Fixing the title and book details from AA (#289)

Should fix #288

app.py

@@ -12,7 +12,7 @@ from flask import url_for as flask_url_for
 import typing
 
 from logger import setup_logger
-from config import _SUPPORTED_BOOK_LANGUAGE, BOOK_LANGUAGE
+from config import _SUPPORTED_BOOK_LANGUAGE, BOOK_LANGUAGE, SUPPORTED_FORMATS
 from env import FLASK_HOST, FLASK_PORT, APP_ENV, CWA_DB_PATH, DEBUG, USING_EXTERNAL_BYPASSER, BUILD_VERSION, RELEASE_VERSION
 import backend
 
@@ -106,6 +106,7 @@ def index() -> str:
     return render_template('index.html', 
                            book_languages=_SUPPORTED_BOOK_LANGUAGE, 
                            default_language=BOOK_LANGUAGE, 
+                           supported_formats=SUPPORTED_FORMATS,
                            debug=DEBUG,
                            build_version=BUILD_VERSION,
                            release_version=RELEASE_VERSION,

backend.py

@@ -81,6 +81,12 @@ def queue_status() -> Dict[str, Dict[str, Any]]:
         Dict: Queue status organized by status type
     """
     status = book_queue.get_status()
+    for _, books in status.items():
+        for _, book_info in books.items():
+            if book_info.download_path:
+                if not os.path.exists(book_info.download_path):
+                    book_info.download_path = None
+
     # Convert Enum keys to strings and properly format the response
     return {
         status_type.value: books

book_manager.py

@@ -10,7 +10,7 @@ from bs4 import BeautifulSoup, Tag, NavigableString, ResultSet
 import downloader
 from logger import setup_logger
 from config import SUPPORTED_FORMATS, BOOK_LANGUAGE, AA_BASE_URL
-from env import AA_DONATOR_KEY, USE_CF_BYPASS, PRIORITIZE_WELIB
+from env import AA_DONATOR_KEY, USE_CF_BYPASS, PRIORITIZE_WELIB, ALLOW_USE_WELIB
 from models import BookInfo, SearchFilters
 logger = setup_logger(__name__)
 
@@ -169,21 +169,6 @@ def _parse_book_info_page(soup: BeautifulSoup, book_id: str) -> BookInfo:
 
     data = soup.find_all("div", {"class": "main-inner"})[0].find_next("div")
     divs = list(data.children)
-    _details = divs[13].text.strip().lower().split(" · ")
-    format = ""
-    size = ""
-    for f in _details:
-        if format == "" and f.strip().lower() in SUPPORTED_FORMATS:
-            format = f.strip().lower()
-        if size == "" and any(u in f.strip().lower() for u in ["mb", "kb", "gb"]):
-            size = f.strip().lower()
-
-    if format == "" or size == "":
-        for f in _details:
-            if f == "" and not " " in f.strip().lower():
-                format = f.strip().lower()
-            if size == "" and "." in f.strip().lower():
-                size = f.strip().lower()
 
     every_url = soup.find_all("a")
     slow_urls_no_waitlist = set()
@@ -237,20 +222,49 @@ def _parse_book_info_page(soup: BeautifulSoup, book_id: str) -> BookInfo:
     # Remove empty urls
     urls = [url for url in urls if url != ""]
 
+    # Filter out divs that are not text
+    original_divs = divs
+    divs = [div.text.strip() for div in divs if div.text.strip() != ""]
+
+    separator_index = 6
+    for i, div in enumerate(divs):
+        if "·" in div.strip():
+            separator_index = i
+            break
+            
+    _details = divs[separator_index].lower().split(" · ")
+    format = ""
+    size = ""
+    for f in _details:
+        if format == "" and f.strip().lower() in SUPPORTED_FORMATS:
+            format = f.strip().lower()
+        if size == "" and any(u in f.strip().lower() for u in ["mb", "kb", "gb"]):
+            size = f.strip().lower()
+
+    if format == "" or size == "":
+        for f in _details:
+            if f == "" and not " " in f.strip().lower():
+                format = f.strip().lower()
+            if size == "" and "." in f.strip().lower():
+                size = f.strip().lower()
+
+    
+    book_title = divs[separator_index-3].strip("🔍")
+
     # Extract basic information
     book_info = BookInfo(
         id=book_id,
         preview=preview,
-        title=divs[7].next.strip(),
-        publisher=divs[11].text.strip(),
-        author=divs[9].text.strip(),
+        title=book_title,
+        publisher=divs[separator_index-1],
+        author=divs[separator_index-2],
         format=format,
         size=size,
         download_urls=urls,
     )
 
