def next_page(self, response: scrapy.http.Response) -> scrapy.Request:
"""
Goes to next page.
:param response: response object
:return: request for next page
"""
# go to next page
next_url = response.xpath("//a[@title='下一页']/@href").extract_first()
if next_url is not None:
self.log('Next page {}'.format(next_url), level=logging.INFO)
time.sleep(random.random())
return response.follow(
url=next_url,
callback=self.parse,
# reuse the current proxy
meta={'proxy': response.request.meta['proxy']},
errback=self.handle_failure)
else:
# try to build the page by ourself
arguments = self.decode_url(response.request.url)
arguments['page'] += 1
url = self.format_url(arguments)
self.log('Next page (manually) {}'.format(url), level=logging.INFO)
return response.follow(
url=url,
callback=self.parse,
# reuse the current proxy
meta={'proxy': response.request.meta['proxy']},
errback=self.handle_failure)
def parse_docs(self, response: scrapy.http.Response):
pdfs: List[str] = []
for url in response.css('a::attr(href)'):
full = response.urljoin(url.extract())
if full.endswith('.pdf'):
pdfs.append(full)
yield {'from': response.url, 'file_urls': pdfs}
def parseCityAttractionsListPage(self, response: scrapy.http.Response):
# example page: https://www.viator.com/Mumbai/d953
print(
'PARSING ATTRACTION LIST ####################################################################################'
)
print(response.url)
self.incrementRequestCount()
hrefs = response.css('div.ptm *> h2 > a')
for href in hrefs:
pointURL = href.css('::attr(href)').extract_first().strip()
pointName = href.css('::text').extract_first().strip()
yield response.follow(pointURL,
callback=self.parseAttractionsPage,
meta={
'countryName':
response.meta['countryName'],
'cityName': response.meta['cityName'],
'pointName': pointName
})
nextPageLink = response.css(
'div.ptm > div:nth-child(1) > div:nth-child(2) > p > a:last-child::attr(href)'
).extract_first()
if nextPageLink:
yield response.follow(nextPageLink,
callback=self.parseCityAttractionsListPage,
meta=response.meta)
def parse(self, response: scrapy.http.Response):
"""
Parses content from a html page response.
"""
listings = response.xpath('//li[@class="result-row"]')
for listing in listings:
# Relative matching
date = listing.xpath(
'.//*[@class="result-date"]/@datetime').extract_first()
url = listing.xpath(
'.//a[@class="result-title hdrlnk"]/@href').extract_first()
title = listing.xpath(
'.//a[@class="result-title hdrlnk"]/text()').extract_first()
yield scrapy.Request(url,
callback=self.parse_listing,
meta=dict(date=date, url=url, title=title))
# Move to the next page of data.
next_page_url = response.xpath(
'//*[@class="button next"]/@href').extract_first()
if next_page_url:
# url must be absolute.
abs_next_page_url = response.urljoin(next_page_url)
yield scrapy.Request(url=abs_next_page_url, callback=self.parse)
def parse(self, response: scrapy.http.Response, **kwargs):
form_data = {
'__VIEWSTATE':
response.xpath('//input[@id="__VIEWSTATE"]/@value').get(),
'__VIEWSTATEGENERATOR':
response.xpath('//input[@id="__VIEWSTATEGENERATOR"]/@value').get(),
'__EVENTVALIDATION':
response.xpath('//input[@id="__EVENTVALIDATION"]/@value').get(),
}
next_page_btn = response.xpath(
'//a[contains(@href, "Page$Next")]').get()
if next_page_btn is not None:
data = form_data.copy()
data['__EVENTTARGET'] = 'grdSQL'
data['__EVENTARGUMENT'] = 'Page$Next'
yield scrapy.FormRequest(f'{self.base_url}/Events.aspx',
formdata=data)
entries = response.xpath(
'//table[@id="grdSQL"]//tr[@onmouseover]').getall()
for i, entry in enumerate(entries):
data = form_data.copy()
data['__EVENTTARGET'] = 'grdSQL'
data['__EVENTARGUMENT'] = f'SysRowSelector${i}'
yield scrapy.FormRequest(f'{self.base_url}/Events.aspx',
formdata=data,
callback=self.parse_entry,
meta={'row_index': i})
def parse(self, response: scrapy.http.Response):
findform = response.xpath("//form[@name='form1']")
form = self.build_form(findform)
if "kirjaamo" not in form:
raise ValueError("kirjaamo not found")
if not isinstance(form["kirjaamo"], list):
raise ValueError("kirjaamo is not list")
method = findform.xpath("./@method").get()
action = response.urljoin(findform.xpath("./@action").get())
alist = form["kirjaamo"]
del form["kirjaamo"]
for param in alist:
val = param["value"]
