From 8d7171c81af2f0224db58d2781101ce0b5baf4e3 Mon Sep 17 00:00:00 2001 From: Ronald Tse Date: Mon, 16 Nov 2026 22:14:00 +0800 Subject: [PATCH] feat(WO20): exhaustive per-row tests for six thin ALA-LC small-table maps (coverage depth) MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Rows transcribed from the held ALA-LC PDFs (bulgarian 2013, santali 2017, manipuri 2022, azerbaijani 2017 Cyrillic section, bengali 2017, georgian 2011 Khutsuri) into row-source provenance artifacts, generated into embedded tests with scripts/gen-row-tests.mjs. No map rules changed. 358 rows transcribed; 327 generated, all passing end-to-end; 31 source-anchored findings recorded (aze-Cyrl: nine letters unencoded or divergent; ben-2017: eleven cells diverge from the printed 2017 table; sat: the ᱱ -> n rule is missing). alalc-lif recorded as not deepenable: the held PDF carries no Limbu glyphs. --- maps/alalc-aze-Cyrl-Latn-1997.isc | 113 +++++++ maps/alalc-ben-Beng-Latn-2017.isc | 77 +++++ maps/alalc-bul-Cyrl-Latn-1997.isc | 91 +++++ maps/alalc-kat-Geok-Latn-1997.isc | 95 ++++++ maps/alalc-mni-Mtei-Latn-2022.isc | 73 +++++ maps/alalc-sat-Olck-Latn-2017.isc | 53 +++ .../row-sources/alalc-aze-Cyrl-Latn-1997.json | 262 +++++++++++++++ .../row-sources/alalc-ben-Beng-Latn-2017.json | 170 ++++++++++ .../row-sources/alalc-bul-Cyrl-Latn-1997.json | 254 ++++++++++++++ .../row-sources/alalc-kat-Geok-Latn-1997.json | 310 ++++++++++++++++++ .../row-sources/alalc-mni-Mtei-Latn-2022.json | 218 ++++++++++++ .../row-sources/alalc-sat-Olck-Latn-2017.json | 130 ++++++++ 12 files changed, 1846 insertions(+) create mode 100644 scripts/row-sources/alalc-aze-Cyrl-Latn-1997.json create mode 100644 scripts/row-sources/alalc-ben-Beng-Latn-2017.json create mode 100644 scripts/row-sources/alalc-bul-Cyrl-Latn-1997.json create mode 100644 scripts/row-sources/alalc-kat-Geok-Latn-1997.json create mode 100644 scripts/row-sources/alalc-mni-Mtei-Latn-2022.json create mode 100644 scripts/row-sources/alalc-sat-Olck-Latn-2017.json diff --git a/maps/alalc-aze-Cyrl-Latn-1997.isc b/maps/alalc-aze-Cyrl-Latn-1997.isc index f7e1e8a..786ec6e 100644 --- a/maps/alalc-aze-Cyrl-Latn-1997.isc +++ b/maps/alalc-aze-Cyrl-Latn-1997.isc @@ -22,6 +22,119 @@ tests { "Азәрбајҹан дили" -> "Azărbaı̐jan dili" "Бүтүн инсанлар ләјагәт вә һүгугларына ҝөрә азад вә бәрабәр доғулурлар.\nОнларын шүурлары вә виҹданлары вар вә бир-бирләринә мүнасибәтдә гардашлыг руһунда давранмалыдырлар." -> "Bu̇tu̇n insanlar lăı̐agăt vă ḣu̇guglaryna ġȯră azad vă bărabăr doghulurlar.\nOnlaryn shu̇urlary vă vijdanlary var vă bir-birlărină mu̇nasibătdă gardashlyg ruḣunda davranmalydyrlar." "Азәрбајҹан! Азәрбајҹан!\nЕј гәһрәман өвладын шанлы Вәтәни!\nСәндән өтрү ҹан вермәјә ҹүмлә һазырыз!\nСәндән өтрү ган төкмәјә ҹүмлә гадириз!\nҮчрәнҝли бајрағынла мәсуд јаша!\nҮчрәнҝли бајрағынла мәсуд јаша!\nМинләрлә ҹан гурбан олду,\nСинән һәрбә мејдан олду!\nҺүгугундан кечән әсҝәр,\nҺәрә бир гәһрәман олду!\nСән оласан ҝүлүстан,\nСәнә һәр ан ҹан гурбан!\nСәнә мин бир мәһәббәт\nСинәмдә тутмуш мәкан!\nНамусуну һифз етмәјә,\nБајрағыны јүксәлтмәјә\nНамусуну һифз етмәјә,\nҸүмлә ҝәнҹләр мүштагдыр!\nШанлы Вәтән! Шанлы Вәтән!\nАзәрбајҹан! Азәрбајҹан!\nАзәрбајҹан! Азәрбајҹан!" -> "Azărbaı̐jan! Azărbaı̐jan!\nEı̐ găḣrăman ȯvladyn shanly Vătăni!\nSăndăn ȯtru̇ jan vermăı̐ă ju̇mlă ḣazyryz!\nSăndăn ȯtru̇ gan tȯkmăı̐ă ju̇mlă gadiriz!\nU̇chrănġli baı̐raghynla măsud ı̐asha!\nU̇chrănġli baı̐raghynla măsud ı̐asha!\nMinlărlă jan gurban oldu,\nSinăn ḣărbă meı̐dan oldu!\nḢu̇gugundan kechăn ăsġăr,\nḢără bir găḣrăman oldu!\nSăn olasan ġu̇lu̇stan,\nSănă ḣăr an jan gurban!\nSănă min bir măḣăbbăt\nSinămdă tutmush măkan!\nNamusunu ḣifz etmăı̐ă,\nBaı̐raghyny ı̐u̇ksăltmăı̐ă\nNamusunu ḣifz etmăı̐ă,\nJu̇mlă ġănjlăr mu̇shtagdyr!\nShanly Vătăn! Shanly Vătăn!\nAzărbaı̐jan! Azărbaı̐jan!\nAzărbaı̐jan! Azărbaı̐jan!" +# Exhaustive per-row tests for alalc-aze-Cyrl-Latn-1997, generated from the source +# document's table. +# Source: ALA-LC Azerbaijani Romanization Table, 2017 version, pp. 6-7 +# (Cyrillic script section); held source +# interscript-private-references/alalc/alalc-azerbaijani.pdf. Extraction: +# PDF text layer, every value cell cross-checked against the rendered page +# image. Tesh-ligature digraphs (TS͡/ts͡, IU͡/iu͡, IA͡/ia͡, U+0361) and +# candrabindu values (I̐/i̐, I + U+0310) transcribed as printed. The +# lower-case ҝ vernacular cell carries a parenthetical variant glyph that is +# illegible at 600 dpi and smudged in the print itself; it belongs to the +# vernacular column, not the romanization, and is documented here rather +# than transcribed. The map is the 1997 table; the only locally-held printed +# source is the 2017 version (the table page lists "Earlier versions: 2011, +# 1997"); the letter table matches the map's encoding and the version +# difference is recorded here. Skipped printed rows: ъ/ь — the printed value +# cells read "ʺ (hard sign)" / "ʹ (soft sign)"; the descriptors are stripped +# and the sign characters transcribed as printed. Canonical-equivalence +# notes: Ө: printed glyph "Ȯ"; the runtime emits the canonically equivalent +# byte form 'Ȯ'; the row is stored in the runtime's byte form so the +# byte-exact embedded test passes. | ө: printed glyph "ȯ"; the runtime +# emits the canonically equivalent byte form 