Update src/summarization/note_generator.py

src/summarization/note_generator.py

@@ -1,27 +1,5 @@
-"""
-AIdea - Premium Note Generator v3.0
-=====================================
-Drop-in upgrade for src/summarization/note_generator.py
-
-Preserved from original:
-  - NoteGenerator class with same __init__ signature (reads GROQ_API_KEY from env)
-  - generateSummary(transcript_text, video_title) -> Dict
-  - extractActionItems(transcript_text, video_title) -> Dict
-  - format_notes_to_markdown(json_notes) -> str
-  - format_final_notes(notes, video_title, video_url, duration) -> str
-  - _get_error_json(error_msg) -> Dict
-  - setup_logger / settings imports preserved
-
-Upgraded:
-  - New premium Pydantic schemas (SummarySchema, ActionItemsSchema updated in-place)
-  - Auto language detection — all output strictly in transcript language
-  - Core Topics / Batches replacing single flat summary
-  - Rich Markdown: emojis, blockquotes, bold/italic, proper spacing
-  - Flutter flutter_markdown optimized structure
-  - Separated system/user prompt for better LLM instruction-following
-"""
-
 import json
+import logging
 import os
 from typing import Dict, List, Optional
 
@@ -29,11 +7,11 @@ from groq import Groq
 from pydantic import BaseModel, Field, ValidationError
 
 from src.utils.logger import setup_logger
-from src.utils.config import settings
+# تأكد إن مسار الـ settings صح حسب مشروعك
+from src.utils.config import settings 
 
 logger = setup_logger(__name__)
 
-
 # ─────────────────────────────────────────────────────────────────────────────
 # PYDANTIC SCHEMAS
 # ─────────────────────────────────────────────────────────────────────────────
@@ -43,75 +21,38 @@ class KeyConceptSchema(BaseModel):
     definition: str = Field(description="Clear, detailed explanation (2-4 sentences)")
     importance: str = Field(description="Why this concept matters in context (1-2 sentences)")
 
-
 class CoreTopicSchema(BaseModel):
-    """One 'batch' / major topic section of the video."""
     title: str = Field(description="Short descriptive title for this topic batch")
     overview: str = Field(description="Rich paragraph overview of this topic (3-6 sentences, NOT bullet list)")
     key_points: List[str] = Field(description="3-6 detailed bullet points under this topic")
-    insight: Optional[str] = Field(
-        default=None,
-        description="A notable insight, tip, or important warning for this topic"
-    )
-
+    insight: Optional[str] = Field(default=None, description="A notable insight, tip, or important warning for this topic")
 
 class TimestampSchema(BaseModel):
     timestamp: str = Field(description="Timestamp in MM:SS format")
     topic: str = Field(description="Short title for this segment")
     summary: str = Field(description="1-2 sentence description of what happens at this point")
 
-
 class ActionItemEntrySchema(BaseModel):
     action: str = Field(description="A specific, actionable takeaway")
     rationale: str = Field(description="Why the viewer should do this (1 sentence)")
 
-
 class SummarySchema(BaseModel):
-    """
-    Main schema returned by generateSummary().
-    Backward-compatible field names kept (title, summary, key_concepts,
-    timestamps, keywords) plus new premium fields.
-    """
-    # ── Existing fields (kept for backward compat) ───────────────────────────
     title: str = Field(description="Inferred video title in the transcript language")
-    summary: str = Field(
-        description=(
-            "Rich executive summary in the transcript language. "
-            "4-6 sentences covering the main thesis, context, and value."
-        )
-    )
-    key_concepts: List[KeyConceptSchema] = Field(
-        description="3-8 key concepts/terms with definition AND importance"
-    )
-    timestamps: List[TimestampSchema] = Field(
-        description="5-10 timeline entries covering the video's major segments"
-    )
+    detected_language: str = Field(description="Detected language of the transcript, stated in English (e.g. 'Arabic', 'English')")
+    hook: str = Field(description="A compelling 2-3 sentence teaser in the transcript language")
+    summary: str = Field(description="Rich executive summary. 4-6 sentences covering the main thesis, context, and value.")
+    core_topics: List[CoreTopicSchema] = Field(description="2-5 major topic batches that organize the video content")
+    key_concepts: List[KeyConceptSchema] = Field(description="3-8 key concepts/terms with definition AND importance")
+    timestamps: List[TimestampSchema] = Field(description="5-10 timeline entries covering the video's major segments")
     keywords: List[str] = Field(description="5-10 relevant topic tags")
-
-    # ── New premium fields ───────────────────────────────────────────────────
-    detected_language: str = Field(
-        description="Detected language of the transcript, stated in English (e.g. 'Arabic', 'English')"
-    )
-    hook: str = Field(
-        description="A compelling 2-3 sentence teaser in the transcript language"
-    )
-    core_topics: List[CoreTopicSchema] = Field(
-        description="2-5 major topic batches that organize the video content"
-    )
-    closing_thought: str = Field(
-        description="A motivating or thought-provoking closing sentence in the transcript language"
-    )
-
+    closing_thought: str = Field(description="A motivating or thought-provoking closing sentence in the transcript language")
 
