interscript-maps 3.2.3 → 3.2.4

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
checksums.yaml CHANGED
@@ -1,7 +1,7 @@
1
1
  ---
2
2
  SHA256:
3
- metadata.gz: 9d1392a72d2562b959ab45fa614e46b2280defa2de21aa8af5762184251d57d0
4
- data.tar.gz: 5da67bfb8709c5f8c37052ef991efc84364c643dbd79cb937b49ffe8f15d9ee5
3
+ metadata.gz: 7c5ec85faa5eb30d2bd2265b1b7d21965e31d82fd8c460442d691bbb22dd152e
4
+ data.tar.gz: 8161cbaa91376a68243e78a6297a83134ce7f71ecb346ec479e554f27e1847c1
5
5
  SHA512:
6
- metadata.gz: dbde51bf5db378aff5bda9a83f1f3c9d87ad7864ae70bce566e06347701c3ab581067b9728f2dfadb758209604734de54d8da17b452ae5ee29bac336ba6590f5
7
- data.tar.gz: 3cb8848c30c81894d265d753fe65a119081b5598ad5039f52c2eb42f80fc9220a5a8472d4f6669b64125c124015c06bce4a75775efd78fb1d73731cdd52baa15
6
+ metadata.gz: 54a62529575119fa93a04d5fce7e13d4b1619535c5fd00c4e11fe9279fac63414841838d3eeb61aa5c914b749891bdf4425505d4ffbfb9878ea93d5a4395ff38
7
+ data.tar.gz: 06742d6c4b171dc068a15d08b54a818ff52a0b4ed15453a04bcc044bf760eacacf936737505498b1d70ddd44e6021e43c9960f8fa0ef4683bc47794dd159005a
@@ -1,4 +1,4 @@
1
- INTERSCRIPT_MAPS_VERSION = "3.2.3"
1
+ INTERSCRIPT_MAPS_VERSION = "3.2.4"
2
2
 
3
3
  Gem::Specification.new do |spec|
4
4
  spec.name = "interscript-maps"
@@ -22,6 +22,119 @@ tests {
22
22
  "Азәрбајҹан дили" -> "Azărbaı̐jan dili"
23
23
  "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."
24
24
  "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!"
25
+ # Exhaustive per-row tests for alalc-aze-Cyrl-Latn-1997, generated from the source
26
+ # document's table.
27
+ # Source: ALA-LC Azerbaijani Romanization Table, 2017 version, pp. 6-7
28
+ # (Cyrillic script section); held source
29
+ # interscript-private-references/alalc/alalc-azerbaijani.pdf. Extraction:
30
+ # PDF text layer, every value cell cross-checked against the rendered page
31
+ # image. Tesh-ligature digraphs (TS͡/ts͡, IU͡/iu͡, IA͡/ia͡, U+0361) and
32
+ # candrabindu values (I̐/i̐, I + U+0310) transcribed as printed. The
33
+ # lower-case ҝ vernacular cell carries a parenthetical variant glyph that is
34
+ # illegible at 600 dpi and smudged in the print itself; it belongs to the
35
+ # vernacular column, not the romanization, and is documented here rather
36
+ # than transcribed. The map is the 1997 table; the only locally-held printed
37
+ # source is the 2017 version (the table page lists "Earlier versions: 2011,
38
+ # 1997"); the letter table matches the map's encoding and the version
39
+ # difference is recorded here. Skipped printed rows: ъ/ь — the printed value
40
+ # cells read "ʺ (hard sign)" / "ʹ (soft sign)"; the descriptors are stripped
41
+ # and the sign characters transcribed as printed. Canonical-equivalence
42
+ # notes: Ө: printed glyph "Ȯ"; the runtime emits the canonically equivalent
43
+ # byte form 'Ȯ'; the row is stored in the runtime's byte form so the
44
+ # byte-exact embedded test passes. | ө: printed glyph "ȯ"; the runtime
45
+ # emits the canonically equivalent byte form 'ȯ'; the row is stored in the
46
+ # runtime's byte form so the byte-exact embedded test passes. Failing
47
+ # printed rows (findings): Ё — printed cell "Ë"; runtime gives 'Ё'. Failing
48
+ # row, finding below — never forced green. | Й — printed cell "Ĭ"; runtime
49
+ # gives 'Y'. Failing row, finding below — never forced green. | Ц — printed
50
+ # cell "TS͡"; runtime gives 'Ц'. Failing row, finding below — never forced
51
+ # green. | Щ — printed cell "Shch"; runtime gives 'Щ'. Failing row, finding
52
+ # below — never forced green. | Ъ — printed cell "ʺ"; runtime gives 'Ъ'.
