fix css, add wiktionary debug

app.py

@@ -1410,21 +1410,19 @@ def _wiktionary_build_report_for_entry(entry_id: int, conn: sqlite3.Connection)
     return report
 
 def _wiktionary_find_all_entries(word: str, conn: sqlite3.Connection) -> List[Dict[str, Any]]:
-    """ Finds all entries related to an English word (Case-Insensitive Search). """
+    """ Finds entries with verbose debugging if lookup fails. """
     log(f"Wiktionary (EN): Querying for '{word}'...")
     found_entry_ids: Set[int] = set()
     
     lang_query = 'English'
     form_titles = ("Inflected form", "verb form", "noun form", "adjective form", "Comparative", "Superlative") 
 
-    # --- FIX: Search for Original, Lowercase, and Title Case ---
-    # This ensures we find "Ready" when input is "ready"
+    # Search variants: input, lowercase, title-case
     search_variants = list(set([word, word.lower(), word.title()]))
     placeholders = ', '.join('?' for _ in search_variants)
 
     # 1. Search Lemmatized Entries
     sql_lemma = f"SELECT id, pos_title, word FROM entries WHERE word IN ({placeholders}) AND lang = ?"
-    # flatten params: [var1, var2, ..., 'English']
     params_lemma = list(search_variants) + [lang_query]
     
     lemma_q = conn.execute(sql_lemma, params_lemma).fetchall()
@@ -1433,59 +1431,81 @@ def _wiktionary_find_all_entries(word: str, conn: sqlite3.Connection) -> List[Di
     
     for row in lemma_q:
         entry_id = row["id"]
-        pos_title = row["pos_title"] or "" # Safe string
+        pos_title = row["pos_title"] or ""
         found_entry_ids.add(entry_id)
         
-        # If it's a form entry (e.g. "running"), try to find the parent ("run")
+        # Check for parent lemma in "form_of" field
         if any(ft in pos_title for ft in form_titles):
-            log(f"Wiktionary: '{row['word']}' is an inflected entry (ID {entry_id}). Looking for parent...")
-            form_of_q = conn.execute(
-                "SELECT form_of FROM senses WHERE entry_id = ?", (entry_id,)
-            ).fetchall()
-            
+            form_of_q = conn.execute("SELECT form_of FROM senses WHERE entry_id = ?", (entry_id,)).fetchall()
             for form_row in form_of_q:
                 form_of_json = form_row["form_of"]
                 if not form_of_json: continue
                 try:
                     form_of_data = json.loads(form_of_json)
                     if isinstance(form_of_data, list) and form_of_data:
-                        parent_lemma_word = form_of_data[0].get("word")
-                        if parent_lemma_word:
-                            parent_lemmas_to_find.add(parent_lemma_word)
-                except json.JSONDecodeError:
-                    pass
-
-    # 2. Search Inflected Forms (Reverse lookup)
-    # We also apply the case variants here
+                        parent = form_of_data[0].get("word")
+                        if parent: parent_lemmas_to_find.add(parent)
+                except json.JSONDecodeError: pass
+
+    # 2. Search Inflected Forms
     sql_form = f"""
         SELECT DISTINCT e.id
         FROM forms f
         JOIN entries e ON f.entry_id = e.id
         WHERE f.form_text IN ({placeholders}) AND e.lang = ?
         AND f.id NOT IN (
-            SELECT ft.form_id
-            FROM form_tags ft
-            JOIN tags t ON ft.tag_id = t.id
+            SELECT ft.form_id FROM form_tags ft JOIN tags t ON ft.tag_id = t.id
             WHERE t.tag IN ('variant', 'auxiliary')
         )
     """
     params_form = list(search_variants) + [lang_query]
-    
     form_q = conn.execute(sql_form, params_form).fetchall()
     for row in form_q:
         found_entry_ids.add(row["id"])
         
-    # 3. Add Parent Lemmas (if any found in step 1)
+    # 3. Add Parent Lemmas
     if parent_lemmas_to_find:
-        log(f"Wiktionary: Found parent lemmas to add: {parent_lemmas_to_find}")
         for lemma_word in parent_lemmas_to_find:
-            # Recursively check exact match for parent
-            parent_id_q = conn.execute(
-                "SELECT id FROM entries WHERE word = ? AND lang = ?", (lemma_word, lang_query)
+            parent_id_q = conn.execute("SELECT id FROM entries WHERE word = ? AND lang = ?", (lemma_word, lang_query)).fetchall()
+            for row in parent_id_q: found_entry_ids.add(row["id"])
+    
+    # =========================================================
+    # 🔍 VERBOSE DEBUG DETECTIVE (Triggered on Failure)
+    # =========================================================
+    if not found_entry_ids:
+        log(f"⚠ [DEBUG-VERBOSE] Zero results for '{word}'. Running diagnostics...")
+        try:
+            # Check 1: Does it exist in ANY language?
+            any_lang = conn.execute(
+                f"SELECT lang, word FROM entries WHERE word IN ({placeholders}) LIMIT 5", 
+                list(search_variants)
             ).fetchall()
-            for row in parent_id_q:
-                found_entry_ids.add(row["id"])
-    
+            if any_lang:
+                found_langs = [f"{r['word']} ({r['lang']})" for r in any_lang]
+                log(f"   -> FOUND in other languages/cases: {found_langs}")
+                log(f"   -> CONCLUSION: Language filter '{lang_query}' might be too strict.")
+            else:
+                log(f"   -> NOT FOUND in 'entries' table (any language).")
+                
+                # Check 2: Does it exist as a form?
+                any_form = conn.execute(
+                    f"SELECT form_text FROM forms WHERE form_text IN ({placeholders}) LIMIT 1", 
+                    list(search_variants)
+                ).fetchone()
+                if any_form:
+                    log(f"   -> FOUND in 'forms' table as '{any_form['form_text']}'! (But failed to link to an English entry)")
+                else:
+                    log(f"   -> NOT FOUND in 'forms' table either.")
+
+                # Check 3: Is it there but with whitespace issues?
+                fuzzy = conn.execute("SELECT word FROM entries WHERE word LIKE ? LIMIT 1", (f"%{word}%",)).fetchone()
+                if fuzzy:
+                    log(f"   -> PARTIAL MATCH found: '{fuzzy['word']}'. (Check for whitespace/punctuation?)")
+                else:
+                    log(f"   -> COMPLETELY MISSING from DB.")
+        except Exception as e:
+            log(f"   -> Detective crashed: {e}")
+
     log(f"Wiktionary: Found {len(found_entry_ids)} unique matching entries.")
     
     all_reports = []
@@ -2257,36 +2277,96 @@ def analyze_word_encyclopedia(word: str, top_n_value: Optional[float] = 0, engin
 
 HTML_CSS = """
 <style>
-    .ling-card { font-family: 'Source Sans Pro', sans-serif; border: 1px solid #e5e7eb; border-radius: 8px; padding: 15px; margin-bottom: 15px; background: #fff; box-shadow: 0 1px 3px rgba(0,0,0,0.1); }
-    .ling-header { display: flex; align-items: baseline; margin-bottom: 10px; border-bottom: 2px solid #f3f4f6; padding-bottom: 5px; }
-    .ling-lemma { font-size: 1.5em; font-weight: bold; color: #1f2937; margin-right: 10px; }
-    .ling-pos { font-size: 0.9em; text-transform: uppercase; font-weight: bold; padding: 2px 6px; border-radius: 4px; color: #fff; }
-    .pos-noun { background-color: #3b82f6; } /* Blue */
-    .pos-verb { background-color: #10b981; } /* Green */
-    .pos-adj { background-color: #f59e0b; }  /* Amber */
-    .pos-adv { background-color: #8b5cf6; }  /* Purple */
-    
-    .ling-section { margin-top: 12px; }
-    .ling-subtitle { font-size: 0.95em; font-weight: bold; color: #6b7280; text-transform: uppercase; margin-bottom: 5px; letter-spacing: 0.05em; }
-    
-    .inflection-table { width: 100%; font-size: 0.9em; border-collapse: collapse; }
-    .inflection-table td { padding: 4px 8px; border-bottom: 1px solid #f3f4f6; }
-    .inflection-label { color: #6b7280; font-weight: 600; width: 40%; }
-    
-    .sense-item { margin-bottom: 6px; line-height: 1.4; font-size: 0.95em; }
-    .source-badge { display: inline-block; font-size: 0.7em; padding: 0 4px; border-radius: 3px; border: 1px solid #e5e7eb; margin-right: 5px; vertical-align: middle; }
-    .src-wikt { background: #fff1f2; color: #be123c; border-color: #fda4af; }
-    .src-oewn { background: #eff6ff; color: #1d4ed8; border-color: #93c5fd; }
-    
-    .rel-chip { display: inline-block; background: #f3f4f6; padding: 2px 8px; border-radius: 12px; font-size: 0.85em; margin: 2px; border: 1px solid #e5e7eb; }
-    .rel-type { color: #6b7280; font-size: 0.8em; margin-right: 3px; font-weight: bold;}
-    
-    .grammar-alert { padding: 10px; border-radius: 6px; margin-bottom: 20px; border-left: 4px solid; }
-    .alert-green { background: #f0fdf4; border-color: #22c55e; color: #166534; }
-    .alert-red { background: #fef2f2; border-color: #ef4444; color: #991b1b; }
-    
-    details > summary { cursor: pointer; padding: 10px; background: #f9fafb; border-radius: 6px; font-weight: 600; margin-bottom: 5px; }
-    details[open] > summary { background: #e5e7eb; }
+    /* Card Container - High Contrast */
+    .ling-card { 
+        font-family: 'Segoe UI', Roboto, Helvetica, Arial, sans-serif; 
+        border: 1px solid #d1d5db; /* Darker border */
+        border-radius: 8px; 
+        padding: 20px; 
+        margin-bottom: 20px; 
+        background: #ffffff; 
+        box-shadow: 0 4px 6px -1px rgba(0, 0, 0, 0.1), 0 2px 4px -1px rgba(0, 0, 0, 0.06); 
+    }
+
+    /* Header Section */
+    .ling-header { 
+        display: flex; 
+        align-items: center; 
+        margin-bottom: 15px; 
+        border-bottom: 2px solid #e5e7eb; 
+        padding-bottom: 10px; 
+    }
+    .ling-lemma { 
+        font-size: 1.8em; 
+        font-weight: 800; 
+        color: #111827; /* Almost Black */
+        margin-right: 12px; 
+    }
+    .ling-pos { 
+        font-size: 0.85em; 
+        text-transform: uppercase; 
+        font-weight: 700; 
+        padding: 4px 10px; 
+        border-radius: 6px; 
+        color: #fff; 
+        letter-spacing: 0.05em;
+    }
+    
+    /* POS Colors */
+    .pos-noun { background-color: #2563eb; }      /* Blue */
+    .pos-verb { background-color: #059669; }      /* Green */
+    .pos-adj { background-color: #d97706; }       /* Amber */
+    .pos-adv { background-color: #7c3aed; }       /* Purple */
+    .pos-name { background-color: #db2777; }      /* Pink (Proper Noun) */
+    .pos-other { background-color: #4b5563; }     /* Gray */
+
+    /* Section Headers */
+    .ling-section { margin-top: 15px; }
+    .ling-subtitle { 
+        font-size: 0.85em; 
+        font-weight: 700; 
+        color: #374151; /* Dark Gray */
+        text-transform: uppercase; 
+        margin-bottom: 8px; 
+        border-left: 4px solid #3b82f6;
+        padding-left: 8px;
+    }
+    
+    /* Tables */
+    .inflection-table { width: 100%; font-size: 0.95em; border-collapse: collapse; margin-bottom: 10px; }
+    .inflection-table td { padding: 6px 10px; border-bottom: 1px solid #e5e7eb; color: #1f2937; }
+    .inflection-label { color: #6b7280; font-weight: 600; width: 35%; background: #f9fafb; }
+    
+    /* Senses */
+    .sense-item { margin-bottom: 8px; line-height: 1.5; font-size: 1em; color: #1f2937; }
+    .source-badge { 
+        display: inline-block; font-size: 0.75em; font-weight: bold; padding: 2px 6px; 
+        border-radius: 4px; border: 1px solid; margin-right: 8px; vertical-align: middle; 
+    }
+    .src-wikt { background: #fff1f2; color: #9f1239; border-color: #fda4af; }
+    .src-oewn { background: #eff6ff; color: #1e40af; border-color: #93c5fd; }
+    
+    /* Relations Chips */
+    .rel-chip { 
+        display: inline-block; 
+        background: #f3f4f6; 
+        color: #1f2937; /* Force Dark Text */
+        padding: 4px 10px; 
+        border-radius: 15px; 
+        font-size: 0.9em; 
+        margin: 3px; 
+        border: 1px solid #d1d5db; 
+        font-weight: 500;
+    }
+    .rel-type { color: #6b7280; font-size: 0.8em; margin-right: 4px; font-weight: 700; text-transform: lowercase;}
+    
+    /* Collapsible */
+    .kg-details > summary { 
+        cursor: pointer; color: #2563eb; font-size: 0.9em; font-weight: 600; 
+        margin-top: 10px; padding: 6px; border-radius: 4px; width: fit-content;
+    }
+    .kg-details > summary:hover { text-decoration: underline; background: #eff6ff; }
+    .kg-content { margin-top: 10px; padding: 10px; background: #f9fafb; border-radius: 8px; border: 1px solid #e5e7eb; }
 </style>
 """
 
