#!/usr/bin/env python3
"""
📚 ScribdHarvester v2.0 - Valideret Metode
==========================================

Kombinerer:
1. Cookie extraction fra Chrome browser
2. Officiel scribd-downloader bibliotek
3. Web scraping for favorites/library
4. Neo4j cloud storage

Kør: pip install -r scribd_requirements.txt
     python scribd_harvester_v2.py

@author WidgeTDC Neural Network
"""

import os
import sys
import json
import hashlib
import requests
import re
import subprocess
from pathlib import Path
from datetime import datetime
from typing import List, Dict, Optional, Any
from dataclasses import dataclass, asdict
from urllib.parse import urljoin, urlparse
import time
from dotenv import load_dotenv

# Load environment variables
load_dotenv()

# Neo4j
from neo4j import GraphDatabase

# HTML parsing
from bs4 import BeautifulSoup

# Cookie extractor
from scribd_cookie_extractor import ScribdCookieExtractor

# Image processing
try:
    from PIL import Image
    import io
    HAS_PIL = True
except ImportError:
    HAS_PIL = False


@dataclass
class ScribdDocument:
    id: str
    title: str
    author: str
    url: str
    doc_type: str
    thumbnail: str
    description: str
    content_hash: str
    saved_at: str
    local_path: str = ""

@dataclass
class ExtractedImage:
    id: str
    source_doc_id: str
    url: str
    caption: str
    content_hash: str
    local_path: str
    width: int
    height: int


class ScribdHarvesterV2:
    """
    Valideret Scribd harvester med cookie-baseret authentication
    """
    
    # Neo4j AuraDB Cloud
    NEO4J_URI = os.getenv("NEO4J_URI", "bolt://localhost:7687")
    NEO4J_USER = os.getenv("NEO4J_USER", "neo4j")
    NEO4J_PASSWORD = os.getenv("NEO4J_PASSWORD", "password")
    
    # Scribd endpoints
    SCRIBD_BASE = "https://www.scribd.com"
    SCRIBD_API = "https://www.scribd.com/api"
    
    HEADERS = {
        "User-Agent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36",
        "Accept": "application/json, text/html, */*",
        "Accept-Language": "en-US,en;q=0.9,da;q=0.8",
    }
    
    def __init__(self, output_dir: str = None):
        self.output_dir = Path(output_dir or "data/scribd_harvest")
        self.image_dir = self.output_dir / "images"
        self.docs_dir = self.output_dir / "documents"
        
        for d in [self.output_dir, self.image_dir, self.docs_dir]:
            d.mkdir(parents=True, exist_ok=True)
        
        # Session med cookies
        self.session = requests.Session()
        self.session.headers.update(self.HEADERS)
        
        # Neo4j
        self.driver = GraphDatabase.driver(
            self.NEO4J_URI,
            auth=(self.NEO4J_USER, self.NEO4J_PASSWORD)
        )
        
        # Stats
        self.stats = {
            "documents_found": 0,
            "documents_downloaded": 0,
            "documents_skipped": 0,
            "images_extracted": 0
        }
        
        print("📚 ScribdHarvester v2.0 - Valideret Metode")
        print(f"   📁 Output: {self.output_dir.absolute()}")
    
    def authenticate(self) -> bool:
        """Hent og anvend cookies fra browser eller fil"""
        print("\n🔐 AUTHENTICATION")
        print("-" * 40)
        
        cookies = None
        
        # FØRST: Check for manuel cookie fil
        cookie_file = self.output_dir / "scribd_cookies.json"
        if cookie_file.exists():
            print(f"   📄 Finder cookie fil: {cookie_file}")
            try:
                with open(cookie_file, 'r') as f:
                    data = json.load(f)
                
                session_cookie = data.get('_scribd_session', '')
                expire_cookie = data.get('_scribd_expire', '')
                
                if session_cookie and 'INDSÆT' not in session_cookie:
                    cookies = {
                        '_scribd_session': session_cookie,
                        '_scribd_expire': expire_cookie
                    }
                    print("   ✅ Cookies loaded fra fil!")
                else:
                    print("   ⚠️  Cookie fil ikke udfyldt - prøver automatisk extraction...")
            except Exception as e:
                print(f"   ⚠️  Fejl ved læsning af cookie fil: {e}")
        