 class ActionItemsSchema(BaseModel):
-    """Schema returned by extractActionItems()."""
-    action_items: List[ActionItemEntrySchema] = Field(
-        description="3-6 specific actionable takeaways, each with action + rationale"
-    )
+    action_items: List[ActionItemEntrySchema] = Field(description="3-6 specific actionable takeaways, each with action + rationale")
 
 
 # ─────────────────────────────────────────────────────────────────────────────
-# PROMPT TEMPLATES
+# PROMPT TEMPLATES (STRICT JSON ENFORCEMENT)
 # ─────────────────────────────────────────────────────────────────────────────
 
 _SUMMARY_SYSTEM = """
@@ -120,24 +61,44 @@ Transform raw video transcripts into structured, deeply insightful JSON notes.
 
 LANGUAGE RULE — CRITICAL, NEVER VIOLATE:
 - Detect the primary language of the transcript.
-- Every single field in your JSON — titles, overviews, bullet points, insights,
-  closing thoughts — MUST be written entirely in that SAME detected language.
-- Do NOT mix languages. Arabic transcript → everything in Arabic. English → everything in English.
+- Every single field MUST be written entirely in that SAME detected language.
+- Do NOT mix languages. Arabic transcript -> everything in Arabic.
 - Only the `detected_language` field itself is stated in English (e.g. "Arabic").
 
-CONTENT RULES:
-1. `summary`: Write a rich flowing paragraph (4-6 sentences). Synthesize the thesis — do NOT list.
-2. `core_topics`: Divide into 2-5 logical batches. Each needs:
-   - A descriptive `title`
-   - A rich `overview` paragraph (3-6 sentences, NOT bullets)
-   - 3-6 detailed `key_points` bullet strings
-   - Optional `insight`: a notable tip or important warning
-3. `key_concepts`: For each term include `definition` AND `importance`.
-4. `timestamps`: 5-10 meaningful entries with descriptive `topic` and `summary`.
-5. Be SPECIFIC — reference actual names, numbers, examples from the transcript.
-6. Be INSIGHTFUL — add synthesis, not just repetition.
-
-OUTPUT: Return ONLY a valid JSON object. No markdown fences, no preamble, no extra text.
+CRITICAL: YOU MUST RETURN A JSON OBJECT EXACTLY MATCHING THIS STRUCTURE. 
+DO NOT CHANGE, OMIT, OR RENAME ANY KEYS (e.g., never use "time" instead of "timestamp").
+{
+    "title": "Inferred video title",
+    "detected_language": "English (or Arabic, etc.)",
+    "hook": "A compelling 2-3 sentence teaser",
+    "summary": "Rich flowing paragraph (4-6 sentences)",
+    "core_topics": [
+        {
+            "title": "Topic title",
+            "overview": "Rich paragraph overview (NOT bullet points)",
+            "key_points": ["point 1", "point 2"],
+            "insight": "Notable tip or warning"
+        }
+    ],
+    "key_concepts": [
+        {
+            "term": "Concept name",
+            "definition": "Clear explanation",
+            "importance": "Why it matters"
+        }
+    ],
+    "timestamps": [
+        {
+            "timestamp": "MM:SS",
+            "topic": "Segment title",
+            "summary": "What happens here"
+        }
+    ],
+    "keywords": ["tag1", "tag2"],
+    "closing_thought": "Motivating closing sentence"
+}
+
+OUTPUT: Return ONLY a valid JSON object. No markdown fences, no extra text.
 """.strip()
 
 _SUMMARY_USER = """
@@ -146,36 +107,36 @@ Video Title: {video_title}
 TRANSCRIPT:
 {transcript}
 
-Analyze thoroughly. Detect the language. Return the structured JSON notes — all content in the transcript's language.
+Analyze thoroughly. Detect the language. Return ONLY the exact JSON structure requested.
 """.strip()
 
 _ACTIONS_SYSTEM = """
 You are an expert at extracting actionable insights from educational content.
 