     # Extract additional metadata
-    info = _extract_book_metadata(divs[-6])
+    info = _extract_book_metadata(original_divs[-6])
     book_info.info = info
 
     # Set language and year from metadata if available
@@ -262,6 +276,8 @@ def _parse_book_info_page(soup: BeautifulSoup, book_id: str) -> BookInfo:
     return book_info
 
 def _get_download_urls_from_welib(book_id: str) -> set[str]:
+    if ALLOW_USE_WELIB == False:
+        return set()
     """Get download urls from welib.org."""
     url = f"https://welib.org/md5/{book_id}"
     logger.info(f"Getting download urls from welib.org for {book_id}. While this uses the bypasser, it will not start downloading them yet.")

env.py

@@ -28,6 +28,7 @@ FLASK_PORT = int(os.getenv("FLASK_PORT", "8084"))
 DEBUG = string_to_bool(os.getenv("DEBUG", "false"))
 APP_ENV = os.getenv("APP_ENV", "N/A").lower()
 PRIORITIZE_WELIB = string_to_bool(os.getenv("PRIORITIZE_WELIB", "false"))
+ALLOW_USE_WELIB = string_to_bool(os.getenv("ALLOW_USE_WELIB", "true"))
 
 # Version information from Docker build
 BUILD_VERSION = os.getenv("BUILD_VERSION", "N/A")

readme.md

@@ -83,6 +83,7 @@ Note that if using TOR, the TZ will be calculated automatically based on IP.
 | `AA_DONATOR_KEY`       | Optional Donator key for Anna's Archive fast download API | ``                                |
 | `USE_BOOK_TITLE`       | Use book title as filename instead of ID                  | `false`                           |
 | `PRIORITIZE_WELIB`     | When downloading, download from WELIB first instead of AA | `false`                           |
+| `ALLOW_USE_WELIB`       | Allow usage of welib for downloading books if found there | `true`                            |
 
 If you change `BOOK_LANGUAGE`, you can add multiple comma separated languages, such as `en,fr,ru` etc.  
 

static/js/main.js

@@ -229,6 +229,10 @@
       for (const [name, items] of Object.entries(data || {})) {
         if (!items || Object.keys(items).length === 0) continue;
         const rows = Object.values(items).map((b) => {
+          const titleText = utils.e(b.title) || '-';
+          const maybeLinkedTitle = b.download_path
+            ? `<a href="/request/api/localdownload?id=${encodeURIComponent(b.id)}" class="text-blue-600 hover:underline">${titleText}</a>`
+            : titleText;
           const actions = (name === 'queued' || name === 'downloading')
             ? `<button class="px-2 py-1 rounded border text-xs" data-cancel="${utils.e(b.id)}" style="border-color: var(--border-muted);">Cancel</button>`
             : '';
@@ -236,7 +240,7 @@
             ? `<div class="h-2 bg-black/10 rounded overflow-hidden"><div class="h-2 bg-blue-600" style="width:${Math.round(b.progress)}%"></div></div>`
             : '';
           return `<li class="p-3 rounded border flex flex-col gap-2" style="border-color: var(--border-muted); background: var(--bg-soft)">
-            <div class="text-sm"><span class="opacity-70">${utils.e(name)}</span> • <strong>${utils.e(b.title || '-') }</strong></div>
+            <div class="text-sm"><span class="opacity-70">${utils.e(name)}</span> • <strong>${maybeLinkedTitle}</strong></div>
             ${progress}
             <div class="flex items-center gap-2">${actions}</div>
           </li>`;

