|
1
|
|
|
#! /usr/bin/python3 |
|
2
|
|
|
# vim: set expandtab tabstop=4 shiftwidth=4 : |
|
3
|
|
|
"""Module with functions wrapping urllib""" |
|
4
|
|
|
|
|
5
|
|
|
import http.client |
|
6
|
|
|
import urllib.request |
|
7
|
|
|
import urllib.parse |
|
8
|
|
|
import json |
|
9
|
|
|
import shutil |
|
10
|
|
|
import gzip |
|
11
|
|
|
from bs4 import BeautifulSoup |
|
12
|
|
|
import inspect |
|
13
|
|
|
import logging |
|
14
|
|
|
import time |
|
15
|
|
|
|
|
16
|
|
|
|
|
17
|
|
|
def log(string): |
|
18
|
|
|
"""Dirty logging function.""" |
|
19
|
|
|
# TODO: https://docs.python.org/2/library/logging.html#logrecord-attributes |
|
20
|
|
|
# we do not need to retrieve the function name manually |
|
21
|
|
|
logging.debug(inspect.stack()[1][3] + " " + string) |
|
22
|
|
|
|
|
23
|
|
|
|
|
24
|
|
|
def convert_iri_to_plain_ascii_uri(uri): |
|
25
|
|
|
"""Convert IRI to plain ASCII URL |
|
26
|
|
|
Based on http://stackoverflow.com/questions/4389572/how-to-fetch-a-non-ascii-url-with-python-urlopen.""" |
|
27
|
|
|
lis = list(urllib.parse.urlsplit(uri)) |
|
28
|
|
|
lis[2] = urllib.parse.quote(lis[2]) |
|
29
|
|
|
url = urllib.parse.urlunsplit(lis) |
|
30
|
|
|
if False and url != uri: |
|
31
|
|
|
print(uri, '->', url) |
|
32
|
|
|
return url |
|
33
|
|
|
|
|
34
|
|
|
|
|
35
|
|
|
def urlopen_wrapper(url, referer=None): |
|
36
|
|
|
"""Wrapper around urllib.request.urlopen (user-agent, etc). |
|
37
|
|
|
|
|
38
|
|
|
url is a string |
|
39
|
|
|
referer is an optional string |
|
40
|
|
|
Returns a byte object.""" |
|
41
|
|
|
log('(url : %s)' % url) |
|
42
|
|
|
user_agent = 'Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/534.30 (KHTML, like Gecko) Ubuntu/11.04 Chromium/12.0.742.112 Chrome/12.0.742.112 Safari/534.30' |
|
43
|
|
|
try: |
|
44
|
|
|
req = urllib.request.Request(url, headers={'User-Agent': user_agent, 'Accept': '*/*'}) |
|
45
|
|
|
if referer: |
|
46
|
|
|
req.add_header('Referer', referer) |
|
47
|
|
|
response = urllib.request.urlopen(req) |
|
48
|
|
|
if response.info().get('Content-Encoding') == 'gzip': |
|
49
|
|
|
return gzip.GzipFile(fileobj=response) |
|
50
|
|
|
return response |
|
51
|
|
|
except (urllib.error.HTTPError, urllib.error.URLError) as e: |
|
52
|
|
|
print("Exception %s for url %s" % (e, url)) |
|
53
|
|
|
raise |
|
54
|
|
|
|
|
55
|
|
|
|
|
56
|
|
|
def urljoin_wrapper(base, url): |
|
57
|
|
|
"""Wrapper around urllib.parse.urljoin. |
|
58
|
|
|
Construct a full ("absolute") URL by combining a "base URL" (base) with |
|
59
|
|
|
another URL (url).""" |
|
60
|
|
|
return urllib.parse.urljoin(base, url) |
|
61
|
|
|
|
|
62
|
|
|
|
|
63
|
|
|
def get_content(url): |
|
64
|
|
|
"""Get content at url. |
|
65
|
|
|
|
|
66
|
|
|
url is a string |
|
67
|
|
|
Returns a string""" |
|
68
|
|
|
log('(url : %s)' % url) |
|
69
|
|
|
try: |
|
70
|
|
|
return urlopen_wrapper(url).read() |
|
71
|
|
|
except http.client.IncompleteRead as e: |
|
72
|
|
|
print("%s for %s" % (e, url)) |
|
73
|
|
|
return e.partial |
|
74
|
|
|
|
|
75
|
|
|
|
|
76
|
|
|
def extensions_are_equivalent(ext1, ext2): |
|
77
|
|
|
"""Return whether file extensions can be considered as equivalent.""" |
|
78
|
|
|
synonyms = [{'jpg', 'jpeg'}] |
|
79
|
|
|
ext1, ext2 = ext1.lower(), ext2.lower() |
|
80
|
|
|
return ext1 == ext2 or any((ext1 in s and ext2 in s) for s in synonyms) |
|
81
|
|
|
|
|
82
|
|
|
|
|
83
|
|
|
def add_extension_to_filename_if_needed(ext, filename): |
|
84
|
|
|
"""Given an extension and a filename, add the extension to the filename |
|
85
|
|
|
if the filename does not already have this extension (or an extension |
|
86
|
|
|
considered to be equivalent.""" |
|
87
|
|
|
filename_ext = filename.split('.')[-1] |