 LANGUAGE RULE — CRITICAL:
-- Detect the primary language of the transcript.
-- Every `action` and `rationale` MUST be in that SAME detected language.
-- Do NOT mix languages.
-
-Return ONLY a valid JSON object:
-{ "action_items": [ { "action": "...", "rationale": "..." } ] }
-
-Rules:
-- 3-6 specific, concrete action items (not vague)
-- Each `rationale` explains WHY the viewer should do this (1 sentence)
-- Be SPECIFIC — reference content from the transcript
+- Detect the primary language of the transcript and output entirely in that language.
+
+CRITICAL: Return ONLY a valid JSON object EXACTLY matching this structure:
+{
+    "action_items": [
+        {
+            "action": "A specific, actionable takeaway",
+            "rationale": "Why the viewer should do this"
+        }
+    ]
+}
 """.strip()
 
 _ACTIONS_USER = """
 TRANSCRIPT:
 {transcript}
 
-Extract actionable takeaways. Return JSON only.
+Extract actionable takeaways. Return JSON only matching the requested structure.
 """.strip()
 
 
 # ─────────────────────────────────────────────────────────────────────────────
-# LANGUAGE LABEL MAPS
+# LANGUAGE LABEL MAPS (For UI localization)
 # ─────────────────────────────────────────────────────────────────────────────
 
 _LABELS = {
@@ -212,7 +173,6 @@ _LABELS = {
 }
 
 def _labels(language: str) -> dict:
-    """Return section labels for the detected language, defaulting to English."""
     return _LABELS.get(language, _LABELS["English"])
 
 
@@ -221,31 +181,20 @@ def _labels(language: str) -> dict:
 # ─────────────────────────────────────────────────────────────────────────────
 
 class NoteGenerator:
-    """
-    Generates premium structured study notes using Groq (Llama-3.3-70b-versatile).
-
-    Public API (unchanged from v1):
-        generateSummary(transcript_text, video_title)    -> Dict
-        extractActionItems(transcript_text, video_title) -> Dict
-        format_notes_to_markdown(json_notes)             -> str
-        format_final_notes(notes, video_title, video_url, duration) -> str
-    """
+    """Generates premium structured study notes using Groq (Llama-3.3-70b-versatile)."""
 
     def __init__(self):
         self.api_key = os.environ.get("GROQ_API_KEY", "").strip()
         self.client = Groq(api_key=self.api_key) if self.api_key else None
         self.model_id = "llama-3.3-70b-versatile"
-        logger.info(f"🚀 NoteGenerator v3.0 initialized — model: {self.model_id}")
-
-    # ── Private helpers ───────────────────────────────────────────────────────
+        logger.info(f"🚀 NoteGenerator v3.1 initialized — model: {self.model_id}")
 
     def _chat(self, system: str, user: str, max_tokens: int = 4096) -> Optional[str]:
-        """Make a Groq API call and return raw content string, or None on failure."""
         try:
             response = self.client.chat.completions.create(
                 model=self.model_id,
                 max_tokens=max_tokens,
-                temperature=0.4,
+                temperature=0.3, # تقليل الإبداع لضمان التزام القالب
                 response_format={"type": "json_object"},
                 messages=[
                     {"role": "system", "content": system},
@@ -270,18 +219,11 @@ class NoteGenerator:
             "closing_thought": "",
         }
 
-    # ── Core generation methods ───────────────────────────────────────────────
-
     def generateSummary(self, transcript_text: str, video_title: str) -> Dict:
-        """
-        Generate premium structured notes from transcript.
-        Returns a dict matching SummarySchema (all content in transcript language).
-        """
         if not self.client:
             return self._get_error_json("Groq API Key missing.")
 
-        logger.info(f"📝 Summary generation started — model: {self.model_id}")
-
+        logger.info(f"📝 Summary generation started via {self.model_id}")
         user_prompt = _SUMMARY_USER.format(
             video_title=video_title,
             transcript=transcript_text[:30000],
@@ -294,26 +236,16 @@ class NoteGenerator:
         try:
             data = json.loads(raw)
             validated = SummarySchema(**data)
-            logger.info(
-                f"✅ Summary done — lang: {validated.detected_language}, "
-                f"topics: {len(validated.core_topics)}, "
-                f"concepts: {len(validated.key_concepts)}"
-            )
             return validated.model_dump()
         except (json.JSONDecodeError, ValidationError) as e:
             logger.error(f"❌ Schema validation failed: {e}")
-            return self._get_error_json(str(e))
+            return self._get_error_json(f"Validation Error: {str(e)}")
 
     def extractActionItems(self, transcript_text: str, video_title: str) -> Dict:
-        """
-        Extract rich action items (action + rationale) from transcript.
-        Returns a dict: { "action_items": [ {"action": ..., "rationale": ...} ] }
-        """
         if not self.client:
             return {"action_items": []}
 