'ȯ'; the row is stored in the +# runtime's byte form so the byte-exact embedded test passes. Failing +# printed rows (findings): Ё — printed cell "Ë"; runtime gives 'Ё'. Failing +# row, finding below — never forced green. | Й — printed cell "Ĭ"; runtime +# gives 'Y'. Failing row, finding below — never forced green. | Ц — printed +# cell "TS͡"; runtime gives 'Ц'. Failing row, finding below — never forced +# green. | Щ — printed cell "Shch"; runtime gives 'Щ'. Failing row, finding +# below — never forced green. | Ъ — printed cell "ʺ"; runtime gives 'Ъ'. +# Failing row, finding below — never forced green. | Ь — printed cell "ʹ"; +# runtime gives 'Ь'. Failing row, finding below — never forced green. | Э — +# printed cell "Ė"; runtime gives 'Э'. Failing row, finding below — never +# forced green. | Ю — printed cell "IU͡"; runtime gives 'Ю'. Failing row, +# finding below — never forced green. | Я — printed cell "IA͡"; runtime +# gives 'Я'. Failing row, finding below — never forced green. | ё — printed +# cell "ë"; runtime gives 'ё'. Failing row, finding below — never forced +# green. | й — printed cell "ĭ"; runtime gives 'y'. Failing row, finding +# below — never forced green. | ј — printed cell "i̐"; runtime gives 'ı̐'. +# Failing row, finding below — never forced green. | ц — printed cell "ts͡"; +# runtime gives 'ц'. Failing row, finding below — never forced green. | щ — +# printed cell "shch"; runtime gives 'щ'. Failing row, finding below — never +# forced green. | ъ — printed cell "ʺ"; runtime gives 'ъ'. Failing row, +# finding below — never forced green. | ь — printed cell "ʹ"; runtime gives +# 'ь'. Failing row, finding below — never forced green. | э — printed cell +# "ė"; runtime gives 'э'. Failing row, finding below — never forced green. | +# ю — printed cell "iu͡"; runtime gives 'ю'. Failing row, finding below — +# never forced green. | я — printed cell "ia͡"; runtime gives 'я'. Failing +# row, finding below — never forced green. +# Provenance artifact: scripts/row-sources/alalc-aze-Cyrl-Latn-1997.json (64 rows). +# generated-by: maps/scripts/gen-row-tests.mjs (2026-11-16); regenerate from the artifact, do not hand-edit. + "А" -> "A" + "Б" -> "B" + "В" -> "V" + "Г" -> "G" + "Ғ" -> "Gh" + "Д" -> "D" + "Е" -> "E" + "Ә" -> "Ă" + "Ж" -> "Zh" + "З" -> "Z" + "И" -> "I" + "Ј" -> "I̐" + "К" -> "K" + "Ҝ" -> "Ġ" + "Л" -> "L" + "М" -> "M" + "Н" -> "N" + "О" -> "O" + "Ө" -> "Ȯ" + "П" -> "P" + "Р" -> "R" + "С" -> "S" + "Т" -> "T" + "У" -> "U" + "Ү" -> "U̇" + "Ф" -> "F" + "Х" -> "Kh" + "Һ" -> "Ḣ" + "Ч" -> "Ch" + "Ҹ" -> "J" + "Ш" -> "Sh" + "Ы" -> "Y" + "а" -> "a" + "б" -> "b" + "в" -> "v" + "г" -> "g" + "ғ" -> "gh" + "д" -> "d" + "е" -> "e" + "ә" -> "ă" + "ж" -> "zh" + "з" -> "z" + "и" -> "i" + "к" -> "k" + "ҝ" -> "ġ" + "л" -> "l" + "м" -> "m" + "н" -> "n" + "о" -> "o" + "ө" -> "ȯ" + "п" -> "p" + "р" -> "r" + "с" -> "s" + "т" -> "t" + "у" -> "u" + "ү" -> "u̇" + "ф" -> "f" + "х" -> "kh" + "һ" -> "ḣ" + "ч" -> "ch" + "ҹ" -> "j" + "ш" -> "sh" + "ы" -> "y" + "’" -> "’" } stage main { diff --git a/maps/alalc-ben-Beng-Latn-2017.isc b/maps/alalc-ben-Beng-Latn-2017.isc index faff02a..a57ee7a 100644 --- a/maps/alalc-ben-Beng-Latn-2017.isc +++ b/maps/alalc-ben-Beng-Latn-2017.isc @@ -30,6 +30,83 @@ tests { "র্ক" -> "rka" "গ্র" -> "gra" "ত্য" -> "tya" +# Exhaustive per-row tests for alalc-ben-Beng-Latn-2017, generated from the source +# document's table. +# Source: ALA-LC Bengali Romanization Table, 2017 version, p. 1 (vowels and +# diphthongs, consonants, anusvara/bisarga/candrabindu/abagraha) + p. 2 +# (notes); held source +# interscript-private-references/alalc/alalc-bengali.pdf. Extraction: PDF +# text layer, every value cell cross-checked against the rendered page +# image. The vowel table prints ই as capital "I" (long ī lowercase with +# macron) — transcribed as printed. ঋ -> ṛ, ৠ -> ṝ, ঌ -> ḹ have no +# text-layer values; visually identified from the rendered page. ৎ (khanda +# ta) -> t as printed. The ড়/ঢ়/য় keys are the printed nukta composites. +# Abagraha ঽ -> ’ (apostrophe) as printed. Skipped printed rows: ব second +# listing — the printed table lists ব twice (labial column and semivowel +# column), both "ba (see Note 3)"; Note 3's cluster-position value va is +# context-dependent and un-expressible as a plain row. One ব -> ba row is +# transcribed; the duplicate printed listing is documented here. | ঁ +# (candrabindu) — printed value cell "n̐, m̐ (see Note 4)": Note 4 splits +# the value by the following consonant class — position-dependent, +# un-expressible as a plain row; skipped. Failing printed rows (findings): ই +# — printed cell "I"; runtime gives 'i'. Failing row, finding below — never +# forced green. | চ — printed cell "ca"; runtime gives 'cha'. Failing row, +# finding below — never forced green. | ছ — printed cell "cha"; runtime +# gives 'chha'. Failing row, finding below — never forced green. | ড় — +# printed cell "ṛa"; runtime gives 'ṙa'. Failing row, finding below — never +# forced green. | ঢ় — printed cell "ṛha"; runtime gives 'ṙha'. Failing row, +# finding below — never forced green. | য — printed cell "ya"; runtime gives +# 'j̱a'. Failing row, finding below — never forced green. | য় — printed +# cell "ẏ"; runtime gives 'ya'. Failing row, finding below — never forced +# green. | শ — printed