feat(api): Add image downloading

src/api/requirements.txt

@@ -9,6 +9,8 @@ asttokens==2.4.1
 async-timeout==4.0.3
 attrs==23.1.0
 backoff==2.2.1
+beautifulsoup4==4.12.3
+bs4==0.0.2
 click==8.1.7
 comm==0.2.0
 debugpy==1.8.0
@@ -48,6 +50,7 @@ pyzmq==25.1.2
 rich==13.7.0
 six==1.16.0
 sniffio==1.3.0
+soupsieve==2.5
 stack-data==0.6.3
 starlette==0.32.0.post1
 tornado==6.4

src/api/src/create_book.py

@@ -6,6 +6,7 @@ import backoff
 from aiohttp import ClientResponseError
 from aiohttp_client_cache.session import CachedSession
 from aiohttp_client_cache import FileBackend
+from bs4 import BeautifulSoup
 
 
 headers = {
@@ -116,19 +117,38 @@ async def set_cover(book, data):
     book.set_cover("cover.jpg", await fetch_cover(data["cover"]))
 
 
-async def add_chapters(book, data):
+async def add_chapters(book, data, download_images: bool = False):
     chapters = []
 
     for part in data["parts"]:
         content = await fetch_part_content(part["id"])
         title = part["title"]
+        clean_title = slugify(title)
 
         # Thanks https://eu17.proxysite.com/process.php?d=5VyWYcoQl%2BVF0BYOuOavtvjOloFUZz2BJ%2Fepiusk6Nz7PV%2B9i8rs7cFviGftrBNll%2B0a3qO7UiDkTt4qwCa0fDES&b=1
         chapter = epub.EpubHtml(
             title=title,
-            file_name=f"{slugify(title)}.xhtml",
+            file_name=f"{clean_title}.xhtml",
             lang=data["language"]["name"],
         )
+
+        if download_images:
+            soup = BeautifulSoup(content, "lxml")
+            async with CachedSession(cache=cache, headers=headers) as session:
+                for idx, image in enumerate(soup.find_all("img")):
+                    if not image["src"]:
+                        continue
+                    async with session.get(image["src"]) as response:
+                        img = epub.EpubImage(
+                            media_type="image/jpeg",
+                            content=await response.read(),
+                            file_name=f"static/{clean_title}/{idx}.jpeg",
+                        )
+                        book.add_item(img)
+                        content = content.replace(
+                            str(image), f'<img src="static/{clean_title}/{idx}.jpeg"/>'
+                        )
+
         chapter.set_content(f"<h1>{title}</h1>" + content)
 
         chapters.append(chapter)

src/api/src/main.py

@@ -17,7 +17,7 @@ def home():
 
 
 @app.get("/download/{story_id}")
-async def download_book(story_id: int):
+async def download_book(story_id: int, download_images: bool = False):
     data = await retrieve_story(story_id)
     book = epub.EpubBook()
 
@@ -27,7 +27,7 @@ async def download_book(story_id: int):
     # print("Metadata Downloaded")
 
     # Chapters are downloaded
-    async for title in add_chapters(book, data):
+    async for title in add_chapters(book, data, download_images=download_images):
         # print(f"Part ({title}) downloaded")
         ...
 
@@ -57,4 +57,4 @@ app.mount("/", StaticFiles(directory=BUILD_PATH), "static")
 if __name__ == "__main__":
     import uvicorn
 
-    uvicorn.run(app, host="0.0.0.0", port=80)
+    uvicorn.run(app, host="0.0.0.0", port=1112)