Update options

Author: johnsamuelwrites

src/main/TRANSLATION_README.md

@@ -90,6 +90,7 @@ python batch_translate.py glossary-sync --target-langs es pt         # Export gl
 **Step-by-Step:**
 ```bash
 python batch_translate.py sync --source-dir en/travel --target-langs es pt    # Extract changed files
+python batch_translate.py sync --target-langs it --force-incomplete          # Force extraction for incomplete languages
 python batch_translate.py validate-csv                               # Pre-import validation
 python batch_translate.py import                                     # Import validated CSVs
 python batch_translate.py generate --target-langs es --require-complete       # Generate HTML
@@ -130,10 +131,20 @@ GLOSSARY_DIR = "../photography"
 
 Pipeline Options:
 ```bash
---target-langs es pt    # Languages to process
---auto-import           # Skip "Import now?" prompt, auto-import if valid
---skip-generation       # Stop after import, don't generate HTML
---force                 # Force re-extract all files (ignore change detection)
+--target-langs es pt         # Languages to process
+--auto-import                # Skip "Import now?" prompt, auto-import if valid
+--skip-generation            # Stop after import, don't generate HTML
+--force                      # Force re-extract all files (ignore change detection)
+```
+
+Sync Options:
+```bash
+--target-langs es pt         # Languages to process
+--source-dir PATH            # Source directory (default: en/photography)
+--workers N                  # Parallel workers (default: 4)
+--force                      # Force re-extract all files (ignore change detection)
+--force-incomplete           # Force extraction for incomplete languages (below coverage threshold)
+--coverage-threshold PCT     # Coverage %% below which --force-incomplete triggers (default: 80.0)
 ```
 
 ## Path Mappings (Required for Generation)
@@ -210,7 +221,14 @@ Output: Detailed pipeline report with all metrics
 
 1. **Extract Changed Files**:
    ```bash
+   # Normal sync: Extract only changed files (uses SHA256 change detection)
+   python batch_translate.py sync --source-dir en/travel --target-langs es pt
+   
+   # Force sync: Extract all files regardless of changes
    python batch_translate.py sync --source-dir en/travel --target-langs es pt --force
+   
+   # Force incomplete: Extract all files for languages below coverage threshold (useful for new languages)
+   python batch_translate.py sync --target-langs it --force-incomplete --coverage-threshold 80.0
    ```
    Output: CSVs in `translations_pending/`
 
@@ -510,8 +528,31 @@ Glossaries include context awareness via suggestion system.
 
 **Pipeline stops at import prompt**: Type "yes" to import, "no" to review CSVs first
 
+**"No changes detected" error on new language**: Use `--force-incomplete` flag
+```bash
+python batch_translate.py sync --target-langs it --force-incomplete
+# This forces extraction even when source files haven't changed
+# Useful for new languages with low coverage
+```
+
 ## Advanced Workflows
 
+**New language onboarding (incomplete translation coverage):**
+```bash
+# When adding a new language with low coverage, use --force-incomplete
+# Example: Italian language only has 32% coverage, normal sync would skip it
+python batch_translate.py sync --target-langs it --force-incomplete
+# Output:
+# [WARNING] [SYNC] it: coverage 32.0% < 80.0% => forcing extraction
+# [INFO] Changed files: 0
+# [INFO] Files extracted: 150
+# [INFO] Translations: 3520
+
+# You can customize the coverage threshold:
+python batch_translate.py sync --target-langs it --force-incomplete --coverage-threshold 95.0
+# Forces extraction if coverage is below 95%
+```
+
 **Skip import, review only:**
 ```bash
 python batch_translate.py pipeline --skip-generation
@@ -530,6 +571,19 @@ python batch_translate.py pipeline --force
 # Extracts all files, even unchanged ones
 ```
 
+**Mixed: Force incomplete + manual edits:**
+```bash
+# 1. Register manual edits done outside the pipeline
+python batch_translate.py register-manual-edit it/index.html --lang it
+
+# 2. Force extraction for remaining untranslated strings
+python batch_translate.py sync --target-langs it --force-incomplete --coverage-threshold 95.0
+
+# 3. Continue with normal pipeline workflow
+python batch_translate.py serve --port 8000  # Review in dashboard
+python batch_translate.py pipeline --target-langs it
+```
+
 ## Performance Notes
 
 - **Sync Command**: ~20-30 files/second with 4 workers

src/main/batch_translate.py