cell "śa"; runtime gives 'sha'. Failing row, finding +# below — never forced green. | ষ — printed cell "sha"; runtime gives 'ṣha'. +# Failing row, finding below — never forced green. | ং — printed cell "ṃ"; +# runtime gives 'ṁ'. Failing row, finding below — never forced green. | ঽ — +# printed cell "’"; runtime gives 'ঽ'. Failing row, finding below — never +# forced green. +# Provenance artifact: scripts/row-sources/alalc-ben-Beng-Latn-2017.json (41 rows). +# generated-by: maps/scripts/gen-row-tests.mjs (2026-11-16); regenerate from the artifact, do not hand-edit. + "অ" -> "a" + "আ" -> "ā" + "ঈ" -> "ī" + "উ" -> "u" + "ঊ" -> "ū" + "ঋ" -> "ṛ" + "এ" -> "e" + "ঐ" -> "ai" + "ও" -> "o" + "ঔ" -> "au" + "ৠ" -> "ṝ" + "ঌ" -> "ḹ" + "ক" -> "ka" + "খ" -> "kha" + "গ" -> "ga" + "ঘ" -> "gha" + "ঙ" -> "ṅa" + "জ" -> "ja" + "ঝ" -> "jha" + "ঞ" -> "ña" + "ট" -> "ṭa" + "ঠ" -> "ṭha" + "ড" -> "ḍa" + "ঢ" -> "ḍha" + "ণ" -> "ṇa" + "ত" -> "ta" + "ৎ" -> "t" + "থ" -> "tha" + "দ" -> "da" + "ধ" -> "dha" + "ন" -> "na" + "প" -> "pa" + "ফ" -> "pha" + "ব" -> "ba" + "ভ" -> "bha" + "ম" -> "ma" + "র" -> "ra" + "ল" -> "la" + "স" -> "sa" + "হ" -> "ha" + "ঃ" -> "ḥ" } dependency "un-ben-Beng-Latn-2016" as benglatn diff --git a/maps/alalc-bul-Cyrl-Latn-1997.isc b/maps/alalc-bul-Cyrl-Latn-1997.isc index dce3169..dff45c9 100644 --- a/maps/alalc-bul-Cyrl-Latn-1997.isc +++ b/maps/alalc-bul-Cyrl-Latn-1997.isc @@ -18,6 +18,97 @@ metadata { tests { # missing +# Exhaustive per-row tests for alalc-bul-Cyrl-Latn-1997, generated from the source +# document's table. +# Source: ALA-LC Bulgarian Romanization Table, 2013 version, p. 1 (letter +# table) + p. 2 (notes); held source +# interscript-private-references/alalc/alalc-bulgarian.pdf. Extraction: PDF +# text layer, every value cell cross-checked against the rendered page +# image. The map is the 1997 table; the only locally-held printed source is +# the 2013 version (the table page lists "Earlier versions: 2012, 2011, +# 1997"). The letter table and notes match what the map encodes, so the rows +# pin the map against the held printed page; the version difference is +# recorded here. The value cells for Ц, Ю, Я, Ѣ print the tesh-ligature +# digraphs (t͡s, i͡u, i͡a, i͡e, U+0361); the text layer stores lookalike +# Cyrillic glyphs for these — transcribed from the printed image. The upper +# Ѫ value cell is blank in the text layer and visually identified at 900 dpi +# as U̐ (U + U+0310), matching the printed lower u̐. Skipped printed rows: +# Ъ/ъ — the table gives Ъ two value cells: "Ŭ (see note 1)" (the vowel +# reading, initial/medial) and "ʺ (hard sign)" (the pre-1945 word-final +# reading). The map implements note 1 with a word-final rule, so a plain-row +# test of the standalone letter takes the word-final path; both cells are +# position-dependent per note 1 and not expressible as plain rows. Only the +# note-1 context is documented here; the Ŭ/ŭ cells were transcribed and +# failed only because the standalone string is word-final. +# Canonical-equivalence notes: Й: printed glyph "Ĭ"; the runtime emits the +# canonically equivalent byte form 'Ĭ'; the row is stored in the runtime's +# byte form so the byte-exact embedded test passes. | й: printed glyph "ĭ"; +# the runtime emits the canonically equivalent byte form 'ĭ'; the row is +# stored in the runtime's byte form so the byte-exact embedded test passes. +# Provenance artifact: scripts/row-sources/alalc-bul-Cyrl-Latn-1997.json (62 rows). +# generated-by: maps/scripts/gen-row-tests.mjs (2026-11-16); regenerate from the artifact, do not hand-edit. + "А" -> "A" + "Б" -> "B" + "В" -> "V" + "Г" -> "G" + "Д" -> "D" + "Е" -> "E" + "Ж" -> "Zh" + "З" -> "Z" + "И" -> "I" + "Й" -> "Ĭ" + "К" -> "K" + "Л" -> "L" + "М" -> "M" + "Н" -> "N" + "О" -> "O" + "П" -> "P" + "Р" -> "R" + "С" -> "S" + "Т" -> "T" + "У" -> "U" + "Ф" -> "F" + "Х" -> "Kh" + "Ц" -> "T͡S" + "Ч" -> "Ch" + "Ш" -> "Sh" + "Щ" -> "Sht" + "Ь" -> "ʹ" + "Ѣ" -> "I͡E" + "Ю" -> "I͡U" + "Я" -> "I͡A" + "Ѫ" -> "U̐" + "а" -> "a" + "б" -> "b" + "в" -> "v" + "г" -> "g" + "д" -> "d" + "е" -> "e" + "ж" -> "zh" + "з" -> "z" + "и" -> "i" + "й" -> "ĭ" + "к" -> "k" + "л" -> "l" + "м" -> "m" + "н" -> "n" + "о" -> "o" + "п" -> "p" + "р" -> "r" + "с" -> "s" + "т" -> "t" + "у" -> "u" + "ф" -> "f" + "х" -> "kh" + "ц" -> "t͡s" + "ч" -> "ch" + "ш" -> "sh" + "щ" -> "sht" + "ь" -> "ʹ" + "ѣ" -> "i͡e" + "ю" -> "i͡u" + "я" -> "i͡a" + "ѫ" -> "u̐" } # This map has been partially converted by the bin/maps_v1_to_v2 script diff --git a/maps/alalc-kat-Geok-Latn-1997.isc b/maps/alalc-kat-Geok-Latn-1997.isc index 506b663..2cfc7f4 100644 --- a/maps/alalc-kat-Geok-Latn-1997.isc +++ b/maps/alalc-kat-Geok-Latn-1997.isc @@ -22,6 +22,101 @@ tests { "ႱႭႪႭႫႭႬ" -> "SOLOMON" "ⴑⴍⴊⴍⴋⴍⴌ" -> "solomon" "ႠႡႰႠჀႠႫ" -> "ABRAHAM" +# Exhaustive per-row tests for alalc-kat-Geok-Latn-1997, generated from the source +# document's table. +# Source: ALA-LC Georgian Romanization Table, 2011 version, pp. 1-2 +# (Khutsuri upper/lower columns with romanization; Mkhedruli column not +# transcribed for this map); held source +# interscript-private-references/alalc/alalc-georgian.pdf. Extraction: PDF +# text layer, every value cell cross-checked against the rendered page +# image. Aspirated/ejective values use the modifier turned comma (ʻ, +# U+02BB): Tʻ, Pʻ, Kʻ, Čʻ, Cʻ and their lower cases, transcribed as printed. +# The table prints Khutsuri upper case, Khutsuri lower case and Mkhedruli +# side by side; this map (Geok = Khutsuri) takes the Khutsuri columns only. +# The Mkhedruli column pairs 1:1 with the Khutsuri lower case in the grid +# (aside from the four tail letters), so the lower-case rows double as the +# Mkhedruli values. Skipped printed rows: ჶ (f), ჷ (ĕ), ჸ (ʻ ayn), უ̌/უ̂ (ŭ) +# — Mkhedruli-only tail rows of the printed table with no Khutsuri letter: +# they belong to the Mkhedruli map, not this Khutsuri (Geok) map; skipped +# here. +# Provenance artifact: scripts/row-sources/alalc-kat-Geok-Latn-1997.json (76 rows). +# generated-by: maps/scripts/gen-row-tests.mjs (2026-11-16); regenerate from the artifact, do not hand-edit. + "Ⴀ" -> "A" + "Ⴁ" -> "B" + "Ⴂ" -> "G" + "Ⴃ" -> "D" + "Ⴄ" -> "E" + "Ⴅ" -> "V" + "Ⴆ" -> "Z" + "Ⴇ" -> "Tʻ" + "Ⴈ" -> "I" + "Ⴉ" -> "K" + "Ⴊ" -> "L" + "Ⴋ" -> "M" + "Ⴌ" -> "N" + "Ⴍ" -> "O" + "Ⴎ" -> "P" + "Ⴏ" -> "Ž" + "Ⴐ" -> "R" + "Ⴑ" -> "S" + "Ⴒ" -> "T" + "Ⴓ" -> "U" + "Ⴔ" -> "Pʻ" + "Ⴕ" -> "Kʻ" + "Ⴖ" -> "Ġ" + "Ⴗ" -> "Q" + "Ⴘ" -> "Š" + "Ⴙ" -> "Čʻ" + "Ⴚ" -> "Cʻ" + "Ⴛ" -> "Ż" + "Ⴜ" -> "C" + "Ⴝ" -> "Č" + "Ⴞ" -> "X" + "Ⴟ" -> "J" + "Ⴠ" -> "H" + "Ⴡ" -> "Ē" + "Ⴢ" -> "Y" + "Ⴣ" -> "W" + "Ⴤ" -> "X̣" + "Ⴥ" -> "Ō" + "ⴀ" -> "a" + "ⴁ" -> "b" + "ⴂ" -> "g" + "ⴃ" -> "d" + "ⴄ" -> "e" + "ⴅ" -> "v" + "ⴆ" -> "z" + "ⴇ" -> "tʻ" + "ⴈ" -> "i" + "ⴉ" -> "k" + "ⴊ" -> "l" + "ⴋ" -> "m" + "ⴌ" -> "n" + "ⴍ" -> "o" + "ⴎ" -> "p" + "ⴏ" -> "ž" + "ⴐ" -> "r" + "ⴑ" -> "s" + "ⴒ" -> "t" + "ⴓ" -> "u" + "ⴔ" -> "pʻ" + "ⴕ" -> "kʻ" + "ⴖ" -> "ġ" + "ⴗ" -> "q" + "ⴘ" -> "š" + "ⴙ" -> "čʻ" + "ⴚ" -> "cʻ" + "ⴛ" -> "ż" + "ⴜ" -> "c" + "ⴝ" -> "č" + "ⴞ" -> "x" + "ⴟ" -> "j" + "ⴠ" -> "h" + "ⴡ" -> "ē" + "ⴢ" -> "y" + "ⴣ" -> "w" + "ⴤ" -> "x̣" + "ⴥ" -> "ō" } stage main { diff --git a/maps/alalc-mni-Mtei-Latn-2022.isc b/maps/alalc-mni-Mtei-Latn-2022.isc index a257ba4..604a5d1 100644 --- a/maps/alalc-mni-Mtei-Latn-2022.isc +++ b/maps/alalc-mni-Mtei-Latn-2022.isc @@ -39,6 +39,79 @@ tests { "\uabf3" -> "3" "\uabdb" -> "k" "\uabc0\uabe5\uabec" -> "k\u0101\u0323" +# Exhaustive per-row tests for alalc-mni-Mtei-Latn-2022, generated from the source +# document's table. +# Source: ALA-LC Modern Manipuri (Meitei script) Romanization Table, June +# 2022, pp. 1-2 (vowels, consonants, finals, visarga, anusvara, numerals); +# held source interscript-private-references/alalc/alalc-manipuri.pdf. +# Extraction: PDF text layer, every value cell cross-checked against the +# rendered page image. Vowel rows are transcribed as printed: the table +# pairs independent signs with romanization. Composite printed cells are +# transcribed as their printed composites: ꯑꯥ -> ā, ꯑꫵ -> aḥ, ꯁꯪ -> saṃ. +# Consonants are printed with inherent vowel "a"; final consonants are +# printed as a separate table (without inherent vowel) and transcribed as +# their own rows. Skipped printed rows: ꯬ (heavy tone) — printed value cell +# "◌̣" (a dot applied to the preceding vowel): a mark, not a standalone +# transliteration string; skipped as un-expressible. | ꯭ (apun iyek / +# "killer") — instruction cell (omits the preceding consonant's inherent +# vowel); no romanization value. | Punctuation marks ꫱, ꫰, ꯯ — the printed +# table names them (question mark, comma, full stop/period) but the value +# pairing in the text layer is layout-dependent; not transcribed as rows. +# Provenance artifact: scripts/row-sources/alalc-mni-Mtei-Latn-2022.json (53 rows). +# generated-by: maps/scripts/gen-row-tests.mjs (2026-11-16); regenerate from the artifact, do not hand-edit. + "ꯑ" -> "a" + "ꯑꯥ" -> "ā" + "ꯏ" -> "i" + "ꯎ" -> "u" + "ꯑꯦ" -> "e" + "ꯑꯩ" -> "ai" + "ꯑꯣ" -> "o" + "ꯑꯧ" -> "au" + "ꯀ" -> "ka" + "ꯁ" -> "sa" + "ꯂ" -> "la" + "ꯃ" -> "ma" + "ꯄ" -> "pa" + "ꯅ" -> "na" + "ꯆ" -> "ca" + "ꯇ" -> "ta" + "ꯈ" -> "kha" + "ꯉ" -> "ṅa" + "ꯊ" -> "tha" + "ꯋ" -> "va" + "ꯌ" -> "ya" + "ꯍ" -> "ha" + "ꯐ" -> "pha" + "ꯒ" -> "ga" + "ꯓ" -> "jha" + "ꯔ" -> "ra" + "ꯕ" -> "ba" + "ꯖ" -> "ja" + "ꯗ" -> "da" + "ꯘ" -> "gha" + "ꯙ" -> "dha" + "ꯚ" -> "bha" + "ꯛ" -> "k" + "ꯜ" -> "l" + "ꯝ" -> "m" + "ꯞ" -> "p" + "ꯟ" -> "n" + "ꯠ" -> "t" + "ꯡ" -> "ṅ" + "ꫵ" -> "ḥ" + "ꯑꫵ" -> "aḥ" + "ꯪ" -> "ṃ" + "ꯁꯪ" -> "saṃ" + "꯰" -> "0" + "꯱" -> "1" + "꯲" -> "2" + "꯳" -> "3" + "꯴" -> "4" + "꯵" -> "5" + "꯶" -> "6" + "꯷" -> "7" + "꯸" -> "8" + "꯹" -> "9" } stage main { parallel { diff --git a/maps/alalc-sat-Olck-Latn-2017.isc b/maps/alalc-sat-Olck-Latn-2017.isc index dd69dc7..a3d7158 100644 --- a/maps/alalc-sat-Olck-Latn-2017.isc +++ b/maps/alalc-sat-Olck-Latn-2017.isc @@ -28,6 +28,59 @@ metadata { tests { "\u1c60" -> "k" +# Exhaustive per-row tests for alalc-sat-Olck-Latn-2017, generated from the source +# document's table. +# Source: ALA-LC Santali (in Ol script) Romanization Table, 2017 version, p. +# 1 (letter table) + p. 2 (other marks); held source +# interscript-private-references/alalc/alalc-santali.pdf. Extraction: PDF +# text layer, every value cell cross-checked against the rendered page +# image. ᱟᱹ is the table's own composite row (ᱟ + the nasalization dot) with +# printed value ạ̄ (a + dot below, U+0323). The ᱽ row's printed value is "’ +# (apostrophe)" with "(see Note 1)" stripped per the proven pattern; Note +# 1's statement that the four checked host consonants romanize plain +# (g/j/d/b) is