Skip to content

feat: add the Cinematic Terminology Glossary (Q141609371) - #57

Open
forzagreen wants to merge 1 commit into
mainfrom
claude/compassionate-cerf-a3n3qc
Open

forzagreen wants to merge 1 commit into
mainfrom
claude/compassionate-cerf-a3n3qc

Conversation

@forzagreen

@forzagreen forzagreen commented Sep 30, 2026 •

Copy link
Copy Markdown
Owner

Adds cinematic_terms: Cinematic Terminology Glossary / معجم المصطلحات السينمائية: للسينما والشاشة والإعلام (Q141609371). It is the English→Arabic glossary of film, screen and media terms compiled by Dr. Muhammad Amer (first edition, Cairo, 2026).

  • 2817 terms, with printed page numbers (2–162).
  • 1146 terms have a description. It holds the Arabic explanatory note that the book prints as a footnote.
  • name_arabic and name_english use the Wikidata ar/en labels, with (2026) added to the Arabic. name_french is NULL: the source has no French column and Wikidata has no fr label.
  • dict_type is terminology, tier is 4 (see the review note below).

Extraction

The source is the PDF linked from the Wikidata item (P953). It is a 262-page Word export, parsed from the table cell geometry: the № / English / Arabic columns, one row per № cell, and footnotes below the table.

  • Serials 1–2817 are complete and contiguous, matching the book's stated entry count.
  • There are 1148 footnotes, matching the stated count of annotated entries; every one is used exactly once.
  • Cross-check against the book's own Arabic→English index (pages 163–247, a separate layout): the Arabic of all 2817 entries matches. The only exceptions are 4 editorial differences in the book itself, and the main glossary's spelling is kept:
    • 378 is missing from the index.
    • 607 إيلينج / إيلينغ
    • 899 الجلفانوجرافيا / الجلفانوغرافيا
    • 1005 الجروتيسك / الغروتيسك
  • The PDF's Arabic text layer needed repairs, applied systematically:
    • lam-alef ligatures come out reversed (االستوديو → الاستوديو);
    • the fathatan glyph maps to an alef (قديماا → قديماً);
    • lines mixing Arabic with Latin words or digits come out in the wrong order ((ضوء بقوة1 )كيلو واط → ضوء إيس (ضوء بقوة 1 كيلو واط)). These were rebuilt from glyph positions.
  • In the notes, a space is added where the source runs an Arabic word straight into a Latin one (شركةSun → شركة Sun).

Source fixes

  • Co- Feature, Co- Principal Photography and Cut- In Reveal lose the stray space after the hyphen.
  • J-Cut (1303): the PDF renders Jقطع بشكل; it is stored as قطع بشكل J.
  • Barndoors (227) and Non-Diegetic Sound (1786) each have two numbered senses with one footnote each. Their description is ١. … ٢. … to match the Arabic.

For review

  • Tier: the ranking spreadsheet isn't in the repo. I picked 4 because this is an individually compiled, self-published glossary, closer to the tier-4 entries than to the academy dictionaries at tier 3. Please adjust if needed.
  • Notes punctuation: some notes keep the punctuation the PDF itself renders out of place around quotes and brackets mixed with Latin words, e.g. "التحريك (Agitation) "تحريك. The Arabic terms are not affected.

Dumps

  • Regenerated with make regenerate_dumps, which also rewrites the README table; make validate passes.
  • The SQLite dump was produced with sqlite3 3.53.4, whose .dump output uses the same unistr(...) format as the committed file. Its diff is exactly 2818 added lines and none removed.
  • The MariaDB dump only adds rows. The other differences are cosmetic header and autocommit lines, because it was made with mariadb-dump 11.8.9 (mariadb:11.8 image) instead of 11.8.5.

🤖 Generated with Claude Code

Add cinematic_terms, the English-Arabic glossary of film, screen and media
terms compiled by Dr. Muhammad Amer (first edition, Cairo, 2026): 2817 terms,
1146 of them with the Arabic explanatory note printed as a footnote.

- extracted from the PDF linked from the Wikidata item (P953); serials
  1-2817 are complete and contiguous, and every Arabic term was checked
  against the book's own Arabic-English index (only 4 editorial differences
  there, e.g. إيلينج / إيلينغ; the main glossary's spelling is kept)
- page = printed page number (2-162)
- description = the entry's footnote(s); the two entries with two numbered
  senses and two footnotes get "١. ... ٢. ..." to match their Arabic
- the PDF's Arabic text needed repairs on extraction: reversed lam-alef
  ligatures, fathatan glyphs mapped to an alef, and the word order of lines
  mixing Arabic with Latin words or digits
- a space is added where the source runs an Arabic word into a Latin one
  in the notes (شركةSun -> شركة Sun)
- source fixes: "Co- Feature", "Co- Principal Photography", "Cut- In Reveal"
  lose the stray space; the Arabic of J-Cut, rendered "Jقطع بشكل" in the
  PDF, is stored as "قطع بشكل J"
- no French column in the source, and no French label on Wikidata
- dict_type terminology, tier 4

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
@forzagreen
forzagreen force-pushed the claude/compassionate-cerf-a3n3qc branch from c209c6f to 4bc527d Compare September 30, 2026 19:23

This branch has not been deployed

No deployments
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants