Merge pull request #183 from openzim/i18_typing

Process / return a new typed Lang class in i18n methods

Author: benoit74

CHANGELOG.md

@@ -19,6 +19,7 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 - New `creator.Creator.convert_and_check_metadata` to convert metadata to bytes or str for known use cases and check proper type is passed to libzim
 - Add svg2png image conversion function #113
 - Add `conversion.convert_svg2png` image conversion function + support for SVG in `probing.format_for` #113
+- Add `i18n.Lang` class used as typed result of i18n operations #151
 
 ## Changed
 
@@ -31,6 +32,8 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 - **BREAKING** `creator.Creator.add_metadata` and `creator.Creator.validate_metadata` now only accepts `bytes | str` as value (it must have been converted before call)
 - **BREAKING** second argument of `creator.Creator.add_metadata` has been renamed to `value` instead of `content` to align with other methods
 - When a type issue arises in metadata checks, wrong value type is displayed in exception
+- **BREAKING** `i18n.get_language_details()`, `i18n.get_iso_lang_data()`, `i18n.find_language_names()` and `i18n.update_with_macro` now process / return a new typed `Lang` class #151
+- **BREAKING** Rename `i18.NotFound` to `i18n.NotFoundError`
 
 ### Fixed
 

src/zimscraperlib/i18n.py

@@ -65,13 +65,63 @@ def setlocale(root_dir: pathlib.Path, locale_name: str):
         ) from exc
 
 
-def get_iso_lang_data(lang: str) -> tuple[dict, dict | None]:
-    """ISO-639-x languages details for lang. Raises NotFound
+class Lang(dict):
 
-    Included keys: iso-639-1, iso-639-2b, iso-639-2t, iso-639-3, iso-639-5
-                   english, iso_types
+    @property
+    def iso_639_1(self) -> str | None:
+        """ISO-639-1 language code"""
+        return self["iso-639-1"]
 
-    See get_language_details() for details"""
+    @property
+    def iso_639_2b(self) -> str | None:
+        """ISO-639-2b language code"""
+        return self["iso-639-2b"]
+
+    @property
+    def iso_639_2t(self) -> str | None:
+        """ISO-639-2t language code"""
+        return self["iso-639-2t"]
+
+    @property
+    def iso_639_3(self) -> str | None:
+        """ISO-639-3 language code"""
+        return self["iso-639-3"]
+
+    @property
+    def iso_639_5(self) -> str | None:
+        """ISO-639-5 language code"""
+        return self["iso-639-5"]
+
+    @property
+    def english(self) -> str:
+        """language name in English"""
+        return self["english"]
+
+    @property
+    def native(self) -> str:
+        """language name in native language"""
+        return self["native"]
+
+    @property
+    def iso_types(self) -> list[str]:
+        """list of supported iso types"""
+        return self["iso_types"]
+
+    @property
+    def query(self) -> list[str]:
+        """Query issued for these language details"""
+        return self["query"]
+
+    @property
+    def querytype(self) -> list[str]:
+        """Type of query issued to retrieve language details"""
+        return self["querytype"]
+
+
+def get_iso_lang_data(lang: str) -> tuple[Lang, Lang | None]:
+    """ISO-639-x languages details for lang. Raises NotFoundError
+
+    Returns a tuple (main_language, macro_language | None)"""
 
     iso_types = []
 
@@ -105,9 +155,9 @@ def replace_types(new_type: str) -> str:
         if str(getattr(isolang, code_type)).lower() == lang.lower():
             iso_types.append(replace_types(code_type))
 
-    lang_data = {
-        f"iso-639-{lang_}": getattr(isolang, f"pt{lang_}") for lang_ in ISO_LEVELS
-    }
+    lang_data = Lang(
+        **{f"iso-639-{lang_}": getattr(isolang, f"pt{lang_}") for lang_ in ISO_LEVELS}
+    )
     lang_data.update({"english": isolang.name, "iso_types": iso_types})
 
     if isolang.macro():
@@ -118,53 +168,51 @@ def replace_types(new_type: str) -> str:
     return lang_data, None
 
 
-def find_language_names(
-    query: str, lang_data: dict | None = None
-) -> tuple[str | None, str | None]:
-    """(native, english) language names for lang with help from language_details dict
+def find_language_names(query: str, lang_data: Lang | None = None) -> tuple[str, str]:
+    """(native, english) language names for lang with help from lang_data
 
     Falls back to English name if available or query if not"""
     if lang_data is None:
-        lang_data = get_language_details(query, failsafe=True) or {}
+        lang_data = get_language_details(query, failsafe=True)
+        if not lang_data:
+            return query, query
+
     try:
         query_locale = babel.Locale.parse(query)
-        return query_locale.get_display_name(), query_locale.get_display_name("en")
+        if native_display_name := query_locale.get_display_name():
+            if english_display_name := query_locale.get_display_name("en"):
+                return native_display_name, english_display_name
     except (babel.UnknownLocaleError, TypeError, ValueError, AttributeError):
         pass
 
     # ISO code lookup order matters (most qualified first)!
     for iso_level in [f"iso-639-{lang_}" for lang_ in reversed(ISO_LEVELS)]:
         try:
             query_locale = babel.Locale.parse(lang_data.get(iso_level))
-            return query_locale.get_display_name(), query_locale.get_display_name("en")
+            if native_display_name := query_locale.get_display_name():
+                if english_display_name := query_locale.get_display_name("en"):
+                    return native_display_name, english_display_name
         except (babel.UnknownLocaleError, TypeError, ValueError, AttributeError):
             pass
-    default = lang_data.get("english", query)
+    default = lang_data.get("english") or query
     return default, default
 
 
-def update_with_macro(lang_data: dict, macro_data: dict | None):
+def update_with_macro(lang_data: Lang, macro_data: Lang | None):
     """update empty keys from lang_data with ones of macro_data"""
     if macro_data:
         for key, value in macro_data.items():
-            if key in lang_data and not lang_data[key]:
+            if key in lang_data and not lang_data.get(key):
                 lang_data[key] = value
     return lang_data
 
 
-def get_language_details(query: str, *, failsafe: bool | None = False) -> dict | None:
+def get_language_details(
+    query: str, failsafe: bool | None = False  # noqa: FBT002
+) -> Lang | None:
     """language details dict from query.
 
-    Raises NotFound or return `und` language details if failsafe
-
-    iso-639-1: str ISO-639-1 language code
-    iso-639-2b: str ISO-639-2b language code
-    iso-639-2t: str ISO-639-2t language code
-    iso-639-3: str ISO-639-3 language code
-    iso-639-5: str ISO-639-5 language code
-    english: str language name in English
-    native: str language name in is native language
-    iso_types: [str] list of supported iso types
+    When query fails, either raises NotFoundError or return None, based on failsafe
 
     """
 

tests/i18n/test_i18n.py

@@ -3,10 +3,12 @@
 
 import locale
 import pathlib
+from unittest.mock import Mock
 
 import pytest
 
 from zimscraperlib.i18n import (
+    Lang,
     NotFoundError,
     _,
     find_language_names,
@@ -190,7 +192,19 @@ def test_lang_details(query, expected):
         with pytest.raises(NotFoundError):
             get_language_details(query)
     else:
-        assert get_language_details(query) == expected
+        result = get_language_details(query)
+        assert result == expected
+        if result:
+            assert result.iso_639_1 == expected.get("iso-639-1")
+            assert result.iso_639_2b == expected.get("iso-639-2b")
+            assert result.iso_639_2t == expected.get("iso-639-2t")
+            assert result.iso_639_3 == expected.get("iso-639-3")
+            assert result.iso_639_5 == expected.get("iso-639-5")
+            assert result.english == expected.get("english")
+            assert result.native == expected.get("native")
+            assert result.iso_types == expected.get("iso_types")
+            assert result.query == expected.get("query")
+            assert result.querytype == expected.get("querytype")
 
 
 @pytest.mark.parametrize(
@@ -201,6 +215,7 @@ def test_lang_details(query, expected):
         ("bm", ("bamanakan", "Bambara")),
         ("zh", ("中文", "Chinese")),
         ("ar", ("العربية", "Arabic")),
+        ("qq", ("qq", "qq")),
     ],
 )
 def test_lang_name(query, expected):
@@ -214,3 +229,46 @@ def test_lang_name(query, expected):
 def test_translation(lang, expected):
     setlocale(pathlib.Path(__file__).parent, lang)
     assert _("Hello World!") == expected
+
+
+@pytest.mark.parametrize(
+    "dict_data",
+    [{}, {"iso-639-1": "ar"}],
+)
+def test_lang_equals(dict_data):
+    assert Lang(dict_data) == Lang(dict_data)
+    assert Lang(dict_data) == Lang({**dict_data})
+
+
+@pytest.mark.parametrize(
+    "dict_data_left, dict_data_right",
+    [
+        ({}, {"iso-639-1": "ar"}),
+        ({"iso-639-1": "ar"}, {"iso-639-1": "ab"}),
+        ({"iso-639-1": "ar"}, {"iso-639-2": "ar"}),
+    ],
+)
+def test_lang_not_equals(dict_data_left, dict_data_right):
+    assert Lang(dict_data_left) != Lang(dict_data_right)
+    assert Lang(dict_data_left) != "foo"
+
+
+@pytest.mark.parametrize(
+    "babel_native_return, babel_english_return, expected_native, expected_english",
+    [
+        ("Native value", "English value", "Native value", "English value"),
+        (None, "English value", "German", "German"),
+        ("Native value", None, "German", "German"),
+    ],
+)
+def test_find_language_names(
+    mocker, babel_native_return, babel_english_return, expected_native, expected_english
+):
+    mock_locale = Mock()
+    mock_locale.get_display_name.side_effect = lambda lang=None: (
+        babel_native_return if lang is None else babel_english_return
+    )
+
+    mocker.patch("babel.Locale.parse", return_value=mock_locale)
+
+    assert find_language_names("de") == (expected_native, expected_english)