instruction prose, not a value cell. ᱦ and ᱷ are two distinct +# printed keys both romanized h. Document order = vowels then consonants, as +# printed. Skipped printed rows: ᱸ — printed value cell "m̐ (see Note 2)": +# the mark romanizes as m̐ after the vowel it modifies — a mark applied to a +# preceding vowel, not a standalone transliteration string; skipped as +# context-dependent. (The composite row ᱟᱹ -> ạ̄ above is the table's own +# printed composite and is tested.) | The marks ᱻ (vowel prolongation, +# omitted) and ᱼ (separator, omitted) — instruction rows on p. 2, no +# romanization values. Failing printed rows (findings): ᱱ — printed cell +# "n"; runtime gives 'ᱱ'. Failing row, finding below — never forced green. +# Provenance artifact: scripts/row-sources/alalc-sat-Olck-Latn-2017.json (31 rows). +# generated-by: maps/scripts/gen-row-tests.mjs (2026-11-16); regenerate from the artifact, do not hand-edit. + "ᱚ" -> "a" + "ᱩ" -> "u" + "ᱟ" -> "ā" + "ᱮ" -> "e" + "ᱟᱹ" -> "ạ̄" + "ᱳ" -> "o" + "ᱤ" -> "i" + "ᱠ" -> "k" + "ᱯ" -> "p" + "ᱜ" -> "g" + "ᱵ" -> "b" + "ᱶ" -> "ṅ" + "ᱢ" -> "m" + "ᱪ" -> "c" + "ᱭ" -> "y" + "ᱡ" -> "j" + "ᱨ" -> "r" + "ᱧ" -> "ñ" + "ᱞ" -> "l" + "ᱴ" -> "ṭ" + "ᱣ" -> "w" + "ᱰ" -> "ḍ" + "ᱥ" -> "s" + "ᱲ" -> "ṛ" + "ᱦ" -> "h" + "ᱬ" -> "ṇ" + "ᱷ" -> "h" + "ᱛ" -> "t" + "ᱫ" -> "d" + "ᱝ" -> "ṃ" + "ᱽ" -> "’" } stage main { diff --git a/scripts/row-sources/alalc-aze-Cyrl-Latn-1997.json b/scripts/row-sources/alalc-aze-Cyrl-Latn-1997.json new file mode 100644 index 0000000..bb63292 --- /dev/null +++ b/scripts/row-sources/alalc-aze-Cyrl-Latn-1997.json @@ -0,0 +1,262 @@ +{ + "system": "alalc-aze-Cyrl-Latn-1997", + "rows": [ + [ + "А", + "A" + ], + [ + "Б", + "B" + ], + [ + "В", + "V" + ], + [ + "Г", + "G" + ], + [ + "Ғ", + "Gh" + ], + [ + "Д", + "D" + ], + [ + "Е", + "E" + ], + [ + "Ә", + "Ă" + ], + [ + "Ж", + "Zh" + ], + [ + "З", + "Z" + ], + [ + "И", + "I" + ], + [ + "Ј", + "I̐" + ], + [ + "К", + "K" + ], + [ + "Ҝ", + "Ġ" + ], + [ + "Л", + "L" + ], + [ + "М", + "M" + ], + [ + "Н", + "N" + ], + [ + "О", + "O" + ], + [ + "Ө", + "Ȯ" + ], + [ + "П", + "P" + ], + [ + "Р", + "R" + ], + [ + "С", + "S" + ], + [ + "Т", + "T" + ], + [ + "У", + "U" + ], + [ + "Ү", + "U̇" + ], + [ + "Ф", + "F" + ], + [ + "Х", + "Kh" + ], + [ + "Һ", + "Ḣ" + ], + [ + "Ч", + "Ch" + ], + [ + "Ҹ", + "J" + ], + [ + "Ш", + "Sh" + ], + [ + "Ы", + "Y" + ], + [ + "а", + "a" + ], + [ + "б", + "b" + ], + [ + "в", + "v" + ], + [ + "г", + "g" + ], + [ + "ғ", + "gh" + ], + [ + "д", + "d" + ], + [ + "е", + "e" + ], + [ + "ә", + "ă" + ], + [ + "ж", + "zh" + ], + [ + "з", + "z" + ], + [ + "и", + "i" + ], + [ + "к", + "k" + ], + [ + "ҝ", + "ġ" + ], + [ + "л", + "l" + ], + [ + "м", + "m" + ], + [ + "н", + "n" + ], + [ + "о", + "o" + ], + [ + "ө", + "ȯ" + ], + [ + "п", + "p" + ], + [ + "р", + "r" + ], + [ + "с", + "s" + ], + [ + "т", + "t" + ], + [ + "у", + "u" + ], + [ + "ү", + "u̇" + ], + [ + "ф", + "f" + ], + [ + "х", + "kh" + ], + [ + "һ", + "ḣ" + ], + [ + "ч", + "ch" + ], + [ + "ҹ", + "j" + ], + [ + "ш", + "sh" + ], + [ + "ы", + "y" + ], + [ + "’", + "’" + ] + ], + "note": "ALA-LC Azerbaijani Romanization Table, 2017 version, pp. 6-7 (Cyrillic script section); held source interscript-private-references/alalc/alalc-azerbaijani.pdf. Extraction: PDF text layer, every value cell cross-checked against the rendered page image. Tesh-ligature digraphs (TS͡/ts͡, IU͡/iu͡, IA͡/ia͡, U+0361) and candrabindu values (I̐/i̐, I + U+0310) transcribed as printed. The lower-case ҝ vernacular cell carries a parenthetical variant glyph that is illegible at 600 dpi and smudged in the print itself; it belongs to the vernacular column, not the romanization, and is documented here rather than transcribed. The map is the 1997 table; the only locally-held printed source is the 2017 version (the table page lists \"Earlier versions: 2011, 1997\"); the letter table matches the map's encoding and the version difference is recorded here. Skipped printed rows: ъ/ь — the printed value cells read \"ʺ (hard sign)\" / \"ʹ (soft sign)\"; the descriptors are stripped and the sign characters transcribed as printed. Canonical-equivalence notes: Ө: printed glyph \"Ȯ\"; the runtime emits the canonically equivalent byte form 'Ȯ'; the row is stored in the runtime's byte form so the byte-exact embedded test passes. | ө: printed glyph \"ȯ\"; the runtime emits the canonically equivalent byte form 'ȯ'; the row is stored in the runtime's byte form so the byte-exact embedded test passes. Failing printed rows (findings): Ё — printed cell \"Ë\"; runtime gives 'Ё'. Failing row, finding below — never forced green. | Й — printed cell \"Ĭ\"; runtime gives 'Y'. Failing row, finding below — never forced green. | Ц — printed cell \"TS͡\"; runtime gives 'Ц'. Failing row, finding below — never forced green. | Щ — printed cell \"Shch\"; runtime gives 'Щ'. Failing row, finding below — never forced green. | Ъ — printed cell \"ʺ\"; runtime gives 'Ъ'. Failing row, finding below — never forced green. | Ь — printed cell \"ʹ\"; runtime gives 'Ь'. Failing row, finding below — never forced green. | Э — printed cell \"Ė\"; runtime gives 'Э'. Failing row, finding below — never forced green. | Ю — printed cell \"IU͡\"; runtime gives 'Ю'. Failing row, finding below — never forced green. | Я — printed cell \"IA͡\"; runtime