interscript-maps 3.3.1 → 3.3.2
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 +4 -4
- data/interscript-maps.gemspec +1 -1
- data/libs/ara-latn-233-2026.isc +97 -0
- data/maps/iso-ara-Arab-Latn-233-2026-modified.isc +318 -0
- data/maps/iso-ara-Arab-Latn-233-2026.isc +188 -0
- data/maps/iso-hye-Armn-Latn-9985-2026.isc +264 -0
- data/maps/iso-kat-Geor-Latn-9984-2026.isc +212 -0
- data/maps/iso-pli-Thai-Latn-15919-2001.isc +18 -11
- data/maps/iso-prs-Arab-Latn-233-3-2026.isc +239 -0
- data/maps/iso-rus-Cyrl-Latn-9-2026.isc +812 -0
- data/maps/iso-tha-Thai-Latn-11940-1998.isc +142 -1
- metadata +8 -1
checksums.yaml
CHANGED
|
@@ -1,7 +1,7 @@
|
|
|
1
1
|
---
|
|
2
2
|
SHA256:
|
|
3
|
-
metadata.gz:
|
|
4
|
-
data.tar.gz:
|
|
3
|
+
metadata.gz: 90b52cb1281dda3fd944fc34e382f7cb8b9b7e4b60ae34b9fd75f0ceff2072c7
|
|
4
|
+
data.tar.gz: 2ea6982ffdb0731dd53afde5dbbec5d5dd29df5e97fa0ea05cd811d302f52a8c
|
|
5
5
|
SHA512:
|
|
6
|
-
metadata.gz:
|
|
7
|
-
data.tar.gz:
|
|
6
|
+
metadata.gz: 99dbf0991a4706aee918534b540f3fdbd7fc50fa3afd279d23727314f8144ed3948353fdd08364e2f5f7ea60b64aac37ee0648e663ca94dbaa5717bfb21acd08
|
|
7
|
+
data.tar.gz: a5af03a155a1eae51dda83225b2d5d63a34d14db34decc83648b52aa3a19f98b5cd2ae5f5e3e653e467afc7380fedbde51d01c73a2cbc954596f692d28207a39
|
data/interscript-maps.gemspec
CHANGED
|
@@ -0,0 +1,97 @@
|
|
|
1
|
+
system "ara-latn-233-2026" {
|
|
2
|
+
|
|
3
|
+
# Shared ISO 233-1:2026 Arabic-to-Latin machinery: rows byte-identical between the
|
|
4
|
+
# strict map (iso-ara-Arab-Latn-233-2026, clause 4 Tables 1-5) and the modified map
|
|
5
|
+
# (iso-ara-Arab-Latn-233-2026-modified, clause 5; per 5.1.2 the strict tables apply
|
|
6
|
+
# unless specified otherwise). The rulesets are spliced into each map's single
|
|
7
|
+
# parallel pass so match semantics see one rule set.
|
|
8
|
+
|
|
9
|
+
ruleset consonants {
|
|
10
|
+
# Table 1 No. 3-28 (the consonant rows shared verbatim by both systems,
|
|
11
|
+
# including tāʾ marbūṭaẗ which 5.2.4 renders ẗ in all positions)
|
|
12
|
+
# bāʾ
|
|
13
|
+
sub "\u0628" "b"
|
|
14
|
+
# tāʾ
|
|
15
|
+
sub "\u062a" "t"
|
|
16
|
+
# ṯāʾ
|
|
17
|
+
sub "\u062b" "\u1e6f"
|
|
18
|
+
# ǧīm
|
|
19
|
+
sub "\u062c" "\u01e7"
|
|
20
|
+
# ḥāʾ
|
|
21
|
+
sub "\u062d" "\u1e25"
|
|
22
|
+
# ẖāʾ
|
|
23
|
+
sub "\u062e" "\u1e96"
|
|
24
|
+
# dāl
|
|
25
|
+
sub "\u062f" "d"
|
|
26
|
+
# ḏāl
|
|
27
|
+
sub "\u0630" "\u1e0f"
|
|
28
|
+
# rāʾ
|
|
29
|
+
sub "\u0631" "r"
|
|
30
|
+
# zāy
|
|
31
|
+
sub "\u0632" "z"
|
|
32
|
+
# sīn
|
|
33
|
+
sub "\u0633" "s"
|
|
34
|
+
# šīn
|
|
35
|
+
sub "\u0634" "\u0161"
|
|
36
|
+
# ṣād
|
|
37
|
+
sub "\u0635" "\u1e63"
|
|
38
|
+
# ḍād
|
|
39
|
+
sub "\u0636" "\u1e0d"
|
|
40
|
+
# ṭāʾ
|
|
41
|
+
sub "\u0637" "\u1e6d"
|
|
42
|
+
# ẓāʾ
|
|
43
|
+
sub "\u0638" "\u1e93"
|
|
44
|
+
# ʿayn
|
|
45
|
+
sub "\u0639" "\u02bf"
|
|
46
|
+
# ġayn
|
|
47
|
+
sub "\u063a" "\u0121"
|
|
48
|
+
# fāʾ
|
|
49
|
+
sub "\u0641" "f"
|
|
50
|
+
# qāf
|
|
51
|
+
sub "\u0642" "q"
|
|
52
|
+
# kāf
|
|
53
|
+
sub "\u0643" "k"
|
|
54
|
+
# lām
|
|
55
|
+
sub "\u0644" "l"
|
|
56
|
+
# mīm
|
|
57
|
+
sub "\u0645" "m"
|
|
58
|
+
# nūn
|
|
59
|
+
sub "\u0646" "n"
|
|
60
|
+
# hāʾ
|
|
61
|
+
sub "\u0647" "h"
|
|
62
|
+
# tāʾ marbūṭaẗ
|
|
63
|
+
sub "\u0629" "\u1e97"
|
|
64
|
+
# wāw
|
|
65
|
+
sub "\u0648" "w"
|
|
66
|
+
# yāʾ
|
|
67
|
+
sub "\u064a" "y"
|
|
68
|
+
}
|
|
69
|
+
|
|
70
|
+
ruleset short_vowels {
|
|
71
|
+
# Table 2 No. 30-32 (bare harakat are identical in both systems)
|
|
72
|
+
sub "\u064e" "a"
|
|
73
|
+
sub "\u064f" "u"
|
|
74
|
+
sub "\u0650" "i"
|
|
75
|
+
}
|
|
76
|
+
|
|
77
|
+
ruleset punctuation {
|
|
78
|
+
# Table 4 (fāṣilaẗ, fāṣilaẗ manqūṭaẗ, ʿalāmaẗ istifhām)
|
|
79
|
+
sub "\u060c" ","
|
|
80
|
+
sub "\u061b" ";"
|
|
81
|
+
sub "\u061f" "?"