@@ -689,11 +689,38 @@ def cmd_sync(args):
         print_info(f"Starting sync check (workers: {args.workers}, force: {args.force})")
         print_info(f"Source directory: {args.source_dir}")
 
+        # Compute per-language coverage if --force-incomplete is set
+        force_langs = set()
+        if getattr(args, 'force_incomplete', False):
+            threshold = getattr(args, 'coverage_threshold', 80.0)
+            from translate_manager import TranslationManager
+            from pathlib import Path
+
+            tm = TranslationManager(source_lang=args.source_lang,
+                                    target_langs=args.target_langs,
+                                    db_path=args.db_path)
+            source_files = [str(f) for f in Path(args.source_dir).rglob('*.html')]
+
+            for lang in args.target_langs:
+                total_all, found_all = 0, 0
+                for source_file in source_files:
+                    result_file = tm.process_file(source_file, lang, check_existing=False, force=True)
+                    total_all += result_file.get('total', 0)
+                    found_all += result_file.get('found', 0)
+
+                pct = (found_all / total_all * 100) if total_all > 0 else 0.0
+                if pct < threshold:
+                    print_warning(f"[SYNC] {lang}: coverage {pct:.1f}% < {threshold}% => forcing extraction")
+                    force_langs.add(lang)
+                else:
+                    print_info(f"[SYNC] {lang}: coverage {pct:.1f}% >= {threshold}% => hash-based sync")
+
         result = sync.check_and_extract(
             args.source_dir,
             args.output_dir,
             target_langs=args.target_langs,
-            force=args.force
+            force=args.force,
+            force_langs=force_langs
         )
 
         if result['status'] == 'no_files':
@@ -1014,6 +1041,10 @@ def main():
                             help='Force extract all files, not just changed')
     sync_parser.add_argument('--workers', type=int, default=4,
                             help='Number of parallel workers (default: 4)')
+    sync_parser.add_argument('--force-incomplete', action='store_true',
+                            help='Force full extraction for languages below coverage threshold')
+    sync_parser.add_argument('--coverage-threshold', type=float, default=80.0,
+                            help='Coverage %% below which --force-incomplete triggers (default: 80.0)')
 
     # validate-csv command
     validate_csv_parser = subparsers.add_parser('validate-csv', help='Validate CSV files for QA issues')

src/main/sync_manager.py

@@ -56,7 +56,8 @@ def check_and_extract(
         source_dir: str,
         output_dir: str,
         target_langs: List[str] = None,
-        force: bool = False
+        force: bool = False,
+        force_langs: set = None
     ) -> Dict[str, any]:
         """
         Check for changed source files and extract translations
@@ -74,19 +75,18 @@ def check_and_extract(
                 'extracted_batches': {}
             }
 
-        # Determine which files changed
+        # Determine which files changed (per-language or global)
+        force_langs = force_langs or set()
+
+        # If force=True globally, extract all files for all langs
+        # Otherwise check per-language via force_langs set
         if force:
             changed_files = all_files
         else:
+            # For now, detect globally (if any lang needs changes, extract)
+            # Per-language filtering happens below
             changed_files = self._detect_changed_files(all_files)
 
-        if not changed_files:
-            return {
-                'status': 'no_changes',
-                'changed_files': [],
-                'extracted_batches': {}
-            }
-
         # Update hashes for all files
         for file_path in all_files:
             hash_val = self._compute_file_hash(file_path)
@@ -108,7 +108,14 @@ def check_and_extract(
                     'csv_files': []
                 }
 
-                for file_path in changed_files:
+                # Per-language force check: use full extraction if lang is in force_langs
+                lang_force = force or (lang in force_langs)
+                if lang_force:
+                    lang_files = all_files
+                else:
+                    lang_files = changed_files
+
+                for file_path in lang_files:
                     future = executor.submit(
                         self._extract_and_prefill,
                         file_path,