Change: switch to curl_cffi, fix oh

.gitignore

@@ -8,3 +8,4 @@ dist
 
 .venv
 test*.*
+temp

comiccrawler/analyzer.py

@@ -138,7 +138,7 @@ def analyze_pages(self):
 			print('Analyzing {}...'.format(url))
 			sleep(getattr(self.mission.module, "rest_analyze", 0))
 			r = urlparse(self.mission.url)
-			self.html = self.grabber.html(url, retry=True, header={
+			self.html = self.grabber.html(url, retry=True, headers={
 				"Referer": self.mission.url,
 				"Origin": f"{r.scheme}://{r.netloc}"
 				})

comiccrawler/crawler.py

@@ -220,7 +220,7 @@ def get_html(self):
 			self.html = True
 		else:
 			r = urlparse(self.mission.url)
-			self.html = self.downloader.html(self.ep.current_url, header={
+			self.html = self.downloader.html(self.ep.current_url, headers={
 				"Referer": self.mission.url,
 				"Origin": f"{r.scheme}://{r.netloc}"
 				})

comiccrawler/error.py

@@ -1,4 +1,5 @@
-from requests import HTTPError
+# from requests import HTTPError
+from curl_cffi.requests.exceptions import HTTPError
 
 class ComicCrawlerSignal(BaseException):
 	"""Extend BaseException."""

comiccrawler/grabber.py

@@ -5,15 +5,15 @@
 from threading import Lock
 from urllib.parse import quote, urlsplit, urlunsplit, urlparse
 import re
-import socket
+# import socket
 import time
 import json
 
 import enlighten
-import requests
 from worker import WorkerExit, async_, await_, sleep, Defer
-from urllib3.util import is_fp_closed
+# from urllib3.util import is_fp_closed
 from urllib3.exceptions import IncompleteRead
+from curl_cffi.requests.exceptions import HTTPError
 
 from .config import setting
 from .io import content_write
@@ -74,24 +74,19 @@ def grabber_log(obj):
 		content = time.strftime("%Y-%m-%dT%H:%M:%S%z") + "\n" + json.dumps(obj, indent=2, sort_keys=True) + "\n\n"
 		content_write(profile("grabber.log"), content, append=True)
 
-def grabber(url, header=None, *, referer=None, cookie=None,
-			retry=False, done=None, proxy=None, **kwargs):
+def grabber(url, *, referer=None, retry=False, done=None, proxy=None, **kwargs):
 	"""Request url, return text or bytes of the content."""
 	s = session_manager.get(url)
 
 	if referer:
 		s.headers['Referer'] = quote_unicode(referer)
 
-	if cookie:
-		quote_unicode_dict(cookie)
-		requests.utils.add_dict_to_cookiejar(s.cookies, cookie)
-
 	if isinstance(proxy, str):
 		proxies = {'http': proxy, 'https': proxy}
 	else:
 		proxies = proxy
 
-	r = await_(do_request, s, url, proxies, retry, headers=header, **kwargs)
+	r = await_(do_request, s, url, proxies, retry, **kwargs)
 
 	if done:
 		done(s, r)
@@ -116,27 +111,13 @@ def do_request(s, url, proxies, retry, **kwargs):
 				})
 
 		if r.status_code in SUCCESS_CODES:
-			content_length = r.headers.get("Content-Length")
-			if not kwargs.get("stream", False) and content_length and int(content_length) != r.raw.tell():
-				raise ValueError(
-					"incomplete response. Content-Length: {content_length}, got: {actual}"
-					.format(content_length=content_length, actual=r.raw.tell())
-					)
 			break
 		if not retry or r.status_code not in RETRYABLE_HTTP_CODES:
 			r.raise_for_status()
 		# 302 error without location header
 		if r.status_code == 302:
 			# pylint: disable=protected-access
-			match = re.search(
-				r"^location:\s*(.+)",
-				str(r.raw._original_response.msg),
-				re.M + re.I
-				)
-			if not match:
-				raise TypeError("status 302 without location header")
-			url = match.group(1)
-			continue
+			raise TypeError("status 302 without location header")
 		print(r)
 		print("retry after {sleep_time} seconds".format(sleep_time=sleep_time))
 		sleep(sleep_time)
@@ -160,19 +141,9 @@ def guess_encoding(r):
 
 def iter_content(r):
 	"""Iterate the content of the response."""
-	# FIXME: requests streaming is so broken wtf
-	# https://github.com/psf/requests/issues/5536
-	# https://github.com/urllib3/urllib3/issues/2123
-	if r.raw.chunked and r.raw.supports_chunked_reads():
-		yield from r.raw.read_chunked(decode_content=True)
-	else:
-		while not is_fp_closed(r.raw._fp) or len(r.raw._decoded_buffer) > 0: # pylint: disable=protected-access
-			b = r.raw.read1(decode_content=True)
-			yield b
-			if not b:
-				sleep(0.1)
+	yield from r.iter_content()
 