|
|
82
|
+
}
|
|
83
|
+
|
|
84
|
+
ruleset numerals {
|
|
85
|
+
# Table 5 (0660-0669 -> 0030-0039)
|
|
86
|
+
sub "\u0660" "0"
|
|
87
|
+
sub "\u0661" "1"
|
|
88
|
+
sub "\u0662" "2"
|
|
89
|
+
sub "\u0663" "3"
|
|
90
|
+
sub "\u0664" "4"
|
|
91
|
+
sub "\u0665" "5"
|
|
92
|
+
sub "\u0666" "6"
|
|
93
|
+
sub "\u0667" "7"
|
|
94
|
+
sub "\u0668" "8"
|
|
95
|
+
sub "\u0669" "9"
|
|
96
|
+
}
|
|
97
|
+
}
|
|
@@ -0,0 +1,318 @@
|
|
|
1
|
+
system "ISO:ara-Arab:Latn:233-2026-modified" {
|
|
2
|
+
|
|
3
|
+
metadata {
|
|
4
|
+
authority_id iso
|
|
5
|
+
id 233-2026-modified
|
|
6
|
+
language iso-639-2:ara
|
|
7
|
+
source_script Arab
|
|
8
|
+
destination_script Latn
|
|
9
|
+
name "ISO 233-1:2026 Information and documentation -- Transliteration of Perso-Arabic characters into Latin characters -- Part 1: Arabic language (modified transliteration, clause 5; FDIS stage)"
|
|
10
|
+
url https://www.iso.org/committee/48820.html
|
|
11
|
+
creation_date 2026
|
|
12
|
+
adoption_date 2026
|
|
13
|
+
description { Modified (clause 5) transliteration of ISO 233-1:2026: Tables 6-8 plus the
|
|
14
|
+
mechanical clauses (5.2.2 initial alif, 5.2.4 tāʾ marbūṭaẗ, 5.3.2 definite
|
|
15
|
+
article) over the strict tables (5.1.2). Not necessarily reversible by design
|
|
16
|
+
(5.1.1). Encoded from ISO/TC 46/WG 3 N129 (FDIS) with the N130 adjudications.
|
|
17
|
+
See notes for the language-knowledge steps that stay outside stateless scope. }
|
|
18
|
+
|
|
19
|
+
notes {
|
|
20
|
+
note "Encoded from ISO/FDIS 233-1:2026 (ISO/TC 46/WG 3 N129, 2026-11-02) clause 5, Tables 6-8; adjudications honored from N130: CN-017 rejected (strict and modified treat alif differently by design), CN-018 rejected (the article stands al- even unpronounced; l- only after the prepositions li-/bi-), CN-019 rejected (4.3.2 and 5.3.2 examples intentionally differ). Verify against the published edition when available."
|
|
21
|
+
note "5.1.2: the strict tables apply unless specified otherwise - the shared rows come from the ara-latn-233-2026 library (byte-identical with the strict map). Map-specific rows: sukūn omitted (5.2.1, Table 8 No. 7), šadda duplication (Table 8 No. 8), tanwīn omitted (Table 6 No. 2/5a/6a), the á/â/ā/ī/ū vowel options (Table 6), hamza options (Table 7), and the article families of 5.3.2."
|
|
22
|
+
note "Language-knowledge steps of the modified system that stay outside stateless scope and are NOT encoded: supplying unwritten vowels (5.2.1 - the printed example li-l-walad corresponds to لِلْوَلَدِ with the flexional final kasra omitted; this map renders the written vowel, li-l-waladi), general 5.3.1 hyphenation of li-/bi-/ka-/wa- beyond the article constructions (a word-initial لِسَان lisān is one word, not li+sān), the article after wa-/ka- (the 5.3.2 l- exception is encoded only for the documented li-/bi- forms; والشَّمْس-class words are ambiguous, e.g. والد wālid), and the šadda of an assimilated article preceded by wa-/ka-."
|
|
23
|
+
note "Table 8 No. 8 duplication is encoded as consonant doubling (28 digraph rules - the corpus idiom of iso-ara-Arab-Latn-233-1984), suppressed in the article assimilation context per the 5.3.2 examples (al-šams, bi-l-layl): the article rendering al-/l- carries the assimilation, and the doubled article lām emits the second l (al-ladhi -> al-laḏī). The suppression is a before-constraint on the preceding word-initial article; the parallel group is therefore megaregexp-ordered (specific rules first), which is cross-runtime consistent - the parity audit's G1 divergence concerns capture/ref only."
|
|
24
|
+
note "Table errata encoded per the governing column/example and flagged against the published edition: Table 6 No. 1 prints hex 0622 where the example (aẖ) and Table 7 No. 4b show the initial alif-with-hamza reading; Table 6 No. 2 prints 0101 in the Hex column of an (omitted) row; Table 7 No. 4 prints sāʾala where the written signs of سَأَلَ yield saʾala."
|
|
25
|
+
note "Reversibility: the modified system is not necessarily reversible (5.1.1); the reverse direction restores strict-sign equivalents of the outputs (e.g. walâkin -> وَلٰكِن without the omitted sukūn). A stateless map cannot supply unwritten vowels (level (b) is documented on the strict map)."
|
|
26
|
+
}
|
|
27
|
+
}
|
|
28
|
+
|
|
29
|
+
dependency "ara-latn-233-2026" as iso233
|
|
30
|
+
|
|
31
|
+
tests {
|
|
32
|
+
# Hand-written worked-example and structure tests.
|
|
33
|
+
"\u0623\u064e\u062e" -> "a\u1e96" # Table 6 No. 1 + example: aẖ