|
88
|
|
|
if extensions_are_equivalent(ext, filename_ext): |
|
89
|
|
|
return filename |
|
90
|
|
|
else: |
|
91
|
|
|
return filename + '.' + ext |
|
92
|
|
|
|
|
93
|
|
|
|
|
94
|
|
|
def get_file_at_url(url, path, referer=None): |
|
95
|
|
|
"""Save content at url in path on file system. |
|
96
|
|
|
In theory, this could have been achieved with urlretrieve but it seems |
|
97
|
|
|
to be about to get deprecated and adding a user-agent seems to be quite |
|
98
|
|
|
awkward. |
|
99
|
|
|
|
|
100
|
|
|
url is a string |
|
101
|
|
|
path is a string corresponding to the file location |
|
102
|
|
|
referer is an optional string |
|
103
|
|
|
Returns the path if the file is retrieved properly, None otherwise.""" |
|
104
|
|
|
log('(url : %s, path : %s)' % (url, path)) |
|
105
|
|
|
try: |
|
106
|
|
|
with urlopen_wrapper(url, referer) as response: |
|
107
|
|
|
content_type = response.info().get('Content-Type', '').split('/') |
|
108
|
|
|
assert 1 <= len(content_type) <= 2 |
|
109
|
|
|
if len(content_type) == 2: |
|
110
|
|
|
data = content_type[1].split(';') |
|
111
|
|
|
path = add_extension_to_filename_if_needed(data[0], path) |
|
112
|
|
|
with open(path, 'wb') as out_file: |
|
113
|
|
|
shutil.copyfileobj(response, out_file) |
|
114
|
|
|
time.sleep(0.4) |
|
115
|
|
|
return path |
|
116
|
|
|
except (urllib.error.HTTPError, urllib.error.URLError): |
|
117
|
|
|
return None |
|
118
|
|
|
|
|
119
|
|
|
|
|
120
|
|
|
def get_filename_from_url(url): |
|
121
|
|
|
"""Get filename from url |
|
122
|
|
|
|
|
123
|
|
|
url is a string |
|
124
|
|
|
Returns a string corresponding to the name of the file.""" |
|
125
|
|
|
return urllib.parse.unquote(url).split('/')[-1] |
|
126
|
|
|
|
|
127
|
|
|
|
|
128
|
|
|
def load_json_at_url(url): |
|
129
|
|
|
"""Get content at url as JSON and return it.""" |
|
130
|
|
|
return json.loads(get_content(url).decode()) |
|
131
|
|
|
|
|
132
|
|
|
|
|
133
|
|
|
def get_soup_at_url(url, detect_meta=False, detect_rel=False, detect_angular=False, save_in_file=False): |
|
134
|
|
|
"""Get content at url as BeautifulSoup. |
|
135
|
|
|
|
|
136
|
|
|
url is a string |
|
137
|
|
|
detect_meta is a hacky flag used to detect comics using similar plugin to |
|
138
|
|
|
be able to reuse code at some point |
|
139
|
|
|
detect_rel is a hacky flag to detect next/first comics automatically |
|
140
|
|
|
detect_rel is a hacky flag to detect if page corresponds to an Angular app |
|
141
|
|
|
save_in_file is a hacky flag to save content in temp file for debugging |
|
142
|
|
|
purposes |
|
143
|
|
|
Returns a BeautifulSoup object.""" |
|
144
|
|
|
time.sleep(0.4) |
|
145
|
|
|
content = get_content(url) |
|
146
|
|
|
soup = BeautifulSoup(content, "html.parser") |
|
147
|
|
|
if detect_meta: |
|
148
|
|
|
for meta_val in ['generator', 'ComicPress', 'Comic-Easel']: |
|
149
|
|
|
meta = soup.find('meta', attrs={'name': meta_val}) |
|
150
|
|
|
if meta is not None: |
|
151
|
|
|
print(meta) |
|
152
|
|
|
if detect_rel: |
|
153
|
|
|
for tag in ['a', 'link']: |
|
154
|
|
|
next_ = soup.find(tag, rel='next') |
|
155
|
|
|
if next_ is not None: |
|
156
|
|
|
print(next_) |
|
157
|
|
|
if detect_angular: |
|
158
|
|
|
html = soup.find('html') |
|
159
|
|
|
if html.has_attr('ng-app'): |
|
160
|
|
|
print(url) |
|
161
|
|
|
if save_in_file: |
|
162
|
|
|
time_ms = time.time() * 1000 |
|
163
|
|
|
prefix = 'get_soup_at_url_' + str(time_ms) + '_' |
|
164
|
|
|
with open(prefix + 'raw', 'wb') as f: |
|
165
|
|
|
f.write(content) |
|
166
|
|
|
with open(prefix + 'content', 'wb') as f: |
|
167
|
|
|
f.write(soup.encode('utf-8')) |
|
168
|
|
|
return soup |
|
169
|
|
|
|
SylvainDe /
ComicBookMaker
