in progress version of script
Browse files- corpuscrawler-irish.py +615 -0
corpuscrawler-irish.py
ADDED
@@ -0,0 +1,615 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
# coding=utf-8
|
2 |
+
# Copyright 2021 The TensorFlow Datasets Authors and the HuggingFace Datasets Authors.
|
3 |
+
# Copyright 2021 Phonetics and Speech Laboratory, Trinity College, Dublin
|
4 |
+
#
|
5 |
+
# Based on Corpus Crawler (utils.py):
|
6 |
+
# Copyright 2017 Google Inc. All rights reserved.
|
7 |
+
#
|
8 |
+
# Based on Corpus Crawler's Irish crawler (crawl_ga.py):
|
9 |
+
# Copyright 2017 Google Inc. All rights reserved.
|
10 |
+
# Copyright 2017 Jim O'Regan
|
11 |
+
#
|
12 |
+
# Licensed under the Apache License, Version 2.0 (the "License");
|
13 |
+
# you may not use this file except in compliance with the License.
|
14 |
+
# You may obtain a copy of the License at
|
15 |
+
#
|
16 |
+
# http://www.apache.org/licenses/LICENSE-2.0
|
17 |
+
#
|
18 |
+
# Unless required by applicable law or agreed to in writing, software
|
19 |
+
# distributed under the License is distributed on an "AS IS" BASIS,
|
20 |
+
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
21 |
+
# See the License for the specific language governing permissions and
|
22 |
+
# limitations under the License.
|
23 |
+
# Lint as: python3
|
24 |
+
"""Corpus Crawler Irish web text dataset."""
|
25 |
+
|
26 |
+
import collections
|
27 |
+
import os
|
28 |
+
import re
|
29 |
+
import struct
|
30 |
+
import unicodedata
|
31 |
+
import base64
|
32 |
+
import hashlib
|
33 |
+
from html.entities import name2codepoint
|
34 |
+
from email import message_from_string as Message
|
35 |
+
from urllib.parse import urlparse
|
36 |
+
from pathlib import Path
|
37 |
+
|
38 |
+
import datasets
|
39 |
+
|
40 |
+
|
41 |
+
_DESCRIPTION = """\
|
42 |
+
Irish web corpus, crawled with Corpus Crawler.
|
43 |
+
|
44 |
+
Uses a list of URLs, collected by the crawler, to
|
45 |
+
retrieve the files from the crawler's cache.
|
46 |
+
"""
|
47 |
+
|
48 |
+
#_SCRAPES = ["20180911", "20191117", "20210810"]
|
49 |
+
_SCRAPES = ["20191117", "20210810"]
|
50 |
+
|
51 |
+
|
52 |
+
logger = datasets.utils.logging.get_logger(__name__)
|
53 |
+
_DATA_URL = 'https://gist.githubusercontent.com/jimregan/66612f4ecb88ed96d41d43266e6d0872/raw/26bd05f11b4c1c31e33d36528ac53dea587be8ef/crawled-{}.txt'
|
54 |
+
|
55 |
+
|
56 |
+
class CorpusCrawlerIrishConfig(datasets.BuilderConfig):
|
57 |
+
"""BuilderConfig for CorpusCrawlerIrish."""
|
58 |
+
|
59 |
+
def __init__(self, **kwargs):
|
60 |
+
super(CorpusCrawlerIrishConfig, self).__init__(version=datasets.Version("2.1.0", ""), **kwargs)
|
61 |
+
|
62 |
+
class CorpusCrawlerIrish(datasets.GeneratorBasedBuilder):
|
63 |
+
"""Corpus Crawler crawled text dataset."""