|
|
34
|
+
"\u0648\u064e\u0644\u064e\u0670\u0643\u0650\u0646" -> "wal\u00e2kin" # Table 6 No. 1a + example: walâkin
|
|
35
|
+
"\u0645\u064e\u0631\u064e\u0627\u064a\u064e\u0627" -> "mar\u0101y\u0101" # Table 6 No. 1b + example: marāyā
|
|
36
|
+
"\u0634\u064e\u062c\u064e\u0631\u064e\u0629\u064b" -> "\u0161a\u01e7ara\u1e97" # Table 6 No. 2 + example: šaǧaraẗ (tanwīn omitted)
|
|
37
|
+
"\u0623\u064e\u064a\u0652\u0636\u064b\u0627" -> "ay\u1e0dan" # Table 6 No. 2a + example: ayḍan
|
|
38
|
+
"\u0623\u064e\u064a\u0652\u0636\u0627\u064b" -> "ay\u1e0dan" # Table 6 No. 2b + example: ayḍan (alif-first order)
|
|
39
|
+
"\u0635\u064e\u062f\u0649" -> "\u1e63ad\u00e1" # Table 6 No. 3 + example: ṣadá
|
|
40
|
+
"\u0635\u064e\u062f\u0649\u0670" -> "\u1e63ad\u00e1" # Table 6 No. 3a: ṣadá
|
|
41
|
+
"\u0635\u064e\u062f\u064e\u0649" -> "\u1e63ad\u00e1" # Table 6 No. 3b: ṣadá
|
|
42
|
+
"\u0635\u064e\u062f\u064e\u0649\u0670" -> "\u1e63ad\u00e1" # Table 6 No. 3c: ṣadá
|
|
43
|
+
"\u0641\u064e\u062a\u064b\u0649" -> "fat\u00e1" # Table 6 No. 4 + example: fatá
|
|
44
|
+
"\u0641\u064e\u062a\u0649\u064b" -> "fat\u00e1" # Table 6 No. 4a: fatá
|
|
45
|
+
"\u062f\u064f\u0631\u064f\u0648\u0633" -> "dur\u016bs" # Table 6 No. 5 + example: durūs
|
|
46
|
+
"\u0634\u064e\u062c\u064e\u0631\u064e\u0629\u064c" -> "\u0161a\u01e7ara\u1e97" # Table 6 No. 5a: šaǧaraẗ
|
|
47
|
+
"\u0633\u064e\u0639\u0650\u064a\u062f" -> "sa\u02bf\u012bd" # Table 6 No. 6 + example: saʿīd
|
|
48
|
+
"\u0634\u064e\u062c\u064e\u0631\u064e\u0629\u064d" -> "\u0161a\u01e7ara\u1e97" # Table 6 No. 6a: šaǧaraẗ
|
|
49
|
+
"\u0621\u064e\u0644\u064e" -> "\u02beala" # Table 7 No. 4 (rule-derived saʾala; the printed sāʾala is an erratum)
|
|
50
|
+
"\u0628\u0650\u0626\u0652\u0631" -> "bi\u02ber" # Table 7 No. 4a + example: biʾr
|
|
51
|
+
"\u0623\u064e\u062e\u0628\u064e\u0627\u064c" -> "a\u1e96b\u0101" # Table 7 No. 4b + example: aḫbār
|
|
52
|
+
"\u0642\u064f\u0631\u0652\u0622\u0646" -> "qur\u02be\u0101n" # Table 7 No. 5 + example: qurʾān
|
|
53
|
+
"\u0622\u062f\u064e\u0627\u0628" -> "\u0101d\u0101b" # Table 7 No. 5a + example: ādāb
|
|
54
|
+
"\u0641\u0650\u064a \u0646\u064f\u0648\u0631\u0650 \u0671\u0644\u0652\u0642\u064e\u0645\u064e\u0631" -> "f\u012b n\u016bri al-qamar" # Table 7 No. 6 + example: fī nūri al-qamar (flexional -i rendered)
|
|
55
|
+
"\u0647\u064e\u0645\u0652\u0629" -> "ham\u1e97" # Table 8 No. 7 + example: hamzaẗ (sukūn omitted)
|
|
56
|
+
"\u0639\u064e\u0628\u0651\u0627\u0633" -> "\u02bfabb\u0101s" # Table 8 No. 8 + example: ʿabbās
|
|
57
|
+
"\u0645\u064e\u062f\u0650\u064a\u064e\u0629\u064c" -> "mad\u012ba\u1e97" # 5.2.4 + example: madīnaẗ
|
|
58
|
+
"\u0627\u0644\u0634\u0651\u064e\u0645\u0652\u0633" -> "al-\u0161ams" # 5.3.2 example: al-šams (assimilation šadda suppressed)
|
|
59
|
+
"\u0627\u0644\u0642\u064e\u0645\u064e\u0631" -> "al-qamar" # moon article: al-qamar
|
|
60
|
+
"\u0628\u0650\u0671\u0644\u0644\ufc60\u064a\u0652\u0644" -> "bi-l-layl" # 5.3.2 example, document encoding (بِٱللَّيْل): bi-l-layl
|
|
61
|
+
"\u0628\u0650\u0627\u0644\u0644\u064e\u0651\u064a\u0652\u0644" -> "bi-l-layl" # 5.3.2 example, NFC canonical order: bi-l-layl
|
|
62
|
+
"\u0628\u0650\u0671\u0644\ufc60\u064a\u0652\u0644" -> "bi-l-layl" # 5.3.2 example, one-lām encoding: bi-l-layl
|
|
63
|
+
"\u0628\u0650\u0671\u0644\u0652\u0642\u064e\u0645\u064e\u0631\u0650" -> "bi-l-qamari" # 5.3.2 bi- moon article: bi-l-qamari
|
|
64
|
+
"\u0644\u0650\u0644\u0652\u0648\u064e\u0644\u064e\u062f\u0650" -> "li-l-waladi" # 5.3.2 example: li-l-waladi (flexional -i rendered)
|
|
65
|
+
"\u0644\u0650\u0644\u0644\ufc60\u0647" -> "li-l-lah" # li- doubled lām (lillāhi-type): li-l-lahi
|
|
66
|
+
"\u0627\u0644\u0644\ufc60\u0630\u0650\u064a" -> "al-la\u1e0f\u012b" # doubled article lām (al-ladhi): al-laḏī
|
|
67
|
+
"\u0644\u0650\u0633\u064e\u0627\u0646" -> "lis\u0101n" # no false prefix: lisān (5.3.1 gap: no standalone li- rule)
|
|
68
|
+
"\u0642\u064e\u0627\u0644\u064e\u0645\u064e" -> "q\u0101lama" # mid-word al- sequence stays per-letter: qālama
|
|
69
|
+
"\u0628\u0627\u0628" -> "b\u0101b" # medial bare alif -> ā (5.2.2)
|
|
70
|
+
"\u062e" -> "\u1e96" # shared consonant ruleset passthrough
|
|
71
|
+
|
|
72
|
+
# Hand-written worked-example and structure tests below; the exhaustive per-row
|
|
73
|
+
# tests follow, generated from the source document's tables (Tables 6-8; the 14 non-empty rows transcribed, the six (omitted) rows covered by the hand tests per the artifact note);
