def l4_parse_poetry(self, response):
'''
Level 4: parse poetry page, extract poetry, and save.
'''
#print response.url
try:
# extract poem
#stanza_selectors = response.xpath("//div[@class='poemContainer']/div[@class='PoemTextHost ']/div[@class='PoemDisplay OrgTextDisplay ']/div")
stanza_selectors = response.xpath("//div[contains(@class,'poemContainer')]/div[contains(@class,'PoemTextHost')]/div[contains(@class,'PoemDisplay') and contains(@class,'OrgTextDisplay')]/div")
poem = ''
for s in stanza_selectors:
line_selectors = s.xpath("./p")
for l in line_selectors:
line = l.xpath(".//text()").extract()
line = ''.join(line)
line = line.strip()
##print line
poem = poem + line + '\n'
poem = poem + '\n'
poem = poem[0:-1]#strip last '\n' from the poem
#print poem
# extract title of the poem
title = response.xpath("//div[@class='shayariContainerDiv']/div[@class='left_pan_shayari']/div[@class='shayari_first']/h1/text()").extract()[0]
# extract poet name
#poet = response.xpath("//div[@class='artist_img_descrpt']/div[@class='about_artist']/h2/text()").extract()[0]
# poet name must be in english, that is why we have to discard the previous one.
poet_href = response.xpath("//div[@class='artist_img']//a/@href").extract()[0]
p = re.compile(ur'poets/(.+)/') #href="/poets/anjum-tarazi/?lang=Hi"
poet = p.search(poet_href).group(1)
poet = poet.replace('-', ' ')
# check response.url for language information: https://.....xyz/?lang=hi
tmp = response.url
tmp = tmp.split('?')
url = tmp[0]
language = tmp[1].split('=')[1]
data = {}
data['poem'] = poem
data['url'] = url
data['title'] = title.strip()
data['author'] = poet.title()
data['language'] = language
# Store these information in DB
save_to_db_poem(data)
except:
print("ERROR: l4_parse_poetry: Unexpected error:", sys.exc_info()[0])
_trace = ''
for frame in traceback.extract_tb(sys.exc_info()[2]):
fname,lineno,fn,text = frame
print ("DBG:: Error in %s on line %d" % (fname, lineno))
_trace = _trace + "Error in %s on line %d" % (fname, lineno)
with open(self.LOGFILE, "a") as outfile:
t = time.asctime( time.localtime(time.time()) )
json.dump({'link': response.url, 'error': 'parsing failed', 'trace': _trace, 'time': t}, outfile, indent=4)
def l4_parse_poetry(self, response):
'''
Level 4: parse poetry page, extract poetry, and save.
'''
#print response.url
try:
# extract data
content = response.xpath("//div[@id='ImageHost']//div[@class='PoemDisplay']/span").extract()[0]
title = response.xpath("//div[@class='shayariContainerDiv']/div[@class='left_pan_shayari']/div[@class='shayari_first']/h1/text()").extract()[0]
#poet = response.xpath("//div[@class='artist_img_descrpt']/div[@class='about_artist']/h2/text()").extract()[0]
# poet name must be in english, that is why we have to discard the previous one.
poet_href = response.xpath("//div[@class='artist_img']//a/@href").extract()[0]
p = re.compile(ur'poets/(.+)/') #href="/poets/anjum-tarazi/?lang=Hi"
poet = p.search(poet_href).group(1)
poet = poet.replace('-', ' ')
# remove outer span tag
poem = remove_tags(content, 'span')
except:
print "ERROR: l4_parse_poetry: failed to extract data"
with open("exception_poetry.txt", "a") as outfile:
json.dump({'link': response.url, 'error': 'parsing failed'}, outfile, indent=4)
try:
# check response.url for language information: https://.....xyz/?lang=hi
tmp = response.url
tmp = tmp.split('?')