if val == "":
continue
fdata = form
fdata["kirjaamo"] = val
yield scrapy.FormRequest(
action,
method=method,
formdata=fdata,
meta={
"name": param["name"],
"dont_cache": True,
},
callback=self.parse_search_result,
)
def parse_subtopic_triangle(self, response: scrapy.http.Response):
# Gathers all subtopics from https://www.walter-fendt.de/html5/mde/tl/tl_start_de.htm
triangle_subtopics = response.xpath(
'/html/body/ul/li/a/@href').getall()
for subtopic_url in triangle_subtopics:
subtopic_url = response.urljoin(subtopic_url)
yield scrapy.Request(url=subtopic_url, callback=self.parse)
def parse(self, response: scrapy.http.Response):
# Next page link
next_page = response.xpath(
"//div[@id='pager_one']/div[@class='subcontainer']/div[contains(@class, 'pageside') and contains(@class, 'pright')]/a/@href"
).get()
if next_page is not None:
yield scrapy.Request(
response.urljoin(next_page),
callback=self.parse,
)
for row in response.xpath("//table[@id='remixtable']/tr[@class]"):
# Get tunes
addeddate = datetime.strptime(
row.xpath("td[1]/text()").get(), "%Y-%m-%d")
link = response.urljoin(row.xpath("td[2]/a/@href").get())
title = row.xpath("td[2]/a/text()").get()
arranger = row.xpath("td[3]/a/text()").get()
composer = row.xpath("td[4]/text()").get()
yield scrapy.Request(link,
callback=self.dl_tune,
meta={
"tune":
Tune(
title=title,
arranger=arranger,
added=addeddate,
composer=composer,
data=None,
),
})
def parseCountryPage(self, response: scrapy.http.Response):
# example page: https://www.viator.com/India/d723-ttd
self.incrementRequestCount()
breadcrumbs = response.css('div.crumbler *> span::text').extract()
countryName = breadcrumbs[1].strip()
countryListing = CountryListing(crawler=self.name,
sourceURL=response.url,
crawlTimestamp=getCurrentTime(),
countryName=countryName)
yield countryListing.jsonify()
if skipNonRequired:
if processName(countryName) not in processedRequiredCountries:
# do not process this country's cities
print('Skipping country: ', countryName)
return
countryId = response.url.split('/')[-1].split('-')[0][1:]
cityListingURL = 'https://www.viator.com/pascities.jspa?country={}'.format(
countryId)
yield response.follow(cityListingURL,
callback=self.parseCountryCities,
meta={'countryName': countryName})
def parse(self, response: scrapy.http.Response):
print("Parsing URL: " + response.url)
# Call Splash only once per page (that contains multiple XML elements).
data = self.getUrlData(response.url)
response.meta["rendered_data"] = data
# We would use .fromstring(response.text) if the response did not include the XML declaration:
# <?xml version="1.0" encoding="utf-8"?>
root = etree.XML(response.body)
tree = etree.ElementTree(root)
# If results are returned.
elements = tree.xpath("/root/items/*")
if len(elements) > 0:
for element in elements:
copyResponse = response.copy()
element_xml_str = etree.tostring(element,
pretty_print=True,
encoding="unicode")
element_dict = xmltodict.parse(element_xml_str)
# Temporary solution for public-only content.
# TODO: remove this when licensed content are enabled!
if not self.is_public(element_dict["data"]):
continue
# TODO: It's probably a pointless attribute.
# del element_dict["data"]["score"]
# Passing the dictionary for easier access to attributes.
copyResponse.meta["item"] = element_dict["data"]
# In case JSON string representation is preferred:
# copyResponse._set_body(json.dumps(copyResponse.meta['item'], indent=1, ensure_ascii=False))
copyResponse._set_body(element_xml_str)
if self.hasChanged(copyResponse):
yield self.handleEntry(copyResponse)
# LomBase.parse() has to be called for every individual instance that needs to be saved to the database.
LomBase.parse(self, copyResponse)
# TODO: To not stress the Rest APIs.
# time.sleep(0.1)
# If the number of returned results is equal to the imposed limit, it means that there are more to be returned.
if len(elements) == self.limit:
self.page += 1
url = self.apiUrl.replace("%start",
str(self.page * self.limit)).replace(
"%anzahl", str(self.limit))
yield scrapy.Request(
url=url,
callback=self.parse,
headers={
"Accept": "application/xml",
"Content-Type": "application/xml",
},
)
def post(
self, response: scrapy.http.Response
) -> Union[Iterator[items.ArticleItem], Iterator[scrapy.Request]]:
"""Get medium posts.
