summaryrefslogtreecommitdiff
path: root/crocoite/cli.py
blob: a97f0ce7e4c1d98e31e3bd628611a58ba3d26dc0 (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
# Copyright (c) 2017 crocoite contributors
# 
# Permission is hereby granted, free of charge, to any person obtaining a copy
# of this software and associated documentation files (the "Software"), to deal
# in the Software without restriction, including without limitation the rights
# to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
# copies of the Software, and to permit persons to whom the Software is
# furnished to do so, subject to the following conditions:
# 
# The above copyright notice and this permission notice shall be included in
# all copies or substantial portions of the Software.
# 
# THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
# IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
# FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
# AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
# LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
# OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
# THE SOFTWARE.

def main ():
    import os, random, logging, argparse
    from io import BytesIO
    import pychrome
    from urllib.parse import urlsplit
    from warcio.warcwriter import WARCWriter
    from warcio.statusandheaders import StatusAndHeaders
    from html5lib.serializer import HTMLSerializer
    from . import html, packageData, packageUrl
    from .warc import WarcLoader
    from .html import StripAttributeFilter, StripTagFilter, ChromeTreeWalker

    def randomString (length=None, chars='abcdefghijklmnopqrstuvwxyz'):
        if length is None:
            length = random.randint (16, 32)
        return ''.join (map (lambda x: random.choice (chars), range (length)))
    
    def getFormattedViewportMetrics (tab):
        layoutMetrics = tab.Page.getLayoutMetrics ()
        # XXX: I’m not entirely sure which one we should use here
        return '{}x{}'.format (layoutMetrics['layoutViewport']['clientWidth'],
                    layoutMetrics['layoutViewport']['clientHeight'])

    def writeDOMSnapshot (tab, writer):
        """
        Get a DOM snapshot of tab and write it to WARC.

        We could use DOMSnapshot.getSnapshot here, but the API is not stable
        yet. Also computed styles are not really necessary here.

        XXX: Currently writes a response, when it should use “resource”. pywb
        can’t handle that though.
        """
        viewport = getFormattedViewportMetrics (tab)
        dom = tab.DOM.getDocument (depth=-1, pierce=True)
        haveUrls = set ()
        for doc in ChromeTreeWalker (dom['root']).split ():
            rawUrl = doc['documentURL']
            if rawUrl in haveUrls:
                # ignore duplicate URLs. they are usually caused by
                # javascript-injected iframes (advertising) with no(?) src
                logger.warning ('have DOM snapshot for URL {}, ignoring'.format (rawUrl))
                continue
            url = urlsplit (rawUrl)
            if url.scheme in ('http', 'https'):
                logger.debug ('saving DOM snapshot for url {}, base {}'.format (doc['documentURL'], doc['baseURL']))
                haveUrls.add (rawUrl)
                walker = ChromeTreeWalker (doc)
                # remove script, to make the page static and noscript, because at the
                # time we took the snapshot scripts were enabled
                disallowedTags = ['script', 'noscript']
                disallowedAttributes = html.eventAttributes
                stream = StripAttributeFilter (StripTagFilter (walker, disallowedTags), disallowedAttributes)
                serializer = HTMLSerializer ()
                httpHeaders = StatusAndHeaders('200 OK', {}, protocol='HTTP/1.1')
                record = writer.create_warc_record (doc['documentURL'], 'response',
                        payload=BytesIO (serializer.render (stream, 'utf8')),
                        http_headers=httpHeaders,
                        warc_headers_dict={'X-DOM-Snapshot': str (True),
                                'X-Chrome-Viewport': viewport})
                writer.write_record (record)

