supervertaler 1.9.331__tar.gz → 1.9.332__tar.gz

{supervertaler-1.9.331 → supervertaler-1.9.332}/CHANGELOG.md

@@ -2,9 +2,27 @@
 
 All notable changes to Supervertaler are documented in this file.
 
-**Current Version:** v1.9.331 (February 26, 2026)
+**Current Version:** v1.9.332 (February 26, 2026)
 
 
+## v1.9.332 - February 26, 2026
+
+### Critical Bug Fixes
+
+- **Fix DOCX import silently dropping paragraphs** — The DOCX import filter used Python `id()` values to distinguish body paragraphs from table cell paragraphs. Because Python reuses memory addresses after garbage collection, body paragraphs could be falsely identified as table cells and silently skipped. The number of dropped paragraphs varied per run (non-deterministic), making this especially hard to notice. Fixed by building a stable element→Paragraph mapping once and removing the fragile `id()`-based table check entirely — body-level `<w:p>` elements can never be table paragraphs by definition.
+
+### Improvements
+
+- **Word count verification on DOCX import** — After importing a DOCX, Supervertaler now compares the raw XML word count against the imported segments. If the difference exceeds 5%, a warning dialog alerts the user that text may have been lost during import. This serves as a safety net against future import bugs.
+- **TMX export now uses standard `<bpt>`/`<ept>` pairs** — Exported TMX files now convert inline formatting tags (`<b>`, `<i>`, `<u>`, etc.) to proper TMX 1.4 `<bpt>`/`<ept>` paired elements instead of escaping them as literal text. This ensures compatibility with Trados Studio, memoQ, and other CAT tools.
+
+### Bug Fixes
+
+- **Fix `<b>`/`<i>` tags not hidden by "hide outer wrapping tags"** — When a formatting tag like `<b>` wrapped the entire segment, the "hide outer wrapping tags in grid" option did not strip it. Formatting tags are now included in the strippable set alongside structural tags.
+- **Fix false `<li-o>` tags on non-list paragraphs** — Paragraphs with `numId=0` in the OOXML (Word's explicit "no numbering" override) were incorrectly treated as list items. Also tightened the fallback list detection to require a proper digit+delimiter pattern (e.g. "1. " or "2) ") rather than just a leading digit.
+
+---
+
 ## v1.9.331 - February 26, 2026
 
 ### Improvements

{supervertaler-1.9.331 → supervertaler-1.9.332}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: supervertaler
-Version: 1.9.331
+Version: 1.9.332
 Summary: Professional AI-enhanced translation workbench with multi-LLM support, glossary system, TM, spellcheck, voice commands, and PyQt6 interface. Batteries included (core).
 Home-page: https://supervertaler.com
 Author: Michael Beijer
@@ -80,7 +80,7 @@ Dynamic: requires-python
 
 **Professional AI-enhanced translation workbench** with multi-LLM support (GPT-4, Claude, Gemini, Ollama), translation memory, glossary management, and seamless CAT tool integration (memoQ, Trados, CafeTran, Phrase, Déjà Vu).
 
-**Latest release:** v1.9.331 - Fix TM/glossary activation leaking between projects; AI prompts now use full language names.
+**Latest release:** v1.9.332 - Fix DOCX import silently dropping paragraphs; add word count verification safety net; TMX export now uses standard bpt/ept pairs.
 
 ---
 

{supervertaler-1.9.331 → supervertaler-1.9.332}/README.md

@@ -6,7 +6,7 @@
 
 **Professional AI-enhanced translation workbench** with multi-LLM support (GPT-4, Claude, Gemini, Ollama), translation memory, glossary management, and seamless CAT tool integration (memoQ, Trados, CafeTran, Phrase, Déjà Vu).
 
-**Latest release:** v1.9.331 - Fix TM/glossary activation leaking between projects; AI prompts now use full language names.
+**Latest release:** v1.9.332 - Fix DOCX import silently dropping paragraphs; add word count verification safety net; TMX export now uses standard bpt/ept pairs.
 
