1 from __future__ import unicode_literals
16 ('>>', operator.rshift),
17 ('<<', operator.lshift),
21 ('/', operator.truediv),
24 _ASSIGN_OPERATORS = [(op + '=', opfunc) for op, opfunc in _OPERATORS]
25 _ASSIGN_OPERATORS.append(('=', lambda cur, right: right))
27 _NAME_RE = r'[a-zA-Z_$][a-zA-Z_$0-9]*'
30 class JSInterpreter(object):
31 def __init__(self, code, objects=None):
36 self._objects = objects
38 def interpret_statement(self, stmt, local_vars, allow_recursion=100):
39 if allow_recursion < 0:
40 raise ExtractorError('Recursion limit reached')
44 stmt_m = re.match(r'var\s', stmt)
46 expr = stmt[len(stmt_m.group(0)):]
48 return_m = re.match(r'return(?:\s+|$)', stmt)
50 expr = stmt[len(return_m.group(0)):]
53 # Try interpreting it as an expression
56 v = self.interpret_expression(expr, local_vars, allow_recursion)
57 return v, should_abort
59 def interpret_expression(self, expr, local_vars, allow_recursion):
61 if expr == '': # Empty expression
64 if expr.startswith('('):
66 for m in re.finditer(r'[()]', expr):
72 sub_expr = expr[1:m.start()]
73 sub_result = self.interpret_expression(
74 sub_expr, local_vars, allow_recursion)
75 remaining_expr = expr[m.end():].strip()
76 if not remaining_expr:
79 expr = json.dumps(sub_result) + remaining_expr
82 raise ExtractorError('Premature end of parens in %r' % expr)
84 for op, opfunc in _ASSIGN_OPERATORS:
86 (?P<out>%s)(?:\[(?P<index>[^\]]+?)\])?
88 (?P<expr>.*)$''' % (_NAME_RE, re.escape(op)), expr)
91 right_val = self.interpret_expression(
92 m.group('expr'), local_vars, allow_recursion - 1)
94 if m.groupdict().get('index'):
95 lvar = local_vars[m.group('out')]
96 idx = self.interpret_expression(
97 m.group('index'), local_vars, allow_recursion)
98 assert isinstance(idx, int)
100 val = opfunc(cur, right_val)
104 cur = local_vars.get(m.group('out'))
105 val = opfunc(cur, right_val)
106 local_vars[m.group('out')] = val
113 r'(?!if|return|true|false)(?P<name>%s)$' % _NAME_RE,
116 return local_vars[var_m.group('name')]
119 return json.loads(expr)
124 r'(?P<in>%s)\[(?P<idx>.+)\]$' % _NAME_RE, expr)
126 val = local_vars[m.group('in')]
127 idx = self.interpret_expression(
128 m.group('idx'), local_vars, allow_recursion - 1)
132 r'(?P<var>%s)(?:\.(?P<member>[^(]+)|\[(?P<member2>[^]]+)\])\s*(?:\(+(?P<args>[^()]*)\))?$' % _NAME_RE,
135 variable = m.group('var')
136 member = remove_quotes(m.group('member') or m.group('member2'))
137 arg_str = m.group('args')
139 if variable in local_vars:
140 obj = local_vars[variable]
142 if variable not in self._objects:
143 self._objects[variable] = self.extract_object(variable)
144 obj = self._objects[variable]
148 if member == 'length':
152 assert expr.endswith(')')
158 self.interpret_expression(v, local_vars, allow_recursion)
159 for v in arg_str.split(',')])
161 if member == 'split':
162 assert argvals == ('',)
165 assert len(argvals) == 1
166 return argvals[0].join(obj)
167 if member == 'reverse':
168 assert len(argvals) == 0
171 if member == 'slice':
172 assert len(argvals) == 1
173 return obj[argvals[0]:]
174 if member == 'splice':
175 assert isinstance(obj, list)
176 index, howMany = argvals
178 for i in range(index, min(index + howMany, len(obj))):
179 res.append(obj.pop(index))
182 return obj[member](argvals)
184 for op, opfunc in _OPERATORS:
185 m = re.match(r'(?P<x>.+?)%s(?P<y>.+)' % re.escape(op), expr)
188 x, abort = self.interpret_statement(
189 m.group('x'), local_vars, allow_recursion - 1)
191 raise ExtractorError(
192 'Premature left-side return of %s in %r' % (op, expr))
193 y, abort = self.interpret_statement(
194 m.group('y'), local_vars, allow_recursion - 1)
196 raise ExtractorError(
197 'Premature right-side return of %s in %r' % (op, expr))
201 r'^(?P<func>%s)\((?P<args>[a-zA-Z0-9_$,]*)\)$' % _NAME_RE, expr)
203 fname = m.group('func')
205 int(v) if v.isdigit() else local_vars[v]
206 for v in m.group('args').split(',')]) if len(m.group('args')) > 0 else tuple()
207 if fname not in self._functions:
208 self._functions[fname] = self.extract_function(fname)
209 return self._functions[fname](argvals)
211 raise ExtractorError('Unsupported JS expression %r' % expr)
213 def extract_object(self, objname):
214 _FUNC_NAME_RE = r'''(?:[a-zA-Z$0-9]+|"[a-zA-Z$0-9]+"|'[a-zA-Z$0-9]+')'''
218 (?<!this\.)%s\s*=\s*{\s*
219 (?P<fields>(%s\s*:\s*function\s*\(.*?\)\s*{.*?}(?:,\s*)?)*)
221 ''' % (re.escape(objname), _FUNC_NAME_RE),
223 fields = obj_m.group('fields')
224 # Currently, it only supports function definitions
225 fields_m = re.finditer(
227 (?P<key>%s)\s*:\s*function\s*\((?P<args>[a-z,]+)\){(?P<code>[^}]+)}
231 argnames = f.group('args').split(',')
232 obj[remove_quotes(f.group('key'))] = self.build_function(argnames, f.group('code'))
236 def extract_function(self, funcname):
239 (?:function\s+%s|[{;,]\s*%s\s*=\s*function|var\s+%s\s*=\s*function)\s*
240 \((?P<args>[^)]*)\)\s*
241 \{(?P<code>[^}]+)\}''' % (
242 re.escape(funcname), re.escape(funcname), re.escape(funcname)),
245 raise ExtractorError('Could not find JS function %r' % funcname)
246 argnames = func_m.group('args').split(',')
248 return self.build_function(argnames, func_m.group('code'))
250 def call_function(self, funcname, *args):
251 f = self.extract_function(funcname)
254 def build_function(self, argnames, code):
256 local_vars = dict(zip(argnames, args))
257 for stmt in code.split(';'):
258 res, abort = self.interpret_statement(stmt, local_vars)