|
|
74
|
+
# the strict-table rows inherited per 5.1.2 carry their provenance in the strict
|
|
75
|
+
# map's artifact).
|
|
76
|
+
# Source: ISO/FDIS 233-1:2026, ISO/TC 46/WG 3 N129 (FDIS for ballot, 2026-11-02),
|
|
77
|
+
# clause 5; adjudications ISO/TC 46/WG 3 N130. Expected strings are the map outputs.
|
|
78
|
+
# Provenance artifact: scripts/row-sources/iso-ara-Arab-Latn-233-2026-modified.json (21 rows).
|
|
79
|
+
# generated-by: maps/scripts/gen-row-tests.mjs (2026-11-19); regenerate from the artifact, do not hand-edit.
|
|
80
|
+
# Exhaustive per-row tests for iso-ara-Arab-Latn-233-2026-modified, generated from the source
|
|
81
|
+
# document's table.
|
|
82
|
+
# Source: ISO 233-1:2026 (FDIS) modified transliteration, clause 5 Tables
|
|
83
|
+
# 6-8: the rows with a non-empty romanization, 14 rows in document order,
|
|
84
|
+
# plus eight (omitted or position-qualified) rows documented here because
|
|
85
|
+
# the generator requires a non-empty output: Table 6 No. 1 (word-initial
|
|
86
|
+
# alif with hamza omitted; the printed hex 0622 is an erratum for the
|
|
87
|
+
# initial-alif-with-hamza reading per the row's own example aẖ), No. 2, No.
|
|
88
|
+
# 5a, No. 6a (tanwīn rows omitted), Table 7 No. 4a (the 0654/0655 rows are
|
|
89
|
+
# position-qualified: non-initial ʾ, initial omitted - biʾr/aḫbār are
|
|
90
|
+
# hand-tested; a standalone mark string is word-initial), No. 6 (hamzat
|
|
91
|
+
# al-waṣl omitted), Table 8 No. 7 (sukūn omitted) - all covered by the map's
|
|
92
|
+
# hand tests (aẖ, šaǧaraẗ, aḫbār, fī nūri al-qamar, hamzaẗ).
|
|
93
|
+
# Position-qualified equivalents of the transcribed rows: No. 1 non-initial
|
|
94
|
+
# alif renders ā (5.2.2), No. 5 non-initial 0622 renders ʾā (qurʾān), No. 4b
|
|
95
|
+
# initial hamza omitted (aḫbār). Table 8 No. 8 'duplication of consonant' is
|
|
96
|
+
# implemented as the 28 consonant+šadda digraph rules of the map, so it has
|
|
97
|
+
# no single-row transcription (ʿabbās is hand-tested). The strict-table rows
|
|
98
|
+
# the modified system inherits (5.1.2) carry their provenance in
|
|
99
|
+
# iso-ara-Arab-Latn-233-2026.json. Source: ISO/TC 46/WG 3 N129 with N130
|
|
100
|
+
# adjudications (CN-017/018/019 rejections honored). Verify against the
|
|
101
|
+
# published edition when available.
|
|
102
|
+
# Provenance artifact: scripts/row-sources/iso-ara-Arab-Latn-233-2026-modified.json (14 rows).
|
|
103
|
+
# generated-by: maps/scripts/gen-row-tests.mjs (2026-11-19); regenerate from the artifact, do not hand-edit.
|
|
104
|
+
"ٰ" -> "â"
|
|
105
|
+
"َا" -> "ā"
|
|
106
|
+
"ًا" -> "an"
|
|
107
|
+
"اً" -> "an"
|
|
108
|
+
"ى" -> "á"
|
|
109
|
+
"ىٰ" -> "á"
|
|
110
|
+
"َى" -> "á"
|
|
111
|
+
"َىٰ" -> "á"
|
|
112
|
+
"ًى" -> "á"
|
|
113
|
+
"ىً" -> "á"
|
|
114
|
+
"ُو" -> "ū"
|
|
115
|
+
"ِي" -> "ī"
|
|
116
|
+
"ء" -> "ʾ"
|
|
117
|
+
"آ" -> "ā"
|
|
118
|
+
}
|
|
119
|
+
|
|
120
|
+
stage main {
|
|
121
|
+
# Pass 1: mark-order and ligature normalization. NFC canonical order places
|
|
122
|
+
# harakat (ccc 30-32) before šadda (ccc 33); the FDIS examples decompose
|
|
123
|
+
# U+FC60 shadda-first. This pass normalizes harakat+šadda to šadda+harakat
|
|
124
|
+
# and decomposes the šadda ligatures, so every source order transliterates
|
|
125
|
+
# identically and the šadda output attaches to the consonant (Table 3
|
|
126
|
+
# footnote a: the macron above ā, ī, ū does not represent šaddaẗ).
|
|
127
|
+
parallel {
|
|
128
|
+
sub "\ufc5f" "\u0651\u064d"
|
|
129
|
+
sub "\ufc60" "\u0651\u064e"
|
|
130
|
+
sub "\ufc61" "\u0651\u064f"
|
|
131
|
+
sub "\ufc62" "\u0651\u0650"
|
|
132
|
+
sub "\ufc63" "\u0651\u0670"
|
|
133
|
+
sub "\u064e\u0651" "\u0651\u064e"
|
|
134
|
+
sub "\u064f\u0651" "\u0651\u064f"
|
|
135
|
+
sub "\u0650\u0651" "\u0651\u0650"
|
|
136
|
+
sub "\u064b\u0651" "\u0651\u064b"
|
|
137
|
+
sub "\u064c\u0651" "\u0651\u064c"
|
|
138
|
+
sub "\u064d\u0651" "\u0651\u064d"
|
|
139
|
+
}
|
|
140
|
+
|
|
141
|
+
# Pass 2: clause 5 over the strict tables (5.1.2). The group carries word-initial
|
|
142
|
+
# article rules (boundary forms) and before-constrained suppressions, so it runs
|
|
143
|
+
# megaregexp-ordered: specific before generic, suppressions before duplication.
|
|
144
|
+
parallel {
|
|
145
|
+
use iso233.consonants
|
|
146
|
+
use iso233.short_vowels
|
|
147
|
+
use iso233.punctuation
|
|
148
|
+
use iso233.numerals
|
|
149
|
+
# 5.3.2 article assimilation: the šadda of the sun-letter assimilation and of the
|
|
150
|
+
# doubled article lām is carried by the article rendering al-/l- (examples al-šams,
|
|
151
|
+
# bi-l-layl, li-l-walad), so the duplication of Table 8 No. 8 is suppressed
|
|
152
|
+
# immediately after the article. Constrained rules come first: the megaregexp
|
|
153
|
+
# group resolves equal-position matches in document order. The constraints are
|
|
154
|
+
# fixed-length look-behinds (ruby Onigmo rejects variable length) over the bare
|
|
155
|
+
# article form; the assimilated article is written bare (the assimilated lam
|
|
156
|
+
# carries the shadda, not a sukun), and sukun-marked article orthography of the
|
|
157
|
+
# al-shadda type falls back to duplication (flagged in the notes).
|
|
158
|
+
sub { from "\u062a\u0651" to "t" before boundary + any("\u0627\u0671") + "\u0644" }
|
|
159
|
+
sub { from "\u062a\u0651" to "t" before boundary + "\u0628\u0650" + any("\u0627\u0671") + "\u0644" }
|
|
160
|
+
sub { from "\u062a\u0651" to "t" before boundary + "\u0644\u0650\u0644" }
|
|
161
|
+
sub { from "\u062b\u0651" to "\u1e6f" before boundary + any("\u0627\u0671") + "\u0644" }
|
|
162
|
+
sub { from "\u062b\u0651" to "\u1e6f" before boundary + "\u0628\u0650" + any("\u0627\u0671") + "\u0644" }
|
|
163
|
+
sub { from "\u062b\u0651" to "\u1e6f" before boundary + "\u0644\u0650\u0644" }
|
|
164
|
+
sub { from "\u062f\u0651" to "d" before boundary + any("\u0627\u0671") + "\u0644" }
|
|
165
|
+
sub { from "\u062f\u0651" to "d" before boundary + "\u0628\u0650" + any("\u0627\u0671") + "\u0644" }
|
|
166
|
+
sub { from "\u062f\u0651" to "d" before boundary + "\u0644\u0650\u0644" }
|
|
167
|
+
sub { from "\u0630\u0651" to "\u1e0f" before boundary + any("\u0627\u0671") + "\u0644" }
|
|
168
|
+
sub { from "\u0630\u0651" to "\u1e0f" before boundary + "\u0628\u0650" + any("\u0627\u0671") + "\u0644" }
|
|
169
|
+
sub { from "\u0630\u0651" to "\u1e0f" before boundary + "\u0644\u0650\u0644" }
|
|
170
|