gives 'Я'. Failing row, finding below — never forced green. | ё — printed cell \"ë\"; runtime gives 'ё'. Failing row, finding below — never forced green. | й — printed cell \"ĭ\"; runtime gives 'y'. Failing row, finding below — never forced green. | ј — printed cell \"i̐\"; runtime gives 'ı̐'. Failing row, finding below — never forced green. | ц — printed cell \"ts͡\"; runtime gives 'ц'. Failing row, finding below — never forced green. | щ — printed cell \"shch\"; runtime gives 'щ'. Failing row, finding below — never forced green. | ъ — printed cell \"ʺ\"; runtime gives 'ъ'. Failing row, finding below — never forced green. | ь — printed cell \"ʹ\"; runtime gives 'ь'. Failing row, finding below — never forced green. | э — printed cell \"ė\"; runtime gives 'э'. Failing row, finding below — never forced green. | ю — printed cell \"iu͡\"; runtime gives 'ю'. Failing row, finding below — never forced green. | я — printed cell \"ia͡\"; runtime gives 'я'. Failing row, finding below — never forced green." +} diff --git a/scripts/row-sources/alalc-ben-Beng-Latn-2017.json b/scripts/row-sources/alalc-ben-Beng-Latn-2017.json new file mode 100644 index 0000000..2a5592a --- /dev/null +++ b/scripts/row-sources/alalc-ben-Beng-Latn-2017.json @@ -0,0 +1,170 @@ +{ + "system": "alalc-ben-Beng-Latn-2017", + "rows": [ + [ + "অ", + "a" + ], + [ + "আ", + "ā" + ], + [ + "ঈ", + "ī" + ], + [ + "উ", + "u" + ], + [ + "ঊ", + "ū" + ], + [ + "ঋ", + "ṛ" + ], + [ + "এ", + "e" + ], + [ + "ঐ", + "ai" + ], + [ + "ও", + "o" + ], + [ + "ঔ", + "au" + ], + [ + "ৠ", + "ṝ" + ], + [ + "ঌ", + "ḹ" + ], + [ + "ক", + "ka" + ], + [ + "খ", + "kha" + ], + [ + "গ", + "ga" + ], + [ + "ঘ", + "gha" + ], + [ + "ঙ", + "ṅa" + ], + [ + "জ", + "ja" + ], + [ + "ঝ", + "jha" + ], + [ + "ঞ", + "ña" + ], + [ + "ট", + "ṭa" + ], + [ + "ঠ", + "ṭha" + ], + [ + "ড", + "ḍa" + ], + [ + "ঢ", + "ḍha" + ], + [ + "ণ", + "ṇa" + ], + [ + "ত", + "ta" + ], + [ + "ৎ", + "t" + ], + [ + "থ", + "tha" + ], + [ + "দ", + "da" + ], + [ + "ধ", + "dha" + ], + [ + "ন", + "na" + ], + [ + "প", + "pa" + ], + [ + "ফ", + "pha" + ], + [ + "ব", + "ba" + ], + [ + "ভ", + "bha" + ], + [ + "ম", + "ma" + ], + [ + "র", + "ra" + ], + [ + "ল", + "la" + ], + [ + "স", + "sa" + ], + [ + "হ", + "ha" + ], + [ + "ঃ", + "ḥ" + ] + ], + "note": "ALA-LC Bengali Romanization Table, 2017 version, p. 1 (vowels and diphthongs, consonants, anusvara/bisarga/candrabindu/abagraha) + p. 2 (notes); held source interscript-private-references/alalc/alalc-bengali.pdf. Extraction: PDF text layer, every value cell cross-checked against the rendered page image. The vowel table prints ই as capital \"I\" (long ī lowercase with macron) — transcribed as printed. ঋ -> ṛ, ৠ -> ṝ, ঌ -> ḹ have no text-layer values; visually identified from the rendered page. ৎ (khanda ta) -> t as printed. The ড়/ঢ়/য় keys are the printed nukta composites. Abagraha ঽ -> ’ (apostrophe) as printed. Skipped printed rows: ব second listing — the printed table lists ব twice (labial column and semivowel column), both \"ba (see Note 3)\"; Note 3's cluster-position value va is context-dependent and un-expressible as a plain row. One ব -> ba row is transcribed; the duplicate printed listing is documented here. | ঁ (candrabindu) — printed value cell \"n̐, m̐ (see Note 4)\": Note 4 splits the value by the following consonant class — position-dependent, un-expressible as a plain row; skipped. Failing printed rows (findings): ই — printed cell \"I\"; runtime gives 'i'. Failing row, finding below — never forced green. | চ — printed cell \"ca\"; runtime gives 'cha'. Failing row, finding below — never forced green. | ছ — printed cell \"cha\"; runtime gives 'chha'. Failing row, finding below — never forced green. | ড় — printed cell \"ṛa\"; runtime gives 'ṙa'. Failing row, finding below — never forced green. | ঢ় — printed cell \"ṛha\"; runtime gives 'ṙha'. Failing row, finding below — never forced green. | য — printed cell \"ya\"; runtime gives 'j̱a'. Failing row, finding below — never forced green. | য় — printed cell \"ẏ\"; runtime gives 'ya'. Failing row, finding below — never forced green. | শ — printed cell \"śa\"; runtime gives 'sha'. Failing row, finding below — never forced green. | ষ — printed cell \"sha\"; runtime gives 'ṣha'. Failing row, finding below — never forced green. | ং — printed cell \"ṃ\"; runtime gives 'ṁ'. Failing row, finding below — never forced green. | ঽ — printed cell \"’\"; runtime gives 'ঽ'. Failing row, finding below — never forced green." +} diff --git a/scripts/row-sources/alalc-bul-Cyrl-Latn-1997.json b/scripts/row-sources/alalc-bul-Cyrl-Latn-1997.json new file mode 100644 index 0000000..5dbff26 --- /dev/null +++ b/scripts/row-sources/alalc-bul-Cyrl-Latn-1997.json @@ -0,0 +1,254 @@ +{ + "system": "alalc-bul-Cyrl-Latn-1997", + "rows": [ + [ + "А", + "A" + ], + [ + "Б", + "B" + ], + [ + "В", + "V" + ], + [ + "Г", + "G" + ], + [ + "Д", + "D" + ], + [ + "Е", + "E" + ], + [ + "Ж", + "Zh" + ], + [ + "З", + "Z" + ], + [ + "И", + "I" + ], + [ + "Й", + "Ĭ" + ], + [ + "К", + "K" + ], + [ + "Л", + "L" + ], + [ + "М", + "M" + ], + [ + "Н", + "N" + ], + [ + "О", + "O" + ], + [ + "П", + "P" + ], + [ + "Р", + "R" + ], + [ + "С", + "S" + ], + [ + "Т", + "T" + ], + [ + "У", + "U" + ], + [ + "Ф", + "F" + ], + [ + "Х", + "Kh" + ], + [ + "Ц", + "T͡S" + ], + [ + "Ч", + "Ch" + ], + [ + "Ш", + "Sh" + ], + [ + "Щ", + "Sht" + ], + [ + "Ь", + "ʹ" + ], + [ + "Ѣ", + "I͡E" + ], + [ + "Ю", + "I͡U" + ], + [ + "Я", + "I͡A" + ], + [ + "Ѫ", + "U̐" + ], + [ + "а", + "a" + ], + [ + "б", + "b" + ], + [ + "в", + "v" + ], + [ + "г", + "g" + ], + [ + "д", + "d" + ], + [ + "е", + "e" + ], + [ + "ж", + "zh" + ], + [ + "з", + "z" + ], + [ + "и", + "i" + ], + [ + "й", + "ĭ" + ], + [ + "к", + "k" + ], + [ + "л", + "l" + ], + [ + "м", + "m" + ], + [ + "н", + "n" + ], + [ + "о", + "o" + ], + [ + "п", + "p" + ], + [ + "р", + "r" + ], + [ + "с", + "s" + ], + [ + "т", + "t" + ], + [ + "у", + "u" + ], + [ + "ф", + "f" + ], + [ + "х", + "kh" + ], + [ + "ц", + "t͡s" + ], + [ + "ч", + "ch" + ], + [ + "ш", + "sh" + ], + [ + "щ", + "sht" + ], + [ + "ь", + "ʹ" + ], + [ + "ѣ", + "i͡e" + ], + [ + "ю", + "i͡u" + ], + [ + "я", + "i͡a" + ], + [ + "ѫ", + "u̐" + ] + ], + "note": "ALA-LC Bulgarian Romanization Table, 2013 version, p. 1 (letter table) + p. 2 (notes); held source interscript-private-references/alalc/alalc-bulgarian.pdf. Extraction: PDF text layer, every value cell cross-checked against the rendered page image. The map is the 1997 table; the only locally-held printed source is the 2013 version (the table page lists \"Earlier versions: 2012, 2011, 1997\"). The letter table and notes match what the map encodes, so the rows pin the map against the held printed page; the version difference is recorded here. The value cells for Ц, Ю, Я, Ѣ print the tesh-ligature digraphs (t͡s, i͡u, i͡a, i͡e, U+0361); the text layer stores lookalike Cyrillic glyphs for these — transcribed from the printed image. The upper Ѫ value cell is blank in the text layer and visually identified at 900 dpi as U̐ (U + U+0310), matching the printed lower u̐. Skipped printed rows: Ъ/ъ — the table gives Ъ two value cells: \"Ŭ (see note 1)\" (the vowel reading, initial/medial) and \"ʺ (hard sign)\" (the pre-1945 word-final reading). The map implements note 1 with a word-final rule, so a plain-row test of the standalone letter takes the word-final path; both cells are position-dependent per note 1 and not expressible as plain rows. Only the note-1 context is documented here; the Ŭ/ŭ cells were transcribed and failed only because the standalone string is word-final. Canonical-equivalence notes: Й: printed glyph \"Ĭ\"; the runtime emits the canonically equivalent byte form 'Ĭ'; the row is stored in the runtime's byte form so the byte-exact embedded test passes. | й: printed glyph \"ĭ\"; the runtime emits the canonically equivalent byte form 'ĭ'; the row is stored in the runtime's byte form so the byte-exact embedded test passes." +} diff --git a/scripts/row-sources/alalc-kat-Geok-Latn-1997.json b/scripts/row-sources/alalc-kat-Geok-Latn-1997.json new file mode 100644 index 0000000..6368247 --- /dev/null +++ b/scripts/row-sources/alalc-kat-Geok-Latn-1997.json @@ -0,0 +1,310 @@ +{ + "system": "alalc-kat-Geok-Latn-1997", + "rows": [ + [ + "Ⴀ", + "A" + ], + [ + "Ⴁ", + "B" + ], + [ + "Ⴂ", + "G" + ], + [ + "Ⴃ", + "D" + ], + [ + "Ⴄ", + "E" + ], + [ + "Ⴅ", + "V" + ], + [ + "Ⴆ", + "Z" + ], + [ + "Ⴇ", + "Tʻ" + ], + [ + "Ⴈ", + "I" + ], + [ + "Ⴉ", + "K" + ], + [ + "Ⴊ", + "L" + ], + [ + "Ⴋ", + "M" + ], + [ + "Ⴌ", + "N" + ], + [ + "Ⴍ", + "O" + ], + [ + "Ⴎ", + "P" + ], + [ + "Ⴏ", + "Ž" + ], + [ + "Ⴐ", + "R" + ], + [ + "Ⴑ", + "S" + ], + [ + "Ⴒ", + "T" + ], + [ + "Ⴓ", + "U" + ], + [ + "Ⴔ", + "Pʻ" + ], + [ + "Ⴕ", + "Kʻ" + ], + [ + "Ⴖ", + "Ġ" + ], + [ + "Ⴗ", + "Q" + ], + [ + "Ⴘ", + "Š" + ], + [ + "Ⴙ", + "Čʻ" + ], + [ + "Ⴚ", + "Cʻ" + ], + [ + "Ⴛ", + "Ż" + ], + [ + "Ⴜ", + "C" + ], + [ + "Ⴝ", + "Č" + ], + [ + "Ⴞ", + "X" + ], + [ + "Ⴟ", + "J" + ], + [ + "Ⴠ", + "H" + ], + [ + "Ⴡ", + "Ē" + ], + [ + "Ⴢ", + "Y" + ], + [ + "Ⴣ", + "W" + ], + [ + "Ⴤ", + "X̣" + ], + [ + "Ⴥ", + "Ō" + ], + [ + "ⴀ", + "a" + ], + [ + "ⴁ", + "b" + ], + [ + "ⴂ", + "g" + ], + [ + "ⴃ", + "d" + ], + [ + "ⴄ", + "e" + ], + [ + "ⴅ", + "v" + ], + [ + "ⴆ", + "z" + ], + [ + "ⴇ", + "tʻ" + ], + [ + "ⴈ", + "i" + ], + [ + "ⴉ", + "k" + ], + [ + "ⴊ", + "l" + ], + [ + "ⴋ", + "m" + ], + [ + "ⴌ", + "n" + ], + [ + "ⴍ", + "o" + ], + [ + "ⴎ", + "p" + ], + [ + "ⴏ", + "ž" + ], + [ + "ⴐ", + "r" + ], + [ + "ⴑ", + "s" + ], + [ + "ⴒ", + "t" + ], + [ + "ⴓ", + "u" + ], + [ + "ⴔ", + "pʻ" + ], + [ + "ⴕ", + "kʻ" + ], + [ + "ⴖ", + "ġ" + ], + [ + "ⴗ", + "q" + ], + [ + "ⴘ", + "š" + ], + [ + "ⴙ", + "čʻ" + ], + [ + "ⴚ", + "cʻ" + ], + [ + "ⴛ", + "ż" + ], + [ + "ⴜ", + "c" + ], + [ + "ⴝ", + "č" + ], + [ + "ⴞ", + "x" + ], + [ + "ⴟ", + "j" + ], + [ + "ⴠ", + "h" + ], + [ + "ⴡ", + "ē" + ], + [ + "ⴢ", + "y" + ], + [ + "ⴣ", + "w" + ], + [ + "ⴤ", + "x̣" + ], + [ + "ⴥ", + "ō" + ] + ], + "note": "ALA-LC Georgian Romanization Table, 2011 version, pp. 1-2 (Khutsuri upper/lower columns with romanization; Mkhedruli column not transcribed for this map); held source interscript-private-references/alalc/alalc-georgian.pdf. Extraction: PDF text layer, every value cell cross-checked against the rendered page image. Aspirated/ejective values use the modifier turned comma (ʻ, U+02BB): Tʻ, Pʻ, Kʻ, Čʻ, Cʻ and their lower cases, transcribed as printed. The table prints Khutsuri upper case, Khutsuri lower case and Mkhedruli side by side; this map (Geok = Khutsuri) takes the Khutsuri columns only. The Mkhedruli column pairs 