add mypy type annotations

Author: pudo

.gitignore

@@ -1,6 +1,7 @@
 *.egg-info
 __pycache__
 .pytest_cache
+.mypy_cache
 .vscode
 build
 dist

MANIFEST.in

@@ -1,2 +1,3 @@
 include README.md
 include LICENSE
+include languagecodes/py.typed

languagecodes/__init__.py

@@ -1,32 +1,37 @@
+from typing import Iterable, Optional, Set
 from banal import ensure_list
 
 from languagecodes.iso639 import ISO3_ALL, ISO2_MAP, ISO3_MAP
 from languagecodes.synonyms import expand_synonyms
 from languagecodes.util import normalize_code
 
 
-def iso_639_alpha3(code):
+def iso_639_alpha3(code: str) -> Optional[str]:
     """Convert a given language identifier into an ISO 639 Part 2 code, such
     as "eng" or "deu". This will accept language codes in the two- or three-
     letter format, and some language names. If the given string cannot be
     converted, ``None`` will be returned.
     """
-    code = normalize_code(code)
-    code = ISO3_MAP.get(code, code)
-    if code in ISO3_ALL:
-        return code
+    norm = normalize_code(code)
+    if norm is not None:
+        norm = ISO3_MAP.get(norm, norm)
+    if norm not in ISO3_ALL:
+        return None
+    return norm
 
 
-def iso_639_alpha2(code):
+def iso_639_alpha2(code: str) -> Optional[str]:
     """Convert a language identifier to an ISO 639 Part 1 code, such as "en"
     or "de". For languages which do not have a two-letter identifier, or
     invalid language codes, ``None`` will be returned.
     """
-    code = iso_639_alpha3(code)
-    return ISO2_MAP.get(code)
+    alpha3 = iso_639_alpha3(code)
+    if alpha3 is None:
+        return None
+    return ISO2_MAP.get(alpha3)
 
 
-def list_to_alpha3(languages, synonyms=True):
+def list_to_alpha3(languages: Iterable[str], synonyms: bool = True) -> Set[str]:
     """Parse all the language codes in a given list into ISO 639 Part 2 codes
     and optionally expand them with synonyms (i.e. other names for the same
     language)."""

languagecodes/py.typed

@@ -1,6 +1,9 @@
 # This is a set of synonyms for pragmatic usage in NLP. It is based on
 # working with Tesseract 3.04, but should be applicable elsewhere.
 
+from typing import Iterable
+
+
 LANG_SYNONYMS = [
     ("srp", "hbs", "hrv", "bos"),
     ("sli", "alb"),
@@ -30,7 +33,7 @@
 ]
 
 
-def expand_synonyms(language):
+def expand_synonyms(language: str) -> Iterable[str]:
     """Expand a language code into a set of codes."""
     for synonyms in LANG_SYNONYMS:
         if language in synonyms:

languagecodes/synonyms.py

@@ -1,4 +1,7 @@
-def normalize_code(code):
+from typing import Optional
+
+
+def normalize_code(code: Optional[str]) -> Optional[str]:
     if code is None:
         return None
     code = str(code)

languagecodes/util.py

@@ -1,4 +1,4 @@
-from setuptools import setup, find_packages
+from setuptools import setup
 
 setup(
     name="languagecodes",
@@ -19,16 +19,18 @@
     author_email="[email protected]",
     url="http://github.com/alephdata/languagecodes",
     license="MIT",
-    packages=find_packages(exclude=["ez_setup", "examples", "test"]),
+    packages=["languagecodes"],
     namespace_packages=[],
     include_package_data=True,
+    package_data={"languagecodes": ["py.typed"]},
     zip_safe=False,
     test_suite="nose.collector",
     install_requires=[],
     extras_require={
         "dev": [
             "wheel>=0.29.0",
             "twine",
+            "mypy",
             "flake8>=2.6.0",
             "nose",
             "banal",