+
sub { from "\u0631\u0651" to "r" before boundary + any("\u0627\u0671") + "\u0644" }
|
|
171
|
+
sub { from "\u0631\u0651" to "r" before boundary + "\u0628\u0650" + any("\u0627\u0671") + "\u0644" }
|
|
172
|
+
sub { from "\u0631\u0651" to "r" before boundary + "\u0644\u0650\u0644" }
|
|
173
|
+
sub { from "\u0632\u0651" to "z" before boundary + any("\u0627\u0671") + "\u0644" }
|
|
174
|
+
sub { from "\u0632\u0651" to "z" before boundary + "\u0628\u0650" + any("\u0627\u0671") + "\u0644" }
|
|
175
|
+
sub { from "\u0632\u0651" to "z" before boundary + "\u0644\u0650\u0644" }
|
|
176
|
+
sub { from "\u0633\u0651" to "s" before boundary + any("\u0627\u0671") + "\u0644" }
|
|
177
|
+
sub { from "\u0633\u0651" to "s" before boundary + "\u0628\u0650" + any("\u0627\u0671") + "\u0644" }
|
|
178
|
+
sub { from "\u0633\u0651" to "s" before boundary + "\u0644\u0650\u0644" }
|
|
179
|
+
sub { from "\u0634\u0651" to "\u0161" before boundary + any("\u0627\u0671") + "\u0644" }
|
|
180
|
+
sub { from "\u0634\u0651" to "\u0161" before boundary + "\u0628\u0650" + any("\u0627\u0671") + "\u0644" }
|
|
181
|
+
sub { from "\u0634\u0651" to "\u0161" before boundary + "\u0644\u0650\u0644" }
|
|
182
|
+
sub { from "\u0635\u0651" to "\u1e63" before boundary + any("\u0627\u0671") + "\u0644" }
|
|
183
|
+
sub { from "\u0635\u0651" to "\u1e63" before boundary + "\u0628\u0650" + any("\u0627\u0671") + "\u0644" }
|
|
184
|
+
sub { from "\u0635\u0651" to "\u1e63" before boundary + "\u0644\u0650\u0644" }
|
|
185
|
+
sub { from "\u0636\u0651" to "\u1e0d" before boundary + any("\u0627\u0671") + "\u0644" }
|
|
186
|
+
sub { from "\u0636\u0651" to "\u1e0d" before boundary + "\u0628\u0650" + any("\u0627\u0671") + "\u0644" }
|
|
187
|
+
sub { from "\u0636\u0651" to "\u1e0d" before boundary + "\u0644\u0650\u0644" }
|
|
188
|
+
sub { from "\u0637\u0651" to "\u1e6d" before boundary + any("\u0627\u0671") + "\u0644" }
|
|
189
|
+
sub { from "\u0637\u0651" to "\u1e6d" before boundary + "\u0628\u0650" + any("\u0627\u0671") + "\u0644" }
|
|
190
|
+
sub { from "\u0637\u0651" to "\u1e6d" before boundary + "\u0644\u0650\u0644" }
|
|
191
|
+
sub { from "\u0638\u0651" to "\u1e93" before boundary + any("\u0627\u0671") + "\u0644" }
|
|
192
|
+
sub { from "\u0638\u0651" to "\u1e93" before boundary + "\u0628\u0650" + any("\u0627\u0671") + "\u0644" }
|
|
193
|
+
sub { from "\u0638\u0651" to "\u1e93" before boundary + "\u0644\u0650\u0644" }
|
|
194
|
+
sub { from "\u0644\u0651" to "l" before boundary + any("\u0627\u0671") + "\u0644" }
|
|
195
|
+
sub { from "\u0644\u0651" to "l" before boundary + "\u0628\u0650" + any("\u0627\u0671") + "\u0644" }
|
|
196
|
+
sub { from "\u0644\u0651" to "l" before boundary + "\u0644\u0650\u0644" }
|
|
197
|
+
sub { from "\u0646\u0651" to "n" before boundary + any("\u0627\u0671") + "\u0644" }
|
|
198
|
+
sub { from "\u0646\u0651" to "n" before boundary + "\u0628\u0650" + any("\u0627\u0671") + "\u0644" }
|
|
199
|
+
sub { from "\u0646\u0651" to "n" before boundary + "\u0644\u0650\u0644" }
|
|
200
|
+
# doubled article lām (al-ladhi-type): emits the second l
|
|
201
|
+
sub { from "\u0644\u0651" to "l" before boundary + any("\u0627\u0671") + "\u0644" }
|
|
202
|
+
sub { from "\u0644\u0651" to "l" before boundary + "\u0628\u0650" + any("\u0627\u0671") + "\u0644" }
|
|
203
|
+
sub { from "\u0644\u0651" to "l" before boundary + "\u0644\u0650\u0644" }
|
|
204
|
+
# 5.3.2 article families, word-initial (specific before generic).
|
|
205
|
+
# bi- article, doubled lām (bi-l-layl):
|
|
206
|
+
sub { from boundary + "\u0628\u0650" + any("\u0627\u0671") + maybe("\u0652") + "\u0644" + "\u0651\u064e" to "bi-l-la" }
|
|
207
|
+
sub { from boundary + "\u0628\u0650" + any("\u0627\u0671") + maybe("\u0652") + "\u0644" + "\u0651\u064f" to "bi-l-lu" }
|
|
208
|
+
sub { from boundary + "\u0628\u0650" + any("\u0627\u0671") + maybe("\u0652") + "\u0644" + "\u0651\u0650" to "bi-l-li" }
|
|
209
|
+
sub { from boundary + "\u0628\u0650" + any("\u0627\u0671") + maybe("\u0652") + "\u0644" + "\u0651" to "bi-l-l" }
|
|
210
|
+
# bi- article, moon letters (bi-l-qamar):
|
|
211
|
+
sub { from boundary + "\u0628\u0650" + any("\u0627\u0671") + "\u0644" to "bi-l-" }
|
|
212
|
+
# li- article, doubled lām:
|
|
213
|
+
sub { from boundary + "\u0644\u0650\u0644" + "\u0651\u064e" to "li-l-la" }
|
|
214
|
+
sub { from boundary + "\u0644\u0650\u0644" + "\u0651\u064f" to "li-l-lu" }
|
|
215
|
+
sub { from boundary + "\u0644\u0650\u0644" + "\u0651\u0650" to "li-l-li" }
|
|
216
|
+
sub { from boundary + "\u0644\u0650\u0644" + "\u0651" to "li-l-l" }
|
|
217
|
+
# li- article, moon letters (li-l-walad):
|
|
218
|
+
sub { from boundary + "\u0644\u0650\u0644" to "li-l-" }
|
|
219
|
+
# bare article, doubled lām (al-ladhi):
|
|
220
|
+
sub { from boundary + any("\u0627\u0671") + maybe("\u0652") + "\u0644" + "\u0651\u064e" to "al-la" }
|
|
221
|
+
sub { from boundary + any("\u0627\u0671") + maybe("\u0652") + "\u0644" + "\u0651\u064f" to "al-lu" }
|
|
222
|
+
sub { from boundary + any("\u0627\u0671") + maybe("\u0652") + "\u0644" + "\u0651\u0650" to "al-li" }
|
|
223
|
+
sub { from boundary + any("\u0627\u0671") + maybe("\u0652") + "\u0644" + "\u0651" to "al-l" }
|
|
224
|
+
# bare article (al-qamar, al-šams):
|
|
225
|
+
sub { from boundary + any("\u0627\u0671") + "\u0644" to "al-" }
|
|
226
|
+
#
|
|
227
|
+
# Table 6 - modified transliteration options for vowels (rows 1-6a).
|
|
228
|
+
# No. 1a: alif ẖanǧariyyaẗ -> â (walâkin); the fatḥa the dagger alif supplies is absorbed
|
|
229
|
+
sub "\u064e\u0670" "\u00e2"
|
|
230
|
+
sub "\u0670" "\u00e2"
|
|
231
|
+
# No. 1b: fatḥa + lengthening alif -> ā (marāyā; 5.2.2)
|
|
232
|
+
sub "\u064e\u0627" "\u0101"
|
|
233
|
+
# No. 2/2a/2b: tanwīn al-fatḥ omitted; an before/following alif (ayḍan)
|
|
234
|
+
sub "\u064b\u0627" "an"
|
|
235
|
+
sub "\u0627\u064b" "an"
|
|
236
|
+
sub "\u064b" ""
|
|
237
|
+
# No. 3-4a: alif maqṣūraẗ -> á (ṣadá, fatá)
|
|
238
|
+
sub "\u064e\u0649\u0670" "\u00e1"
|
|
239
|
+
sub "\u0649\u0670" "\u00e1"
|
|
240
|
+
sub "\u064b\u0649" "\u00e1"
|
|
241
|
+
sub "\u0649\u064b" "\u00e1"
|
|
242
|
+
sub "\u064e\u0649" "\u00e1"
|
|
243
|
+
sub "\u0649" "\u00e1"
|
|
244
|
+
# No. 5: ḍamma + wāw -> ū (durūs); No. 6: kasra + yāʾ -> ī (saʿīd)
|
|
245
|
+
sub "\u064f\u0648" "\u016b"
|
|
246
|
+
sub "\u0650\u064a" "\u012b"
|
|
247
|
+
# No. 5a/6a: tanwīn al-ḍamm / al-kasr omitted
|
|
248
|
+
sub "\u064c" ""
|
|
249
|
+
sub "\u064d" ""