Args:
response (scrapy.http.Response): scrapy response
Yields:
items.ArticleItem: ArticleItem object
scrapy.Request: scrapy request object
"""
data = response.text.replace('])}while(1);</x>', '', 1)
obj = json.loads(data)['payload']
post_record = self.parse_post_item(post=obj)
yield post_record
if post_record['comment_count'] > 0:
post_id = obj['value']['id']
response.meta['post_id'] = post_id
response.meta['post_record'] = post_record
url = (f'https://medium.com/_/api/posts/{post_id}/responsesStream')
yield scrapy.Request(url=url,
meta=response.meta,
callback=self.comment)
def parse(self, response: scrapy.http.Response):
"""
Get list of tunes
"""
u: SplitResult = urlsplit(response.url)
q: dict = dict(queryparse(u.query))
for tune in response.xpath(
"//div[@id='result']/table/tr/th[@colspan='6']/../../tr[@class]"
):
artist = "".join(tune.xpath("./td[2]//text()").getall()).strip()
title = "".join(tune.xpath("./td[1]//text()").getall()).strip()
link = tune.xpath("./td[1]/a/@href").get().strip()
fileformat = "".join(
tune.xpath("./td[3]//text()").getall()).strip().lower()
# Download tune
yield scrapy.Request(
response.urljoin(link),
callback=self.download_mod,
meta={
"tune": {
"id": q['view'],
"artist": artist,
"title": title,
"format": fileformat,
}
},
)
def parse_apollonian_subtopic(self, response: scrapy.http.Response):
# Gathers variant-URLs to crawl from https://www.walter-fendt.de/html5/mde/apolloniosproblem_de.htm
apollonios_subtopics = response.xpath(
'//table/tbody/tr/td/a/@href').getall()
for apollo_url in apollonios_subtopics:
apollo_url = response.urljoin(apollo_url)
yield scrapy.Request(url=apollo_url, callback=self.parse)
def parse(self, response: scrapy.http.Response):
# example page: https://www.viator.com/Amsterdam/d525-ttd
countryMenuBox = response.css(
'#countryMenuBox > div.menu-dropdown-box.small > div > div:nth-child(1)'
)
hrefs = countryMenuBox.css('a::attr(durl)').extract()
for href in hrefs:
yield response.follow(href, callback=self.parseCountryPage)
def parse_category_overview_for_topics_and_subpages(
self, response: scrapy.http.Response):
"""
Crawls an overview page of a "type"-category (e.g. "Hintergrund", "Bilderserie" etc.) for subpages and topics.
If the overview has subpages, it will recursively yield additional scrapy.Requests to the overview-subpages.
Afterwards it yields the (10) individual topic_urls (per overview page) to the parse()-method.
Scrapy Contracts:
@url https://www.umwelt-im-unterricht.de/suche/?tx_solr%5Bfilter%5D%5B0%5D=type%3Alessons
@returns requests 10
"""
topic_urls_raw: list = response.xpath(
'//a[@class="internal-link readmore"]/@href').getall()
for url_ending in topic_urls_raw:
self.topic_urls.add(response.urljoin(url_ending))
# if there's a "Letzte"-Button in the overview, there's more topic_urls to be gathered than the initially
# displayed 10 elements
last_page_button_url: str = response.xpath(
'//li[@class="tx-pagebrowse-last last"]/a/@href').get()
# the string last_page_button_url typically looks like this:
# "/suche/?tx_solr%5Bfilter%5D%5B0%5D=type%3Amaterials_images&tx_solr%5Bpage%5D=8"
page_number_regex = re.compile(
r'(?P<url_with_parameters>.*&tx_solr%5Bpage%5D=)(?P<nr>\d+)')
overview_urls_parsed: set = set(
) # temporary set used for checking off already visited URLs
if last_page_button_url is not None:
page_number_dict: dict = page_number_regex.search(
last_page_button_url).groupdict()
url_without_page_parameter: str = response.urljoin(
page_number_dict.get('url_with_parameters'))
last_page_number = int(page_number_dict.get('nr'))
for i in range(2, last_page_number + 1):
# the initial url from start_urls already counts as page 1, therefore we're iterating
# from page 2 to the last page
next_overview_subpage_to_crawl = str(
url_without_page_parameter + str(i))
if next_overview_subpage_to_crawl not in self.overview_urls_already_parsed:
yield scrapy.Request(
url=next_overview_subpage_to_crawl,
callback=self.
