From 076022d1d77424d99a01ea09cf7ad13c55fadc08 Mon Sep 17 00:00:00 2001
From: bram <bram@pescheck.io>
Date: Mon, 23 Dec 2024 16:29:35 +0100
Subject: [PATCH 1/2] Added data class

---
 .gitignore                     |  3 ++-
 python_gpt_po/po_translator.py | 23 +++++++++++++++--------
 2 files changed, 17 insertions(+), 9 deletions(-)

diff --git a/.gitignore b/.gitignore
index 05ab914..2c4675e 100644
--- a/.gitignore
+++ b/.gitignore
@@ -4,4 +4,5 @@ dist/
 *.egg-info/
 __pycache__
 build/
-db.sqlite3
\ No newline at end of file
+db.sqlite3
+.venv
\ No newline at end of file
diff --git a/python_gpt_po/po_translator.py b/python_gpt_po/po_translator.py
index bdaaaed..e090c96 100644
--- a/python_gpt_po/po_translator.py
+++ b/python_gpt_po/po_translator.py
@@ -7,6 +7,7 @@
 import json
 import logging
 import os
+from dataclasses import dataclass
 
 import polib
 from dotenv import load_dotenv
@@ -99,14 +100,14 @@ def update_po_entry(po_file, original_text, translated_text):
             logging.warning("Original text '%s' not found in the .po file.", original_text)
 
 
+@dataclass
 class TranslationConfig:
     """ Class to hold configuration parameters for the translation service. """
-    def __init__(self, client, model, bulk_mode=False, fuzzy=False, folder_language=False):  # pylint: disable=R0913
-        self.client = client
-        self.model = model
-        self.bulk_mode = bulk_mode
-        self.fuzzy = fuzzy
-        self.folder_language = folder_language
+    client: object
+    model: str
+    bulk_mode: bool = False
+    fuzzy: bool = False
+    folder_language: bool = False
 
 
 class TranslationService:
@@ -473,8 +474,14 @@ def main():
     else:
         detail_langs = [None] * len(lang_codes)  # If no detailed language is provided, default to None
 
-    # Create a configuration object
-    config = TranslationConfig(client, args.model, args.bulk, args.fuzzy, args.folder_language)
+    # And in main():
+    config = TranslationConfig(
+        client=client,
+        model=args.model,
+        bulk_mode=args.bulk,  # Changed bulk to bulk_mode
+        fuzzy=args.fuzzy,
+        folder_language=args.folder_language
+    )
 
     # Initialize the translation service with the configuration object
     translation_service = TranslationService(config, args.bulksize)

From 3f94342f8b75dad88e7ce9f8f64b109ea7dc4ef2 Mon Sep 17 00:00:00 2001
From: bram <bram@pescheck.io>
Date: Mon, 23 Dec 2024 16:45:18 +0100
Subject: [PATCH 2/2] Added language translation

---
 python_gpt_po/po_translator.py | 46 ++++++++++++++++++++++++++--------
 requirements.txt               |  5 ++--
 setup.py                       |  2 +-
 3 files changed, 40 insertions(+), 13 deletions(-)

diff --git a/python_gpt_po/po_translator.py b/python_gpt_po/po_translator.py
index e090c96..2d4e312 100644
--- a/python_gpt_po/po_translator.py
+++ b/python_gpt_po/po_translator.py
@@ -10,6 +10,7 @@
 from dataclasses import dataclass
 
 import polib
+import pycountry
 from dotenv import load_dotenv
 from openai import OpenAI
 from pkg_resources import DistributionNotFound, get_distribution
@@ -58,18 +59,43 @@ def disable_fuzzy_translations(po_file_path):
     @staticmethod
     def get_file_language(po_file_path, po_file, languages, folder_language):
         """Determines the language for a .po file."""
-        # Attempt to get language from the file metadata first
         file_lang = po_file.metadata.get('Language', '')
+        normalized_lang = POFileHandler.normalize_language_code(file_lang)
 
-        # If the file's language is not valid, infer it from the folder structure
-        if not file_lang or file_lang not in languages:
-            if folder_language:
-                inferred_lang = next((part for part in po_file_path.split(os.sep) if part in languages), None)
-                if inferred_lang:
-                    logging.info("Inferred language for .po file: %s as %s", po_file_path, inferred_lang)
-                    return inferred_lang
-            return None
-        return file_lang
+        if normalized_lang in languages:
+            return normalized_lang
+
+        if folder_language:
+            for part in po_file_path.split(os.sep):
+                norm_part = POFileHandler.normalize_language_code(part)
+                if norm_part in languages:
+                    logging.info("Inferred language for .po file: %s as %s", po_file_path, norm_part)
+                    return norm_part
+
+        return None
+
+    @staticmethod
+    def normalize_language_code(lang):
+        """Convert language name or code to ISO 639-1 code."""
+        # Try direct lookup
+        if len(lang) == 2:
+            try:
+                return pycountry.languages.get(alpha_2=lang.lower()).alpha_2
+            except AttributeError:
+                pass
+
+        # Try by name
+        try:
+            return pycountry.languages.get(name=lang.title()).alpha_2
+        except AttributeError:
+            pass
+
+        # Try by native name
+        for language in pycountry.languages:
+            if hasattr(language, 'inverted_name') and language.inverted_name.lower() == lang.lower():
+                return language.alpha_2
+
+        return None
 
     @staticmethod
     def log_translation_status(po_file_path, original_texts, translations):
diff --git a/requirements.txt b/requirements.txt
index 59408a2..8c9fa1c 100644
--- a/requirements.txt
+++ b/requirements.txt
@@ -1,6 +1,7 @@
 polib==1.2.0
-openai==v1.42.0
+openai==1.58.1
 python-dotenv==1.0.0
 pytest==8.2.2
 tenacity==9.0.0
-setuptools-scm==8.1.0
\ No newline at end of file
+setuptools-scm==8.1.0
+pycountry==24.6.1
\ No newline at end of file
diff --git a/setup.py b/setup.py
index 6956bb3..551997a 100644
--- a/setup.py
+++ b/setup.py
@@ -24,7 +24,7 @@
     include_package_data=True,
     install_requires=[
         'polib==1.2.0',
-        'openai==v1.42.0',
+        'openai==1.58.1',
         'python-dotenv==1.0.0',
         'tenacity==9.0.0',
     ],