-def grabimg(*args, on_opened=None, tempfile=None, header=None, **kwargs):
+def grabimg(*args, on_opened=None, tempfile=None, headers=None, **kwargs):
 	"""Grab the image. Return ImgResult"""
 	kwargs["stream"] = True
 	loaded = 0
@@ -182,12 +153,12 @@ def grabimg(*args, on_opened=None, tempfile=None, header=None, **kwargs):
 		except FileNotFoundError:
 			pass
 	if loaded:
-		if not header:
-			header = {}
-		header["Range"] = f"bytes={loaded}-"
+		if not headers:
+			headers = {}
+		headers["Range"] = f"bytes={loaded}-"
 	try:
-		r = grabber(*args, header=header, **kwargs)
-	except requests.HTTPError as err:
+		r = grabber(*args, headers=headers, **kwargs)
+	except HTTPError as err:
 		if err.response.status_code != 416:
 			raise err
 		try:
@@ -227,8 +198,7 @@ def _():
 						counter.update(len(chunk))
 						loaded += len(chunk)
 	except WorkerExit:
-		socket.close(r.raw._fp.fileno()) # pylint: disable=protected-access
-		r.raw.release_conn()
+		r.close()
 		raise
 	if total and loaded < total:
 		raise IncompleteRead(loaded, total - loaded)

comiccrawler/mods/oh.py

@@ -13,7 +13,7 @@
 
 from ..core import Episode, grabhtml
 
-domain = ["www.ohmanhua.com", "www.cocomanhua.com"]
+domain = ["www.ohmanhua.com", "www.cocomanhua.com", "www.colamanga.com"]
 name = "OH漫畫"
 
 def get_title(html, url):
@@ -48,7 +48,7 @@ def __str__(self):
 scripts = ScriptCache()
 
 def get_images(html, url):
-	cdata = re.search("var C_DATA='[^']+'", html).group(0)
+	cdata = re.search("var C_DATA=('[^']+')", html).group(1)
 
 	scripts.fetch(html, url, [
 		"\/l\.js",
@@ -58,31 +58,71 @@ def get_images(html, url):
 	])
 
 	code = """
+const _log = console.log;
+
+Function.prototype.toString = (function(_toString) {
+  return function() {
+    return _toString.apply(this, arguments).replace(/\\r?\\n/g, '');
+  }
+})(Function.prototype.toString);
+
+self.setInterval = function() {};
+
+self.eval = function(_eval) {
+  return function() {
+    _log('eval', arguments[0]);
+    return _eval.apply(this, arguments);
+  };
+}(self.eval);
+
+self.convertWordArrayToUint8Array =
+  self.convertUint8ArrayToWordArray =
+  self.__b_a =
+  self.__cad = 
+  self.__js = 
+  undefined;
+
 	(function() {
 	
+  let _cookies = "";
+
 	function noop(path = "") {
-	  if (path === "document.cookie") return "";
+	  if (path === "document.cookie") return _cookies;
 	  if (path === "$.inArray") return (v, a) => a.indexOf(v);
 	  
 	  return new Proxy(() => {}, {
-		apply: () => noop("?"),
-		get: (target, prop) => noop(`${path}.${prop}`)
+      apply: () => noop(`${path}.called`),
+      get: (target, prop) => {
+        const propPath = typeof prop == "symbol" ? `${path}.${String(prop)}` : `${path}.${prop}`;
+        if (propPath == "document.domain") return "www.colamanga.com";
+        _log("get", propPath);
+        return noop(propPath);
+      },
+      set: (target, prop, value) => {
+        const propPath = `${path}.${prop}`;
+        if (propPath == "document.cookie") {
+          _cookies += value.split(";")[0] + "; ";
+        }
+        _log(propPath, value);
+        return value;
+      }
 	  });
 	}
-	
-	const exports = undefined;
-	const window = global;
-	window.location = {
+
+    self.window = self;
+	self.location = {
     	protocol: "http://",
 		href: '""" + url + """'
 	}
-	const navigator = {
+	self.navigator = {
 	  userAgent: ""
 	};
-	const document = noop("document")
-	const $ = noop("$");
+	self.document = noop("document")
+	self.$ = noop("$");
+    self.devtools = noop("devtools");
+	self.localStorage = noop("localStorage");
 	