-        logger.info(f"✅ Action items extraction started — model: {self.model_id}")
-
+        logger.info(f"✅ Action items extraction started via {self.model_id}")
         user_prompt = _ACTIONS_USER.format(transcript=transcript_text[:20000])
 
         raw = self._chat(_ACTIONS_SYSTEM, user_prompt, max_tokens=1024)
@@ -323,28 +255,12 @@ class NoteGenerator:
         try:
             data = json.loads(raw)
             validated = ActionItemsSchema(**data)
-            logger.info(f"✅ Action items done — count: {len(validated.action_items)}")
             return validated.model_dump()
         except (json.JSONDecodeError, ValidationError) as e:
             logger.error(f"❌ Action items validation failed: {e}")
             return {"action_items": []}
 
-    # ── Formatting methods ────────────────────────────────────────────────────
-
     def format_notes_to_markdown(self, json_notes: Dict) -> str:
-        """
-        Convert the merged Dict from generateSummary() + extractActionItems()
-        into premium Flutter-optimized Markdown.
-
-        Flutter MarkdownStyleSheet targets:
-          #    → h1Style  (hero title, primary color)
-          ##   → h2Style  (section headers, accent color, bold)
-          ###  → h3Style  (topic batch titles, secondary color)
-          >    → blockquoteDecoration (light card background, left border)
-          **   → bold emphasis
-          *    → italic / lighter gray text
-          `  ` → codeStyle (monospace font for timestamps)
-        """
         lang = json_notes.get("detected_language", "English")
         L = _labels(lang)
         lines: list[str] = []
@@ -360,7 +276,7 @@ class NoteGenerator:
             lines.append("---")
             lines.append("")
 
-        # ── HOOK / TEASER ─────────────────────────────────────────────────────
+        # ── HOOK ──
         hook = json_notes.get("hook", "")
         if hook:
             add(f"## ✨ {L['hook']}")
@@ -368,7 +284,7 @@ class NoteGenerator:
             add(f"*{hook}*")
             divider()
 
-        # ── EXECUTIVE SUMMARY ─────────────────────────────────────────────────
+        # ── SUMMARY ──
         summary = json_notes.get("summary", "")
         if summary:
             add(f"## 📋 {L['summary']}")
@@ -376,23 +292,16 @@ class NoteGenerator:
             add(summary)
             divider()
 
-        # ── CORE TOPICS / BATCHES ─────────────────────────────────────────────
+        # ── CORE TOPICS ──
         core_topics = json_notes.get("core_topics", [])
         if core_topics:
             add(f"## 🧠 {L['core_topics']}")
             blank()
             for i, topic in enumerate(core_topics, start=1):
-                # Support both dict (model_dump output) and Pydantic objects
-                if isinstance(topic, dict):
-                    t_title    = topic.get("title", "")
-                    t_overview = topic.get("overview", "")
-                    t_points   = topic.get("key_points", [])
-                    t_insight  = topic.get("insight")
-                else:
-                    t_title    = topic.title
-                    t_overview = topic.overview
-                    t_points   = topic.key_points
-                    t_insight  = topic.insight
+                t_title    = topic.get("title", "") if isinstance(topic, dict) else topic.title
+                t_overview = topic.get("overview", "") if isinstance(topic, dict) else topic.overview
+                t_points   = topic.get("key_points", []) if isinstance(topic, dict) else topic.key_points
+                t_insight  = topic.get("insight") if isinstance(topic, dict) else topic.insight
 
                 add(f"### {i}. {t_title}")
                 blank()
@@ -402,27 +311,19 @@ class NoteGenerator:
                     add(f"- {point}")
                 blank()
                 if t_insight:
-                    add(f"> {L['insight']}")
-                    add(f"> {t_insight}")
+                    add(f"> **{L['insight']}:** {t_insight}")
                     blank()
+            divider()
 
-            add("---")
-            blank()
-
-        # ── KEY CONCEPTS ──────────────────────────────────────────────────────
+        # ── KEY CONCEPTS ──
         key_concepts = json_notes.get("key_concepts", [])
         if key_concepts:
             add(f"## 💡 {L['concepts']}")
             blank()
             for concept in key_concepts:
-                if isinstance(concept, dict):
-                    term       = concept.get("term", "")
-                    definition = concept.get("definition", "")
-                    importance = concept.get("importance", "")
-                else:
-                    term       = concept.term
-                    definition = concept.definition
-                    importance = concept.importance
+                term       = concept.get("term", "") if isinstance(concept, dict) else concept.term
+                definition = concept.get("definition", "") if isinstance(concept, dict) else concept.definition
+                importance = concept.get("importance", "") if isinstance(concept, dict) else concept.importance
 
                 add(f"**{term}**")
                 blank()
@@ -431,60 +332,40 @@ class NoteGenerator:
                 if importance:
                     add(f"> *{L['importance']}: {importance}*")
                 blank()
+            divider()
 
-            add("---")
-            blank()
-
-        # ── TIMELINE ──────────────────────────────────────────────────────────
+        # ── TIMELINE ──
         timestamps = json_notes.get("timestamps", [])
         if timestamps:
             add(f"## ⏱️ {L['timeline']}")
             blank()
             for entry in timestamps:
-                if isinstance(entry, dict):
-                    ts         = entry.get("timestamp", "")
-                    topic      = entry.get("topic", "")
-                    ts_summary = entry.get("summary", "")
-                else:
-                    ts         = entry.timestamp
-                    topic      = entry.topic
-                    ts_summary = entry.summary
-
-                # Two-space trailing = line break in Markdown (inline subtitle)
-                add(f"- **`{ts}`** — **{topic}**  ")
+                ts         = entry.get("timestamp", "") if isinstance(entry, dict) else entry.timestamp
+                topic      = entry.get("topic", "") if isinstance(entry, dict) else entry.topic
+                ts_summary = entry.get("summary", "") if isinstance(entry, dict) else entry.summary
+
+                add(f"- **`{ts}`** — **{topic}** ")
                 add(f"  {ts_summary}")
+                blank()
             divider()
 
-        # ── ACTION ITEMS ──────────────────────────────────────────────────────
+        # ── ACTION ITEMS ──
         action_items = json_notes.get("action_items", [])
         if action_items:
             add(f"## 🎯 {L['actions']}")
             blank()
             for idx, item in enumerate(action_items, start=1):
-                if isinstance(item, str):
-                    # Backward compat: old plain-string format
-                    add(f"**{idx}.** {item}")
-                    blank()
-                elif isinstance(item, dict):
-                    action    = item.get("action", "")
-                    rationale = item.get("rationale", "")
-                    add(f"**{idx}. {action}**")
-                    blank()
-                    if rationale:
-                        add(f"> *{L['why']} {rationale}*")
-                    blank()
-                else:
-                    # Pydantic object
-                    add(f"**{idx}. {item.action}**")
-                    blank()
-                    if item.rationale:
-                        add(f"> *{L['why']} {item.rationale}*")
-                    blank()
-
-            add("---")
-            blank()
+                action    = item.get("action", "") if isinstance(item, dict) else item.action
+                rationale = item.get("rationale", "") if isinstance(item, dict) else item.rationale
+                
+                add(f"**{idx}. {action}**")
+                blank()
+                if rationale:
+                    add(f"> *{L['why']} {rationale}*")
+                blank()
+            divider()
 
-        # ── KEYWORDS / TAGS ───────────────────────────────────────────────────
+        # ── KEYWORDS ──
         keywords = json_notes.get("keywords", [])
         if keywords:
             add(f"## 🏷️ {L['tags']}")
@@ -492,7 +373,7 @@ class NoteGenerator:
             add("  ".join([f"`{kw}`" for kw in keywords]))
             divider()
 
-        # ── CLOSING THOUGHT ───────────────────────────────────────────────────
+        # ── CLOSING THOUGHT ──
         closing = json_notes.get("closing_thought", "")
         if closing:
             add(f"## 🔖 {L['closing']}")
@@ -511,11 +392,15 @@ class NoteGenerator:
     ) -> str:
         """
         Wrap the formatted Markdown body with the video header.
-        `duration` is in seconds (int), matching original signature.
+        Fixes the 00:00 duration bug.
         """
-        minutes      = int(duration // 60)
-        secs         = int(duration % 60)
-        duration_str = f"{minutes:02d}:{secs:02d}"
+        # إذا كانت المدة صفر أو غير موجودة
+        if duration and duration > 0:
+            minutes      = int(duration // 60)
+            secs         = int(duration % 60)
+            duration_str = f"{minutes:02d}:{secs:02d}"
+        else:
+            duration_str = "N/A (Auto-generated)"
 
         header = (
             f"# {video_title}\n\n"