|
|
250
|
+
# No. 1 + 5.2.2: initial alif omitted, T7 No. 4b: initial hamza omitted (aḫbār);
|
|
251
|
+
# word-initial only - the multi-char Table 6 rules above take precedence in
|
|
252
|
+
# megaregexp document order (e.g. اً renders an, not the omission)
|
|
253
|
+
sub { from "\u0627" to "" before boundary }
|
|
254
|
+
sub { from "\u0623" to "" before boundary }
|
|
255
|
+
sub { from "\u0625" to "" before boundary }
|
|
256
|
+
sub { from "\u0654" to "" before boundary }
|
|
257
|
+
sub { from "\u0655" to "" before boundary }
|
|
258
|
+
sub { from "\u0624" to "" before boundary }
|
|
259
|
+
sub { from "\u0626" to "" before boundary }
|
|
260
|
+
# Table 7 - hamzaẗ and alif maddaẗ options.
|
|
261
|
+
# No. 5a/5: 0622 word-initial -> ā (ādāb), non-initial -> ʾā (qurʾān)
|
|
262
|
+
sub { from boundary + "\u0622" to "\u0101" }
|
|
263
|
+
sub "\u0622" "\u02be\u0101"
|
|
264
|
+
# No. 4: hamzaẗ non-initial -> ʾ (saʾala; the printed sāʾala is an erratum, see notes)
|
|
265
|
+
sub "\u0621" "\u02be"
|
|
266
|
+
# No. 4a: combining hamzaẗ non-initial -> ʾ (biʾr); initial omitted above (No. 4b)
|
|
267
|
+
sub "\u0654" "\u02be"
|
|
268
|
+
sub "\u0655" "\u02be"
|
|
269
|
+
# 4.1.4 precomposed hamza seats: non-initial -> ʾ (5.2.3: the vocalic-context hamza
|
|
270
|
+
# is not represented; the seat is not separately represented)
|
|
271
|
+
sub "\u0623" "\u02be"
|
|
272
|
+
sub "\u0625" "\u02be"
|
|
273
|
+
sub "\u0624" "\u02be"
|
|
274
|
+
sub "\u0626" "\u02be"
|
|
275
|
+
# No. 6: hamzaẗ al-waṣl omitted (fī nūr al-qamar); the word-initial wasla article
|
|
276
|
+
# family above renders al- per the table's own example
|
|
277
|
+
sub "\u0671" ""
|
|
278
|
+
# 5.2.2 non-initial bare alif: the lengthening alif renders ā
|
|
279
|
+
sub "\u0627" "\u0101"
|
|
280
|
+
# Table 8 - other signs.
|
|
281
|
+
# No. 7: sukūn omitted (hamzaẗ)
|
|
282
|
+
sub "\u0652" ""
|
|
283
|
+
# No. 8: šaddaẗ -> duplication of consonant (ʿabbās; the corpus idiom of
|
|
284
|
+
# iso-ara-Arab-Latn-233-1984); a bare orphan šadda with no consonant to double is dropped
|
|
285
|
+
sub "\u0651" ""
|
|
286
|
+
sub "\u0628\u0651" "bb"
|
|
287
|
+
sub "\u062a\u0651" "tt"
|
|
288
|
+
sub "\u062b\u0651" "\u1e6f\u1e6f"
|
|
289
|
+
sub "\u062c\u0651" "\u01e7\u01e7"
|
|
290
|
+
sub "\u062d\u0651" "\u1e25\u1e25"
|
|
291
|
+
sub "\u062e\u0651" "\u1e96\u1e96"
|
|
292
|
+
sub "\u062f\u0651" "dd"
|
|
293
|
+
sub "\u0630\u0651" "\u1e0f\u1e0f"
|
|
294
|
+
sub "\u0631\u0651" "rr"
|
|
295
|
+
sub "\u0632\u0651" "zz"
|
|
296
|
+
sub "\u0633\u0651" "ss"
|
|
297
|
+
sub "\u0634\u0651" "\u0161\u0161"
|
|
298
|
+
sub "\u0635\u0651" "\u1e63\u1e63"
|
|
299
|
+
sub "\u0636\u0651" "\u1e0d\u1e0d"
|
|
300
|
+
sub "\u0637\u0651" "\u1e6d\u1e6d"
|
|
301
|
+
sub "\u0638\u0651" "\u1e93\u1e93"
|
|
302
|
+
sub "\u0639\u0651" "\u02bf\u02bf"
|
|
303
|
+
sub "\u063a\u0651" "\u0121\u0121"
|
|
304
|
+
sub "\u0641\u0651" "ff"
|
|
305
|
+
sub "\u0642\u0651" "qq"
|
|
306
|
+
sub "\u0643\u0651" "kk"
|
|
307
|
+
sub "\u0644\u0651" "ll"
|
|
308
|
+
sub "\u0645\u0651" "mm"
|
|
309
|
+
sub "\u0646\u0651" "nn"
|
|
310
|
+
sub "\u0647\u0651" "hh"
|
|
311
|
+
sub "\u0629\u0651" "\u1e97\u1e97"
|
|
312
|
+
sub "\u0648\u0651" "ww"
|
|
313
|
+
sub "\u064a\u0651" "yy"
|
|
314
|
+
# 5.3.1 general prefix hyphenation, vowel supplying, and flexional-omission are
|
|
315
|
+
# language-knowledge steps outside stateless scope (see notes).
|
|
316
|
+
}
|
|
317
|
+
}
|
|
318
|
+
}
|
|
@@ -0,0 +1,188 @@
|
|
|
1
|
+
system "ISO:ara-Arab:Latn:233-2026" {
|
|
2
|
+
|
|
3
|
+
metadata {
|
|
4
|
+
authority_id iso
|
|
5
|
+
id 233-2026
|
|
6
|
+
language iso-639-2:ara
|
|
7
|
+
source_script Arab
|
|
8
|
+
destination_script Latn
|
|
9
|
+
name "ISO 233-1:2026 Information and documentation -- Transliteration of Perso-Arabic characters into Latin characters -- Part 1: Arabic language (strict transliteration, clause 4; FDIS stage)"
|
|
10
|
+
url https://www.iso.org/committee/48820.html
|
|
11
|
+
creation_date 2026
|
|
12
|
+
adoption_date 2026
|
|
13
|
+
description { Strict (fully reversible, level (a)) transliteration of ISO 233-1:2026, clause 4,
|
|
14
|
+
Tables 1-5: written signs only, diacritical vowels transliterated only where written
|
|
15
|
+
in the source. Encoded from the committee ballot ISO/TC 46/WG 3 N129 (FDIS,
|
|
16
|
+
2026-11-02; ballot closed 2026-11-07) with the comment adjudications of N130.
|
|
17
|
+
See notes for provenance, errata and reversibility. }
|
|
18
|
+
|
|
19
|
+
notes {
|
|
20
|
+
note "Encoded from ISO/FDIS 233-1:2026, ISO/TC 46/WG 3 N129 (2026-11-02, FDIS ballot closed 2026-11-07), clause 4 Tables 1-5; adjudications honored from the collated comments N130: DE-002 accepted (hamzaẗ al-waṣl renders 02BE+0331), CN-017 rejected (alif treatment intentionally differs between strict and modified), CN-016 rejected (Table 4 intentionally lists only Arabic-encoded punctuation). Source documents archived in interscript/interscript-private-references (iso/ISO-TC46-WG3-N129-*, N130-*). Verify against the published edition when available. The FDIS cancels and replaces ISO 233:1984 and ISO 233-2:1993: new work uses this map while the legacy corpus maps iso-ara-Arab-Latn-233-1984 and iso-ara-Arab-Latn-233-2-1993 remain for existing data. ISO 233-3:2026 (Persian) is a separate part, already published; its legacy corpus coverage is iso-prs-Arab-Latn-233-3-1999."
