[jsinterp] Implement splice and general improvement

I still get 403s on YouTube though.
This commit is contained in:
Philipp Hagemeister 2014-07-24 10:41:12 +02:00
parent 8944ec0109
commit 825abb8175
1 changed files with 59 additions and 29 deletions

View File

@ -1,5 +1,6 @@
from __future__ import unicode_literals from __future__ import unicode_literals
import json
import re import re
from .utils import ( from .utils import (
@ -40,8 +41,9 @@ class JSInterpreter(object):
assign = lambda v: v assign = lambda v: v
expr = stmt[len('return '):] expr = stmt[len('return '):]
else: else:
raise ExtractorError( # Try interpreting it as an expression
'Cannot determine left side of statement in %r' % stmt) expr = stmt
assign = lambda v: v
v = self.interpret_expression(expr, local_vars, allow_recursion) v = self.interpret_expression(expr, local_vars, allow_recursion)
return assign(v) return assign(v)
@ -53,35 +55,62 @@ class JSInterpreter(object):
if expr.isalpha(): if expr.isalpha():
return local_vars[expr] return local_vars[expr]
m = re.match(r'^(?P<in>[a-z]+)\.(?P<member>.*)$', expr) try:
if m: return json.loads(expr)
member = m.group('member') except ValueError:
variable = m.group('in') pass
if variable not in local_vars: m = re.match(
r'^(?P<var>[a-z]+)\.(?P<member>[^(]+)(?:\(+(?P<args>[^()]*)\))?$',
expr)
if m:
variable = m.group('var')
member = m.group('member')
arg_str = m.group('args')
if variable in local_vars:
obj = local_vars[variable]
else:
if variable not in self._objects: if variable not in self._objects:
self._objects[variable] = self.extract_object(variable) self._objects[variable] = self.extract_object(variable)
obj = self._objects[variable] obj = self._objects[variable]
key, args = member.split('(', 1)
args = args.strip(')')
argvals = [int(v) if v.isdigit() else local_vars[v]
for v in args.split(',')]
return obj[key](argvals)
val = local_vars[variable] if arg_str is None:
if member == 'split("")': # Member access
return list(val)
if member == 'join("")':
return ''.join(val)
if member == 'length': if member == 'length':
return len(val) return len(obj)
if member == 'reverse()': return obj[member]
return val[::-1]
slice_m = re.match(r'slice\((?P<idx>.*)\)', member) assert expr.endswith(')')
if slice_m: # Function call
idx = self.interpret_expression( if arg_str == '':
slice_m.group('idx'), local_vars, allow_recursion - 1) argvals = tuple()
return val[idx:] else:
argvals = tuple([
self.interpret_expression(v, local_vars, allow_recursion)
for v in arg_str.split(',')])
if member == 'split':
assert argvals == ('',)
return list(obj)
if member == 'join':
assert len(argvals) == 1
return argvals[0].join(obj)
if member == 'reverse':
assert len(argvals) == 0
return obj[::-1]
if member == 'slice':
assert len(argvals) == 1
return obj[argvals[0]:]
if member == 'splice':
assert isinstance(obj, list)
index, howMany = argvals
res = []
for i in range(index, min(index + howMany, len(obj))):
res.append(obj.pop(i))
return res
return obj[member](argvals)
m = re.match( m = re.match(
r'^(?P<in>[a-z]+)\[(?P<idx>.+)\]$', expr) r'^(?P<in>[a-z]+)\[(?P<idx>.+)\]$', expr)
@ -100,13 +129,14 @@ class JSInterpreter(object):
return a % b return a % b
m = re.match( m = re.match(
r'^(?P<func>[a-zA-Z$]+)\((?P<args>[a-z0-9,]+)\)$', expr) r'^(?P<func>[.a-zA-Z$]+)\((?P<args>[a-z0-9,]+)\)$', expr)
if m: if m:
fname = m.group('func') fname = m.group('func')
argvals = tuple([
int(v) if v.isdigit() else local_vars[v]
for v in m.group('args').split(',')])
if fname not in self._functions: if fname not in self._functions:
self._functions[fname] = self.extract_function(fname) self._functions[fname] = self.extract_function(fname)
argvals = [int(v) if v.isdigit() else local_vars[v]
for v in m.group('args').split(',')]
return self._functions[fname](argvals) return self._functions[fname](argvals)
raise ExtractorError('Unsupported JS expression %r' % expr) raise ExtractorError('Unsupported JS expression %r' % expr)