|
64 |
+
|
65 |
+
BUILDER_CONFIGS = [
|
66 |
+
CorpusCrawlerIrishConfig(name=scrape) for scrape in _SCRAPES
|
67 |
+
]
|
68 |
+
|
69 |
+
def _info(self):
|
70 |
+
return datasets.DatasetInfo(
|
71 |
+
description=_DESCRIPTION,
|
72 |
+
features=datasets.Features(
|
73 |
+
{
|
74 |
+
"url": datasets.Value("string"),
|
75 |
+
"genre": datasets.Value("string"),
|
76 |
+
"publication_date": datasets.Value("string"),
|
77 |
+
"title": datasets.Value("string"),
|
78 |
+
"text": datasets.Value("string"),
|
79 |
+
"video_url": datasets.Value("string"),
|
80 |
+
}
|
81 |
+
),
|
82 |
+
)
|
83 |
+
|
84 |
+
def _split_generators(self, dl_manager):
|
85 |
+
if not self.config.data_dir:
|
86 |
+
raise ValueError(f"Path to Corpus Crawler cache directory must be specified, but got data_dir={self.config.data_dir}")
|
87 |
+
cc_cache = self.config.data_dir
|
88 |
+
|
89 |
+
if not self.config.name:
|
90 |
+
raise ValueError(f"Scrape set must be specified, but got name={self.config.name}")
|
91 |
+
scrape_set = self.config.name
|
92 |
+
dl_path = dl_manager.download(_DATA_URL.format(self.config.name))
|
93 |
+
|
94 |
+
return [
|
95 |
+
datasets.SplitGenerator(
|
96 |
+
name=datasets.Split.TRAIN,
|
97 |
+
gen_kwargs={
|
98 |
+
"name": scrape_set,
|
99 |
+
"data_dir": cc_cache,
|
100 |
+
"data_file": dl_path,
|
101 |
+
})
|
102 |
+
]
|
103 |
+
|
104 |
+
def _generate_examples(self, name, data_dir, data_file):
|
105 |
+
"""Generate examples from a Corpus Crawl cache."""
|
106 |
+
logger.info("generating examples from = %s", name)
|
107 |
+
links = _get_links(data_file)
|
108 |
+
if not self.config.data_dir:
|
109 |
+
self.config.data_dir = data_dir
|
110 |
+
dd_path = Path(data_dir)
|
111 |
+
if not dd_path.is_dir():
|
112 |
+
raise Exception('No directory: ' + data_dir)
|
113 |
+
|
114 |
+
_id = 1
|
115 |
+
for link in links:
|
116 |
+
res = self._fetch_page(link, data_dir)
|
117 |
+
for para in res['text']:
|
118 |
+
example = {
|
119 |
+
"genre": res.get('genre', ''),
|
120 |
+
"url": res['location'],
|
121 |
+
"publication_date": res.get('publication-date', ''),
|
122 |
+
"video_url": res.get('video', ''),
|
123 |
+
"title": res.get('title', ''),
|
124 |
+
"text": para
|
125 |
+
}
|
126 |
+
yield _id, example
|
127 |
+
_id += 1
|
128 |
+
|
129 |
+
def _fetch_page(self, url, data_dir):
|
130 |
+
_EXTRATORS = {
|
131 |
+
'www.unicode.org': do_udhr,
|
132 |
+
'tuairisc.ie': do_tuairisc_ie,
|
133 |
+
'www.rte.ie': do_nuachtrte,
|
134 |
+
'www.irishtimes.com': do_irishtimes,
|
135 |
+
'www.chg.gov.ie': do_chg,
|
136 |
+
'www.ainm.ie': do_ainm_ie,
|
137 |
+
'gaeltacht21.blogspot.com': do_blogspot,
|
138 |
+
'aonghus.blogspot.com': do_blogspot,
|
139 |
+
'nimill.blogspot.com': do_blogspot,
|
140 |
+
'turasailse.blogspot.com': do_blogspot,
|
141 |
+
'caomhach.blogspot.com': do_blogspot,
|
142 |
+
'breacleabhar.blogspot.com': do_blogspot,
|
143 |
+
'gearoid.blogspot.com': do_blogspot,
|
144 |
+