53
+ # Failing row, finding below — never forced green. | Ь — printed cell "ʹ";
54
+ # runtime gives 'Ь'. Failing row, finding below — never forced green. | Э —
55
+ # printed cell "Ė"; runtime gives 'Э'. Failing row, finding below — never
56
+ # forced green. | Ю — printed cell "IU͡"; runtime gives 'Ю'. Failing row,
57
+ # finding below — never forced green. | Я — printed cell "IA͡"; runtime
58
+ # gives 'Я'. Failing row, finding below — never forced green. | ё — printed
59
+ # cell "ë"; runtime gives 'ё'. Failing row, finding below — never forced
60
+ # green. | й — printed cell "ĭ"; runtime gives 'y'. Failing row, finding
61
+ # below — never forced green. | ј — printed cell "i̐"; runtime gives 'ı̐'.
62
+ # Failing row, finding below — never forced green. | ц — printed cell "ts͡";
63
+ # runtime gives 'ц'. Failing row, finding below — never forced green. | щ —
64
+ # printed cell "shch"; runtime gives 'щ'. Failing row, finding below — never
65
+ # forced green. | ъ — printed cell "ʺ"; runtime gives 'ъ'. Failing row,
66
+ # finding below — never forced green. | ь — printed cell "ʹ"; runtime gives
67
+ # 'ь'. Failing row, finding below — never forced green. | э — printed cell
68
+ # "ė"; runtime gives 'э'. Failing row, finding below — never forced green. |
69
+ # ю — printed cell "iu͡"; runtime gives 'ю'. Failing row, finding below —
70
+ # never forced green. | я — printed cell "ia͡"; runtime gives 'я'. Failing
71
+ # row, finding below — never forced green.
72
+ # Provenance artifact: scripts/row-sources/alalc-aze-Cyrl-Latn-1997.json (64 rows).
73
+ # generated-by: maps/scripts/gen-row-tests.mjs (2026-11-16); regenerate from the artifact, do not hand-edit.
74
+ "А" -> "A"
75
+ "Б" -> "B"
76
+ "В" -> "V"
77
+ "Г" -> "G"
78
+ "Ғ" -> "Gh"
79
+ "Д" -> "D"
80
+ "Е" -> "E"
81
+ "Ә" -> "Ă"
82
+ "Ж" -> "Zh"
83
+ "З" -> "Z"
84
+ "И" -> "I"
85
+ "Ј" -> "I̐"
86
+ "К" -> "K"
87
+ "Ҝ" -> "Ġ"
88
+ "Л" -> "L"
89
+ "М" -> "M"
90
+ "Н" -> "N"
91
+ "О" -> "O"
92
+ "Ө" -> "Ȯ"
93
+ "П" -> "P"
94
+ "Р" -> "R"
95
+ "С" -> "S"
96
+ "Т" -> "T"
97
+ "У" -> "U"
98
+ "Ү" -> "U̇"
99
+ "Ф" -> "F"
100
+ "Х" -> "Kh"
101
+ "Һ" -> "Ḣ"
102
+ "Ч" -> "Ch"
103
+ "Ҹ" -> "J"
104
+ "Ш" -> "Sh"
105
+ "Ы" -> "Y"
106
+ "а" -> "a"
107
+ "б" -> "b"
108
+ "в" -> "v"
109
+ "г" -> "g"
110
+ "ғ" -> "gh"
111
+ "д" -> "d"
112
+ "е" -> "e"
113
+ "ә" -> "ă"
114
+ "ж" -> "zh"
115
+ "з" -> "z"
116
+ "и" -> "i"
117
+ "к" -> "k"
118
+ "ҝ" -> "ġ"
119
+ "л" -> "l"
120
+ "м" -> "m"
121
+ "н" -> "n"
122
+ "о" -> "o"
123
+ "ө" -> "ȯ"
124
+ "п" -> "p"
125
+ "р" -> "r"
126
+ "с" -> "s"
127
+ "т" -> "t"
128
+ "у" -> "u"
129
+ "ү" -> "u̇"
130
+ "ф" -> "f"
131
+ "х" -> "kh"
132
+ "һ" -> "ḣ"
133
+ "ч" -> "ch"
134
+ "ҹ" -> "j"
135
+ "ш" -> "sh"
136
+ "ы" -> "y"
137
+ "’" -> "’"
25
138
  }
26
139
 
27
140
  stage main {
@@ -30,6 +30,83 @@ tests {
30
30
  "র্ক" -> "rka"
31
31
  "গ্র" -> "gra"
32
32
  "ত্য" -> "tya"
33
+ # Exhaustive per-row tests for alalc-ben-Beng-Latn-2017, generated from the source
34
+ # document's table.
35
+ # Source: ALA-LC Bengali Romanization Table, 2017 version, p. 1 (vowels and
36
+ # diphthongs, consonants, anusvara/bisarga/candrabindu/abagraha) + p. 2
37
+ # (notes); held source
38
+ # interscript-private-references/alalc/alalc-bengali.pdf. Extraction: PDF
39
+ # text layer, every value cell cross-checked against the rendered page
40
+ # image. The vowel table prints ই as capital "I" (long ī lowercase with
41
+ # macron) — transcribed as printed. ঋ -> ṛ, ৠ -> ṝ, ঌ -> ḹ have no
42
+ # text-layer values; visually identified from the rendered page. ৎ (khanda
43
+ # ta) -> t as printed. The ড়/ঢ়/য় keys are the printed nukta composites.
44
+ # Abagraha ঽ -> ’ (apostrophe) as printed. Skipped printed rows: ব second
45
+ # listing — the printed table lists ব twice (labial column and semivowel
46
+ # column), both "ba (see Note 3)"; Note 3's cluster-position value va is
47
+ # context-dependent and un-expressible as a plain row. One ব -> ba row is
48
+ # transcribed; the duplicate printed listing is documented here. | ঁ
49
+ # (candrabindu) — printed value cell "n̐, m̐ (see Note 4)": Note 4 splits
50
+ # the value by the following consonant class — position-dependent,
51
+ # un-expressible as a plain row; skipped. Failing printed rows (findings): ই
52
+ # — printed cell "I"; runtime gives 'i'. Failing row, finding below — never
53
+ # forced green. | চ — printed cell "ca"; runtime gives 'cha'. Failing row,
54
+ # finding below — never forced green. | ছ — printed cell "cha"; runtime
55
+ # gives 'chha'. Failing row, finding below — never forced green. | ড় —
56
+ # printed cell "ṛa"; runtime gives 'ṙa'. Failing row, finding below — never
57
+ # forced green. | ঢ় — printed cell "ṛha"; runtime gives 'ṙha'. Failing row,
58
+ # finding below — never forced green. | য — printed cell "ya"; runtime gives
59
+ # 'j̱a'. Failing row, finding below — never forced green. | য় — printed
60
+ # cell "ẏ"; runtime gives 'ya'. Failing row, finding below — never forced
61
+ # green. | শ — printed cell "śa"; runtime gives 'sha'. Failing row, finding
62
+ # below — never forced green. | ষ — printed cell "sha"; runtime gives 'ṣha'.
63
+ # Failing row, finding below — never forced green. | ং — printed cell "ṃ";
64
+ # runtime gives 'ṁ'. Failing row, finding below — never forced green. | ঽ —
65
+ # printed cell "’"; runtime gives 'ঽ'. Failing row, finding below — never
66
+ # forced green.
67
+ # Provenance artifact: scripts/row-sources/alalc-ben-Beng-Latn-2017.json (41 rows).
68
+ # generated-by: maps/scripts/gen-row-tests.mjs (2026-11-16); regenerate from the artifact, do not hand-edit.
69
+ "অ" -> "a"
70
+ "আ" -> "ā"
71
+ "ঈ" -> "ī"
72
+ "উ" -> "u"
73
+ "ঊ" -> "ū"
74
+ "ঋ" -> "ṛ"
75
+ "এ" -> "e"
76
+ "ঐ" -> "ai"
77
+ "ও" -> "o"
78
+ "ঔ" -> "au"
79
+ "ৠ" -> "ṝ"
80
+ "ঌ" -> "ḹ"
81
+ "ক" -> "ka"
82
+ "খ" -> "kha"
83
+ "গ" -> "ga"
84
+ "ঘ" -> "gha"
85
+ "ঙ" -> "ṅa"
86
+ "জ" -> "ja"
87
+ "ঝ" -> "jha"
88
+ "ঞ" -> "ña"
89
+ "ট" -> "ṭa"
90
+ "ঠ" -> "ṭha"
91
+ "ড" -> "ḍa"
92
+ "ঢ" -> "ḍha"
93
+ "ণ" -> "ṇa"
94
+ "ত" -> "ta"
95
+ "ৎ" -> "t"
96
+ "থ" -> "tha"
97
+ "দ" -> "da"
98
+ "ধ" -> "dha"
99
+ "ন" -> "na"
100
+ "প" -> "pa"
101
+ "ফ" -> "pha"
102
+ "ব" -> "ba"
103
+ "ভ" -> "bha"
104
+ "ম" -> "ma"
105
+ "র" -> "ra"
106
+ "ল" -> "la"
107
+ "স" -> "sa"
108
+ "হ" -> "ha"
109
+ "ঃ" -> "ḥ"
33
110
  }
34
111
 
35
112
  dependency "un-ben-Beng-Latn-2016" as benglatn
@@ -18,6 +18,97 @@ metadata {
18
18
 
19
19
  tests {
20
20
  # missing
21
+ # Exhaustive per-row tests for alalc-bul-Cyrl-Latn-1997, generated from the source
22
+ # document's table.