url = tmp[0]
language = tmp[1].split('=')[1]
# Correct the line order of the poem; and save
refined_poem = refine_poetry(poem, url, language)
if refined_poem:
data = {}
data['poem'] = refined_poem
data['url'] = url
data['title'] = title
data['author'] = poet.title()
data['language'] = language
# Store these information in DB
save_to_db_poem(data)
else:
with open("exception_poetry.txt", "a") as outfile:
json.dump({'link': response.url, 'error': 'refine_poetry failed'}, outfile, indent=4)
except:
print("ERROR: l4_parse_poetry: Unexpected error:", sys.exc_info()[0])
for frame in traceback.extract_tb(sys.exc_info()[2]):
fname,lineno,fn,text = frame
print ("DBG:: Error in %s on line %d" % (fname, lineno))
with open("exception_poetry.txt", "a") as outfile:
json.dump({'link': response.url, 'error': sys.exc_info()[0]}, outfile, indent=4)
def l3_parse_article_page(self,response):
"""
First check for the page containing div[@class='poem'] in the XPATH
1. If found then extract the poem and save in the database
2. If not found call l2_parse_author_page again because it contains list of poems in a journal
"""
try:
print "DBG:: l3_parse_article_page: Extracting poem from Article page"
p = response.xpath("//div[@id='mw-content-text']/div[@class='poem']//p").extract()
poem = " ".join(x.encode('utf-8') for x in p)
try:
h1 = response.xpath("//h1[@id='firstHeading']//text()").extract()[0].encode('utf-8')
title = h1
author = h1.split('/')[-1]
data = {}
data['poem'] = poem
data['url'] = response.url.encode('utf-8')
data['title'] = title
data['author'] = author.title()
data['language'] = 'hi'# Content of this site are in hindi
# Store these information in DB
save_to_db_poem(data)
except:
print "ERROR:: l3_parse_article_page: Title not found"
except:
# Extract article links from the Author page and generate request for each
try:
print "DBG:: l3_parse_article_page: Extracting poem links from Author page"
articles = response.xpath("//div[@id='mw-content-text']/ul/li/a/@href").extract()
articles_links = [self.domain_name+x for x in articles]
for url in articles_links:
# Check if the entry for ``url`` exist in db,
# Also find out the list of languages in which the content is there.
lang_list = get_language_list_for_url(url)
# Now crawl poetry page only for remaining langauge
for lang in (x for x in self.LANGUAGES if x not in lang_list):
#print "Visiting Article: ", url
yield scrapy.Request(url, callback=self.l3_parse_article_page)
except:
print "DBG:: Nothing found in Author page!!!"
print("ERROR: l3_parse_article_page: Unexpected error:", sys.exc_info()[0])
for frame in traceback.extract_tb(sys.exc_info()[2]):
fname,lineno,fn,text = frame
print ("DBG:: Error in %s on line %d" % (fname, lineno))
def parse_article_page(self, response):
self.count_visit_article = self.count_visit_article + 1
try:
print "Extracting poem ", self.count_visit_article, " from Article page"
p = response.xpath("/html/body/table//tr[2]/td/table//tr/td/pre").extract()
poem = " ".join(x.encode('utf-8') for x in p)
if poem:
title = response.xpath("/html/body/table//tr[2]/td/table//tr/td/p/strong/text()").extract()[0].encode('utf-8')
self.count_articles = self.count_articles + 1
data = {}
data['index'] = self.count_articles
data['title'] = title
data['author'] = ''
data['poem'] = poem
data['url'] = response.url.encode('utf-8')
# Store these information in DB
save_to_db_poem(data)
else:
print "First method failed trying another xpath"
p = response.xpath("/html/body/center/p").extract()
poem = " ".join(x.encode('utf-8') for x in p)
if poem:
title = response.xpath("/html/body/center/p[1]/strong/text()").extract()[0].encode('utf-8')
author = response.xpath("/html/body/center/p[1]/a/em/text()").extract()[0].encode('utf-8')
self.count_articles = self.count_articles + 1
data = {}
data['index'] = self.count_articles
data['title'] = "".join(x.encode('utf-8') for x in title)
data['author'] = "".join(x.encode('utf-8') for x in author)
data['poem'] = poem
data['url'] = response.url.encode('utf-8')
# Store these information in DB
save_to_db_poem(data)
else:
print "Both method failed write it in file for further processing"
except:
print "Error Article page!!!"
