Add support for a proxy host PAC url on the General Config/Advanced Settings page.

CHANGES.md

@@ -23,6 +23,7 @@
 * Change corrected spelling & better clarified various log messages
 * Change minor PEP8 tweaks in sab.py
 * Add api disabled error code for newznab providers
+* Add support for a proxy host PAC url on the General Config/Advanced Settings page
 
 [develop changelog]
 

gui/slick/interfaces/default/config_general.tmpl

@@ -551,7 +551,8 @@
 								<span class="component-title">Proxy host</span>
 								<span class="component-desc">
 									<input type="text" name="proxy_setting" value="$sickbeard.PROXY_SETTING" class="form-control input-sm input300" />
-									<div class="clear-left"><p>blank to disable or proxy to use when connecting to providers</p></div>
+									<p>blank to disable</p>
+									<div class="clear-left"><p>proxy address for connecting to providers (use 'PAC:Url' for PAC support)</p></div>
 							</label>
 						</div>
 

sickbeard/helpers.py

@@ -1236,6 +1236,60 @@ def _getTempDir():
 
     return os.path.join(tempfile.gettempdir(), "SickGear-%s" % (uid))
 
+
+def proxy_setting(proxy_setting, request_url, force=False):
+    """
+    Returns a list of a) proxy_setting address value or a PAC is fetched and parsed if proxy_setting
+    starts with "PAC:" (case-insensitive) and b) True/False if "PAC" is found in the proxy_setting.
+
+    The PAC data parser is crude, javascript is not eval'd. The first "PROXY URL" found is extracted with a list
+    of "url_a_part.url_remaining", "url_b_part.url_remaining", "url_n_part.url_remaining" and so on.
+    Also, PAC data items are escaped for matching therefore regular expression items will not match a request_url.
+
+    If force is True or request_url contains a PAC parsed data item then the PAC proxy address is returned else False.
+    None is returned in the event of an error fetching PAC data.
+
+    """
+
+    # check for "PAC" usage
+    match = re.search(r'^\s*PAC:\s*(.*)', proxy_setting, re.I)
+    if not match:
+        return proxy_setting, False
+    pac_url = match.group(1)
+
+    # prevent a recursive test with existing proxy setting when fetching PAC url
+    proxy_setting_backup = sickbeard.PROXY_SETTING
+    sickbeard.PROXY_SETTING = ''
+
+    resp = ''
+    try:
+        resp = getURL(pac_url)
+    except:
+        pass
+    sickbeard.PROXY_SETTING = proxy_setting_backup
+
+    if not resp:
+        return None, False
+
+    proxy_address = None
+    request_url_match = False
+    for pac_data in re.finditer(r"""(?:[^'"]*['"])([^\.]+\.[^'"]*)(?:['"])""", resp, re.I):
+        data = re.search(r"""PROXY\s+([^'"]+)""", pac_data.group(1), re.I)
+        if data:
+            if force:
+                return data.group(1), True
+            proxy_address = (proxy_address, data.group(1))[None is proxy_address]
+        elif re.search(re.escape(pac_data.group(1)), request_url, re.I):
+            request_url_match = True
+            if None is not proxy_address:
+                break
+
+    if None is proxy_address:
+        return None, True
+
+    return (False, proxy_address)[request_url_match], True
+
+
 def getURL(url, post_data=None, params=None, headers=None, timeout=30, session=None, json=False):
     """
     Returns a byte-string retrieved from the url provider.
@@ -1265,10 +1319,16 @@ def getURL(url, post_data=None, params=None, headers=None, timeout=30, session=N
 
         # request session proxies
         if sickbeard.PROXY_SETTING:
-            logger.log("Using proxy for url: " + url, logger.DEBUG)
+            (proxy_address, pac_found) = proxy_setting(sickbeard.PROXY_SETTING, url)
+            msg = '%sproxy for url: %s' % (('', 'PAC parsed ')[pac_found], url)
+            if None is proxy_address:
+                logger.log('Proxy error, aborted the request using %s' % msg, logger.DEBUG)
+                return
+            elif proxy_address:
+                logger.log('Using %s' % msg, logger.DEBUG)
                 session.proxies = {
-                "http": sickbeard.PROXY_SETTING,
+                    'http': proxy_address,
-                "https": sickbeard.PROXY_SETTING,
+                    'https': proxy_address
                 }
 
         # decide if we get or post data to server
@@ -1316,10 +1376,16 @@ def download_file(url, filename, session=None):
 
     # request session proxies
     if sickbeard.PROXY_SETTING:
-        logger.log("Using proxy for url: " + url, logger.DEBUG)
+        (proxy_address, pac_found) = proxy_setting(sickbeard.PROXY_SETTING, url)
+        msg = '%sproxy for url: %s' % (('', 'PAC parsed ')[pac_found], url)
+        if None is proxy_address:
+            logger.log('Proxy error, aborted the request using %s' % msg, logger.DEBUG)
+            return
+        elif proxy_address:
+            logger.log('Using %s' % msg, logger.DEBUG)
             session.proxies = {
-            "http": sickbeard.PROXY_SETTING,
+                'http': proxy_address,
-            "https": sickbeard.PROXY_SETTING,
+                'https': proxy_address
             }
 
     try:

sickbeard/indexers/indexer_api.py

@@ -19,7 +19,7 @@ import os
 import sickbeard
 
 from indexer_config import initConfig, indexerConfig
-
+from sickbeard.helpers import proxy_setting
 
 class indexerApi(object):
     def __init__(self, indexerID=None):
@@ -49,7 +49,9 @@ class indexerApi(object):
             if sickbeard.CACHE_DIR:
                 indexerConfig[self.indexerID]['api_params']['cache'] = os.path.join(sickbeard.CACHE_DIR, 'indexers', self.name)
             if sickbeard.PROXY_SETTING and sickbeard.PROXY_INDEXERS:
-                indexerConfig[self.indexerID]['api_params']['proxy'] = sickbeard.PROXY_SETTING
+                (proxy_address, pac_found) = proxy_setting(sickbeard.PROXY_SETTING, indexerConfig[self.indexerID]['base_url'], force=True)
+                if proxy_address:
+                    indexerConfig[self.indexerID]['api_params']['proxy'] = proxy_address
 
             return indexerConfig[self.indexerID]['api_params']