parse_category_overview_for_topics_and_subpages)
overview_urls_parsed.add(next_overview_subpage_to_crawl)
self.overview_urls_already_parsed.update(
overview_urls_parsed
) # checking off the (10) URLs that we yielded
parsed_urls: set = set(
) # temporary set used for checking off already visited topics
for url in self.topic_urls:
if url not in self.topic_urls_parsed:
# making sure that we don't accidentally crawl individual pages more than once
yield scrapy.Request(url=url, callback=self.parse)
parsed_urls.add(url)
self.topic_urls_parsed.update(parsed_urls)
def parse_landing_page(self, response: scrapy.http.Response):
# On a landing page, we can extract all the documents, or infer the JSON link and use that.
# yield {'title': pub.css('h1 ::text').extract_first().strip()}
for pub in response.css('.publication'):
# This is a publication, so let's infer the API link:
lp_url = list(urlsplit(response.url))
lp_url[2] = "/api/content%s" % lp_url[2]
api_json_url = urlunsplit(lp_url)
yield response.follow(api_json_url, self.parse_content_api_json)
def parse(self, response: scrapy.http.Response):
for relative_url in response.xpath('//h3/a/@href').extract():
absolute_url = response.urljoin(relative_url)
yield scrapy.Request(absolute_url, callback=self.parse_book)
next_page_url = response.xpath(
'//*[@class="next"]/a/@href').extract_first()
absolute_url = response.urljoin(next_page_url)
yield scrapy.Request(absolute_url)
def parse_posts_list(self, response: scrapy.http.Response):
# Fetch the posts
for href in response.css("#posts a::attr(href)"):
if href.get().startswith("/p"):
yield response.follow(href, self.parse_thread)
# Fetch all pages
for href in response.css(".pagination a::attr(href)"):
yield response.follow(href, self.parse_posts_list)
def parse(self, response: scrapy.http.Response):
# Extract every link to a landing page:
for title in response.css('.document-row > h3 > a'):
yield response.follow(title, self.parse_landing_page)
# Extract the link to the next page of results:
for next_page in response.css('.next > a'):
yield response.follow(next_page, self.parse)
def parse_listing(self, response: scrapy.http.Response):
i = {}
i['url'] = response.url
i['expire'] = response.xpath(
'//a[@class="expire"]/span/text()').extract_first()
i['job-title'] = response.css('span#main-job-title *::text').extract()
i['main'] = response.css('div#main-lang-block *::text').extract()
i['job-details'] = response.css('div.jobdetails *::text').extract()
return i
def parse_page(self, response: scrapy.http.Response):
image_url = response.css(
'div#all div.text-center img.img-fluid::attr(src)').get()
image_url = response.urljoin(image_url)
image = ImageItem()
image['comic_id'] = response.meta['comic_id']
image['vol_id'] = response.meta['vol_id']
image['page'] = response.meta['page']
image['url'] = image_url
yield image
def parse_content(self, response: scrapy.http.Response):
item = PttItem()
item['content'] = response.xpath(
"//div[@id='main-content']/text()").get().replace('\n', '')
meta = response.xpath("//span[@class='article-meta-value']")
item['author'] = meta[0].xpath('text()').get()
item['title'] = meta[2].xpath('text()').get()
item['date'] = meta[3].xpath('text()').get()
item['url'] = response.url
yield item
def get_next_vimeo_overview_page(self, response: scrapy.http.Response):
"""
if there is a "next"-button at the bottom of the vimeo-user's overview page:
grabs the url from it and yields it
"""
# next_vimeo_overview_page = response.xpath('//*[@id="pagination"]/ol/li[9]').get()
next_vimeo_overview_page = response.css(
'#pagination > ol > li.pagination_next a::attr(href)').get()
if next_vimeo_overview_page is not None:
yield response.follow(next_vimeo_overview_page, self.parse)
def parse_ad_page(self, resp: scrapy.http.Response) -> None:
'''
Yields Ad objects if search phrase is found in response.
