scriptconv 0.0.4a13__tar.gz → 0.0.4a14__tar.gz

{scriptconv-0.0.4a13/scriptconv.egg-info → scriptconv-0.0.4a14}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: scriptconv
-Version: 0.0.4a13
+Version: 0.0.4a14
 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-0.0.4a13 → scriptconv-0.0.4a14}/pyproject.toml

@@ -76,6 +76,10 @@ scriptconv = ["py.typed"]
 "scriptconv.phonemizers._thirdparty" = ["ko_tables/*.csv", "ko_tables/*.tsv"]
 "scriptconv.phonemizers._vendored.kog2p" = ["rulebook.txt", "LICENSE.md"]
 "scriptconv.phonemizers._vendored.mantoq" = ["LICENSE.md"]
+"scriptconv.phonemizers._vendored.africa_g2p" = [
+    "LICENSE.md", "DATA_LICENSE.md",
+    "languages/*.json", "data/*.json",
+]
 
 [tool.pytest.ini_options]
 testpaths = ["tests"]

{scriptconv-0.0.4a13 → scriptconv-0.0.4a14}/scriptconv/phonemizers/_vendored/__init__.py

@@ -8,6 +8,15 @@ distributed under ITS OWN license, stated in its directory:
   Arabic-Phonetiser) — non-commercial use only. See ``mantoq/LICENSE.md``.
 * ``kog2p/`` — Korean G2P (github.com/scarletcho/KoG2P), **GPL-3.0**.
   See ``kog2p/LICENSE.md``.
+* ``africa_g2p/`` — African-language G2P (github.com/AfriSpeech/africa-g2p),
+  400+ ISO 639-3 rule files. Vendored (not depended on) because it is not
+  published to PyPI and this project does not take ``git+`` dependencies.
+  Its **code** is Apache-2.0 — same license as the rest of this repository,
+  unlike the other two entries here — but its **data** (the ``languages/``
+  rule files and ``data/registry.json``) is derived from Omniglot script
+  charts (© Simon Ager) and Hartell (ed.), *Alphabets of Africa* (UNESCO/SIL,
+  1993), and carries its own attribution requirements. See
+  ``africa_g2p/LICENSE.md`` and ``africa_g2p/DATA_LICENSE.md``.
 
 Nothing in scriptconv imports these at package import time; they load only
 when a caller explicitly requests the corresponding phonemizer, accepting the

scriptconv-0.0.4a14/scriptconv/phonemizers/_vendored/africa_g2p/DATA_LICENSE.md

@@ -0,0 +1,36 @@
+# Data sources & licensing
+
+The **code** in this repository is licensed under Apache-2.0 (see `LICENSE`). The
+**linguistic data** in `src/africa_g2p/languages/` is derived from third-party sources and
+is provided for research and language-technology use. Please retain the attributions below.
+
+## Sources
+
+### Omniglot script charts — © Simon Ager
+Language rule files with `"confidence": "omniglot"` are derived from the writing-system
+charts published on [Omniglot](https://www.omniglot.com/) (© Simon Ager). Each such file
+records its source chart in its `"source"` field. We extract only the factual
+grapheme ↔ IPA ↔ romanisation correspondences (not Omniglot's prose, images, or page
+layout). Omniglot content is copyrighted; this project uses these factual correspondences
+with attribution and is not affiliated with or endorsed by Omniglot. If you are the
+rights-holder and have any concern about this use, please open an issue and we will respond
+promptly.
+
+### Alphabets of Africa (Hartell, ed.) — © UNESCO 1993
+Language rule files sourced from Rhonda L. Hartell (ed.), *Alphabets of Africa*
+(UNESCO / SIL, Dakar, 1993). Recorded in each file's `"source"` field as "Hartell 1993".
+
+### afriso — ISO 639-3 + Glottolog
+Language metadata (canonical codes, families, regions, alternative names) via
+[afriso](https://github.com/AfriSpeech/afriso), built from SIL ISO 639-3 tables and
+Glottolog (CC-BY 4.0).
+
+### africa-corpus — public Bible translations
+Example sentences in the README showcase come from
+[africa-corpus](https://github.com/AfriSpeech/africa-corpus-builder).
+
+## Using the data
+
+If you redistribute the language data, keep these attributions and this notice. For any
+use beyond research/attention that the original rights-holders' terms may restrict, obtain
+permission from the respective source.

scriptconv-0.0.4a14/scriptconv/phonemizers/_vendored/africa_g2p/LICENSE.md

@@ -0,0 +1,15 @@
+# Third-party license notice — africa-g2p
+
+Vendored from https://github.com/AfriSpeech/africa-g2p (not published to
+PyPI, so pinned in-tree rather than depended on). The **code**
+(`g2p.py`, `loader.py`, `normalizer.py`, `pipeline.py`, `cli.py`,
+`__init__.py`) is **Apache-2.0**, same as the rest of this repository —
+unlike `mantoq/` and `kog2p/`, this subpackage's code carries no extra
+restriction beyond scriptconv's own license.
+
+The **data** under `languages/` and `data/registry.json` is NOT
+Apache-2.0: it is derived from third-party sources (Omniglot script
+charts, © Simon Ager; Hartell 1993/UNESCO's *Alphabets of Africa*) and
+carries its own attribution requirements — see `DATA_LICENSE.md` in this
+directory. Redistributing the language data must keep those
+attributions.

scriptconv-0.0.4a14/scriptconv/phonemizers/_vendored/africa_g2p/__init__.py

@@ -0,0 +1,15 @@
+"""africa-g2p: rule-based grapheme-to-phoneme conversion for African languages."""
+from .g2p import G2P, g2p
+from .pipeline import AfricaPipeline
+from .loader import available_languages, registry, LanguageNotFoundError
+
+__version__ = "0.1.0"
+
+__all__ = [
+    "AfricaPipeline",
+    "G2P",
+    "g2p",
+    "available_languages",
+    "registry",
+    "LanguageNotFoundError",
+]

scriptconv-0.0.4a14/scriptconv/phonemizers/_vendored/africa_g2p/cli.py

@@ -0,0 +1,37 @@
+"""Command-line interface: ``africa-g2p <lang> "text"`` or ``--list``."""
+from __future__ import annotations
+
+import argparse
+import sys
+
+from .g2p import G2P
+from .loader import available_languages, registry, LanguageNotFoundError
+
+
+def main(argv=None) -> int:
+    parser = argparse.ArgumentParser(prog="africa-g2p", description=__doc__)
+    parser.add_argument("lang", nargs="?", help="ISO 639-3 language code")
+    parser.add_argument("text", nargs="?", help="text to convert (or read stdin)")
+    parser.add_argument("--sep", default=" ", help="separator between phonemes")
+    parser.add_argument("--list", action="store_true", help="list available languages")
+    args = parser.parse_args(argv)
+
+    if args.list or not args.lang:
+        reg = registry()
+        for code in available_languages():
+            name = reg.get(code, {}).get("name", "")
+            print(f"{code}\t{name}")
+        return 0
+
+    text = args.text if args.text is not None else sys.stdin.read()
+    try:
+        conv = G2P(args.lang)
+    except LanguageNotFoundError as e:
+        print(e, file=sys.stderr)
+        return 1
+    print(conv.convert(text, sep=args.sep))
+    return 0
+
+
+if __name__ == "__main__":
+    raise SystemExit(main())