summaryrefslogtreecommitdiffstats
path: root/tools/site_compare/scrapers/firefox/firefox2.py
blob: fa0d6201200ef472941030d48dc1f10bd09afb36 (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
#!/usr/bin/python2.4
# Copyright (c) 2006-2008 The Chromium Authors. All rights reserved.
# Use of this source code is governed by a BSD-style license that can be
# found in the LICENSE file.

"""Does scraping for Firefox 2.0."""

import pywintypes
import time
import types

from drivers import keyboard
from drivers import mouse
from drivers import windowing

# Default version
version = "2.0.0.6"

DEFAULT_PATH = r"c:\program files\mozilla firefox\firefox.exe"

# TODO(jhaas): the Firefox scraper is a bit rickety at the moment. Known
# issues: 1) won't work if the default profile puts toolbars in different
# locations, 2) uses sleep() statements rather than more robust checks,
# 3) fails badly if an existing Firefox window is open when the scrape
# is invoked. This needs to be fortified at some point.

def GetBrowser(path):
  """Invoke the Firefox browser and return the process and window.

  Args:
    path: full path to browser

  Returns:
    A tuple of (process handle, render pane)
  """
  if not path: path = DEFAULT_PATH

  # Invoke Firefox
  (proc, wnd) = windowing.InvokeAndWait(path)

  # Get the content pane
  render_pane = windowing.FindChildWindow(
    wnd,
    "MozillaWindowClass/MozillaWindowClass/MozillaWindowClass")

  return (proc, wnd, render_pane)


def InvokeBrowser(path):
  """Invoke the Firefox browser.

  Args:
    path: full path to browser

  Returns:
    A tuple of (main window, process handle, render pane)
  """
  # Reuse an existing instance of the browser if we can find one. This
  # may not work correctly, especially if the window is behind other windows.
  wnds = windowing.FindChildWindows(0, "MozillaUIWindowClass")
  if len(wnds):
    wnd = wnds[0]
    proc = None
  else:
    # Invoke Firefox
    (proc, wnd) = windowing.InvokeAndWait(path)

  # Get the content pane
  render_pane = windowing.FindChildWindow(
    wnd,
    "MozillaWindowClass/MozillaWindowClass/MozillaWindowClass")

  return (wnd, proc, render_pane)


def Scrape(urls, outdir, size, pos, timeout=20, **kwargs):
  """Invoke a browser, send it to a series of URLs, and save its output.

  Args:
    urls: list of URLs to scrape
    outdir: directory to place output
    size: size of browser window to use
    pos: position of browser window
    timeout: amount of time to wait for page to load
    kwargs: miscellaneous keyword args

  Returns:
    None if success, else an error string
  """
  if "path" in kwargs and kwargs["path"]: path = kwargs["path"]
  else: path = DEFAULT_PATH

  (wnd, proc, render_pane) = InvokeBrowser(path)

  # Resize and reposition the frame
  windowing.MoveAndSizeWindow(wnd, pos, size, render_pane)

  time.sleep(3)

  # Firefox is a bit of a pain: it doesn't use standard edit controls,
  # and it doesn't display a throbber when there's no tab. Let's make
  # sure there's at least one tab, then select the first one

  mouse.ClickInWindow(wnd)
  keyboard.TypeString("[t]", True)
  mouse.ClickInWindow(wnd, (30, 115))
  time.sleep(2)

  timedout = False

  # Visit each URL we're given
  if type(urls) in types.StringTypes: urls = [urls]

  for url in urls:

    # Use keyboard shortcuts
    keyboard.TypeString("{d}", True)
    keyboard.TypeString(url)
    keyboard.TypeString("\n")

    # Wait for the page to finish loading
    load_time = windowing.WaitForThrobber(wnd, (10, 96, 26, 112), timeout)
    timedout = load_time < 0

    if timedout:
      break

    # Scrape the page
    image = windowing.ScrapeWindow(render_pane)

    # Save to disk
    if "filename" in kwargs:
      if callable(kwargs["filename"]):
        filename = kwargs["filename"](url)
      else:
        filename = kwargs["filename"]
    else:
      filename = windowing.URLtoFilename(url, outdir, ".bmp")
    image.save(filename)

  # Close all the tabs, cheesily
  mouse.ClickInWindow(wnd)

  while len(windowing.FindChildWindows(0, "MozillaUIWindowClass")):
    keyboard.TypeString("[w]", True)
    time.sleep(1)

  if timedout:
    return "timeout"


def Time(urls, size, timeout, **kwargs):
  """Measure how long it takes to load each of a series of URLs

  Args:
    urls: list of URLs to time
    size: size of browser window to use
    timeout: amount of time to wait for page to load
    kwargs: miscellaneous keyword args

  Returns:
    A list of tuples (url, time). "time" can be "crashed" or "timeout"
  """
  if "path" in kwargs and kwargs["path"]: path = kwargs["path"]
  else: path = DEFAULT_PATH
  proc = None

  # Visit each URL we're given
  if type(urls) in types.StringTypes: urls = [urls]

  ret = []
  for url in urls:
    try:
      # Invoke the browser if necessary
      if not proc:
        (wnd, proc, render_pane) = InvokeBrowser(path)

        # Resize and reposition the frame
        windowing.MoveAndSizeWindow(wnd, (0,0), size, render_pane)

        time.sleep(3)

        # Firefox is a bit of a pain: it doesn't use standard edit controls,
        # and it doesn't display a throbber when there's no tab. Let's make
        # sure there's at least one tab, then select the first one

        mouse.ClickInWindow(wnd)
        keyboard.TypeString("[t]", True)
        mouse.ClickInWindow(wnd, (30, 115))
        time.sleep(2)

      # Use keyboard shortcuts
      keyboard.TypeString("{d}", True)
      keyboard.TypeString(url)
      keyboard.TypeString("\n")

      # Wait for the page to finish loading
      load_time = windowing.WaitForThrobber(wnd, (10, 96, 26, 112), timeout)
      timedout = load_time < 0

      if timedout:
        load_time = "timeout"

        # Try to close the browser; if this fails it's probably a crash
        mouse.ClickInWindow(wnd)

        count = 0
        while (len(windowing.FindChildWindows(0, "MozillaUIWindowClass"))
          and count < 5):
          keyboard.TypeString("[w]", True)
          time.sleep(1)
          count = count + 1

        if len(windowing.FindChildWindows(0, "MozillaUIWindowClass")):
          windowing.EndProcess(proc)
          load_time = "crashed"

        proc = None
    except pywintypes.error:
      proc = None
      load_time = "crashed"

    ret.append( (url, load_time) )

  if proc:
    count = 0
    while (len(windowing.FindChildWindows(0, "MozillaUIWindowClass"))
      and count < 5):
      keyboard.TypeString("[w]", True)
      time.sleep(1)
      count = count + 1
  return ret


if __name__ == "__main__":
  # We're being invoked rather than imported, so run some tests
  path = r"c:\sitecompare\scrapes\Firefox\2.0.0.6"
  windowing.PreparePath(path)

  # Scrape three sites and save the results
  Scrape(
    ["http://www.microsoft.com", "http://www.google.com",
     "http://www.sun.com"],
    path, (1024, 768), (0, 0))