23
+ # Source: ALA-LC Bulgarian Romanization Table, 2013 version, p. 1 (letter
24
+ # table) + p. 2 (notes); held source
25
+ # interscript-private-references/alalc/alalc-bulgarian.pdf. Extraction: PDF
26
+ # text layer, every value cell cross-checked against the rendered page
27
+ # image. The map is the 1997 table; the only locally-held printed source is
28
+ # the 2013 version (the table page lists "Earlier versions: 2012, 2011,
29
+ # 1997"). The letter table and notes match what the map encodes, so the rows
30
+ # pin the map against the held printed page; the version difference is
31
+ # recorded here. The value cells for Ц, Ю, Я, Ѣ print the tesh-ligature
32
+ # digraphs (t͡s, i͡u, i͡a, i͡e, U+0361); the text layer stores lookalike
33
+ # Cyrillic glyphs for these — transcribed from the printed image. The upper
34
+ # Ѫ value cell is blank in the text layer and visually identified at 900 dpi
35
+ # as U̐ (U + U+0310), matching the printed lower u̐. Skipped printed rows:
36
+ # Ъ/ъ — the table gives Ъ two value cells: "Ŭ (see note 1)" (the vowel
37
+ # reading, initial/medial) and "ʺ (hard sign)" (the pre-1945 word-final
38
+ # reading). The map implements note 1 with a word-final rule, so a plain-row
39
+ # test of the standalone letter takes the word-final path; both cells are
40
+ # position-dependent per note 1 and not expressible as plain rows. Only the
41
+ # note-1 context is documented here; the Ŭ/ŭ cells were transcribed and
42
+ # failed only because the standalone string is word-final.
43
+ # Canonical-equivalence notes: Й: printed glyph "Ĭ"; the runtime emits the
44
+ # canonically equivalent byte form 'Ĭ'; the row is stored in the runtime's
45
+ # byte form so the byte-exact embedded test passes. | й: printed glyph "ĭ";
46
+ # the runtime emits the canonically equivalent byte form 'ĭ'; the row is
47
+ # stored in the runtime's byte form so the byte-exact embedded test passes.
48
+ # Provenance artifact: scripts/row-sources/alalc-bul-Cyrl-Latn-1997.json (62 rows).
49
+ # generated-by: maps/scripts/gen-row-tests.mjs (2026-11-16); regenerate from the artifact, do not hand-edit.