templates/index.html

@@ -132,14 +132,38 @@
                     <div class="md:col-span-2 lg:col-span-3">
                         <label class="block text-sm mb-1 opacity-80">Formats</label>
                         <div class="flex flex-wrap gap-3 text-sm">
-                            <label class="inline-flex items-center gap-2"><input type="checkbox" id="format-pdf" value="pdf"> PDF</label>
-                            <label class="inline-flex items-center gap-2"><input type="checkbox" id="format-epub" value="epub" checked> EPUB</label>
-                            <label class="inline-flex items-center gap-2"><input type="checkbox" id="format-mobi" value="mobi" checked> MOBI</label>
-                            <label class="inline-flex items-center gap-2"><input type="checkbox" id="format-azw3" value="azw3" checked> AZW3</label>
-                            <label class="inline-flex items-center gap-2"><input type="checkbox" id="format-fb2" value="fb2" checked> FB2</label>
-                            <label class="inline-flex items-center gap-2"><input type="checkbox" id="format-djvu" value="djvu" checked> DJVU</label>
-                            <label class="inline-flex items-center gap-2"><input type="checkbox" id="format-cbz" value="cbz" checked> CBZ</label>
-                            <label class="inline-flex items-center gap-2"><input type="checkbox" id="format-cbr" value="cbr" checked> CBR</label>
+                            <label class="inline-flex items-center gap-2 {{ 'opacity-50 cursor-not-allowed' if 'pdf' not in supported_formats else '' }}">
+                                <input type="checkbox" id="format-pdf" value="pdf" {% if 'pdf' not in supported_formats %}disabled{% endif %}>
+                                PDF
+                            </label>
+                            <label class="inline-flex items-center gap-2 {{ 'opacity-50 cursor-not-allowed' if 'epub' not in supported_formats else '' }}">
+                                <input type="checkbox" id="format-epub" value="epub" {% if 'epub' in supported_formats %}checked{% endif %} {% if 'epub' not in supported_formats %}disabled{% endif %}>
+                                EPUB
+                            </label>
+                            <label class="inline-flex items-center gap-2 {{ 'opacity-50 cursor-not-allowed' if 'mobi' not in supported_formats else '' }}">
+                                <input type="checkbox" id="format-mobi" value="mobi" {% if 'mobi' in supported_formats %}checked{% endif %} {% if 'mobi' not in supported_formats %}disabled{% endif %}>
+                                MOBI
+                            </label>
+                            <label class="inline-flex items-center gap-2 {{ 'opacity-50 cursor-not-allowed' if 'azw3' not in supported_formats else '' }}">
+                                <input type="checkbox" id="format-azw3" value="azw3" {% if 'azw3' in supported_formats %}checked{% endif %} {% if 'azw3' not in supported_formats %}disabled{% endif %}>
+                                AZW3
+                            </label>
+                            <label class="inline-flex items-center gap-2 {{ 'opacity-50 cursor-not-allowed' if 'fb2' not in supported_formats else '' }}">
+                                <input type="checkbox" id="format-fb2" value="fb2" {% if 'fb2' in supported_formats %}checked{% endif %} {% if 'fb2' not in supported_formats %}disabled{% endif %}>
+                                FB2
+                            </label>
+                            <label class="inline-flex items-center gap-2 {{ 'opacity-50 cursor-not-allowed' if 'djvu' not in supported_formats else '' }}">
+                                <input type="checkbox" id="format-djvu" value="djvu" {% if 'djvu' in supported_formats %}checked{% endif %} {% if 'djvu' not in supported_formats %}disabled{% endif %}>
+                                DJVU
+                            </label>
+                            <label class="inline-flex items-center gap-2 {{ 'opacity-50 cursor-not-allowed' if 'cbz' not in supported_formats else '' }}">
+                                <input type="checkbox" id="format-cbz" value="cbz" {% if 'cbz' in supported_formats %}checked{% endif %} {% if 'cbz' not in supported_formats %}disabled{% endif %}>
+                                CBZ
+                            </label>
+                            <label class="inline-flex items-center gap-2 {{ 'opacity-50 cursor-not-allowed' if 'cbr' not in supported_formats else '' }}">
+                                <input type="checkbox" id="format-cbr" value="cbr" {% if 'cbr' in supported_formats %}checked{% endif %} {% if 'cbr' not in supported_formats %}disabled{% endif %}>
+                                CBR
+                            </label>
                         </div>
                     </div>
                     <div class="md:col-span-2 lg:col-span-3 flex justify-end">