1:1 with the Khutsuri lower case in the grid (aside from the four tail letters), so the lower-case rows double as the Mkhedruli values. Skipped printed rows: ჶ (f), ჷ (ĕ), ჸ (ʻ ayn), უ̌/უ̂ (ŭ) — Mkhedruli-only tail rows of the printed table with no Khutsuri letter: they belong to the Mkhedruli map, not this Khutsuri (Geok) map; skipped here." +} diff --git a/scripts/row-sources/alalc-mni-Mtei-Latn-2022.json b/scripts/row-sources/alalc-mni-Mtei-Latn-2022.json new file mode 100644 index 0000000..b4a60b3 --- /dev/null +++ b/scripts/row-sources/alalc-mni-Mtei-Latn-2022.json @@ -0,0 +1,218 @@ +{ + "system": "alalc-mni-Mtei-Latn-2022", + "rows": [ + [ + "ꯑ", + "a" + ], + [ + "ꯑꯥ", + "ā" + ], + [ + "ꯏ", + "i" + ], + [ + "ꯎ", + "u" + ], + [ + "ꯑꯦ", + "e" + ], + [ + "ꯑꯩ", + "ai" + ], + [ + "ꯑꯣ", + "o" + ], + [ + "ꯑꯧ", + "au" + ], + [ + "ꯀ", + "ka" + ], + [ + "ꯁ", + "sa" + ], + [ + "ꯂ", + "la" + ], + [ + "ꯃ", + "ma" + ], + [ + "ꯄ", + "pa" + ], + [ + "ꯅ", + "na" + ], + [ + "ꯆ", + "ca" + ], + [ + "ꯇ", + "ta" + ], + [ + "ꯈ", + "kha" + ], + [ + "ꯉ", + "ṅa" + ], + [ + "ꯊ", + "tha" + ], + [ + "ꯋ", + "va" + ], + [ + "ꯌ", + "ya" + ], + [ + "ꯍ", + "ha" + ], + [ + "ꯐ", + "pha" + ], + [ + "ꯒ", + "ga" + ], + [ + "ꯓ", + "jha" + ], + [ + "ꯔ", + "ra" + ], + [ + "ꯕ", + "ba" + ], + [ + "ꯖ", + "ja" + ], + [ + "ꯗ", + "da" + ], + [ + "ꯘ", + "gha" + ], + [ + "ꯙ", + "dha" + ], + [ + "ꯚ", + "bha" + ], + [ + "ꯛ", + "k" + ], + [ + "ꯜ", + "l" + ], + [ + "ꯝ", + "m" + ], + [ + "ꯞ", + "p" + ], + [ + "ꯟ", + "n" + ], + [ + "ꯠ", + "t" + ], + [ + "ꯡ", + "ṅ" + ], + [ + "ꫵ", + "ḥ" + ], + [ + "ꯑꫵ", + "aḥ" + ], + [ + "ꯪ", + "ṃ" + ], + [ + "ꯁꯪ", + "saṃ" + ], + [ + "꯰", + "0" + ], + [ + "꯱", + "1" + ], + [ + "꯲", + "2" + ], + [ + "꯳", + "3" + ], + [ + "꯴", + "4" + ], + [ + "꯵", + "5" + ], + [ + "꯶", + "6" + ], + [ + "꯷", + "7" + ], + [ + "꯸", + "8" + ], + [ + "꯹", + "9" + ] + ], + "note": "ALA-LC Modern Manipuri (Meitei script) Romanization Table, June 2022, pp. 1-2 (vowels, consonants, finals, visarga, anusvara, numerals); held source interscript-private-references/alalc/alalc-manipuri.pdf. Extraction: PDF text layer, every value cell cross-checked against the rendered page image. Vowel rows are transcribed as printed: the table pairs independent signs with romanization. Composite printed cells are transcribed as their printed composites: ꯑꯥ -> ā, ꯑꫵ -> aḥ, ꯁꯪ -> saṃ. Consonants are printed with inherent vowel \"a\"; final consonants are printed as a separate table (without inherent vowel) and transcribed as their own rows. Skipped printed rows: ꯬ (heavy tone) — printed value cell \"◌̣\" (a dot applied to the preceding vowel): a mark, not a standalone transliteration string; skipped as un-expressible. | ꯭ (apun iyek / \"killer\") — instruction cell (omits the preceding consonant's inherent vowel); no romanization value. | Punctuation marks ꫱, ꫰, ꯯ — the printed table names them (question mark, comma, full stop/period) but the value pairing in the text layer is layout-dependent; not transcribed as rows." +} diff --git a/scripts/row-sources/alalc-sat-Olck-Latn-2017.json b/scripts/row-sources/alalc-sat-Olck-Latn-2017.json new file mode 100644 index 0000000..3f69ab0 --- /dev/null +++ b/scripts/row-sources/alalc-sat-Olck-Latn-2017.json @@ -0,0 +1,130 @@ +{ + "system": "alalc-sat-Olck-Latn-2017", + "rows": [ + [ + "ᱚ", + "a" + ], + [ + "ᱩ", + "u" + ], + [ + "ᱟ", + "ā" + ], + [ + "ᱮ", + "e" + ], + [ + "ᱟᱹ", + "ạ̄" + ], + [ + "ᱳ", + "o" + ], + [ + "ᱤ", + "i" + ], + [ + "ᱠ", + "k" + ], + [ + "ᱯ", + "p" + ], + [ + "ᱜ", + "g" + ], + [ + "ᱵ", + "b" + ], + [ + "ᱶ", + "ṅ" + ], + [ + "ᱢ", + "m" + ], + [ + "ᱪ", + "c" + ], + [ + "ᱭ", + "y" + ], + [ + "ᱡ", + "j" + ], + [ + "ᱨ", + "r" + ], + [ + "ᱧ", + "ñ" + ], + [ + "ᱞ", + "l" + ], + [ + "ᱴ", + "ṭ" + ], + [ + "ᱣ", + "w" + ], + [ + "ᱰ", + "ḍ" + ], + [ + "ᱥ", + "s" + ], + [ + "ᱲ", + "ṛ" + ], + [ + "ᱦ", + "h" + ], + [ + "ᱬ", + "ṇ" + ], + [ + "ᱷ", + "h" + ], + [ + "ᱛ", + "t" + ], + [ + "ᱫ", + "d" + ], + [ + "ᱝ", + "ṃ" + ], + [ + "ᱽ", + "’" + ] + ], + "note": "ALA-LC Santali (in Ol script) Romanization Table, 2017 version, p. 1 (letter table) + p. 2 (other marks); held source interscript-private-references/alalc/alalc-santali.pdf. Extraction: PDF text layer, every value cell cross-checked against the rendered page image. ᱟᱹ is the table's own composite row (ᱟ + the nasalization dot) with printed value ạ̄ (a + dot below, U+0323). The ᱽ row's printed value is \"’ (apostrophe)\" with \"(see Note 1)\" stripped per the proven pattern; Note 1's statement that the four checked host consonants romanize plain (g/j/d/b) is instruction prose, not a value cell. ᱦ and ᱷ are two distinct printed keys both romanized h. Document order = vowels then consonants, as printed. Skipped printed rows: ᱸ — printed value cell \"m̐ (see Note 2)\": the mark romanizes as m̐ after the vowel it modifies — a mark applied to a preceding vowel, not a standalone transliteration string; skipped as context-dependent. (The composite row ᱟᱹ -> ạ̄ above is the table's own printed composite and is tested.) | The marks ᱻ (vowel prolongation, omitted) and ᱼ (separator, omitted) — instruction rows on p. 2, no romanization values. Failing printed rows (findings): ᱱ — printed cell \"n\"; runtime gives 'ᱱ'. Failing row, finding below — never forced green." +}