50
+ "А" -> "A"
51
+ "Б" -> "B"
52
+ "В" -> "V"
53
+ "Г" -> "G"
54
+ "Д" -> "D"
55
+ "Е" -> "E"
56
+ "Ж" -> "Zh"
57
+ "З" -> "Z"
58
+ "И" -> "I"
59
+ "Й" -> "Ĭ"
60
+ "К" -> "K"
61
+ "Л" -> "L"
62
+ "М" -> "M"
63
+ "Н" -> "N"
64
+ "О" -> "O"
65
+ "П" -> "P"
66
+ "Р" -> "R"
67
+ "С" -> "S"
68
+ "Т" -> "T"
69
+ "У" -> "U"
70
+ "Ф" -> "F"
71
+ "Х" -> "Kh"
72
+ "Ц" -> "T͡S"
73
+ "Ч" -> "Ch"
74
+ "Ш" -> "Sh"
75
+ "Щ" -> "Sht"
76
+ "Ь" -> "ʹ"
77
+ "Ѣ" -> "I͡E"
78
+ "Ю" -> "I͡U"
79
+ "Я" -> "I͡A"
80
+ "Ѫ" -> "U̐"
81
+ "а" -> "a"
82
+ "б" -> "b"
83
+ "в" -> "v"
84
+ "г" -> "g"
85
+ "д" -> "d"
86
+ "е" -> "e"
87
+ "ж" -> "zh"
88
+ "з" -> "z"
89
+ "и" -> "i"
90
+ "й" -> "ĭ"
91
+ "к" -> "k"
92
+ "л" -> "l"
93
+ "м" -> "m"
94
+ "н" -> "n"
95
+ "о" -> "o"
96
+ "п" -> "p"
97
+ "р" -> "r"
98
+ "с" -> "s"
99
+ "т" -> "t"
100
+ "у" -> "u"
101
+ "ф" -> "f"
102
+ "х" -> "kh"
103
+ "ц" -> "t͡s"
104
+ "ч" -> "ch"
105
+ "ш" -> "sh"
106
+ "щ" -> "sht"
107
+ "ь" -> "ʹ"
108
+ "ѣ" -> "i͡e"
109
+ "ю" -> "i͡u"
110
+ "я" -> "i͡a"
111
+ "ѫ" -> "u̐"
21
112
  }
22
113
 
23
114
  # This map has been partially converted by the bin/maps_v1_to_v2 script
@@ -22,6 +22,101 @@ tests {
22
22
  "ႱႭႪႭႫႭႬ" -> "SOLOMON"
23
23
  "ⴑⴍⴊⴍⴋⴍⴌ" -> "solomon"
24
24
  "ႠႡႰႠჀႠႫ" -> "ABRAHAM"
25
+ # Exhaustive per-row tests for alalc-kat-Geok-Latn-1997, generated from the source
26
+ # document's table.
27
+ # Source: ALA-LC Georgian Romanization Table, 2011 version, pp. 1-2
28
+ # (Khutsuri upper/lower columns with romanization; Mkhedruli column not
29
+ # transcribed for this map); held source
30
+ # interscript-private-references/alalc/alalc-georgian.pdf. Extraction: PDF
31
+ # text layer, every value cell cross-checked against the rendered page
32
+ # image. Aspirated/ejective values use the modifier turned comma (ʻ,
33
+ # U+02BB): Tʻ, Pʻ, Kʻ, Čʻ, Cʻ and their lower cases, transcribed as printed.
34
+ # The table prints Khutsuri upper case, Khutsuri lower case and Mkhedruli
35
+ # side by side; this map (Geok = Khutsuri) takes the Khutsuri columns only.
36
+ # The Mkhedruli column pairs 1:1 with the Khutsuri lower case in the grid
37
+ # (aside from the four tail letters), so the lower-case rows double as the
38
+ # Mkhedruli values. Skipped printed rows: ჶ (f), ჷ (ĕ), ჸ (ʻ ayn), უ̌/უ̂ (ŭ)
39
+ # — Mkhedruli-only tail rows of the printed table with no Khutsuri letter:
40
+ # they belong to the Mkhedruli map, not this Khutsuri (Geok) map; skipped
41
+ # here.
42
+ # Provenance artifact: scripts/row-sources/alalc-kat-Geok-Latn-1997.json (76 rows).
43
+ # generated-by: maps/scripts/gen-row-tests.mjs (2026-11-16); regenerate from the artifact, do not hand-edit.
