Zippyshare bug #1191 correction

This version use the internal `Beatifulsoup`. When you decide to get rid of it I suggest to add a dependency to `lxml`.

I've found a way to handle deliberate errors in the JS scripts without relying on PyV8: I've added JS `try/catch` statements around zippyshare scripts.

1 files changed, 36 insertions, 13 deletions

diff --git a/module/plugins/hoster/ZippyshareCom.py b/module/plugins/hoster/ZippyshareCom.py
index 615559989..ad4688bac 100644
--- a/module/plugins/hoster/ZippyshareCom.py
+++ b/module/plugins/hoster/ZippyshareCom.py
@@ -5,17 +5,18 @@ import re
 from module.plugins.internal.CaptchaService import ReCaptcha
 from module.plugins.internal.SimpleHoster import SimpleHoster, create_getInfo
 
+from module.lib.BeautifulSoup import BeautifulSoup
 
 class ZippyshareCom(SimpleHoster):
     __name__    = "ZippyshareCom"
     __type__    = "hoster"
-    __version__ = "0.73"
+    __version__ = "0.74"
 
     __pattern__ = r'http://www\d{0,2}\.zippyshare\.com/v(/|iew\.jsp.*key=)(?P<KEY>[\w^_]+)'
 
     __description__ = """Zippyshare.com hoster plugin"""
     __license__     = "GPLv3"
-    __authors__     = [("Walter Purcaro", "vuolter@gmail.com")]
+    __authors__     = [("Walter Purcaro", "vuolter@gmail.com", "sebdelsol")]
 
 
     COOKIES = [("zippyshare.com", "ziplocale", "en")]
@@ -46,20 +47,42 @@ class ZippyshareCom(SimpleHoster):
                 self.error(e)
 
         else:
-            self.link = '/'.join(("d", self.info['pattern']['KEY'], str(self.get_checksum()), self.pyfile.name))
+            self.link = self.get_link()
 
-
-    def get_checksum(self):
+    def get_link(self):
         try:
-            b1 = eval(re.search(r'\.omg = (.+?);', self.html).group(1))
-            b2 = eval(re.search(r'\* \((.+?)\)', self.html).group(1))
-            checksum = b1 * b2 + 18
+            # get all the scripts inside the html body
+            soup = BeautifulSoup(self.html)
+            scripts = (s.getText() for s in soup.body.findAll('script', type='text/javascript'))
+
+            # meant to be populated with the initialization of all the DOM elements found in the scripts
+            initScripts = set()
+            
+            def replElementById(element):
+                id, attr = element.group(1), element.group(4)      # attr might be None
+            
+                varName = '%s_%s' %(id, attr)
+            
+                initValues = (elt.get(attr, None) for elt in soup.findAll(id=id))
+                initValues = [v for v in initValues if v is not None]
+                initValue = '"%s"' %initValues[-1] if initValues else 'null'
+            
+                initScripts.add('var %s = %s;' %(varName, initValue))
+                return varName
+            
+            # handle all getElementById
+            reVar = r'document.getElementById\([\'"](\w+)[\'"]\)(\.)?(getAttribute\([\'"])?(\w+)?([\'"]\))?'
+            scripts = [re.sub(reVar, replElementById, script) for script in scripts]
+            
+            # add try/catch in JS to handle deliberate errors
+            tryJS, catchJS =  u'try{', u'} catch(err){}' # '', '' to see where the script fails
+            scripts = ['\n'.join((tryJS, script, catchJS)) for script in scripts if script.strip()]
+
+            # get the file's url by evaluating all the scripts
+            scripts = '\n'.join(list(initScripts) + scripts + ['dlbutton_href'])
+            return self.js.eval(scripts)
 
         except Exception:
-            self.error(_("Unable to calculate checksum"))
-
-        else:
-            return checksum
-
+            self.error(_("Unable to calculate the link"))
 
 getInfo = create_getInfo(ZippyshareCom)