        # DEREFTER: Prøv automatisk extraction
        if not cookies:
            extractor = ScribdCookieExtractor()
            cookies = extractor.get_cookies()
        
        if not cookies:
            return False
        
        # Anvend cookies til session
        for name, value in cookies.items():
            self.session.cookies.set(name, value, domain=".scribd.com")
        
        # Verificer
        return self._verify_session()
    
    def _verify_session(self) -> bool:
        """Verificer at vi er logget ind"""
        try:
            # Prøv at hente bruger info
            response = self.session.get(
                f"{self.SCRIBD_BASE}/account",
                allow_redirects=False
            )
            
            if response.status_code == 200:
                if 'login' not in response.url.lower():
                    print("✅ Session verificeret - logget ind!")
                    return True
            
            # Prøv alternativ endpoint
            response = self.session.get(f"{self.SCRIBD_BASE}/saved")
            if response.status_code == 200:
                soup = BeautifulSoup(response.text, 'html.parser')
                # Check for logged-in indicators
                if soup.find('a', href=re.compile(r'/logout')):
                    print("✅ Session verificeret via /saved")
                    return True
            
            print("⚠️  Session ikke verificeret - cookies kan være udløbet")
            return False
            
        except Exception as e:
            print(f"❌ Verification fejl: {e}")
            return False


    def fetch_library(self) -> List[Dict]:
        """Hent brugerens bibliotek/gemte dokumenter"""
        print("\n📖 FETCHING LIBRARY")
        print("-" * 40)
        
        all_items = []
        
        # Endpoints at prøve
        endpoints = [
            "/saved",
            "/library", 
            "/your-library",
            "/account/saved",
            "/lists"
        ]
        
        for endpoint in endpoints:
            url = f"{self.SCRIBD_BASE}{endpoint}"
            print(f"   Prøver: {endpoint}")
            
            try:
                response = self.session.get(url)
                if response.status_code != 200:
                    continue
                
                soup = BeautifulSoup(response.text, 'html.parser')
                
                # Find dokumenter med forskellige selectors
                items = self._extract_items_from_html(soup)
                
                for item in items:
                    if not any(i['url'] == item['url'] for i in all_items):
                        all_items.append(item)
                        print(f"      📄 {item['title'][:50]}...")
                
                # Prøv også at finde JSON data
                json_items = self._extract_items_from_scripts(soup)
                for item in json_items:
                    if not any(i['url'] == item['url'] for i in all_items):
                        all_items.append(item)
                
            except Exception as e:
                print(f"      ⚠️  Fejl: {e}")
        
        print(f"\n   📚 Fandt {len(all_items)} dokumenter total")
        self.stats["documents_found"] = len(all_items)
        return all_items
    
    def _extract_items_from_html(self, soup: BeautifulSoup) -> List[Dict]:
        """Ekstraher dokumenter fra HTML"""
        items = []
        
        # Forskellige link patterns
        patterns = [
            ('a[href*="/document/"]', 'document'),
            ('a[href*="/book/"]', 'book'),
            ('a[href*="/read/"]', 'book'),
            ('a[href*="/audiobook/"]', 'audiobook'),
            ('.doc-list-item', 'document'),
            ('[data-doc-id]', 'document'),
        ]
        
        for selector, doc_type in patterns:
            try:
                elements = soup.select(selector)
                for el in elements:
                    href = el.get('href', '')
                    if not href:
                        # Prøv at finde link i children
                        link = el.find('a')
                        if link:
                            href = link.get('href', '')
                    
                    if not href or '/login' in href:
                        continue
                    
                    if not href.startswith('http'):
                        href = urljoin(self.SCRIBD_BASE, href)
                    
                    # Ekstraher ID
                    match = re.search(r'/(document|book|read|audiobook)/(\d+)', href)
                    doc_id = match.group(2) if match else None
                    
                    if not doc_id:
                        continue
                    
                    # Find titel
                    title = el.get_text(strip=True)
                    if not title or len(title) < 3:
                        title_el = el.find(['h1', 'h2', 'h3', 'h4', '.title', '[class*="title"]'])
                        if title_el:
                            title = title_el.get_text(strip=True)
                    
                    # Find thumbnail
                    thumbnail = ''
                    img = el.find('img')
                    if img:
                        thumbnail = img.get('src', '') or img.get('data-src', '')
                    
                    items.append({
                        'id': doc_id,
                        'url': href,
                        'title': title or f"Document {doc_id}",
                        'type': doc_type,
                        'thumbnail': thumbnail,
                    })
            except:
                pass
        
        return items
    
    def _extract_items_from_scripts(self, soup: BeautifulSoup) -> List[Dict]:
        """Ekstraher dokumenter fra JSON scripts i HTML"""
        items = []
        
        scripts = soup.find_all('script')
        for script in scripts:
            text = script.string or ''
            
            # Prøv at finde JSON data
            patterns = [
                r'window\.__INITIAL_STATE__\s*=\s*({.*?});',
                r'window\.Scribd\..*?=\s*({.*?});',
                r'"documents"\s*:\s*(\[.*?\])',
            ]
            
            for pattern in patterns:
                try:
                    match = re.search(pattern, text, re.DOTALL)
                    if match:
                        data = json.loads(match.group(1))
                        extracted = self._traverse_json_for_docs(data)
                        items.extend(extracted)
                except:
                    pass
        
        return items
    
    def _traverse_json_for_docs(self, obj, depth=0) -> List[Dict]:
        """Traverser JSON for at finde dokumenter"""
        items = []
        
        if depth > 8:
            return items
        
        if isinstance(obj, dict):
            # Check om dette er et dokument
            if 'id' in obj and ('title' in obj or 'name' in obj):
                doc_id = str(obj.get('id', ''))
                if doc_id.isdigit():
                    doc_type = obj.get('type', 'document').lower()
                    if doc_type in ['book', 'audiobook']:
                        url = f"{self.SCRIBD_BASE}/{doc_type}/{doc_id}"
                    else:
                        url = f"{self.SCRIBD_BASE}/document/{doc_id}"
                    
                    items.append({
                        'id': doc_id,
                        'url': url,
                        'title': obj.get('title') or obj.get('name', f'Document {doc_id}'),
                        'type': doc_type,
                        'thumbnail': obj.get('thumbnail_url', obj.get('cover_url', '')),
                        'author': obj.get('author', {}).get('name', '') if isinstance(obj.get('author'), dict) else obj.get('author', ''),
                    })
            
            for v in obj.values():
                items.extend(self._traverse_json_for_docs(v, depth + 1))
        
        elif isinstance(obj, list):
            for item in obj:
                items.extend(self._traverse_json_for_docs(item, depth + 1))
        
        return items


    def download_document(self, item: Dict) -> Optional[Path]:
        """Download dokument med scribdl eller direkte"""
        doc_id = item.get('id', '')
        url = item.get('url', '')
        title = item.get('title', f'doc_{doc_id}')
        
        # Sanitize filename
        safe_title = re.sub(r'[<>:"/\\|?*]', '_', title)[:100]
        
        print(f"   📥 Downloader: {title[:50]}...")
        
        # Metode 1: Brug scribdl CLI
        output_path = self.docs_dir / f"{doc_id}_{safe_title}"
        
        try:
            # Prøv scribdl først
            result = subprocess.run(
                ['scribdl', '-i', url],
                cwd=str(self.docs_dir),
                capture_output=True,
                text=True,
                timeout=120
            )
            
            if result.returncode == 0:
                # Find downloaded files
                for f in self.docs_dir.glob(f"*{doc_id}*"):
                    print(f"      ✅ Downloaded: {f.name}")
                    return f
                    
        except FileNotFoundError:
            print("      ⚠️  scribdl ikke installeret, bruger alternativ metode")
        except subprocess.TimeoutExpired:
            print("      ⚠️  Timeout på download")
        except Exception as e:
            print(f"      ⚠️  scribdl fejl: {e}")
        
        # Metode 2: Download direkte
        return self._direct_download(item)
    
    def _direct_download(self, item: Dict) -> Optional[Path]:
        """Direkte download af dokument sider"""
        doc_id = item['id']
        url = item['url']
        
        try:
            response = self.session.get(url)
            if response.status_code != 200:
                return None
            
            soup = BeautifulSoup(response.text, 'html.parser')
            
            # Find dokument reader
            reader = soup.find('div', class_=re.compile(r'reader|document|pages'))
            if not reader:
                # Gem HTML som fallback
                html_path = self.docs_dir / f"{doc_id}.html"
                with open(html_path, 'w', encoding='utf-8') as f:
                    f.write(response.text)
                return html_path
            