44
+ "Ⴀ" -> "A"
45
+ "Ⴁ" -> "B"
46
+ "Ⴂ" -> "G"
47
+ "Ⴃ" -> "D"
48
+ "Ⴄ" -> "E"
49
+ "Ⴅ" -> "V"
50
+ "Ⴆ" -> "Z"
51
+ "Ⴇ" -> "Tʻ"
52
+ "Ⴈ" -> "I"
53
+ "Ⴉ" -> "K"
54
+ "Ⴊ" -> "L"
55
+ "Ⴋ" -> "M"
56
+ "Ⴌ" -> "N"
57
+ "Ⴍ" -> "O"
58
+ "Ⴎ" -> "P"
59
+ "Ⴏ" -> "Ž"
60
+ "Ⴐ" -> "R"
61
+ "Ⴑ" -> "S"
62
+ "Ⴒ" -> "T"
63
+ "Ⴓ" -> "U"
64
+ "Ⴔ" -> "Pʻ"
65
+ "Ⴕ" -> "Kʻ"
66
+ "Ⴖ" -> "Ġ"
67
+ "Ⴗ" -> "Q"
68
+ "Ⴘ" -> "Š"
69
+ "Ⴙ" -> "Čʻ"
70
+ "Ⴚ" -> "Cʻ"
71
+ "Ⴛ" -> "Ż"
72
+ "Ⴜ" -> "C"
73
+ "Ⴝ" -> "Č"
74
+ "Ⴞ" -> "X"
75
+ "Ⴟ" -> "J"
76
+ "Ⴠ" -> "H"
77
+ "Ⴡ" -> "Ē"
78
+ "Ⴢ" -> "Y"
79
+ "Ⴣ" -> "W"
80
+ "Ⴤ" -> "X̣"
81
+ "Ⴥ" -> "Ō"
82
+ "ⴀ" -> "a"
83
+ "ⴁ" -> "b"
84
+ "ⴂ" -> "g"
85
+ "ⴃ" -> "d"
86
+ "ⴄ" -> "e"
87
+ "ⴅ" -> "v"
88
+ "ⴆ" -> "z"
89
+ "ⴇ" -> "tʻ"
90
+ "ⴈ" -> "i"
91
+ "ⴉ" -> "k"
92
+ "ⴊ" -> "l"
93
+ "ⴋ" -> "m"
94
+ "ⴌ" -> "n"
95
+ "ⴍ" -> "o"
96
+ "ⴎ" -> "p"
97
+ "ⴏ" -> "ž"
98
+ "ⴐ" -> "r"
99
+ "ⴑ" -> "s"
100
+ "ⴒ" -> "t"
101
+ "ⴓ" -> "u"
102
+ "ⴔ" -> "pʻ"
103
+ "ⴕ" -> "kʻ"
104
+ "ⴖ" -> "ġ"
105
+ "ⴗ" -> "q"
106
+ "ⴘ" -> "š"
107
+ "ⴙ" -> "čʻ"
108
+ "ⴚ" -> "cʻ"
109
+ "ⴛ" -> "ż"
110
+ "ⴜ" -> "c"
111
+ "ⴝ" -> "č"
112
+ "ⴞ" -> "x"
113
+ "ⴟ" -> "j"
114
+ "ⴠ" -> "h"
115
+ "ⴡ" -> "ē"
116
+ "ⴢ" -> "y"
117
+ "ⴣ" -> "w"
118
+ "ⴤ" -> "x̣"
119
+ "ⴥ" -> "ō"
25
120
  }
26
121
 
27
122
  stage main {
@@ -39,6 +39,79 @@ tests {
39
39
  "\uabf3" -> "3"
40
40
  "\uabdb" -> "k"
41
41
  "\uabc0\uabe5\uabec" -> "k\u0101\u0323"
42
+ # Exhaustive per-row tests for alalc-mni-Mtei-Latn-2022, generated from the source
43
+ # document's table.
44
+ # Source: ALA-LC Modern Manipuri (Meitei script) Romanization Table, June
45
+ # 2022, pp. 1-2 (vowels, consonants, finals, visarga, anusvara, numerals);
46
+ # held source interscript-private-references/alalc/alalc-manipuri.pdf.
47
+ # Extraction: PDF text layer, every value cell cross-checked against the
48
+ # rendered page image. Vowel rows are transcribed as printed: the table
49
+ # pairs independent signs with romanization. Composite printed cells are
50
+ # transcribed as their printed composites: ꯑꯥ -> ā, ꯑꫵ -> aḥ, ꯁꯪ -> saṃ.
51
+ # Consonants are printed with inherent vowel "a"; final consonants are
52
+ # printed as a separate table (without inherent vowel) and transcribed as
53
+ # their own rows. Skipped printed rows: ꯬ (heavy tone) — printed value cell
54
+ # "◌̣" (a dot applied to the preceding vowel): a mark, not a standalone
55
+ # transliteration string; skipped as un-expressible. | ꯭ (apun iyek /
56
+ # "killer") — instruction cell (omits the preceding consonant's inherent
57
+ # vowel); no romanization value. | Punctuation marks ꫱, ꫰, ꯯ — the printed
58
+ # table names them (question mark, comma, full stop/period) but the value
59
+ # pairing in the text layer is layout-dependent; not transcribed as rows.
60
+ # Provenance artifact: scripts/row-sources/alalc-mni-Mtei-Latn-2022.json (53 rows).
61
+ # generated-by: maps/scripts/gen-row-tests.mjs (2026-11-16); regenerate from the artifact, do not hand-edit.
