLE_N, '\n').replace(SINGLE_N, '\n')
+        rows.append({
+            'no': idx,
+            'entry': fname,
+            'type': ftype if ftype is not None else '(not transcribed)',
+            'text': text,
+            'words': len(text.split()) if text else 0,
+        })
+    return rows
+
+
+def load_actual():
+    rows = json.load(open(ACTUAL_JSON, encoding='utf-8'))
+    md = open(ACTUAL_MD, encoding='utf-8').read()
+    return rows, md
+
+
+def md_table_rows(md):
+    rows = []
+    for line in md.splitlines():
+        if line.startswith('|') and '---' not in line and 'No.' not in line:
+            cells = [c.strip() for c in line.strip('|').split('|')]
+            if cells:
+                rows.append(cells)
+    return rows
+
+
+def main():
+    errors = []
+    expected = reextract()
+    actual, md = load_actual()
+
+    if len(actual) != len(expected):
+        errors.append(f'row count: actual={len(actual)} expected={len(expected)}')
+
+    for i, (a, e) in enumerate(zip(actual, expected), 1):
+        for key in ('no', 'entry', 'type', 'text', 'words'):
+            if a.get(key) != e[key]:
+                errors.append(f'row {i} field {key}: actual={a.get(key)!r} expected={e[key]!r}')
+
+    # markdown sanity: every entry must appear, with a type, and word counts must
+    # match the numbers in the table.
+    seen = set()
+    for cells in md_table_rows(md):
+        if len(cells) == 5 and cells[1].startswith('`'):
+            seen.add(cells[1].strip('`'))
+            if cells[3] == '—':
+                if actual[int(cells[0]) - 1]['text'] is not None:
+                    errors.append(f'row {cells[0]} shows — but source text present')
+            else:
+                text = actual[int(cells[0]) - 1]['text']
+                if text is None:
+                    errors.append(f'row {cells[0]} preview missing but source text absent')
+                else:
+                    single = ' '.join(text.split())
+                    exp_preview = single[:140].rstrip() + ('…' if len(single) > 140 else '')
+                    if cells[3] != exp_preview:
+                        errors.append(f'row {cells[0]} preview mismatch: {cells[3]!r} vs {exp_preview!r}')
+            if cells[2] != actual[int(cells[0]) - 1]['type']:
+                errors.append(f'row {cells[0]} type: {cells[2]} vs {actual[int(cells[0]) - 1]["type"]}')
+    expected_entries = {r['entry'] for r in expected}
+    if seen != expected_entries:
+        errors.append('md entries missing/extra: ' + ', '.join(sorted(seen ^ expected_entries)))
+
+    if errors:
+        print('CHECK FAILED:')
+        for e in errors:
+            print(' -', e)
+        sys.exit(1)
+    print(f'CHECK OK: {len(actual)} rows match the source exactly '
+          f'({sum(1 for r in actual if r["text"])} transcribed, '
+          f'{sum(1 for r in actual if not r["text"])} empty).')
+    sys.exit(0)
+
+
+if __name__ == '__main__':
+    main()