scriptconv 0.0.3a9__tar.gz → 0.0.3a10__tar.gz

{scriptconv-0.0.3a9 → scriptconv-0.0.3a10}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: scriptconv
-Version: 0.0.3a9
+Version: 0.0.3a10
 Summary: Zero-dependency script & phoneme-notation core — ISO-15924 detection & metadata, IPA↔ARPABET/X-SAMPA/Lexique/Kirshenbaum/Cotovía/RFE, Buckwalter↔Arabic, Hangul→jamo, kana
 License: Apache-2.0
 Project-URL: Homepage, https://github.com/TigreGotico/scriptconv

{scriptconv-0.0.3a9 → scriptconv-0.0.3a10}/scriptconv/scripts.py

@@ -652,6 +652,34 @@ _LANG_TO_SCRIPT: dict[str, str] = {
     "rw": "Latn",   # Kinyarwanda
     "tk": "Latn",   # Turkmen (Latin since 1993)
     "eo": "Latn",   # Esperanto
+    # Languages with no ISO 639-1 code (keyed by their 639-3 code directly)
+    "kjh": "Cyrl",  # Khakas
+    "xal": "Cyrl",  # Kalmyk (Cyrillic since 1938)
+    "kbd": "Cyrl",  # Kabardian
+    "erz": "Cyrl",  # Erzya
+    "mdf": "Cyrl",  # Moksha
+}
+
+# ISO 639-2/3 (incl. bibliographic variants) -> the 639-1 key used above.
+# Reference: Library of Congress ISO 639-2 code list.
+_ISO639_ALIASES: dict[str, str] = {
+    "rus": "ru", "ukr": "uk", "bel": "be", "bul": "bg", "srp": "sr",
+    "mkd": "mk", "mac": "mk", "kaz": "kk", "kir": "ky", "tgk": "tg",
+    "mon": "mn", "tat": "tt", "bak": "ba", "chv": "cv", "oss": "os",
+    "che": "ce", "hye": "hy", "arm": "hy", "kat": "ka", "geo": "ka",
+    "aze": "az", "uzb": "uz", "tuk": "tk", "eng": "en", "deu": "de",
+    "ger": "de", "fra": "fr", "fre": "fr", "spa": "es", "por": "pt",
+    "ita": "it", "nld": "nl", "dut": "nl", "ell": "el", "gre": "el",
+    "ara": "ar", "heb": "he", "zho": "zh", "chi": "zh", "jpn": "ja",
+    "kor": "ko", "tur": "tr", "fas": "fa", "per": "fa", "urd": "ur",
+    "hin": "hi", "tha": "th", "vie": "vi", "pol": "pl", "ces": "cs",
+    "cze": "cs", "slk": "sk", "slo": "sk", "ron": "ro", "rum": "ro",
+    "hun": "hu", "fin": "fi", "swe": "sv", "nor": "no", "dan": "da",
+    "isl": "is", "ice": "is", "mya": "my", "bur": "my", "khm": "km",
+    "lao": "lo", "amh": "am", "tam": "ta", "tel": "te", "ben": "bn",
+    "guj": "gu", "kan": "kn", "mal": "ml", "mar": "mr", "pan": "pa",
+    "sin": "si", "nep": "ne", "eus": "eu", "baq": "eu", "cat": "ca",
+    "glg": "gl", "cym": "cy", "wel": "cy", "gle": "ga", "epo": "eo",
 }
 
 
@@ -674,14 +702,21 @@ def lang_to_script(lang: str) -> Optional[str]:
     'Latn'
     """
     subtags = lang.replace("_", "-").split("-")
-    # An explicit script subtag (4 letters, e.g. "Latn") overrides the
-    # language's default script.
+    # An explicit script subtag overrides the language's default script:
+    # 4-letter ISO-15924 codes ("sr-Latn"), and the widespread informal
+    # 3-letter forms "cyr"/"lat" ("uzb_cyr", "aze_lat").
     for sub in subtags[1:]:
+        low = sub.lower()
         if len(sub) == 4 and sub.isalpha():
             resolved = normalize_script_tag(sub)
             if resolved is not None:
                 return resolved
-    return _LANG_TO_SCRIPT.get(subtags[0].lower())
+        elif low in ("cyr", "lat"):
+            return "Cyrl" if low == "cyr" else "Latn"
+    key = subtags[0].lower()
+    if key in _LANG_TO_SCRIPT:
+        return _LANG_TO_SCRIPT[key]
+    return _LANG_TO_SCRIPT.get(_ISO639_ALIASES.get(key, ""))
 