 ---
 

{supervertaler-1.9.331 → supervertaler-1.9.332}/Supervertaler.py

@@ -471,11 +471,12 @@ def strip_outer_wrapping_tags(text: str) -> tuple:
     """
     Strip outer wrapping tags from text if the entire segment is wrapped in a single tag pair.
 
-    This handles structural tags like <li-o>...</li-o>, <p>...</p>, <td>...</td>, etc.
-    Inner formatting tags like <b>...</b> are preserved.
+    This handles structural tags like <li-o>...</li-o>, <p>...</p>, <td>...</td>,
+    as well as formatting tags like <b>...</b> when they wrap the *entire* segment.
+    Inner formatting (e.g. "Some <b>bold</b> text") is never affected.
 
     Args:
-        text: Text that may be wrapped in outer structural tags
+        text: Text that may be wrapped in outer tags
 
     Returns:
         Tuple of (stripped_text, tag_name) where tag_name is the outer tag that was stripped,
@@ -485,7 +486,7 @@ def strip_outer_wrapping_tags(text: str) -> tuple:
         "<li-o>Hello <b>world</b></li-o>" -> ("Hello <b>world</b>", "li-o")
         "<p>Simple text</p>" -> ("Simple text", "p")
         "No tags here" -> ("No tags here", None)
-        "<b>Bold text</b>" -> ("<b>Bold text</b>", None)  # <b> is formatting, not structural
+        "<b>Bold text</b>" -> ("Bold text", "b")  # <b> wraps entire segment → stripped
     """
     import re
 
@@ -494,12 +495,19 @@ def strip_outer_wrapping_tags(text: str) -> tuple:
 
     text = text.strip()
 
-    # Structural tags that wrap entire segments (not inline formatting)
-    structural_tags = {
+    # Tags that wrap entire segments — structural AND formatting.
+    # Formatting tags (b, i, u, …) are included because when they wrap
+    # the *entire* segment they act as structural wrappers; the function
+    # already guarantees it only strips the single outermost pair, so
+    # inner formatting like "Some <b>bold</b> text" is never affected.
+    strippable_tags = {
+        # Structural / layout
         'li-o', 'li-b', 'li', 'p', 'td', 'th', 'tr', 'div', 'span',
         'h1', 'h2', 'h3', 'h4', 'h5', 'h6', 'title', 'caption',
         'blockquote', 'pre', 'code', 'dt', 'dd', 'header', 'footer',
-        'article', 'section', 'aside', 'nav', 'main', 'figure', 'figcaption'
+        'article', 'section', 'aside', 'nav', 'main', 'figure', 'figcaption',
+        # Formatting (only stripped when wrapping the entire segment)
+        'b', 'i', 'u', 'em', 'strong', 's', 'strike', 'sub', 'sup', 'mark',
     }
 
     # Pattern to match opening tag at start: <tag> or <tag attr="...">
@@ -512,8 +520,8 @@ def strip_outer_wrapping_tags(text: str) -> tuple:
     tag_name = opening_match.group(1).lower()
     rest = opening_match.group(2)
 
-    # Only strip structural tags, not inline formatting like <b>, <i>, <u>, <em>, <strong>
-    if tag_name not in structural_tags:
+    # Only strip recognised tags
+    if tag_name not in strippable_tags:
         return (text, None)
 
