farahidi 0.3.0__py3-none-any.whl

farahidi/__init__.py

@@ -0,0 +1,57 @@
+"""farahidi — Arabic morphological analyzer (a pure-Python port of AlKhalil Morpho Sys 2).
+
+Quick start::
+
+    import farahidi
+
+    for a in farahidi.analyze("لأنهم"):
+        print(a.voweled_word, a.lemma, a.root, a.part_of_speech)
+
+:func:`analyze` returns a list of :class:`Analysis` candidates sorted by
+``priority`` (most frequent first). For repeated use, build an :class:`Analyzer`
+once and reuse it — the bundled lexicon loads lazily and is shared across calls.
+"""
+
+from __future__ import annotations
+
+from functools import lru_cache
+
+from .analyzer import Analyzer
+from .disambiguate import Disambiguator
+from .models import Analysis, TokenResult
+
+__all__ = [
+    "Analyzer",
+    "Analysis",
+    "Disambiguator",
+    "TokenResult",
+    "analyze",
+    "analyze_text",
+    "__version__",
+]
+__version__ = "0.3.0"
+
+
+@lru_cache(maxsize=1)
+def _default_analyzer() -> Analyzer:
+    return Analyzer()
+
+
+@lru_cache(maxsize=1)
+def _default_disambiguator() -> Disambiguator:
+    return Disambiguator(analyzer=_default_analyzer())
+
+
+def analyze(word: str) -> list[Analysis]:
+    """Return all analyses of ``word``, sorted by ``priority`` (descending).
+
+    Uses a shared module-level :class:`Analyzer`.
+    """
+    return _default_analyzer().analyze(word)
+
+
+def analyze_text(text: str) -> list[TokenResult]:
+    """In-context analysis: one chosen lemma/stem/root per word token across the
+    sentence(s) in ``text``. Uses a shared module-level :class:`Disambiguator`.
+    """
+    return _default_disambiguator().analyze_text(text)