    def emulateScreenMetrics (l):
        """
        Emulate different screen sizes, causing the site to fetch assets (img
        srcset and css, for example) for different screen resolutions.
        """
        cssPpi = 96
        sizes = [
                {'width': 1920, 'height': 1080, 'deviceScaleFactor': 1.5, 'mobile': False},
                {'width': 1920, 'height': 1080, 'deviceScaleFactor': 2, 'mobile': False},
                # very dense display
                {'width': 1920, 'height': 1080, 'deviceScaleFactor': 4, 'mobile': False},
                # just a few samples:
                # 1st gen iPhone (portrait mode)
                {'width': 320, 'height': 480, 'deviceScaleFactor': 163/cssPpi, 'mobile': True},
                # 6th gen iPhone (portrait mode)
                {'width': 750, 'height': 1334, 'deviceScaleFactor': 326/cssPpi, 'mobile': True},
                # and reset
                {'width': 1920, 'height': 1080, 'deviceScaleFactor': 1, 'mobile': False},
                ]
        for s in sizes:
            l.tab.Emulation.setDeviceMetricsOverride (**s)
            l.wait (1)
        # XXX: this seems to be broken, it does not clear the override
        #tab.Emulation.clearDeviceMetricsOverride ()
        # wait until assets finished loading
        l.waitIdle (2, 10)

    def loadScripts (paths, scripts=[]):
        for p in paths:
            if not os.path.exists (p):
                # search for defaults scripts in package data directory
                p = packageData (p)
            with open (p, 'r') as fd:
                scripts.append (fd.read ())
        return '\n'.join (scripts)

    def writeScript (path, source, writer):
        record = writer.create_warc_record (packageUrl (path), 'metadata',
                payload=BytesIO (source.encode ('utf8')),
                warc_headers_dict={'Content-Type': 'application/javascript; charset=utf-8'})
        writer.write_record (record)

    logger = logging.getLogger(__name__)
    logging.basicConfig (level=logging.DEBUG)

    parser = argparse.ArgumentParser(description='Save website to WARC using Google Chrome.')
    parser.add_argument('--browser', default='http://127.0.0.1:9222', help='DevTools URL')
    parser.add_argument('--timeout', default=10, type=int, help='Maximum time for archival')
    parser.add_argument('--idle-timeout', default=2, type=int, help='Maximum idle seconds (i.e. no requests)', dest='idleTimeout')
    parser.add_argument('--log-buffer', default=1000, type=int, dest='logBuffer')
    parser.add_argument('--max-body-size', default=10*1024*1024, type=int, dest='maxBodySize', help='Max body size in bytes')
    #parser.add_argument('--keep-tab', action='store_true', default=False, dest='keepTab', help='Keep tab open')
    parser.add_argument('--onload', default=[], action='append', help='Inject JavaScript file before loading page')
    parser.add_argument('--onsnapshot', default=[], action='append', help='Run JavaScript files before creating DOM snapshot')
    parser.add_argument('url', help='Website URL')
    parser.add_argument('output', help='WARC filename')

    args = parser.parse_args ()

    stopVarname = '__' + __package__ + '_stop__'
    # avoid sites messing with our scripts by using a random stop variable name
    newStopVarname = randomString ()
    onload = loadScripts (args.onload, ['var {} = false;\n'.format (stopVarname)]).replace (stopVarname, newStopVarname)
    stopVarname = newStopVarname

    browser = pychrome.Browser(url=args.browser)

    fd = open (args.output, 'wb')
    writer = WARCWriter (fd, gzip=True)

    with WarcLoader (browser, args.url, writer, logBuffer=args.logBuffer,
            maxBodySize=args.maxBodySize) as l:
        version = l.tab.Browser.getVersion ()
        payload = {
                'software': __package__,
                'browser': version['product'],
                'useragent': version['userAgent'],
                'viewport': getFormattedViewportMetrics (l.tab),
                }
        warcinfo = writer.create_warcinfo_record (filename=None, info=payload)
        writer.write_record (warcinfo)
        # save onload script as well
        writeScript ('onload', onload, writer)

        # inject our custom javascript to the page before loading
        l.tab.Page.addScriptToEvaluateOnNewDocument (source=onload)
        l.start ()

        l.waitIdle (args.idleTimeout, args.timeout)

        # get ready for snapshot: stop loading and scripts, disable events
        l.tab.Runtime.evaluate (expression='{} = true; window.scrollTo (0, 0);'.format (stopVarname), returnByValue=True)
        # if we stopped due to timeout, wait for remaining assets
        l.waitIdle (2, 10)

        emulateScreenMetrics (l)

        l.stop ()

        script = loadScripts (args.onsnapshot)
        writeScript ('onsnapshot', script, writer)
        l.tab.Runtime.evaluate (expression=script, returnByValue=True)
        writeDOMSnapshot (l.tab, writer)

    return True