     # Check if text ends with matching closing tag
@@ -26773,7 +26781,27 @@ class SupervertalerQt(QMainWindow):
             
             paragraphs = self.docx_handler.import_docx(file_path)
             self.original_docx = file_path
-            
+
+            # ── Word-count safety check ──────────────────────────────
+            # Compare raw DOCX word count against imported paragraphs to
+            # detect any silently dropped content.
+            raw_wc = self.docx_handler.get_raw_word_count()
+            imported_wc = self.docx_handler.get_imported_word_count(paragraphs)
+            wc_diff = raw_wc - imported_wc
+            wc_pct = (abs(wc_diff) / raw_wc * 100) if raw_wc > 0 else 0
+            self.log(f"📊 Word count check: DOCX={raw_wc}, imported={imported_wc} (Δ {wc_diff:+d}, {wc_pct:.1f}%)")
+            if wc_pct > 5:   # >5 % difference = likely missing content
+                QMessageBox.warning(
+                    self,
+                    "Import Word Count Mismatch",
+                    f"⚠️ The imported text may be incomplete.\n\n"
+                    f"Word count in DOCX file: {raw_wc:,}\n"
+                    f"Word count after import: {imported_wc:,}\n"
+                    f"Difference: {abs(wc_diff):,} words ({wc_pct:.1f}%)\n\n"
+                    f"If the difference is large, some text may have been "
+                    f"lost during import. Please verify the imported segments."
+                )
+
             # Segment paragraphs
             self.log("Segmenting text...")
             

{supervertaler-1.9.331 → supervertaler-1.9.332}/docs/index.html

@@ -70,7 +70,7 @@
                     Supervertaler
                     <span class="gradient-text">The Ultimate Translation Workbench</span>
                 </h1>
-                <p class="version-badge">v1.9.331</p>
+                <p class="version-badge">v1.9.332</p>
                 <p class="hero-subtitle">
                     AI-enhanced translation workbench. Work with ChatGPT, Claude, and Gemini simultaneously.
                     Built by translators, for translators and writers.
@@ -707,7 +707,7 @@
             <div style="text-align: center; margin: 40px 0;">
                 <div class="download-card featured"
                     style="border: 3px solid #6366F1; max-width: 700px; margin: 0 auto;">
-                    <div class="download-badge" style="background: #6366F1;">v1.9.331</div>
+                    <div class="download-badge" style="background: #6366F1;">v1.9.332</div>
                     <h3>🎨 Supervertaler</h3>
                     <p class="download-description" style="font-weight: 600; color: #6366F1; margin-bottom: 10px;">
                         Modern Professional Interface
@@ -820,16 +820,16 @@
 
             <div class="screenshots-grid">
                 <div class="screenshot-card" style="border: 3px solid #10B981;">
-                    <a href="screenshots/Supervertaler_v1.9.331_Term_and_NT_insert_popup.png" class="screenshot-link" target="_blank">
-                        <img src="screenshots/Supervertaler_v1.9.331_Term_and_NT_insert_popup.png"
-                            alt="Supervertaler v1.9.331 Term and NT insert popup" loading="lazy">
+                    <a href="screenshots/Supervertaler_v1.9.332_Term_and_NT_insert_popup.png" class="screenshot-link" target="_blank">
+                        <img src="screenshots/Supervertaler_v1.9.332_Term_and_NT_insert_popup.png"
+                            alt="Supervertaler v1.9.332 Term and NT insert popup" loading="lazy">
                         <div class="screenshot-overlay">
                             <span class="zoom-icon">🔍 Click to enlarge</span>
                         </div>
                     </a>
                     <div class="screenshot-caption">
                         <h3>🔤 Term &amp; NT Insert Popup <span class="badge"
-                                style="background: #10B981; color: white; padding: 3px 8px; border-radius: 12px; font-size: 11px; margin-left: 5px;">v1.9.331</span>
+                                style="background: #10B981; color: white; padding: 3px 8px; border-radius: 12px; font-size: 11px; margin-left: 5px;">v1.9.332</span>
                         </h3>
                         <p>Press Ctrl once to open a memoQ-style floating popup listing all glossary matches and non-translatables for the current segment. Pick with a number key, ↑↓+Enter, or Esc to dismiss. <a href="https://www.youtube.com/watch?v=O-4q4uDF9q0" target="_blank" style="color: #10B981; font-weight: 600;">▶ Watch screencast</a></p>
                     </div>