debian/control: Update list of supported sites.

[youtubedl] / youtube_dl / jsinterp.py

diff --git a/youtube_dl/jsinterp.py b/youtube_dl/jsinterp.py
index 453e2732cc4faa453a98b153356c2188feef1d35..24cdec28c6cb2332232212d6bcf39d03edc27c7a 100644 (file)
--- a/youtube_dl/jsinterp.py
+++ b/youtube_dl/jsinterp.py
@@ -30,13 +30,10 @@ class JSInterpreter(object):
     def __init__(self, code, objects=None):
         if objects is None:
             objects = {}
-        self.code = self._remove_comments(code)
+        self.code = code
         self._functions = {}
         self._objects = objects
 
-    def _remove_comments(self, code):
-        return re.sub(r'(?s)/\*.*?\*/', '', code)
-
     def interpret_statement(self, stmt, local_vars, allow_recursion=100):
         if allow_recursion < 0:
             raise ExtractorError('Recursion limit reached')
@@ -201,12 +198,12 @@ class JSInterpreter(object):
             return opfunc(x, y)
 
         m = re.match(
-            r'^(?P<func>%s)\((?P<args>[a-zA-Z0-9_$,]+)\)$' % _NAME_RE, expr)
+            r'^(?P<func>%s)\((?P<args>[a-zA-Z0-9_$,]*)\)$' % _NAME_RE, expr)
         if m:
             fname = m.group('func')
             argvals = tuple([
                 int(v) if v.isdigit() else local_vars[v]
-                for v in m.group('args').split(',')])
+                for v in m.group('args').split(',')]) if len(m.group('args')) > 0 else tuple()
             if fname not in self._functions:
                 self._functions[fname] = self.extract_function(fname)
             return self._functions[fname](argvals)
@@ -216,8 +213,8 @@ class JSInterpreter(object):
     def extract_object(self, objname):
         obj = {}
         obj_m = re.search(
-            (r'(?:var\s+)?%s\s*=\s*\{' % re.escape(objname)) +
-            r'\s*(?P<fields>([a-zA-Z$0-9]+\s*:\s*function\(.*?\)\s*\{.*?\})*)' +
+            (r'(?<!this\.)%s\s*=\s*\{' % re.escape(objname)) +
+            r'\s*(?P<fields>([a-zA-Z$0-9]+\s*:\s*function\(.*?\)\s*\{.*?\}(?:,\s*)?)*)' +
             r'\}\s*;',
             self.code)
         fields = obj_m.group('fields')
@@ -235,10 +232,10 @@ class JSInterpreter(object):
     def extract_function(self, funcname):
         func_m = re.search(
             r'''(?x)
-                (?:function\s+%s|[{;]%s\s*=\s*function)\s*
+                (?:function\s+%s|[{;,]\s*%s\s*=\s*function|var\s+%s\s*=\s*function)\s*
                 \((?P<args>[^)]*)\)\s*
                 \{(?P<code>[^}]+)\}''' % (
-                re.escape(funcname), re.escape(funcname)),
+                re.escape(funcname), re.escape(funcname), re.escape(funcname)),
             self.code)
         if func_m is None:
             raise ExtractorError('Could not find JS function %r' % funcname)