-	""" + cdata + "\n" + str(scripts) + """
+	self.C_DATA = """ + cdata + "\n" + str(scripts) + """
 	
 	window.use_domain = {
 	},
@@ -108,8 +148,10 @@ class Image {
 		__cr.preLoadImg(i++)
 	} while (dirty);
 	return imgs;
-	}).call(global);
+	}).call(self);
 	"""
 
+	# import pathlib
+	# pathlib.Path("oh0.mjs").write_text(code, encoding="utf-8")
 	imgs = eval(code)
 	return [urljoin(url, i) for i in imgs]

comiccrawler/mods/setnmh.py

@@ -31,7 +31,7 @@ def get_episodes(html, url):
 			"order_by": "1",
 			"chapter_type": "1"
 		},
-		header = {
+		headers = {
 			"X-Requested-With": "XMLHttpRequest"
 		}
 	)
@@ -74,7 +74,7 @@ def get_images(html, url):
 			"chapter_id": chapter_id,
 			"page": page
 		},
-		header = {
+		headers = {
 			"X-Requested-With": "XMLHttpRequest"
 		}
 	)

comiccrawler/module_grabber.py

@@ -1,4 +1,4 @@
-from requests.utils import dict_from_cookiejar
+# from requests.utils import dict_from_cookiejar
 
 from .grabber import grabhtml, grabimg
 
@@ -19,8 +19,8 @@ def img(self, url, **kwargs):
 
 	def grab(self, grab_method, url=None, **kwargs):
 		new_kwargs = {
-			"header": self.get_header(),
-			"cookie": purify_cookie(self.get_cookie()),
+			"headers": self.get_header(),
+			"cookies": purify_cookie(self.get_cookie()),
 			"done": self.handle_grab,
 			"proxy": self.mod.config.get("proxy"),
 			"verify": self.mod.config.getboolean("verify", True)
@@ -50,13 +50,14 @@ def get_cookie(self):
 		return cookie
 
 	def handle_grab(self, session, _response):
-		cookie = dict_from_cookiejar(session.cookies)
-		config = getattr(self.mod, "config", None)
-		if not config:
-			return
-
-		for key in config:
-			if key.startswith("cookie_"):
-				name = key[7:]
-				if name in cookie:
-					config[key] = cookie[name]
+		pass
+		# cookie = dict_from_cookiejar(session.cookies)
+		# config = getattr(self.mod, "config", None)
+		# if not config:
+		# 	return
+		# 	
+		# for key in config:
+		# 	if key.startswith("cookie_"):
+		# 		name = key[7:]
+		# 		if name in cookie:
+		# 			config[key] = cookie[name]

comiccrawler/session_manager.py

@@ -2,7 +2,7 @@
 from threading import Lock
 from typing import Callable, Any
 
-from requests import Session as RequestsSession
+from curl_cffi.requests import Session as RequestsSession
 
 from .util import extract_curl
 

comiccrawler/util.py

@@ -2,9 +2,10 @@
 import string
 from functools import total_ordering
 from pathlib import Path
-from requests.cookies import RequestsCookieJar
+from http.cookiejar import CookieJar
 
 import uncurl
+import curl_cffi.requests.cookies
 
 def dump(html):
 	Path("dump.html").write_text(html, encoding="utf-8")
@@ -100,7 +101,9 @@ def balance(s: str, index: int, left="(", right=")", skip=0):
 
 	return s[start:end]
 
-def get_cookie(cookie_jar: RequestsCookieJar, name, domain=None) -> str:
+def get_cookie(cookie_jar: CookieJar | curl_cffi.requests.cookies.Cookies, name, domain=None) -> str:
+	if hasattr(cookie_jar, "jar"):
+		cookie_jar = cookie_jar.jar
 	l = [cookie for cookie in cookie_jar if cookie.name == name]
 	def key(cookie):
 		if not domain or not cookie.domain: