shishantbiswas/aniscrap
1
0
Fork 0
mirror of https://huggingface.co/spaces/Catapang1989/aniscrap synced 2026-04-17 15:51:46 +00:00
Code Issues Packages Projects Releases Wiki Activity

Update main.py

Browse Source
This commit is contained in:
Aira Catapang
2026-03-18 09:53:57 +00:00
committed by system
parent 99a9cdb91a
commit 07048d71c4
1 changed files with 58 additions and 144 deletions
Download Patch File Download Diff File Expand all files Collapse all files

202
main.py
View File

@@ -55,12 +55,13 @@ class AnimePahe:
async def _intercept_assets(self, route):
url = route.request.url.lower()
# Allow all requests from aniwatchtv so season posters can load
if "aniwatchtv.to" in url:
# Allow requests from aniwatchtv & kwik (video host) so players/posters load correctly
if "aniwatchtv.to" in url or "kwik" in url:
await route.continue_()
return
if any(ad in url for ad in self.ad_domains) or url.endswith(
(".png", ".jpg", ".jpeg", ".webp", ".woff")
(".png", ".jpg", ".jpeg", ".webp", ".woff", ".gif")
):
await route.abort()
else:
@@ -82,33 +83,31 @@ class AnimePahe:
) -> Optional[str]:
if not m3u8_url:
return None
match = re.search(r"(https?://[^.]+)[^/]*/stream/(.*?)/[^/]+\.m3u8", m3u8_url)
# Example: https://na-02.kwik.cx/stream/abc123def/index.m3u8
match = re.search(r"(https?://[^/]+)/stream/([^/]+)/", m3u8_url)
if match:
clean_name = re.sub(r"[^\w\s]", "", anime_name).strip().replace(" ", "_")
if not clean_name:
clean_name = "Anime"
filename = f"{clean_name}_EP{episode}_{res}P.mp4"
return f"{match.group(1)}.kwik.cx/mp4/{match.group(2)}?file={filename}"
domain = match.group(1) # e.g. https://na-02.kwik.cx
token = match.group(2) # e.g. abc123def
return f"{domain}/mp4/{token}?file={filename}"
return None
async def _scrape_play_meta(self, page) -> tuple:
meta = await page.evaluate("""() => {
const titleEl = document.querySelector('.theatre-info h2 a, .anime-title, h2 a[href*="/anime/"]')
const epEl = document.querySelector('.theatre-info h2, .episode-title, h2')
const titleEl = document.querySelector('.theatre-info h1 a, .theatre-info h2 a, .anime-title, h1, h2');
let title = titleEl ? titleEl.innerText.trim() : '';
let episode = '';
let title = titleEl ? titleEl.innerText.trim() : ''
let episode = ''
if (epEl) {
const m = epEl.innerText.match(/episode\\s*(\\d+)/i)
if (m) episode = m[1]
}
if (!title || !episode) {
const t = document.title || ''
const m = t.match(/^(.+?)\\s*[-\\u2013]\\s*Episode\\s*(\\d+)/i)
if (m) {
if (!title) title = m[1].trim()
if (!episode) episode = m[2].trim()
}
const t = document.title || '';
// Match exactly: "Anime Name - 01 - AnimePahe" OR "Anime Name - Episode 01 - AnimePahe"
const m = t.match(/^(.+?)\\s*[-\\u2013]\\s*(?:Episode\\s*)?(\\d+(?:\\.\\d+)?)/i);
if (m) {
if (!title || title.length < 2) title = m[1].trim();
if (!episode) episode = m[2].trim();
}
return { title, episode }
@@ -129,7 +128,6 @@ class AnimePahe:
timeout=30000,
)
# Wait for the anime info block to render
try:
await page.wait_for_selector(
"div.anime-info, div.anime-summary, aside, main", timeout=10000
@@ -137,15 +135,8 @@ class AnimePahe:
except:
pass
# Extra wait for JS-rendered content
await asyncio.sleep(2)
# Debug: log all hrefs found on page
all_links = await page.evaluate("""() => {
return Array.from(document.querySelectorAll('a[href]')).map(a => a.href)
