1#!/usr/bin/env python
2# Copyright 2013 Google Inc. All Rights Reserved.
3#
4# Licensed under the Apache License, Version 2.0 (the "License");
5# you may not use this file except in compliance with the License.
6# You may obtain a copy of the License at
7#
8#      http://www.apache.org/licenses/LICENSE-2.0
9#
10# Unless required by applicable law or agreed to in writing, software
11# distributed under the License is distributed on an "AS IS" BASIS,
12# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
13# See the License for the specific language governing permissions and
14# limitations under the License.
15
16"""Inject javascript into html page source code."""
17
18import datetime
19import logging
20import os
21import re
22import util
23import third_party.jsmin as jsmin
24
25DOCTYPE_RE = re.compile(r'^.{,256}?(<!--.*-->)?.{,256}?<!doctype html>',
26                        re.IGNORECASE | re.DOTALL)
27HTML_RE = re.compile(r'^.{,256}?(<!--.*-->)?.{,256}?<html.*?>',
28                     re.IGNORECASE | re.DOTALL)
29HEAD_RE = re.compile(r'^.{,256}?(<!--.*-->)?.{,256}?<head.*?>',
30                     re.IGNORECASE | re.DOTALL)
31
32# Occurences of this marker in injected scripts will be replaced with
33# recording time in javascripts' Date().toValue() format.  This allows
34# to properly set deterministic date in JS code.  See
35# https://github.com/chromium/web-page-replay/issues/71 for details.
36TIME_SEED_MARKER = '{{WPR_TIME_SEED_TIMESTAMP}}'
37
38
39def GetScriptInjector(scripts):
40  """Loads |scripts| from disk and returns an injector of their content."""
41  lines = []
42  if scripts:
43    if not isinstance(scripts, list):
44      scripts = scripts.split(',')
45    for script in scripts:
46      if os.path.exists(script):
47        with open(script) as f:
48          lines.extend(f.read())
49      elif util.resource_exists(script):
50        lines.extend(util.resource_string(script))
51      else:
52        raise Exception('Script does not exist: %s', script)
53
54  script_template = jsmin.jsmin(''.join(lines), quote_chars="'\"`")
55  def injector(record_time):
56    delta = record_time - datetime.datetime(1970, 1, 1)
57    js_timestamp = \
58        int(delta.total_seconds()) * 1000 + delta.microseconds / 1000
59    return script_template.replace(TIME_SEED_MARKER, str(js_timestamp))
60  return injector
61
62
63def _IsHtmlContent(content):
64  content = content.strip()
65  return content.startswith('<') and content.endswith('>')
66
67
68def InjectScript(text_chunks, content_type, script_to_inject):
69  """Inject |script_to_inject| into |content| if |content_type| is 'text/html'.
70
71  Inject |script_to_inject| into |text_chunks| immediately after <head>,
72  <html> or <!doctype html>, if one of them is found. Otherwise, inject at
73  the beginning.
74
75  Returns:
76    text_chunks, already_injected
77    |text_chunks| is the new content if script is injected, otherwise
78      the original.  If the script was injected, exactly one chunk in
79      |text_chunks| will have changed.
80    |just_injected| indicates if |script_to_inject| was just injected in
81      the content.
82  """
83  if not content_type or content_type != 'text/html':
84    return text_chunks, False
85  content = "".join(text_chunks)
86  if not content or not _IsHtmlContent(content) or script_to_inject in content:
87    return text_chunks, False
88  for regexp in (HEAD_RE, HTML_RE, DOCTYPE_RE):
89    matchobj = regexp.search(content)
90    if matchobj:
91      pos = matchobj.end(0)
92      for i, chunk in enumerate(text_chunks):
93        if pos <= len(chunk):
94          result = text_chunks[:]
95          result[i] = '%s<script>%s</script>%s' % (chunk[0:pos],
96                                                   script_to_inject,
97                                                   chunk[pos:])
98          return result, True
99        pos -= len(chunk)
100  result = text_chunks[:]
101  result[0] = '<script>%s</script>%s' % (script_to_inject,
102                                         text_chunks[0])
103  logging.warning('Inject at the very beginning, because no tag of '
104                  '<head>, <html> or <!doctype html> is found.')
105  return result, True
106