Update workflow

Author: johnsamuelwrites

src/main/README.md

@@ -14,6 +14,8 @@ python src/main/main.py <command> [args...]
 Examples:
 
 ```bash
+python src/main/main.py check
+python src/main/main.py build
 python src/main/main.py blog
 python src/main/main.py slides en/slides/example.html
 python src/main/main.py convert-thumbnails input.html output.html
@@ -24,7 +26,9 @@ python src/main/main.py verify-usage . --exclude-dirs templates analysis
 
 - `batch-translate`: translation workflow helper.
 - `blog`: generates blog indexes and feeds.
+- `build`: runs the default static-site build workflow.
 - `build-all-search-indexes`: builds search indexes for all languages.
+- `check`: runs the default static-site validation workflow.
 - `check-db`: prints translation path mappings stored in `translations.db`.
 - `ci-internal-links`: CI-focused internal link verification.
 - `convert-thumbnails`: rewrites Wikimedia thumbnail URLs in HTML.
@@ -43,10 +47,15 @@ python src/main/main.py verify-usage . --exclude-dirs templates analysis
 
 - `main.py`: single dispatcher for executable tools in this folder.
 - `file_rewrite.py`: shared helper for safe text-file rewrites.
+- `config.py` and `paths.py`: shared configuration and repository path helpers.
+- `manifest.py`: tracks source fingerprints so generators can skip unchanged outputs.
+- `translation_rules.py`: shared translation parsing rules used by extraction and generation.
 - Library-style modules such as `analyse.py`, `links.py`, and `translate_manager.py` provide reusable functionality used by the command scripts.
 
 ## Development Notes
 
 - Prefer adding new command-line tools through `main.py` so `src/main` keeps a single stable entry point.
 - Avoid top-level argument parsing in reusable modules; expose `main(argv=None)` instead.
 - Keep file rewrite logic centralized when possible to reduce duplication and accidental partial writes.
+- Prefer offline-first behavior for static-site checks and make network access explicit when needed.
+- Prefer incremental generators that can skip work when inputs and outputs have not changed.

src/main/blog.py

@@ -8,6 +8,7 @@
 # Generate an upto date list of articles in reverse chronological order
 #!/usr/bin/env python3
 
+import argparse
 import sys
 from typing import Dict, List, Set, Tuple
 from dataclasses import dataclass
@@ -20,7 +21,11 @@
 from pytz import timezone
 
 from analyse import HTMLTextAnalysis, WebsiteAnalysis
+from config import SITE_AUTHOR, SITE_URL
 from git import get_first_latest_modification
+from manifest import BuildManifest
+from paths import REPO_ROOT
+from site_text import strip_author_from_title
 
 
 @dataclass
@@ -66,6 +71,7 @@ class BlogGenerator:
     @staticmethod
     def clean_title(title: str) -> str:
         """Remove author name variations from title."""
+        return strip_author_from_title(title)
         replacements = [
             ": John Samuel",
             ": ജോൺ ശമൂവേൽ",
@@ -463,12 +469,12 @@ def generate_feed(df: pd.DataFrame, feed_count: int = 20) -> None:
 
         articleset: Set[str] = set()
         fg = FeedGenerator()
-        fg.id("https://johnsamuel.info")
-        fg.title("John Samuel")
-        fg.description("Personal Blog of John Samuel")
-        fg.author({"name": "John Samuel"})
+        fg.id(SITE_URL)
+        fg.title(SITE_AUTHOR)
+        fg.description(f"Personal Blog of {SITE_AUTHOR}")
+        fg.author({"name": SITE_AUTHOR})
         fg.language("en")
-        fg.link(href="https://johnsamuel.info")
+        fg.link(href=SITE_URL)
 
         for _, row in df.iterrows():
             if row["filepath"] in articleset:
@@ -479,15 +485,15 @@ def generate_feed(df: pd.DataFrame, feed_count: int = 20) -> None:
                 metadata = BlogGenerator.extract_article_metadata(row["filepath"])
 
                 fe = fg.add_entry(order="append")
-                fe.id(f"https://johnsamuel.info/{row['filepath']}")
+                fe.id(f"{SITE_URL}/{row['filepath']}")
                 fe.title(metadata.title)
                 fe.pubDate(
                     datetime.fromtimestamp(
                         metadata.modification_time, tz=timezone("Europe/Amsterdam")
                     )
                 )
                 fe.description(metadata.title)
-                fe.link(href=f"https://johnsamuel.info/{row['filepath']}")
+                fe.link(href=f"{SITE_URL}/{row['filepath']}")
             except Exception as e:
                 print(f"Error adding to feed {row['filepath']}: {e}")
                 continue
@@ -499,21 +505,43 @@ def generate_feed(df: pd.DataFrame, feed_count: int = 20) -> None:
 
 def main(argv=None):
     """Main entry point."""
-    argv = list(sys.argv[1:] if argv is None else argv)
-    if argv:
-        print("This program takes no input")
-        return 1
+    parser = argparse.ArgumentParser(
+        description="Generate blog pages and feeds."
+    )
+    parser.add_argument(
+        "--force",
+        action="store_true",
+        help="Regenerate outputs even when the build manifest says they are current.",
+    )
+    args = parser.parse_args(sys.argv[1:] if argv is None else argv)
 
-    # Get articles dataframe
     df = WebsiteAnalysis.get_articles_list_dataframe()
+    article_sources = [REPO_ROOT / str(filepath) for filepath in df["filepath"].tolist()]
+    template_sources = list((REPO_ROOT / "templates").glob("blog.html"))
+    template_sources.extend((REPO_ROOT / "templates" / "blog").glob("*.html"))
+    manifest = BuildManifest()
+    output_paths = [
+        REPO_ROOT / "blog" / "index.html",
+        REPO_ROOT / "atom.xml",
+        REPO_ROOT / "rss.xml",
+        REPO_ROOT / "en" / "blog.html",
+        REPO_ROOT / "fr" / "blog.html",
+        REPO_ROOT / "ml" / "ബ്ലോഗ്.html",
+        REPO_ROOT / "pa" / "ਬਲਾਗ.html",
+        REPO_ROOT / "hi" / "ब्लॉग.html",
+    ]
+    sources = [Path(__file__), *template_sources, *article_sources]
+
+    if not args.force and manifest.is_current("blog", sources, output_paths):
+        print("[SKIP] Blog outputs are up to date.")
+        return 0
 
-    # Generate complete article list with year organization
     BlogGenerator.generate_complete_list(df)
-
-    # Generate feeds (latest N articles)
     BlogGenerator.generate_feed(df, feed_count=20)
+    manifest.update("blog", sources, output_paths)
 
     print("✓ Blog generation complete!")
+    return 0
 
 
 if __name__ == "__main__":