 
 # ---------------------------------------------------------------------------

{scriptconv-0.0.3a9 → scriptconv-0.0.3a10}/scriptconv/version.py

@@ -2,7 +2,7 @@
 VERSION_MAJOR = 0
 VERSION_MINOR = 0
 VERSION_BUILD = 3
-VERSION_ALPHA = 9
+VERSION_ALPHA = 10
 # END_VERSION_BLOCK
 
 VERSION_STR = f"{VERSION_MAJOR}.{VERSION_MINOR}.{VERSION_BUILD}"

{scriptconv-0.0.3a9 → scriptconv-0.0.3a10}/scriptconv.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: scriptconv
-Version: 0.0.3a9
+Version: 0.0.3a10
 Summary: Zero-dependency script & phoneme-notation core — ISO-15924 detection & metadata, IPA↔ARPABET/X-SAMPA/Lexique/Kirshenbaum/Cotovía/RFE, Buckwalter↔Arabic, Hangul→jamo, kana
 License: Apache-2.0
 Project-URL: Homepage, https://github.com/TigreGotico/scriptconv

{scriptconv-0.0.3a9 → scriptconv-0.0.3a10}/scriptconv.egg-info/SOURCES.txt

@@ -30,4 +30,5 @@ tests/test_notation.py
 tests/test_readings.py
 tests/test_readings_zh.py
 tests/test_scripts.py
+tests/test_scripts_stressonnx_compat.py
 tests/test_translit.py

scriptconv-0.0.3a10/tests/test_scripts_stressonnx_compat.py

@@ -0,0 +1,47 @@
+"""Downstream compat: the script tags scriptconv returns are a wire format.
+
+stressonnx (accentor.py) maps its supported languages to a private script
+taxonomy that mirrors phoonnx's Alphabet enum.  These fixtures pin, key by
+key, that scriptconv resolves every one of those language codes to the
+expected ISO-15924 tag — so stressonnx's migration to scriptconv is a pure
+deletion, and future edits here cannot silently break it.
+"""
+import pytest
+
+from scriptconv import lang_to_script
+
+# stressonnx LANG_SCRIPT, verbatim keys -> the ISO-15924 tag its Script
+# enum value corresponds to (cyrillic->Cyrl, latin->Latn, armenian->Armn,
+# georgian->Geor).
+STRESSONNX_LANGS = {
+    "ru": "Cyrl", "ukr": "Cyrl", "bel": "Cyrl", "bel_simple": "Cyrl",
+    "kaz": "Cyrl", "tat": "Cyrl", "bak": "Cyrl", "chv": "Cyrl",
+    "sah": "Cyrl", "kir": "Cyrl", "kjh": "Cyrl", "tgk": "Cyrl",
+    "udm": "Cyrl", "xal": "Cyrl", "kbd": "Cyrl", "erz": "Cyrl",
+    "mdf": "Cyrl", "uzb_cyr": "Cyrl", "aze_cyr": "Cyrl",
+    "aze_lat": "Latn", "uzb_lat": "Latn",
+    "hye": "Armn",
+    "kat": "Geor",
+}
+
+
+@pytest.mark.parametrize("lang,expected", sorted(STRESSONNX_LANGS.items()))
+def test_stressonnx_language_resolves(lang, expected):
+    assert lang_to_script(lang) == expected
+
+
+def test_iso639_bibliographic_variants():
+    assert lang_to_script("ger") == lang_to_script("deu") == "Latn"
+    assert lang_to_script("gre") == lang_to_script("ell") == "Grek"
+    assert lang_to_script("arm") == lang_to_script("hye") == "Armn"
+
+
+def test_explicit_script_subtag_still_authoritative():
+    assert lang_to_script("uzb-Latn") == "Latn"
+    assert lang_to_script("sr_lat") == "Latn"
+    assert lang_to_script("sr") == "Cyrl"
+
+
+def test_unknown_language_stays_none():
+    assert lang_to_script("zzz") is None
+    assert lang_to_script("qqq_cyr") == "Cyrl"  # variant subtag wins even unknown-lang