            # Find og download billeder af sider
            images = reader.find_all('img', src=True)
            if images:
                doc_folder = self.docs_dir / doc_id
                doc_folder.mkdir(exist_ok=True)
                
                for i, img in enumerate(images):
                    img_url = img['src']
                    if not img_url.startswith('http'):
                        img_url = urljoin(url, img_url)
                    
                    try:
                        img_response = self.session.get(img_url, timeout=30)
                        if img_response.status_code == 200:
                            ext = 'jpg' if 'jpeg' in img_response.headers.get('content-type', '') else 'png'
                            img_path = doc_folder / f"page_{i:03d}.{ext}"
                            with open(img_path, 'wb') as f:
                                f.write(img_response.content)
                    except:
                        pass
                
                return doc_folder
            
            return None
            
        except Exception as e:
            print(f"      ❌ Download fejl: {e}")
            return None
    
    def extract_images_for_presentations(self, item: Dict) -> List[ExtractedImage]:
        """Ekstraher billeder egnet til præsentationer"""
        images = []
        url = item['url']
        doc_id = item['id']
        
        try:
            response = self.session.get(url)
            if response.status_code != 200:
                return images
            
            soup = BeautifulSoup(response.text, 'html.parser')
            
            # Find alle billeder
            for idx, img in enumerate(soup.find_all('img')):
                src = img.get('src', '') or img.get('data-src', '')
                if not src:
                    continue
                
                # Skip ikoner og små billeder
                skip_patterns = ['avatar', 'icon', 'logo', 'button', 'sprite', '1x1', 'tracking']
                if any(p in src.lower() for p in skip_patterns):
                    continue
                
                # Check størrelse
                width = int(img.get('width', 0) or 0)
                height = int(img.get('height', 0) or 0)
                if (width > 0 and width < 150) or (height > 0 and height < 150):
                    continue
                
                # Download billede
                if not src.startswith('http'):
                    src = urljoin(url, src)
                
                try:
                    img_response = self.session.get(src, timeout=30)
                    if img_response.status_code != 200:
                        continue
                    
                    # Check actual size
                    if HAS_PIL:
                        pil_img = Image.open(io.BytesIO(img_response.content))
                        width, height = pil_img.size
                        
                        if width < 200 or height < 150:
                            continue
                    
                    # Gem lokalt
                    content_hash = hashlib.md5(img_response.content).hexdigest()
                    ext = 'jpg' if 'jpeg' in img_response.headers.get('content-type', '') else 'png'
                    local_path = self.image_dir / f"{doc_id}_img_{idx}.{ext}"
                    
                    with open(local_path, 'wb') as f:
                        f.write(img_response.content)
                    
                    # Caption
                    caption = img.get('alt', '') or img.get('title', '')
                    figure = img.find_parent('figure')
                    if figure:
                        figcaption = figure.find('figcaption')
                        if figcaption:
                            caption = figcaption.get_text(strip=True)
                    
                    images.append(ExtractedImage(
                        id=f"{doc_id}_img_{idx}",
                        source_doc_id=doc_id,
                        url=src,
                        caption=caption,
                        content_hash=content_hash,
                        local_path=str(local_path),
                        width=width,
                        height=height
                    ))
                    
                except Exception as e:
                    pass
            
            if images:
                print(f"      🖼️  {len(images)} billeder ekstraheret")
                self.stats["images_extracted"] += len(images)
                
        except Exception as e:
            print(f"      ⚠️  Image extraction fejl: {e}")
        
        return images


    def save_to_neo4j(self, item: Dict, local_path: Optional[Path], images: List[ExtractedImage]):
        """Gem dokument og billeder til Neo4j"""
        doc_id = item['id']
        content_hash = hashlib.md5(f"{item['title']}-{item['url']}".encode()).hexdigest()
        
        with self.driver.session() as session:
            # Check for duplicate
            result = session.run(
                "MATCH (d:ScribdDocument {contentHash: $hash}) RETURN d LIMIT 1",
                hash=content_hash
            )
            if len(list(result)) > 0:
                self.stats["documents_skipped"] += 1
                return
            
            # Gem dokument
            session.run("""
                MERGE (d:ScribdDocument {id: $id})
                SET d.title = $title,
                    d.author = $author,
                    d.url = $url,
                    d.type = $doc_type,
                    d.thumbnail = $thumbnail,
                    d.contentHash = $content_hash,
                    d.localPath = $local_path,
                    d.savedAt = datetime(),
                    d.source = 'Scribd'
                
                MERGE (s:DataSource {name: 'Scribd'})
                SET s.type = 'DocumentPlatform', s.lastHarvest = datetime()
                MERGE (d)-[:HARVESTED_FROM]->(s)
                
                MERGE (cat:Category {name: $doc_type})
                MERGE (d)-[:BELONGS_TO]->(cat)
            """,
                id=doc_id,
                title=item.get('title', ''),
                author=item.get('author', ''),
                url=item.get('url', ''),
                doc_type=item.get('type', 'document'),
                thumbnail=item.get('thumbnail', ''),
                content_hash=content_hash,
                local_path=str(local_path) if local_path else ''
            )
            
            self.stats["documents_downloaded"] += 1
            
            # Gem billeder
            for img in images:
                session.run("""
                    MERGE (i:ScribdImage {id: $id})
                    SET i.url = $url,
                        i.caption = $caption,
                        i.contentHash = $content_hash,
                        i.localPath = $local_path,
                        i.width = $width,
                        i.height = $height,
                        i.usableForPresentations = true,
                        i.savedAt = datetime()
                    
                    WITH i
                    MATCH (d:ScribdDocument {id: $source_doc_id})
                    MERGE (i)-[:EXTRACTED_FROM]->(d)
                    
                    MERGE (cat:AssetCategory {name: 'Presentation Images'})
                    MERGE (i)-[:AVAILABLE_FOR]->(cat)
                """,
                    id=img.id,
                    url=img.url,
                    caption=img.caption,
                    content_hash=img.content_hash,
                    local_path=img.local_path,
                    width=img.width,
                    height=img.height,
                    source_doc_id=img.source_doc_id
                )
    
    def run(self, download_docs: bool = True, extract_images: bool = True):
        """Hovedeksekveringsflow"""
        print("")
        print("╔══════════════════════════════════════════════════════════════╗")
        print("║  📚 SCRIBD HARVESTER v2.0 - VALIDERET METODE                ║")
        print("║  Cookie-based authentication med Neo4j Cloud storage        ║")
        print("╚══════════════════════════════════════════════════════════════╝")
        
        # Step 1: Authentication
        if not self.authenticate():
            print("\n❌ Authentication fejlede!")
            print("   Prøv at:")
            print("   1. Åbn Chrome og log ind på scribd.com")
            print("   2. Luk Chrome helt")
            print("   3. Kør scriptet igen")
            return self.stats
        
        # Step 2: Fetch library
        items = self.fetch_library()
        
        if not items:
            print("\n⚠️  Ingen dokumenter fundet i dit bibliotek")
            print("   Check at du har gemte dokumenter på scribd.com/saved")
            return self.stats
        
        # Step 3: Process documents
        print(f"\n⚙️  PROCESSING {len(items)} DOCUMENTS")
        print("-" * 40)
        
        for i, item in enumerate(items, 1):
            print(f"\n[{i}/{len(items)}] {item.get('title', 'Unknown')[:50]}...")
            
            local_path = None
            images = []
            
            # Download
            if download_docs:
                local_path = self.download_document(item)
            
            # Extract images
            if extract_images:
                images = self.extract_images_for_presentations(item)
            
            # Save to Neo4j
            self.save_to_neo4j(item, local_path, images)
            
            # Rate limiting
            time.sleep(2)
        
        # Summary
        self._print_summary()
        return self.stats
    
    def _print_summary(self):
        """Print summary"""
        print("")
        print("═" * 60)
        print("📊 HARVEST COMPLETE")
        print("═" * 60)
        print(f"   📚 Documents found:      {self.stats['documents_found']}")
        print(f"   ✅ Documents downloaded: {self.stats['documents_downloaded']}")
        print(f"   ⏭️  Documents skipped:    {self.stats['documents_skipped']}")
        print(f"   🖼️  Images extracted:     {self.stats['images_extracted']}")
        print(f"   📁 Output directory:     {self.output_dir.absolute()}")
        print("═" * 60)
    
    def close(self):
        """Cleanup"""
        self.driver.close()


def main():
    """Entry point"""
    import argparse
    
    parser = argparse.ArgumentParser(description='Scribd Harvester v2.0')
    parser.add_argument('--no-download', action='store_true', help='Skip document download')
    parser.add_argument('--no-images', action='store_true', help='Skip image extraction')
    parser.add_argument('--output', type=str, help='Output directory')
    args = parser.parse_args()
    
    harvester = ScribdHarvesterV2(output_dir=args.output)
    
    try:
        harvester.run(
            download_docs=not args.no_download,
            extract_images=not args.no_images
        )
    finally:
        harvester.close()


if __name__ == "__main__":
    main()