|
|
21
|
+
note "Three levels of transliteration (Introduction): this map implements level (a), strict fully reversible, written signs only. Level (b) - strict with the diacritical vowels supplied regardless of the source - is an application mode of the same Tables 1-2, not a separate table; a stateless map cannot supply unwritten vowels, so the level is documented here rather than forced into the map."
|
|
22
|
+
note "Reversibility (4.1.3) holds one-to-one for every row except the merger the standard itself prescribes: Table 1 No. 2a maps both U+0654 and U+0655 to 02C8, so the reverse direction restores 0654. The foreword legacy variants for existing text (02BC, 0329, 02B9, 02BB) are notes, not rules of this map."
|
|
23
|
+
note "4.1.4 precomposed/decomposed equality: the map adds canonical-decomposition equivalents of the precomposed alef forms (0623, 0625, 0624, 0626; 0622 is a table row itself) and decomposes the šadda-haraka ligatures U+FC5F..U+FC63 in the normalization pass. Letter positional presentation forms (U+FE70..U+FEFF; Annex A is informative) pass through unchanged."
|
|
24
|
+
note "Mark order: a first pass normalizes harakat+šadda to šadda+harakat (NFC orders harakat ccc 30-32 before šadda ccc 33; the FDIS examples decompose U+FC60 shadda-first), so every source order transliterates identically and the šadda macron (0304) attaches to the consonant output. Per Table 3 footnote a, the macron above ā, ī, ū does not represent šaddaẗ - the precomposed ā/ī/ū outputs of the vowel rows are vowels, not šadda."
|
|
25
|
+
note "4.3.2 examples: example 1 (أَلشَّمْسُ ʾˈalš̄am̊su) is conformant with the tables and is tested verbatim in NFC-composed form. Printed examples 2-3 are errata: bi-̓llaẙli carries the DIS rendering of U+0671 (U+002D+U+0313) that DE-002 replaced with 02BE+0331 and omits the šadda macron; lilwaladi omits the sukūn ring. The rule-derived forms are tested instead (N130 CN-019: 4.3.2 and 5.3.2 intentionally represent different methods)."
|
|
26
|
+
}
|
|
27
|
+
}
|
|
28
|
+
|
|
29
|
+
dependency "ara-latn-233-2026" as iso233
|
|
30
|
+
|
|
31
|
+
tests {
|
|
32
|
+
# Hand-written worked-example and structure tests.
|
|
33
|
+
"\u0623\u064e\u0644\u0634\u0651\u064e\u0645\u0652\u0633\u064f" -> "\u02be\u02c8al\u0161\u0304am\u030asu" # 4.3.2 example 1 verbatim: ʾˈalš̄am̊su
|
|
34
|
+
"\u0646\u064e\u0642\u0650\u0641\u064f" -> "naqifu" # Introduction worked example: naqifu
|
|
35
|
+
"\u0646\u0642\u0641" -> "nqf" # Introduction worked example: nqf (unvocalized)
|
|
36
|
+
"\u0623\u064e\u062e" -> "\u02be\u02c8a\u1e96" # 4.1.4 precomposed أَ -> ʾˈa
|
|
37
|
+
"\u0627\u0654\u064e\u062e" -> "\u02be\u02c8a\u1e96" # 4.1.4 decomposed ا+0654+َ identical
|
|
38
|
+
"\u0634\ufc60\u0645" -> "\u0161\u0304am" # U+FC60 ligature
|
|
39
|
+
"\u0634\u064e\u0651\u0645" -> "\u0161\u0304am" # NFC order identical
|
|
40
|
+
"\u0634\u0651\u064e\u0645" -> "\u0161\u0304am" # typed order identical
|
|
41
|
+
"\u0628\u0651" -> "b\u0304" # šadda macron attaches to the consonant
|
|
42
|
+
"\u0645\u0652" -> "m\u030a" # sukūn ring attaches to the consonant
|
|
43
|
+
"\u0628\u0651\u064e\u0642\u064f" -> "b\u0304aqu" # šadda macron stays on the consonant before a following vowel
|
|
44
|
+
|
|
45
|
+
# Hand-written worked-example and structure tests below; the exhaustive per-row tests
|
|
46
|
+
# follow, generated from the source document's tables (Tables 1-5, 57 rows: 56 document rows, No. 2a split into its 0654/0655 pair).