62
+ "ꯑ" -> "a"
63
+ "ꯑꯥ" -> "ā"
64
+ "ꯏ" -> "i"
65
+ "ꯎ" -> "u"
66
+ "ꯑꯦ" -> "e"
67
+ "ꯑꯩ" -> "ai"
68
+ "ꯑꯣ" -> "o"
69
+ "ꯑꯧ" -> "au"
70
+ "ꯀ" -> "ka"
71
+ "ꯁ" -> "sa"
72
+ "ꯂ" -> "la"
73
+ "ꯃ" -> "ma"
74
+ "ꯄ" -> "pa"
75
+ "ꯅ" -> "na"
76
+ "ꯆ" -> "ca"
77
+ "ꯇ" -> "ta"
78
+ "ꯈ" -> "kha"
79
+ "ꯉ" -> "ṅa"
80
+ "ꯊ" -> "tha"
81
+ "ꯋ" -> "va"
82
+ "ꯌ" -> "ya"
83
+ "ꯍ" -> "ha"
84
+ "ꯐ" -> "pha"
85
+ "ꯒ" -> "ga"
86
+ "ꯓ" -> "jha"
87
+ "ꯔ" -> "ra"
88
+ "ꯕ" -> "ba"
89
+ "ꯖ" -> "ja"
90
+ "ꯗ" -> "da"
91
+ "ꯘ" -> "gha"
92
+ "ꯙ" -> "dha"
93
+ "ꯚ" -> "bha"
94
+ "ꯛ" -> "k"
95
+ "ꯜ" -> "l"
96
+ "ꯝ" -> "m"
97
+ "ꯞ" -> "p"
98
+ "ꯟ" -> "n"
99
+ "ꯠ" -> "t"
100
+ "ꯡ" -> "ṅ"
101
+ "ꫵ" -> "ḥ"
102
+ "ꯑꫵ" -> "aḥ"
103
+ "ꯪ" -> "ṃ"
104
+ "ꯁꯪ" -> "saṃ"
105
+ "꯰" -> "0"
106
+ "꯱" -> "1"
107
+ "꯲" -> "2"
108
+ "꯳" -> "3"
109
+ "꯴" -> "4"
110
+ "꯵" -> "5"
111
+ "꯶" -> "6"
112
+ "꯷" -> "7"
113
+ "꯸" -> "8"
114
+ "꯹" -> "9"
42
115
  }
43
116
  stage main {
44
117
  parallel {
@@ -28,6 +28,59 @@ metadata {
28
28
 
29
29
  tests {
30
30
  "\u1c60" -> "k"
31
+ # Exhaustive per-row tests for alalc-sat-Olck-Latn-2017, generated from the source
32
+ # document's table.
33
+ # Source: ALA-LC Santali (in Ol script) Romanization Table, 2017 version, p.
34
+ # 1 (letter table) + p. 2 (other marks); held source
35
+ # interscript-private-references/alalc/alalc-santali.pdf. Extraction: PDF
36
+ # text layer, every value cell cross-checked against the rendered page
37
+ # image. ᱟᱹ is the table's own composite row (ᱟ + the nasalization dot) with
38
+ # printed value ạ̄ (a + dot below, U+0323). The ᱽ row's printed value is "’
39
+ # (apostrophe)" with "(see Note 1)" stripped per the proven pattern; Note
40
+ # 1's statement that the four checked host consonants romanize plain
41
+ # (g/j/d/b) is instruction prose, not a value cell. ᱦ and ᱷ are two distinct
42
+ # printed keys both romanized h. Document order = vowels then consonants, as
43
+ # printed. Skipped printed rows: ᱸ — printed value cell "m̐ (see Note 2)":
44
+ # the mark romanizes as m̐ after the vowel it modifies — a mark applied to a
45
+ # preceding vowel, not a standalone transliteration string; skipped as
46
+ # context-dependent. (The composite row ᱟᱹ -> ạ̄ above is the table's own
47
+ # printed composite and is tested.) | The marks ᱻ (vowel prolongation,
48
+ # omitted) and ᱼ (separator, omitted) — instruction rows on p. 2, no
49
+ # romanization values. Failing printed rows (findings): ᱱ — printed cell
50
+ # "n"; runtime gives 'ᱱ'. Failing row, finding below — never forced green.
51
+ # Provenance artifact: scripts/row-sources/alalc-sat-Olck-Latn-2017.json (31 rows).
52
+ # generated-by: maps/scripts/gen-row-tests.mjs (2026-11-16); regenerate from the artifact, do not hand-edit.