}""")
print(f"[scrape_ids] All links found: {all_links}")
ids = await page.evaluate("""() => {
let ids = {}
document.querySelectorAll("a[href]").forEach(a => {
@@ -166,7 +157,6 @@ class AnimePahe:
return ids
}""")
print(f"[scrape_ids] Extracted ids: {ids}")
ids["animepahe"] = session
return ids
@@ -230,17 +220,35 @@ class AnimePahe:
p.on("request", capture)
try:
await p.set_extra_http_headers({"Referer": BASE_URL})
await p.goto(embed_url, wait_until="domcontentloaded")
await p.evaluate(
"document.querySelectorAll('button, video, [class*=play]').forEach(el => el.click())"
)
# Set the exact referer Kwik expects to prevent token rejections
await p.set_extra_http_headers({"Referer": "https://animepahe.si/"})
await p.goto(embed_url, wait_until="domcontentloaded", timeout=15000)
# Click loop: Muting allows browsers to bypass autoplay restrictions safely
for _ in range(6):
if found.is_set():
break
await p.evaluate("""() => {
document.querySelectorAll('video').forEach(v => {
v.muted = true;
const p = v.play();
if (p !== undefined) p.catch(() => {});
});
document.querySelectorAll('button, .vjs-big-play-button').forEach(b => {
try { b.click() } catch(e) {}
});
}""")
await asyncio.sleep(1.5)
try:
await asyncio.wait_for(found.wait(), timeout=5.0)
except asyncio.TimeoutError:
pass
except Exception as e:
print(f"[_embed_to_m3u8] ERROR: {e}")
finally:
await p.close()
return m3u8
# ---------------- ANILIST ----------------
@@ -251,85 +259,21 @@ class AnimePahe:
Media(id: $id, type: ANIME) {
id
idMal
title {
romaji
english
native
}
title { romaji english native }
synonyms
description(asHtml: false)
format
status
episodes
duration
source
countryOfOrigin
isAdult
format status episodes duration source countryOfOrigin isAdult
startDate { year month day }
endDate { year month day }
season
seasonYear
averageScore
meanScore
popularity
favourites
trending
genres
coverImage {
extraLarge
large
medium
color
}
season seasonYear averageScore meanScore popularity favourites trending genres
coverImage { extraLarge large medium color }
bannerImage
trailer {
id
site
}
studios(isMain: true) {
nodes { name siteUrl }
}
relations {
edges {
relationType(version: 2)
node {
id
idMal
title { romaji english }
format
status
episodes
averageScore
coverImage { medium }
siteUrl
}
}
}
recommendations(perPage: 20, sort: RATING_DESC) {
nodes {
rating
mediaRecommendation {
id
idMal
title { romaji english }
format
status
episodes
averageScore
coverImage { medium }
siteUrl
}
}
}
externalLinks {
site
url
type
}
nextAiringEpisode {
airingAt
episode
}
trailer { id site }
studios(isMain: true) { nodes { name siteUrl } }
relations { edges { relationType(version: 2) node { id idMal title { romaji english } format status episodes averageScore coverImage { medium } siteUrl } } }
recommendations(perPage: 20, sort: RATING_DESC) { nodes { rating mediaRecommendation { id idMal title { romaji english } format status episodes averageScore coverImage { medium } siteUrl } } }
externalLinks { site url type }
nextAiringEpisode { airingAt episode }
}
}
"""
@@ -628,21 +572,13 @@ class AnimePahe:
# ---------------- SEASONS ----------------
async def get_seasons(self, anime_id: str) -> dict:
"""
Scrape the 'More Seasons' section from aniwatchtv.to using the
existing Playwright browser context.