@@ -2298,15 +2378,29 @@ def _format_word_analysis_html(data: Dict[str, Any]) -> str:
     html = HTML_CSS
     analysis = data["analysis"]
     
-    # Iterate over POS (noun, verb, etc.)
+    # Iterate over POS
     for pos_key, entries in analysis.items():
         if not entries: continue
         entry = entries[0] # Take best candidate
         
-        # --- Header ---
-        pos_class = f"pos-{pos_key}" if pos_key in ["noun", "verb", "adj", "adv"] else "pos-noun"
-        
-        # Safe extraction of lemma
+        # --- POS Display Logic ---
+        # Map internal keys to nice display names and CSS classes
+        display_pos = pos_key.upper()
+        css_class = "pos-other"
+        
+        if pos_key == 'noun': css_class = "pos-noun"
+        elif pos_key == 'verb': css_class = "pos-verb"
+        elif pos_key == 'adj' or pos_key == 'adjective': 
+            css_class = "pos-adj"
+            display_pos = "ADJECTIVE"
+        elif pos_key == 'adv' or pos_key == 'adverb': 
+            css_class = "pos-adv"
+            display_pos = "ADVERB"
+        elif pos_key == 'name': 
+            css_class = "pos-name"
+            display_pos = "PROPER NOUN"
+        
+        # Data Extraction
         inf_wikt = entry.get("inflections_wiktionary") or {}
         inf_pat = entry.get("inflections_pattern") or {}
         sem_comb = entry.get("semantics_combined") or {}
@@ -2316,11 +2410,12 @@ def _format_word_analysis_html(data: Dict[str, Any]) -> str:
                 sem_comb.get("lemma") or \
                 data.get("input_word") or "?"
 
+        # --- CARD START ---
         html += f"""
         <div class="ling-card">
             <div class="ling-header">
                 <span class="ling-lemma">{lemma}</span>
-                <span class="ling-pos {pos_class}">{pos_key}</span>
+                <span class="ling-pos {css_class}">{display_pos}</span>
             </div>
         """
         
@@ -2328,57 +2423,65 @@ def _format_word_analysis_html(data: Dict[str, Any]) -> str:
         html += "<div class='ling-section'><div class='ling-subtitle'>Morphology & Inflections</div>"
         html += "<table class='inflection-table'>"
         
-        # Noun Logic
+        # We check Pattern data first. If it's empty, we show '-' or rely on Wiktionary forms.
+        has_pattern_data = bool(inf_pat) and "error" not in inf_pat
+        
         if pos_key == 'noun':
-            if inf_pat:
-                html += f"<tr><td class='inflection-label'>Singular</td><td>{inf_pat.get('singular', '-')}</td></tr>"
-                html += f"<tr><td class='inflection-label'>Plural</td><td>{inf_pat.get('plural', '-')}</td></tr>"
+            html += f"<tr><td class='inflection-label'>Singular</td><td>{inf_pat.get('singular', lemma if not has_pattern_data else '-')}</td></tr>"
+            html += f"<tr><td class='inflection-label'>Plural</td><td>{inf_pat.get('plural', '-')}</td></tr>"
+            if has_pattern_data:
                 html += f"<tr><td class='inflection-label'>Context</td><td>{inf_pat.get('article', '-')}</td></tr>"
         
-        # Verb Logic
         elif pos_key == 'verb':
             cj = inf_pat.get('conjugation') or {}
             pres = cj.get('Present') or {}
             past = cj.get('Past') or {}
             parts = inf_pat.get('participles') or {}
-            
             html += f"<tr><td class='inflection-label'>Infinitive</td><td>{inf_pat.get('infinitive', lemma)}</td></tr>"
             html += f"<tr><td class='inflection-label'>3rd Person (He/She)</td><td>{pres.get('he/she (3sg)', '-')}</td></tr>"
             html += f"<tr><td class='inflection-label'>Past Simple</td><td>{past.get('General', '-')}</td></tr>"
             html += f"<tr><td class='inflection-label'>Participle (Ing)</td><td>{parts.get('Present Participle (gerund)', '-')}</td></tr>"
             html += f"<tr><td class='inflection-label'>Participle (Past)</td><td>{parts.get('Past Participle', '-')}</td></tr>"
 