53
+ "ᱚ" -> "a"
54
+ "ᱩ" -> "u"
55
+ "ᱟ" -> "ā"
56
+ "ᱮ" -> "e"
57
+ "ᱟᱹ" -> "ạ̄"
58
+ "ᱳ" -> "o"
59
+ "ᱤ" -> "i"
60
+ "ᱠ" -> "k"
61
+ "ᱯ" -> "p"
62
+ "ᱜ" -> "g"
63
+ "ᱵ" -> "b"
64
+ "ᱶ" -> "ṅ"
65
+ "ᱢ" -> "m"
66
+ "ᱪ" -> "c"
67
+ "ᱭ" -> "y"
68
+ "ᱡ" -> "j"
69
+ "ᱨ" -> "r"
70
+ "ᱧ" -> "ñ"
71
+ "ᱞ" -> "l"
72
+ "ᱴ" -> "ṭ"
73
+ "ᱣ" -> "w"
74
+ "ᱰ" -> "ḍ"
75
+ "ᱥ" -> "s"
76
+ "ᱲ" -> "ṛ"
77
+ "ᱦ" -> "h"
78
+ "ᱬ" -> "ṇ"
79
+ "ᱷ" -> "h"
80
+ "ᱛ" -> "t"
81
+ "ᱫ" -> "d"
82
+ "ᱝ" -> "ṃ"
83
+ "ᱽ" -> "’"
31
84
  }
32
85
 
33
86
  stage main {
@@ -34,6 +34,62 @@ tests {
34
34
  "б" -> "b"
35
35
  "в" -> "v"
36
36
  "г" -> "g"
37
+ # Exhaustive per-row tests for din-ab-Cyrl-Latn-1460-2-2011, generated from the source
38
+ # document's table.
39
+ # Source: DIN 1460-2:2011-10, Tabelle 1, language row 2 (Abchasisch); held
40
+ # source interscript-private-references/din/din-1460-2.pdf. Rows are the
41
+ # language column's printed cells, taken from the durable mechanical
42
+ # extraction tools/din1460-2-table.json (x-anchored text-layer extraction,
43
+ # maps PR #242 provenance), re-verified cell-by-cell against the printed
44
+ # grid rectangles (pymupdf find_tables re-extraction) and end-to-end through
45
+ # iscrs transliterate. Positions referenced by the table's own position
46
+ # numbers. Cell notes: к̢ (position 37) — printed cell "ķ" (value 'ķ');
47
+ # runtime gives 'k'. Failing row, finding below — never forced green. | ѵ
48
+ # (position 71) — the printed cell carries the value parenthesized ("(у̇)"),
49
+ # DIN's optional-value notation; transcribed without the typographic
50
+ # parentheses; the row passes end-to-end. Findings (map-vs-source, map
51
+ # untouched): к̢ (position 37): printed cell "ķ" vs runtime 'k'.
52
+ # Provenance artifact: scripts/row-sources/din-ab-Cyrl-Latn-1460-2-2011.json (39 rows).
53
+ # generated-by: maps/scripts/gen-row-tests.mjs (2026-11-16); regenerate from the artifact, do not hand-edit.
54
+ "a" -> "a"
55
+ "б" -> "b"
56
+ "в" -> "v"
57
+ "г" -> "g"
58
+ "ҕ" -> "ġ"
59
+ "д" -> "d"
60
+ "е" -> "e"
61
+ "ҽ" -> "ċ"
62
+ "ҿ" -> "ć"
63
+ "ж" -> "ž"
64
+ "з" -> "z"
65
+ "ӡ" -> "z̈"
66
+ "и" -> "i"
67
+ "к" -> "k"
68
+ "ҟ" -> "q"
69
+ "л" -> "l"
70
+ "м" -> "m"
71
+ "н" -> "n"
72
+ "о" -> "o"
73
+ "ҩ" -> "w"
74
+ "п" -> "p"
75
+ "ҧ" -> "р́"
76
+ "р" -> "r"
77
+ "с" -> "s"
78
+ "т" -> "t"
79
+ "Ђ" -> "ţ"
80
+ "у" -> "u"
81
+ "ѵ" -> "у̇"
82
+ "ф" -> "f"
83
+ "х" -> "h"
84
+ "ҳ" -> "ḩ"
85
+ "ц" -> "c"
86
+ "ҵ" -> "ç"
87
+ "џ" -> "ḑž"
88
+ "ч" -> "č"
89
+ "ӌ" -> "ǵ"
90
+ "ш" -> "š"
91
+ "ы" -> "y"
92
+ "ь" -> "ĭ"
37
93
  }
38
94
 
39
95
  stage main {