'philo-celtic.blogspot.com': do_blogspot,
|
145 |
+
'iomhannablag.blogspot.com': do_blogspot,
|
146 |
+
'smaointefanacha.blogspot.com': do_blogspot,
|
147 |
+
'imeall.blogspot.com': do_blogspot,
|
148 |
+
'coislife.ie': do_coislife_ie,
|
149 |
+
'meoneile.ie': do_meoneile_ie,
|
150 |
+
'peig.ie': do_peig_ie,
|
151 |
+
'www.forasnagaeilge.ie': do_forasnagaeilge_ie,
|
152 |
+
}
|
153 |
+
parsed_url = urlparse(url)
|
154 |
+
host = parsed_url.netloc
|
155 |
+
extract = _EXTRATORS.get(host)
|
156 |
+
if extract:
|
157 |
+
fr = fetch(data_dir, url)
|
158 |
+
if fr is None:
|
159 |
+
raise Exception("Failed to fetch " + url + " from " + data_dir)
|
160 |
+
return extract(fr)
|
161 |
+
|
162 |
+
|
163 |
+
# Corpus Crawler: utils.py
|
164 |
+
_TAG_REGEX = re.compile(r'\<.+?\>', flags=re.DOTALL)
|
165 |
+
def striptags(s):
|
166 |
+
return _TAG_REGEX.sub('', s)
|
167 |
+
|
168 |
+
|
169 |
+
def unichar(i):
|
170 |
+
try:
|
171 |
+
return chr(i)
|
172 |
+
except ValueError:
|
173 |
+
# non-BMP codepoint in narrow Python build
|
174 |
+
return struct.pack('i', i).decode('utf-32')
|
175 |
+
|
176 |
+
|
177 |
+
def replace_html_entities(html):
|
178 |
+
entities = name2codepoint
|
179 |
+
html = re.sub(r'&#([0-9]+);',
|
180 |
+
lambda z:unichar(int(z.group(1))), html)
|
181 |
+
html = re.sub(r'&#[xX]([0-9a-fA-F]+);',
|
182 |
+
lambda z:unichar(int(z.group(1), 16)), html)
|
183 |
+
html = re.sub(r'&([a-zA-Z]+);',
|
184 |
+
lambda z:unichar(entities.get(z.group(1).lower(), 0x20)), html)
|
185 |
+
return html
|
186 |
+
|
187 |
+
|
188 |
+
def cleantext(html):
|
189 |
+
html = re.sub(r'<script.+?</script>', ' ', html, flags=re.DOTALL)
|
190 |
+
html = replace_html_entities(striptags(html))
|
191 |
+
# Some web sites insert zero-width spaces, possibly as byte order marks
|
192 |
+
# (from Microsoft Notepad) which their scripts failed to recognize as such.
|
193 |
+
html = html.replace('\u200B', '')
|
194 |
+
return unicodedata.normalize('NFC', ' '.join(html.split()))
|
195 |
+
|
196 |
+
|
197 |
+
def clean_paragraphs(html):
|
198 |
+
text = html.replace('\n', ' ')
|
199 |
+
text = re.sub(r'</(?:div|DIV|p|P|[hH][1-6]|table|TABLE|tr|td|article)>',
|
200 |
+
'\n', text)
|
201 |
+
text = re.sub(r'<(?:br|BR)\s*/?>', '\n', text)
|
202 |
+
return list(filter(None, [cleantext(p) for p in text.split('\n')]))
|
203 |
+
|
204 |
+
|
205 |
+
def extract(before, after, html):
|
206 |
+
s = html.split(before, 1)
|
207 |
+
return s[1].split(after)[0] if len(s) == 2 else None
|
208 |
+
|
209 |
+
|
210 |
+
FetchResult = collections.namedtuple('FetchResult',
|
211 |
+
['headers', 'content', 'url', 'filepath'])
|
212 |
+
|
213 |
+
|
214 |
+
def fetch(cache_dir, url):
|
215 |
+
logger.info("fetching url %s from cache %s", url, cache_dir)
|
216 |
+
try:
|
217 |
+
digest = hashlib.sha256(url.encode('utf-8')).digest()
|
218 |
+
filepath = os.path.join(cache_dir,
|
219 |
+
"f" + base64.urlsafe_b64encode(digest).decode('utf-8'))