'''
ensure_response_200(resp)
title = resp.xpath('//div[@id = "adTitle"]//text()').extract_first()
description = ' '.join(
resp.xpath('//div[@id = "adDescription"]//text()').extract())
if title and description:
if self._search_phrase in title or self._search_phrase in description:
yield Ad(url=resp.url, search_phrase=self._search_phrase)
def parseCountryCities(self, response: scrapy.http.Response):
# example page: https://www.viator.com/pascities.jspa?country=723
self.incrementRequestCount()
hrefs = response.css(
'div.unit.size-pas-cities *> a::attr(durl)').extract()
for href in hrefs:
yield response.follow(href,
callback=self.parseCityPage,
meta=response.meta)
def parse_section_overview(self, response: scrapy.http.Response):
# Each section (e.g. "Mathematik Teilgebiete") holds a list of individual topic-categories (e.g. "Kreislehre")
section_urls = response.xpath(
'/html/body/table/tr/td/a/@href').getall()
section_urls.sort()
# print(section_urls)
# print("Section URLs: ", len(section_urls))
for url in section_urls:
current_url = response.urljoin(url)
yield scrapy.Request(url=current_url,
callback=self.parse_topic_overview)
def parse(self, response: scrapy.http.Response, **kwargs):
next_page_url = response.xpath('//a[@rel="next"]/@href').get()
if next_page_url is not None:
yield scrapy.Request(next_page_url)
entry_urls = response.xpath(
f'//div[{util.xpath_class(["type-tribe_events"])}]//*[{util.xpath_class(["tribe-events-list-event-title"])}]/a/@href'
).getall()
for url in entry_urls:
yield scrapy.Request(url, callback=self.parse_entry)
def parse(self, response: scrapy.http.Response, **kwargs):
next_page_url = response.xpath(
f'//li[{xpath_class(["next"])}]/a/@href').get()
if next_page_url is not None:
yield scrapy.Request(f'{self.base_url}{next_page_url}')
entries = response.xpath(
f'//article[{xpath_class(["event"])}]').getall()
for entry in entries:
yield ResponseItem({'body': entry, 'meta': response.meta})
def parse(
self, response: scrapy.http.Response
) -> typing.Generator[scrapy.Request, None, None]:
"""Find all the cases."""
for case_url in response.xpath('//table[@class="cases"]/tbody/tr/td/a/@href'):
url = response.urljoin(case_url.extract())
yield scrapy.Request(
url=url,
callback=self.parse_case,
dont_filter=True,
)
def parse_search_result(self, response: scrapy.http.Response):
tbl = response.xpath(
"//table[@class='table table-striped table-hover table-bordered']")
for rowidx, row in enumerate(tbl.xpath("./tr")):
if rowidx == 0:
continue
obj = {}
for idx, col in enumerate(row.xpath("./td")):
if idx == 0:
rawdate = "".join(col.xpath("./text()").getall()).strip()
rawdate = ' '.join(rawdate.split())
rawdate = rawdate.strip()
rem = re.split(r"^(\d+)\s+/(\d+) (\d+)\.(\d+)\.(\d+)$",
rawdate)[1:]
rem.pop()
vhnum, vhyear, pday, pmonth, pyear = rem
obj["date"] = f"{vhyear}-{vhnum.zfill(3)}__{pyear}-{pmonth.zfill(2)}-{pday.zfill(2)}"
elif idx == 1:
for link in col.xpath("./a"):
txt = link.xpath("./text()").get().strip()
url = response.urljoin(link.xpath("./@href").get())
if txt == '0 kpl':
continue
if 'title' not in obj:
obj["title"] = txt
obj["link"] = url
else:
obj["attach"] = url
dirpath = os.path.join(self.name, )
if "attach" in obj:
yield scrapy.Request(
obj["attach"],
meta={
"name": response.meta["name"],
"id": obj["date"],
},
callback=self.parse_attachments,
)
yield scrapy.Request(
obj["link"],
meta={
"name": response.meta["name"],
"id": obj["date"],
},
callback=self.dl_doc,
)