-        # Adjective Logic
-        elif pos_key == 'adjective':
+        elif pos_key in ['adjective', 'adj']:
             gr = inf_pat.get('grading') or {}
-            html += f"<tr><td class='inflection-label'>Positive</td><td>{gr.get('Positive', '-')}</td></tr>"
+            html += f"<tr><td class='inflection-label'>Positive</td><td>{gr.get('Positive', lemma)}</td></tr>"
             html += f"<tr><td class='inflection-label'>Comparative</td><td>{gr.get('Comparative', '-')}</td></tr>"
             html += f"<tr><td class='inflection-label'>Superlative</td><td>{gr.get('Superlative', '-')}</td></tr>"
 
-        # Wiktionary Forms Fallback
+        # Wiktionary Forms (The "Other Forms" box)
         forms_list = inf_wikt.get("forms_list") or []
         if forms_list:
-            # Safe extraction of form_text
-            forms_str_list = [f.get('form_text', '') for f in forms_list[:5] if f.get('form_text')]
+            # Extract text carefully
+            forms_str_list = []
+            for f in forms_list[:8]: # Show up to 8 forms
+                txt = f.get('form_text')
+                tags = f.get('tags')
+                if txt:
+                    # Append tag if available e.g., "Readys (plural)"
+                    display_txt = f"{txt} <small style='color:#6b7280'>({tags})</small>" if tags else txt
+                    forms_str_list.append(display_txt)
+            
             if forms_str_list:
-                html += f"<tr><td class='inflection-label'>Other Forms (Wikt)</td><td>{', '.join(forms_str_list)}...</td></tr>"
-
+                html += f"<tr><td class='inflection-label'>Forms (DB)</td><td>{', '.join(forms_str_list)}</td></tr>"
+        
         html += "</table></div>"
         
         # --- Semantics Section ---
         html += "<div class='ling-section'><div class='ling-subtitle'>Definitions & Senses</div>"
         
-        # Wiktionary Senses
         wikt_senses = sem_comb.get("wiktionary_senses") or []
+        oewn_senses = sem_comb.get("odenet_senses") or []
+        
+        if not wikt_senses and not oewn_senses:
+             html += "<div class='sense-item'><i>No definitions found.</i></div>"
+
         for s in wikt_senses[:3]: 
-            # FIX: Ensure definition is a string before replacing
             gloss_raw = s.get("definition") or "" 
-            gloss = gloss_raw.replace(";", "<br>")
+            gloss = str(gloss_raw).replace(";", "<br>") # Ensure string
             if gloss:
                 html += f"<div class='sense-item'><span class='source-badge src-wikt'>Wikt</span> {gloss}</div>"
             
-        # OEWN Senses
-        oewn_senses = sem_comb.get("odenet_senses") or []
         for s in oewn_senses[:3]:
             defi = s.get("definition") or ""
             if defi:
@@ -2386,17 +2489,43 @@ def _format_word_analysis_html(data: Dict[str, Any]) -> str:
             
         html += "</div>"
         
-        # --- Relations Section (ConceptNet) ---
+        # --- Relations Section ---
         rels = sem_comb.get("conceptnet_relations") or []
         if rels:
-            html += "<div class='ling-section'><div class='ling-subtitle'>Knowledge Graph (Top 5)</div>"
+            html += "<div class='ling-section'><div class='ling-subtitle'>Knowledge Graph</div>"
+            
+            top_n = 5
+            visible_rels = rels[:top_n]
+            hidden_rels = rels[top_n:]
+            
+            def render_rel(r):
+                # Robust extraction
+                rel_name = r.get("relation", "Rel")
+                # Prefer other_node, fall back to parsing surface, fall back to '?'
+                target = r.get("other_node") or "?"
+                # Clean up surface text if needed
+                if target == "?" and "surface" in r:
+                     parts = str(r["surface"]).split()
+                     if len(parts) > 2: target = parts[-1] 
+                
+                return f"<span class='rel-chip'><span class='rel-type'>{rel_name}:</span> {target}</span>"
+
             html += "<div>"
-            for r in rels[:5]:
-                rel_name = r.get("relation", "Related")
-                # Safe extraction of target
-                target = r.get("other_node") or r.get("surface") or "?"
-                html += f"<span class='rel-chip'><span class='rel-type'>{rel_name}:</span> {target}</span>"
-            html += "</div></div>"
+            for r in visible_rels:
+                html += render_rel(r)
+            html += "</div>"
+
+            if hidden_rels:
+                html += f"""
+                <details class='kg-details'>
+                    <summary>Show {len(hidden_rels)} more relations</summary>
+                    <div class='kg-content'>
+                """
+                for r in hidden_rels:
+                    html += render_rel(r)
+                html += "</div></details>"
+            
+            html += "</div>"
             
         html += "</div>" # End Card