Merge pull request #1478 from jdholtz/115-compatibility

Add support for downloading Chromedriver versions 115+

undetected_chromedriver/patcher.py

@@ -3,9 +3,11 @@
 
 from distutils.version import LooseVersion
 import io
+import json
 import logging
 import os
 import pathlib
+import platform
 import random
 import re
 import shutil
@@ -24,21 +26,9 @@
 
 class Patcher(object):
     lock = Lock()
-    url_repo = "https://chromedriver.storage.googleapis.com"
-    zip_name = "chromedriver_%s.zip"
     exe_name = "chromedriver%s"
 
     platform = sys.platform
-    if platform.endswith("win32"):
-        zip_name %= "win32"
-        exe_name %= ".exe"
-    if platform.endswith(("linux", "linux2")):
-        zip_name %= "linux64"
-        exe_name %= ""
-    if platform.endswith("darwin"):
-        zip_name %= "mac64"
-        exe_name %= ""
-
     if platform.endswith("win32"):
         d = "~/appdata/roaming/undetected_chromedriver"
     elif "LAMBDA_TASK_ROOT" in os.environ:
@@ -72,6 +62,10 @@ def __init__(
         prefix = "undetected"
         self.user_multi_procs = user_multi_procs
 
+        self.is_old_chromedriver = version_main and version_main <= 114
+        # Needs to be called before self.exe_name is accessed
+        self._set_platform_name()
+
         if not os.path.exists(self.data_path):
             os.makedirs(self.data_path, exist_ok=True)
 
@@ -97,9 +91,33 @@ def __init__(
             self._custom_exe_path = True
             self.executable_path = executable_path
 
+        # Set the correct repository to download the Chromedriver from
+        if self.is_old_chromedriver:
+            self.url_repo = "https://chromedriver.storage.googleapis.com"
+        else:
+            self.url_repo = "https://googlechromelabs.github.io/chrome-for-testing"
+
         self.version_main = version_main
         self.version_full = None
 
+    def _set_platform_name(self):
+        """
+        Set the platform and exe name based on the platform undetected_chromedriver is running on
+        in order to download the correct chromedriver.
+        """
+        if self.platform.endswith("win32"):
+            self.platform_name = "win32"
+            self.exe_name %= ".exe"
+        if self.platform.endswith(("linux", "linux2")):
+            self.platform_name = "linux64"
+            self.exe_name %= ""
+        if self.platform.endswith("darwin"):
+            if self.is_old_chromedriver:
+                self.platform_name = "mac64"
+            else:
+                self.platform_name = "mac-x64"
+            self.exe_name %= ""
+
     def auto(self, executable_path=None, force=False, version_main=None, _=None):
         """
 
@@ -217,12 +235,32 @@ def fetch_release_number(self):
         :return: version string
         :rtype: LooseVersion
         """
-        path = "/latest_release"
-        if self.version_main:
-            path += f"_{self.version_main}"
-        path = path.upper()
+        # Endpoint for old versions of Chromedriver (114 and below)
+        if self.is_old_chromedriver:
+            path = f"/latest_release_{self.version_main}"
+            path = path.upper()
+            logger.debug("getting release number from %s" % path)
+            return LooseVersion(urlopen(self.url_repo + path).read().decode())
+
+        # Endpoint for new versions of Chromedriver (115+)
+        if not self.version_main:
+            # Fetch the latest version
+            path = "/last-known-good-versions-with-downloads.json"
+            logger.debug("getting release number from %s" % path)
+            with urlopen(self.url_repo + path) as conn:
+                response = conn.read().decode()
+
+            last_versions = json.loads(response)
+            return LooseVersion(last_versions["channels"]["Stable"]["version"])
+
+        # Fetch the latest minor version of the major version provided
+        path = "/latest-versions-per-milestone-with-downloads.json"
         logger.debug("getting release number from %s" % path)
-        return LooseVersion(urlopen(self.url_repo + path).read().decode())
+        with urlopen(self.url_repo + path) as conn:
+            response = conn.read().decode()
+
+        major_versions = json.loads(response)
+        return LooseVersion(major_versions["milestones"][str(self.version_main)]["version"])
 
     def parse_exe_version(self):
         with io.open(self.executable_path, "rb") as f:
@@ -237,17 +275,29 @@ def fetch_package(self):
 
         :return: path to downloaded file
         """
-        u = "%s/%s/%s" % (self.url_repo, self.version_full.vstring, self.zip_name)
-        logger.debug("downloading from %s" % u)
-        # return urlretrieve(u, filename=self.data_path)[0]
-        return urlretrieve(u)[0]
+        zip_name = f"chromedriver_{self.platform_name}.zip"
+        if self.is_old_chromedriver:
+            download_url = "%s/%s/%s" % (self.url_repo, self.version_full.vstring, zip_name)
+        else:
+            zip_name = zip_name.replace("_", "-", 1)
+            download_url = "https://edgedl.me.gvt1.com/edgedl/chrome/chrome-for-testing/%s/%s/%s"
+            download_url %= (self.version_full.vstring, self.platform_name, zip_name)
+
+        logger.debug("downloading from %s" % download_url)
+        return urlretrieve(download_url)[0]
 
     def unzip_package(self, fp):
         """
         Does what it says
 
         :return: path to unpacked executable
         """
+        exe_path = self.exe_name
+        if not self.is_old_chromedriver:
+            # The new chromedriver unzips into its own folder
+            zip_name = f"chromedriver-{self.platform_name}"
+            exe_path = os.path.join(zip_name, self.exe_name)
+
         logger.debug("unzipping %s" % fp)
         try:
             os.unlink(self.zip_path)
@@ -256,10 +306,10 @@ def unzip_package(self, fp):
 
         os.makedirs(self.zip_path, mode=0o755, exist_ok=True)
         with zipfile.ZipFile(fp, mode="r") as zf:
-            zf.extract(self.exe_name, self.zip_path)
-        os.rename(os.path.join(self.zip_path, self.exe_name), self.executable_path)
+            zf.extractall(self.zip_path)
+        os.rename(os.path.join(self.zip_path, exe_path), self.executable_path)
         os.remove(fp)
-        os.rmdir(self.zip_path)
+        shutil.rmtree(self.zip_path)
         os.chmod(self.executable_path, 0o755)
         return self.executable_path