anime_id is the full slug, e.g. jujutsu-kaisen-the-culling-game-part-1-20401
"""
url = f"{ANIWATCHTV_BASE}/{anime_id}"
page = await self.context.new_page()
try:
await page.goto(url, wait_until="domcontentloaded", timeout=30000)
# Short wait for lazy-loaded images and JS rendering
await asyncio.sleep(1)
# Wait for season block — try common selectors gracefully
for selector in [".os-list", ".seasons-block", "[class*='season']", "main"]:
try:
await page.wait_for_selector(selector, timeout=5000)
@@ -653,10 +589,9 @@ class AnimePahe:
seasons = await page.evaluate(f"""() => {{
const BASE = "{ANIWATCHTV_BASE}";
const currentId = "{anime_id}";
const results = [];
const results =[];
const seen = new Set();
// Strategy 1: dedicated season list block (.os-list or similar)
const block = (
document.querySelector('.os-list') ||
document.querySelector('.seasons-block') ||
@@ -664,7 +599,6 @@ class AnimePahe:
document.querySelector('[class*="season-list"]')
);
// Strategy 2: find a heading that says "More Seasons" and walk up
const fallbackContainer = (() => {{
for (const el of document.querySelectorAll('*')) {{
if (/more\\s+seasons?/i.test(el.innerText?.trim() || '')) {{
@@ -690,7 +624,6 @@ class AnimePahe:
if (!fullUrl) continue;
const slug = fullUrl.replace(/\\/$/, '').split('/').pop();
// Include ALL slugs — current page included — dedupe only
if (!slug || seen.has(slug)) continue;
seen.add(slug);
@@ -700,7 +633,6 @@ class AnimePahe:
const titleEl = a.querySelector('span, [class*="title"], [class*="name"]');
const title = (titleEl?.innerText?.trim() || a.innerText?.trim() || slug);
// Poster is in a sibling/child div.season-poster as a CSS background-image
const posterEl = a.querySelector('.season-poster') || a.closest('li, div')?.querySelector('.season-poster');
let poster = null;
if (posterEl) {{
@@ -752,8 +684,8 @@ async def root():
"/ids/:session",
"/episodes/:session?p=:page&resolve=false|true",
"/resolve/:animeSession/:episodeSession",
"/seasons/:animeId - e.g. /seasons/jujutsu-kaisen-the-culling-game-part-1-20401",
"/poster?url=:cdnImageUrl - proxy hotlink-protected poster images",
"/seasons/:animeId",
"/poster?url=:cdnImageUrl",
],
}
@@ -790,17 +722,6 @@ async def api_resolve(anime: str, episode: str):
@app.get("/seasons/{anime_id:path}")
async def api_seasons(anime_id: str, request: Request):
"""
Scrape the More Seasons section from aniwatchtv.to.
Example:
GET /seasons/jujutsu-kaisen-the-culling-game-part-1-20401
Returns:
id - the slug passed in
total - number of other seasons found
seasons[] - list of { title, id, numericId, url, poster }
"""
data = await pahe.get_seasons(anime_id)
base_url = str(request.base_url).rstrip("/")
for season in data.get("seasons", []):
@@ -813,20 +734,13 @@ async def api_seasons(anime_id: str, request: Request):
@app.get("/poster")
async def api_poster(url: str = Query(..., description="CDN image URL to proxy")):
"""
Proxy a hotlink-protected poster image with the correct Referer header.
Use this to display season/anime posters in the browser.
Example:
GET /poster?url=https://cdn.noitatnemucod.net/thumbnail/100x200/100/abc123.jpg
"""
try:
async with httpx.AsyncClient(timeout=10, follow_redirects=True) as client:
resp = await client.get(
url,
headers={
"Referer": "https://aniwatchtv.to/",
"User-Agent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/122 Safari/537.36",
"User-Agent": "Mozilla/5.0",
},
)
resp.raise_for_status()