|
|
47
|
+
# Source: ISO/FDIS 233-1:2026, ISO/TC 46/WG 3 N129 (FDIS for ballot, 2026-11-02;
|
|
48
|
+
# ballot closed 2026-11-07), clause 4. Adjudications: ISO/TC 46/WG 3 N130 collated
|
|
49
|
+
# comments. Expected strings are the map outputs (precomposed Latin per the hex
|
|
50
|
+
# columns; combining marks 030A/0304/0331 attach to the preceding output character).
|
|
51
|
+
# Provenance artifact: scripts/row-sources/iso-ara-Arab-Latn-233-2026.json (57 rows).
|
|
52
|
+
# generated-by: maps/scripts/gen-row-tests.mjs (2026-11-19); regenerate from the artifact, do not hand-edit.
|
|
53
|
+
# Exhaustive per-row tests for iso-ara-Arab-Latn-233-2026, generated from the source
|
|
54
|
+
# document's table.
|
|
55
|
+
# Source: ISO 233-1:2026 (FDIS) strict transliteration, clause 4 Tables 1-5,
|
|
56
|
+
# 57 rows (56 document rows; the 0654-or-0655 pair of No. 2a split) in
|
|
57
|
+
# document order: Table 1 consonants (No. 1-29a including 1a/2a; the
|
|
58
|
+
# 0654-or-0655 pair of No. 2a encoded as two rows sharing 02C8 per the
|
|
59
|
+
# standard), Table 2 vowels for the fully reversible system (No. 30-32a; the
|
|
60
|
+
# foreword's á-to-ã, í-to-ĩ, ú-to-ũ changes), Table 3 other conventional
|
|
61
|
+
# signs (No. 33 sukūn 030A, No. 34 šadda 0304 with footnote a 'The macron
|
|
62
|
+
# above ā, ī, ū does not represent šaddaẗ', No. 35 hamzat al-waṣl 02BE+0331
|
|
63
|
+
# per the accepted DE-002), Table 4 punctuation (the CN-016 rejection:
|
|
64
|
+
# intentionally only Arabic-encoded marks), Table 5 numerals. Source: ISO/TC
|
|
65
|
+
# 46/WG 3 N129 (ISO_FDIS 233-1 ed.1, 2026-11-02; ballot closed 2026-11-07),
|
|
66
|
+
# hex columns authoritative, cross-checked against the N130 collated
|
|
67
|
+
# comments. Ballot documents archived in
|
|
68
|
+
# interscript/interscript-private-references (iso/ISO-TC46-WG3-N129-* and
|
|
69
|
+
# N130-*). Verify against the published edition when available.
|
|
70
|
+
# Provenance artifact: scripts/row-sources/iso-ara-Arab-Latn-233-2026.json (57 rows).
|
|
71
|
+
# generated-by: maps/scripts/gen-row-tests.mjs (2026-11-19); regenerate from the artifact, do not hand-edit.
|
|
72
|
+
"ا" -> "ʾ"
|
|
73
|
+
"آ" -> "ʾâ"
|
|
74
|
+
"ء" -> "ˌ"
|
|
75
|
+
"ٔ" -> "ˈ"
|
|
76
|
+
"ٕ" -> "ˈ"
|
|
77
|
+
"ب" -> "b"
|
|
78
|
+
"ت" -> "t"
|
|
79
|
+
"ث" -> "ṯ"
|
|
80
|
+
"ج" -> "ǧ"
|
|
81
|
+
"ح" -> "ḥ"
|
|
82
|
+
"خ" -> "ẖ"
|
|
83
|
+
"د" -> "d"
|
|
84
|
+
"ذ" -> "ḏ"
|
|
85
|
+
"ر" -> "r"
|
|
86
|
+
"ز" -> "z"
|
|
87
|
+
"س" -> "s"
|
|
88
|
+
"ش" -> "š"
|
|
89
|
+
"ص" -> "ṣ"
|
|
90
|
+
"ض" -> "ḍ"
|
|
91
|
+
"ط" -> "ṭ"
|
|
92
|
+
"ظ" -> "ẓ"
|
|
93
|
+
"ع" -> "ʿ"
|
|
94
|
+
"غ" -> "ġ"
|
|
95
|
+
"ف" -> "f"
|
|
96
|
+
"ق" -> "q"
|
|
97
|
+
"ك" -> "k"
|
|
98
|
+
"ل" -> "l"
|
|
99
|
+
"م" -> "m"
|
|
100
|
+
"ن" -> "n"
|
|
101
|
+
"ه" -> "h"
|
|
102
|
+
"ة" -> "ẗ"
|
|
103
|
+
"و" -> "w"
|
|
104
|
+
"ي" -> "y"
|
|
105
|
+
"ى" -> "ỳ"
|
|
106
|
+
"َ" -> "a"
|
|
107
|
+
"ٰ" -> "ā"
|
|
108
|
+
"ً" -> "ã"
|
|
109
|
+
"ُ" -> "u"
|
|
110
|
+
"ٌ" -> "ũ"
|
|
111
|
+
"ِ" -> "i"
|
|
112
|
+
"ٍ" -> "ĩ"
|
|
113
|
+
"ْ" -> "̊"
|
|
114
|
+
"ّ" -> "̄"
|
|
115
|
+
"ٱ" -> "ʾ̱"
|
|
116
|
+
"،" -> ","
|
|
117
|
+
"؛" -> ";"
|
|
118
|
+
"؟" -> "?"
|
|
119
|
+
"٠" -> "0"
|
|
120
|
+
"١" -> "1"
|
|
121
|
+
"٢" -> "2"
|
|
122
|
+
"٣" -> "3"
|
|
123
|
+
"٤" -> "4"
|
|
124
|
+
"٥" -> "5"
|
|
125
|
+
"٦" -> "6"
|
|
126
|
+
"٧" -> "7"
|
|
127
|
+
"٨" -> "8"
|
|
128
|
+
"٩" -> "9"
|
|
129
|
+
}
|
|
130
|
+
|
|
131
|
+
stage main {
|
|
132
|
+
# Pass 1: mark-order and ligature normalization. NFC canonical order places
|
|
133
|
+
# harakat (ccc 30-32) before šadda (ccc 33); the FDIS examples decompose
|
|
134
|
+
# U+FC60 shadda-first. This pass normalizes harakat+šadda to šadda+harakat
|
|
135
|
+
# and decomposes the šadda ligatures, so every source order transliterates
|
|
136
|
+
# identically and the šadda output attaches to the consonant (Table 3
|
|
137
|
+
# footnote a: the macron above ā, ī, ū does not represent šaddaẗ).
|
|
138
|
+
parallel {
|
|
139
|
+
sub "\ufc5f" "\u0651\u064d"
|
|
140
|
+
sub "\ufc60" "\u0651\u064e"
|
|
141
|
+
sub "\ufc61" "\u0651\u064f"
|
|
142
|
+
sub "\ufc62" "\u0651\u0650"
|
|
143
|
+
sub "\ufc63" "\u0651\u0670"
|
|
144
|
+
sub "\u064e\u0651" "\u0651\u064e"
|
|
145
|
+
sub "\u064f\u0651" "\u0651\u064f"
|
|
146
|
+
sub "\u0650\u0651" "\u0651\u0650"
|
|
147
|
+
sub "\u064b\u0651" "\u0651\u064b"
|
|
148
|
+
sub "\u064c\u0651" "\u0651\u064c"
|
|
149
|
+
sub "\u064d\u0651" "\u0651\u064d"
|
|
150
|
+
}
|
|
151
|
+
|
|
152
|
+
# Pass 2: Tables 1-5 (clause 4). Single-constraint-free parallel pass (trie semantics).
|
|
153
|
+
parallel {
|
|
154
|
+
use iso233.consonants
|
|
155
|
+
use iso233.short_vowels
|
|
156
|
+
use iso233.punctuation
|
|
157
|
+
use iso233.numerals
|
|
158
|
+
# Table 1 No. 1: alif
|
|
159
|
+
sub "\u0627" "\u02be"
|
|
160
|
+
# Table 1 No. 1a: alif maddaẗ
|
|
161
|
+
sub "\u0622" "\u02be\u00e2"
|
|
162
|
+
# Table 1 No. 2: hamzaẗ (footnote b: legacy combining vertical line below 0329)
|
|
163
|
+
sub "\u0621" "\u02cc"
|
|
164
|
+
# Table 1 No. 2a: hamzaẗ combining above/below (footnote c: legacy modifier letter prime 02B9)
|
|
165
|
+
sub "\u0654" "\u02c8"
|
|
166
|
+
sub "\u0655" "\u02c8"
|
|
167
|
+
# Table 1 No. 29a: alif maqṣūraẗ
|
|
168
|
+
sub "\u0649" "\u1ef3"
|
|
169
|
+
# 4.1.4: canonical-decomposition equivalents of the precomposed alef forms
|
|
170
|
+
sub "\u0623" "\u02be\u02c8"
|
|
171
|
+
sub "\u0625" "\u02be\u02c8"
|
|
172
|
+
sub "\u0624" "w\u02c8"
|
|
173
|
+
sub "\u0626" "y\u02c8"
|
|
174
|
+
# Table 2 No. 30a: alif ẖanǧariyyaẗ
|
|
175
|
+
sub "\u0670" "\u0101"
|
|
176
|
+
# Table 2 No. 30b/31a/32a: tanwīn (foreword: á→ã, í→ĩ, ú→ũ)
|
|
177
|
+
sub "\u064b" "\u00e3"
|
|
178
|
+
sub "\u064c" "\u0169"
|
|
179
|
+
sub "\u064d" "\u0129"
|
|
180
|
+
# Table 3 No. 33: sukūn (combining ring above)
|
|
181
|
+
sub "\u0652" "\u030a"
|
|
182
|
+
# Table 3 No. 34: šaddaẗ (combining macron; footnote a)
|
|
183
|
+
sub "\u0651" "\u0304"
|
|
184
|
+
# Table 3 No. 35: hamzaẗ al-waṣl 02BE+0331 (N130 DE-002 accepted)
|
|
185
|
+
sub "\u0671" "\u02be\u0331"
|
|
186
|
+
}
|
|
187
|
+
}
|
|
188
|
+
}
|