Merge pull request 'Initial langage detection' (#16) from detect-post-language into main

Reviewed-on: #16

.woodpecker/lint.yml

@@ -2,7 +2,7 @@ when:
   branch: main
 steps:
   - name: lint
-    image: python:3-alpine
+    image: python:3-slim
     commands:
       - python -m pip install --upgrade pip
       - python -m pip install -r requirements.txt

config-sample.yml

@@ -3,3 +3,7 @@ mastodon:
   client_secret: 
   access_token: 
   api_base_url: 
+languages:
+  - en
+  - es
+default_language: en

kuow_fetcher.py

@@ -6,6 +6,7 @@ from typing import Optional
 import requests
 import yaml
 from bs4 import BeautifulSoup
+from lingua import IsoCode639_1, Language, LanguageDetectorBuilder
 from mastodon import Mastodon
 from sqlalchemy import create_engine, select
 from sqlalchemy.exc import NoResultFound
@@ -21,13 +22,27 @@ class KuowStory(Base):
     pageview_story_id: Mapped[str] = mapped_column(primary_key=True, unique=True)
     dfp_targeting_id: Mapped[str] = mapped_column()
     article_link: Mapped[str] = mapped_column()
+    article_language: Mapped[Optional[str]] = mapped_column()
     last_updated_time: Mapped[datetime] = mapped_column()
     post_id: Mapped[Optional[str]] = mapped_column()
 
     def __repr__(self) -> str:
-        return f"KuowStory(pageview_story_id={self.pageview_story_id!r}, dfp_targeting_id={self.dfp_targeting_id!r}, article_link={self.article_link!r}, last_updated_time={self.last_updated_time!r}, post_id={self.post_id!r})"
+        return f"KuowStory(pageview_story_id={self.pageview_story_id!r}, dfp_targeting_id={self.dfp_targeting_id!r}, article_link={self.article_link!r}, article_language={self.article_language!r}, last_updated_time={self.last_updated_time!r}, post_id={self.post_id!r})"
 
 
+def get_language_from_iso_code_639_1_str(iso_code_639_1_str: str) -> Language:
+    iso_code_369_1 = getattr(IsoCode639_1, iso_code_639_1_str.upper())
+    return Language.from_iso_code_639_1(iso_code_369_1)
+
+
+def detect_article_language(article_description: str) -> str:
+    detector = LanguageDetectorBuilder.from_languages(*languages).build()
+    try:
+        language = detector.detect_language_of(article_description)
+        return language.iso_code_639_1.name
+    except AttributeError:
+        return default_language.iso_code_639_1.name
+
 engine = create_engine("sqlite:///kuow.db")
 Base.metadata.create_all(engine)
 
@@ -44,6 +59,11 @@ mastodon = Mastodon(
     api_base_url=config["mastodon"]["api_base_url"],
 )
 
+languages = [
+    get_language_from_iso_code_639_1_str(language) for language in config["languages"]
+]
+default_language = get_language_from_iso_code_639_1_str(config["default_language"])
+
 kuow_response = requests.get(url)
 soup = BeautifulSoup(kuow_response.content, "html.parser")
 articles = soup.find_all("span", class_="txt")
@@ -131,6 +151,10 @@ with Session(engine) as session:
                 )
                 if not article_record.post_id:
                     print("Posting to Mastodon")
+
+                    article_language = detect_article_language(article_description)
+                    article_record.article_language = article_language
+
                     mastodon_post_result = mastodon.status_post(
                         status=article_description
                         + "\n"
@@ -138,6 +162,7 @@ with Session(engine) as session:
                         + article_link
                         + "\n#KUOW #News{}".format(additional_tag_string),
                         visibility="public",
+                        language=article_language,
                     )
                     article_record.post_id = mastodon_post_result["id"]
                 else:

requirements.txt

@@ -5,6 +5,7 @@ charset-normalizer==3.3.2
 decorator==5.1.1
 greenlet==3.0.3
 idna==3.6
+lingua-language-detector==2.0.2
 Mastodon.py==1.8.1
 python-dateutil==2.8.2
 python-magic==0.4.27