|
220 |
+
except:
|
221 |
+
digest = hashlib.sha256(url).digest()
|
222 |
+
filepath = os.path.join(cache_dir,
|
223 |
+
"f" + base64.urlsafe_b64encode(digest))
|
224 |
+
|
225 |
+
fp = Path(filepath)
|
226 |
+
if not fp.is_file():
|
227 |
+
raise Exception("No such file: " + fp)
|
228 |
+
|
229 |
+
try:
|
230 |
+
with open(filepath, 'r', encoding='utf-8-sig', newline='') as f:
|
231 |
+
file_content = f.read()
|
232 |
+
if '\r\n\r\n\r\n' in file_content:
|
233 |
+
splitter = '\r\n\r\n\r\n'
|
234 |
+
else:
|
235 |
+
splitter = '\n\n\n'
|
236 |
+
cached = file_content.split(splitter, 1)
|
237 |
+
if len(cached) == 2:
|
238 |
+
headers, content = cached
|
239 |
+
try:
|
240 |
+
content = content.encode('utf-8')
|
241 |
+
except:
|
242 |
+
# already encoded as bytes
|
243 |
+
pass
|
244 |
+
headers = Message(headers)
|
245 |
+
return FetchResult(headers, content, url, filepath)
|
246 |
+
except IOError:
|
247 |
+
raise Exception("fetch() failed")
|
248 |
+
|
249 |
+
def do_udhr(fetchresult):
|
250 |
+
out = {}
|
251 |
+
text = fetchresult.content.decode('utf-8').split('---', 1)[1]
|
252 |
+
out['location'] = fetchresult.url
|
253 |
+
out['genre'] = 'Legal'
|
254 |
+
paras = []
|
255 |
+
for paragraph in text.splitlines():
|
256 |
+
paragraph = paragraph.strip()
|
257 |
+
if len(paragraph) > 0:
|
258 |
+
paras.append(paragraph)
|
259 |
+
out['text'] = paras
|
260 |
+
return out
|
261 |
+
|
262 |
+
|
263 |
+
# corpuscrawler: crawl_ga.py
|
264 |
+
_ENGLISH_MONTHS = {
|
265 |
+
'january': 1,
|
266 |
+
'february': 2,
|
267 |
+
'march': 3,
|
268 |
+
'april': 4,
|
269 |
+
'may': 5,
|
270 |
+
'june': 6,
|
271 |
+
'july': 7,
|
272 |
+
'august': 8,
|
273 |
+
'september': 9,
|
274 |
+
'october': 10,
|
275 |
+
'november': 11,
|
276 |
+
'december': 12
|
277 |
+
}
|
278 |
+
|
279 |
+
|
280 |
+
def _byline_to_pubdate(byline):
|
281 |
+
date = re.search(r'(\d{1,2}) ([^ ]+?) (\d{4})', byline)
|
282 |
+
if not date:
|
283 |
+
return None
|
284 |
+
day = int(date.group(1))
|
285 |
+
year = int(date.group(3))
|
286 |
+
month = _ENGLISH_MONTHS[date.group(2).lower()]
|
287 |
+
if not month:
|
288 |
+
return None
|
289 |
+
out = "{}-{:0>2d}-{:0>2d}".format(year, month, day)
|
290 |
+
return out
|
291 |
+
|
292 |
+
|
293 |
+
def _rte_writable_paragraph(text):
|
294 |
+
if text == '':
|
295 |
+
return False
|
296 |
+
if text.startswith('© RTÉ '):
|
297 |
+
return False
|
298 |
+
if text.startswith('By using this website, you consent'):
|
299 |
+
return False
|
300 |
+
if text.startswith('RTÉ.ie is the website of Raidió Teilifís Éireann'):
|
301 |
+
return False
|
302 |
+
if text.find('is not responsible for the content') >= 0:
|
303 |
+
return False
|
304 |
+
if text.find('RTÉ uses cookies in accordance with our Cookie Policy') >= 0:
|
305 |
+
return False
|
306 |
+
if re.match('^[\*\+]+$', text):
|
307 |
+
return False
|
308 |
+
return True
|
309 |
+
|
310 |
+
|
311 |
+
def _rte_cleanall(html):
|
312 |
+
section_article_regex = re.compile(r'<section[^>]+itemprop="articleBody"[^>]*>')
|
313 |
+
search = section_article_regex.search(html)
|
314 |
+
out = []
|
315 |
+
if search:
|
316 |
+
body = extract(search.group(0), '</section>', html)
|
317 |
+
for para in clean_paragraphs(body):
|
318 |
+
if _rte_writable_paragraph(para):
|
319 |
+
out.append(para)
|
320 |
+
return '\n'.join(out)
|
321 |
+
for paragraph in re.findall(r'<p>(.+?)</p>', html):
|
322 |
+
cleaned = cleantext(paragraph)
|
323 |
+
if _rte_writable_paragraph(cleaned):
|
324 |
+
out.append(cleaned)
|
325 |
+
else:
|
326 |
+
continue
|
327 |
+
return out
|
328 |
+
|
329 |
+
|
330 |
+
def _sceala_clean(paras):
|
331 |
+
out = []
|
332 |
+
for para in paras:
|
333 |
+
if '\n____' not in para:
|
334 |
+
out.append(para)
|
335 |
+
else:
|
336 |
+
out.append(para.split('\n____')[0])
|
337 |
+
break
|
338 |
+
return out
|
339 |
+
|
340 |
+
|
341 |
+
def do_nuachtrte(fetchresult):
|
342 |
+
out = {}
|
343 |
+
pubdate_regex = re.compile(r'name="DC.date" (?:scheme="DCTERMS.URI" )?content="([0-9T:+\-]{19,25})"')
|
344 |
+
html = fetchresult.content.decode('utf-8')
|
345 |
+
pubdate_match = pubdate_regex.search(html)
|
346 |
+
pubdate = pubdate_match.group(1) if pubdate_match else None
|
347 |
+
if pubdate is None: pubdate = fetchresult.headers.get('Last-Modified')
|
348 |
+
out['location'] = fetchresult.url
|
349 |
+
if 'nuacht' in fetchresult.url:
|
350 |
+
out['genre'] = 'News'
|
351 |
+
if pubdate:
|
352 |
+
out['publication-date'] = pubdate
|
353 |
+
title = re.search(r'<title>(.+?)</title>', html)
|
354 |
+
if title:
|
355 |
+
title = striptags(title.group(1).split('- RTÉ')[0]).strip()
|
356 |
+
if title:
|
357 |
+
out['title'] = cleantext(title)
|
358 |
+
cleaned = _rte_cleanall(html)
|
359 |
+
if '/sceala/' in fetchresult.url:
|
360 |
+
cleaned = _sceala_clean(cleaned)
|
361 |
+
out['text'] = cleaned
|
362 |
+
return out
|
363 |
+
|
364 |
+
|
365 |
+
def do_meoneile_ie(fetchresult):
|
366 |
+
out = {}
|
367 |
+
html = fetchresult.content.decode('utf-8')
|
368 |
+
title = extract(r'<title>', '</title>', html).strip()
|
369 |
+
title = title.split('<')[0].strip() if title else ''
|
370 |
+
video = re.search(r"<iframe.*src='(//player.vimeo.com/video/[0-9]+)[^>]*></iframe>", html)
|
371 |
+
body = extract("<div class='article-content'>", '</article>', html) or ''
|
372 |
+
byline = extract("<div class='byline'>", '</span>', html) or ''
|
373 |
+
byline = _byline_to_pubdate(byline)
|
374 |
+
if title:
|
375 |
+
out['title'] = title
|
376 |
+
paras = clean_paragraphs(body)
|
377 |
+
if paras:
|
378 |
+
out['location'] = fetchresult.url
|
379 |
+
out['genre'] = 'News'
|
380 |
+
if video:
|
381 |
+
out['video'] = f'https:{video.group(1)}'
|
382 |
+
if byline:
|
383 |
+
out['publication-date'] = byline
|
384 |
+
paras_filt = []
|
385 |
+
for para in paras:
|
386 |
+
if para == 'Roinn':
|
387 |
+
continue
|
388 |
+
else:
|
389 |
+
paras_filt.append(para)
|
390 |
+
out['text'] = paras_filt
|
391 |
+
return out
|
392 |
+
|
393 |
+
|
394 |
+
def do_irishtimes(fetchresult):
|
395 |
+
out = {}
|
396 |
+
html = fetchresult.content.decode('utf-8')
|
397 |
+
pubdatere1 = re.compile(r'<meta itemprop="datePublished" content="([^"]*)"/>')
|
398 |
+
pubdatere2 = re.compile(r'"datePublished": "([^"])"')
|
399 |
+
out['location'] = fetchresult.url
|
400 |
+
out['genre'] = 'News'
|
401 |
+
title = re.search(r'<title>(.+?)</title>', html)
|
402 |
+
pubdate_match = pubdatere1.search(html)
|
403 |
+
pubdate_match = pubdate_match if pubdate_match else pubdatere2.search(html)
|
404 |
+
pubdate = pubdate_match.group(1) if pubdate_match else None
|
405 |
+
if pubdate is None:
|
406 |
+
pubdate = fetchresult.headers.get('Last-Modified')
|
407 |
+
if pubdate:
|
408 |
+
out['publication-date'] = pubdate
|
409 |
+
if title:
|
410 |
+
out['title'] = cleantext(title.group(1))
|
411 |
+
paras = []
|
412 |
+
for paragraph in re.findall(
|
413 |
+
r'<p class="no_name">(.+?)</p>',
|
414 |
+
html.split('<div class="article_bodycopy">')[1]):
|
415 |
+
cleaned = cleantext(paragraph)
|
416 |
+
paras.append(cleaned)
|
417 |
+
out['text'] = paras
|
418 |
+
|
419 |
+
|
420 |
+
def do_blogspot(fetchresult):
|
421 |
+
out = {}
|
422 |
+
pubdate_regex = re.compile(
|
423 |
+
r"<abbr class='published' title='([^']*)'>[^<]*</abbr>")
|
424 |
+
html = fetchresult.content.decode('utf-8')
|
425 |
+
pubdate_match = pubdate_regex.search(html)
|
426 |
+
pubdate = pubdate_match.group(1) if pubdate_match else None
|
427 |
+
if pubdate is None: pubdate = fetchresult.headers.get('Last-Modified')
|
428 |
+
title = re.search(r"<meta content='([^']+)' property='og:title'/>",
|
429 |
+
html)
|
430 |
+
title = title.group(1) if title else ''
|
431 |
+
post = extract("<div class='post-body entry-content'>",
|
432 |
+
"<div class='post-footer'>", html)
|
433 |
+
if post == None:
|
434 |
+
post = extract("<div class='post-header'>",
|
435 |
+
"<div class='post-footer'>", html)
|
436 |
+
if post == None:
|
437 |
+
post = extract('<div class="post-body">',
|
438 |
+
'<p class="post-footer">', html)
|
439 |
+
paras = clean_paragraphs(post)
|
440 |
+
if paras:
|
441 |
+
out['title'] = title
|
442 |
+
out['location'] = fetchresult.url
|
443 |
+
out['genre'] = 'News'
|
444 |
+
if pubdate:
|
445 |
+
out['publication-date'] = pubdate
|
446 |
+
out['text'] = paras
|
447 |
+
return out
|
448 |
+
|
449 |
+
|
450 |
+
def do_ainm_ie(fetchresult):
|
451 |
+
out = {}
|
452 |
+
html = fetchresult.content.decode('utf-8')
|
453 |
+
title = re.search(r'<title>(.+?)</title>', html)
|
454 |
+
title = title.group(1).split('|')[0] if title else ''
|
455 |
+
body = extract('<div class="article">',
|
456 |
+
'<!-- .contentWrapper-->', html) or ''
|
457 |
+
body = body.split('<div id="machines"')[0]
|
458 |
+
paras = clean_paragraphs(body)
|
459 |
+
pubdate = fetchresult.headers.get('Last-Modified')
|
460 |
+
if paras:
|
461 |
+
out['title'] = title
|
462 |
+
out['location'] = fetchresult.url
|
463 |
+
out['genre'] = 'Biography'
|
464 |
+
if pubdate:
|
465 |
+
out['publication-date'] = pubdate
|
466 |
+
out['text'] = paras
|
467 |
+
return out
|
468 |
+
|
469 |
+
|
470 |
+
def do_tuairisc_ie(fetchresult):
|
471 |
+
out = {}
|
472 |
+
pubdate_regex = re.compile(
|
473 |
+
r'<time datetime="(20\d\d-\d\d-\d\d)\s+(\d\d:\d\d)" '
|
474 |
+
r'itemprop="datePublished">')
|
475 |
+
html = fetchresult.content.decode('utf-8')
|
476 |
+
title = extract('<h1 class="title article--full__title">', '</h1>',
|
477 |
+
html) or ''
|
478 |
+
pubdate_match = pubdate_regex.search(html)
|
479 |
+
if pubdate_match:
|
480 |
+
pubdate = '%sT%s:00Z' % (
|
481 |
+
pubdate_match.group(1), pubdate_match.group(2))
|
482 |
+
body = extract(
|
483 |
+
'<div class="article--full__content" itemprop="articleBody">',
|
484 |
+
'</article>', html)
|
485 |
+
if body:
|
486 |
+
paras = clean_paragraphs(body)
|
487 |
+
if paras:
|
488 |
+
out['title'] = title
|
489 |
+
out['location'] = fetchresult.url
|
490 |
+
out['genre'] = 'News'
|
491 |
+
if pubdate:
|
492 |
+
out['publication-date'] = pubdate
|
493 |
+
out['text'] = paras
|
494 |
+
return out
|
495 |
+
|
496 |
+
|
497 |
+
def do_coislife_ie(fetchresult):
|
498 |
+
out = {}
|
499 |
+
html = fetchresult.content.decode('utf-8')
|
500 |
+
title = re.search(r'<title>(.+?)</title>', html)
|
501 |
+
title = title.group(1).split('–')[0].strip() if title else ''
|
502 |
+
desc = re.search(r'<meta property="og:description" content="([^"]+?)"', html)
|
503 |
+
desc = cleantext(desc.group(1))
|
504 |
+
body = extract('<div class="tab-content">',
|
505 |
+
'<div class="entry-content in fade tab-pane" id="tab-additional_information">', html) or ''
|
506 |
+
paras = clean_paragraphs(title + '<br/>' + body)
|
507 |
+
pubdate = fetchresult.headers.get('Last-Modified')
|
508 |
+
if paras:
|
509 |
+
out['title'] = title
|
510 |
+
out['location'] = fetchresult.url
|
511 |
+
out['genre'] = 'Commerce'
|
512 |
+
if desc:
|
513 |
+
out['description'] = desc
|
514 |
+
if pubdate:
|
515 |
+
out['publication-date'] = pubdate
|
516 |
+
outp = []
|
517 |
+
for para in paras:
|
518 |
+
if para.find('Léigh sliocht as an leabhar') >= 0:
|
519 |
+
continue
|
520 |
+
else:
|
521 |
+
outp.append(para)
|
522 |
+
out['text'] = outp
|
523 |
+
return out
|
524 |
+
|
525 |
+
|
526 |
+
def do_chg(fetchresult):
|
527 |
+
out = {}
|
528 |
+
def _chg_content(page):
|
529 |
+
return page.split('<div class="container" id="article">')[1].split('<!-- /.right columns -->')[0]
|
530 |
+
phtml = fetchresult.content.decode('utf-8')
|
531 |
+
ptext = _chg_content(phtml)
|
532 |
+
title = re.search(r'<title>(.+?)</title>', phtml)
|
533 |
+
if title: title = striptags(title.group(1).split('|')[0]).strip()
|
534 |
+
pubdate = fetchresult.headers.get('Last-Modified')
|
535 |
+
out['location'] = fetchresult.url
|
536 |
+
out['genre'] = 'Government'
|
537 |
+
if pubdate:
|
538 |
+
out['publication-date'] = pubdate
|
539 |
+
if title:
|
540 |
+
out['title'] = title
|
541 |
+
paras = []
|
542 |
+
for paragraph in re.findall(r'<p>(.+?)</p>', ptext):
|
543 |
+
cleaned = cleantext(paragraph)
|
544 |
+
paras.append(cleaned)
|
545 |
+
out['text'] = paras
|
546 |
+
return out
|
547 |
+
|
548 |
+
|
549 |
+
def do_peig_ie(fetchresult):
|
550 |
+
out = {}
|
551 |
+
def peig_cat(page):
|
552 |
+
if page.find('/imeachtai/') >= 0:
|
553 |
+
return 'Events'
|
554 |
+
elif page.find('peig.ie/20') >= 0:
|
555 |
+
return 'News'
|
556 |
+
elif page.find('/fol%C3%BAntais/') >= 0:
|
557 |
+
return 'Job listings'
|
558 |
+
else:
|
559 |
+
return ''
|
560 |
+
# Peig.ie has a lot of posts from other sites
|
561 |
+
html = fetchresult.content.decode('utf-8')
|
562 |
+
title = re.search(r'<title>(.+?)</title>', html)
|
563 |
+
title = title.group(1).split('|')[0].strip() if title else ''
|
564 |
+
if '<meta property="article:modified_time"' in html:
|
565 |
+
date = re.search(r'<meta property="article:modified_time" content="([^"]+)"', html).group(1)
|
566 |
+
else:
|
567 |
+
date = re.search(r'"datePublished":"([^"]+)"', html).group(1)
|
568 |
+
body = extract('<div class="uk-margin-medium-top" property="text">', '<ul class="uk-pagination', html) or ''
|
569 |
+
paras = clean_paragraphs(body)
|
570 |
+
genre = peig_cat(fetchresult.url)
|
571 |
+
if paras:
|
572 |
+
out['location'] = fetchresult.url
|
573 |
+
if title:
|
574 |
+
out['title'] = title
|
575 |
+
if genre:
|
576 |
+
out['genre'] = genre
|
577 |
+
if date:
|
578 |
+
out['publication-date'] = date
|
579 |
+
out['text'] = paras
|
580 |
+
return out
|
581 |
+
|
582 |
+
|
583 |
+
def do_forasnagaeilge_ie(fetchresult):
|
584 |
+
out = {}
|
585 |
+
pubdate_regex = re.compile(r'"datePublished":"([^"]+)",')
|
586 |
+
html = fetchresult.content.decode('utf-8')
|
587 |
+
if '<html class="no-js" lang="en">' in html:
|
588 |
+
return {}
|
589 |
+
title = extract('<title>', ' - www.forasnagaeilge.ie</title>',
|
590 |
+
html) or ''
|
591 |
+
pubdate_match = pubdate_regex.search(html)
|
592 |
+
if pubdate_match:
|
593 |
+
pubdate = pubdate_match.group(1)
|
594 |
+
body = extract(
|
595 |
+
'<div id="main" class="container">',
|
596 |
+
'</div><!-- /.content -->', html)
|
597 |
+
if body:
|
598 |
+
paras = clean_paragraphs(body)
|
599 |
+
if paras:
|
600 |
+
out['location'] = fetchresult.url
|
601 |
+
out['genre'] = 'News'
|
602 |
+
if title:
|
603 |
+
out['title'] = title
|
604 |
+
if pubdate:
|
605 |
+
out['publication-date'] = pubdate
|
606 |
+
out['text'] = paras
|
607 |
+
return out
|
608 |
+
|
609 |
+
|
610 |
+
def _get_links(scrape):
|
611 |
+
links = set()
|
612 |
+
with open(scrape) as f:
|
613 |
+
for url in f.readlines():
|
614 |
+
links.add(url.rstrip())
|
615 |
+
return list(links)
|