with open("up_exception_article.txt", "a") as outfile:
json.dump({'index': self.count_visit_article, 'link': response.url}, outfile, indent=4)
def parse_article_page(self, response):
self.count_visit_article = self.count_visit_article + 1
try:
print "Extracting poem ", self.count_visit_article, " from Article page"
p = response.xpath(
"/html/body/table//tr[2]/td/table//tr/td/pre").extract()
poem = " ".join(x.encode('utf-8') for x in p)
if poem:
title = response.xpath(
"/html/body/table//tr[2]/td/table//tr/td/p/strong/text()"
).extract()[0].encode('utf-8')
self.count_articles = self.count_articles + 1
data = {}
data['index'] = self.count_articles
data['title'] = title
data['author'] = ''
data['poem'] = poem
data['url'] = response.url.encode('utf-8')
# Store these information in DB
save_to_db_poem(data)
else:
print "First method failed trying another xpath"
p = response.xpath("/html/body/center/p").extract()
poem = " ".join(x.encode('utf-8') for x in p)
if poem:
title = response.xpath(
"/html/body/center/p[1]/strong/text()").extract(
)[0].encode('utf-8')
author = response.xpath(
"/html/body/center/p[1]/a/em/text()").extract(
)[0].encode('utf-8')
self.count_articles = self.count_articles + 1
data = {}
data['index'] = self.count_articles
data['title'] = "".join(x.encode('utf-8') for x in title)
data['author'] = "".join(x.encode('utf-8') for x in author)
data['poem'] = poem
data['url'] = response.url.encode('utf-8')
# Store these information in DB
save_to_db_poem(data)
else:
print "Both method failed write it in file for further processing"
except:
print "Error Article page!!!"
with open("up_exception_article.txt", "a") as outfile:
json.dump(
{
'index': self.count_visit_article,
'link': response.url
},
outfile,
indent=4)
def parse_poetry(self, response):
'''
Parse poetry page, extract poetry, and save.
'''
self.logger.debug("parse_poetry: IN.")
try:
# extract poem
stanza_selectors = response.xpath(
"//div[contains(@class,'mainContentBody')]/div[contains(@class,'poemPageContentBody')]/div/div"
)
poem = ''
for s in stanza_selectors:
line_selectors = s.xpath(".//p")
for l in line_selectors:
line = l.xpath(".//text()").extract()
line = ''.join(line)
line = line.strip()
##print line
poem = poem + line + '\n'
poem = poem + '\n'
poem = poem[0:-1] #strip last '\n' from the poem
#print poem
# extract title of the poem
title = response.xpath(
"//div[contains(@class,'mainContentBody')]/div[contains(@class,'poemPageContentHeader')]/h1/text()"
).extract()[0]
# extract poet name
# poet name must be in english.
poet_href = response.xpath(
"//div[contains(@class,'mainContentBody')]/div[contains(@class,'poemPageContentHeader')]//a[contains(@class,'ghazalAuthor')]/@href"
).extract()[0]
p = re.compile(
ur'poets/(.+)/') #href="/poets/anjum-tarazi/?lang=Hi"
poet = p.search(poet_href).group(1)
poet = poet.replace('-', ' ')
# check response.url for language information: https://.....xyz/?lang=hi
tmp = response.url
tmp = tmp.split('?')
url = tmp[0]
language = tmp[1].split('=')[1]
data = {}
data['poem'] = poem
data['url'] = url
data['title'] = title.strip()
data['author'] = poet.title()
data['language'] = language
# Store these information in DB
save_to_db_poem(data)
except:
self.logger.error("parse_poetry: %s", sys.exc_info()[0])
_trace = ''
for frame in traceback.extract_tb(sys.exc_info()[2]):
fname, lineno, fn, text = frame
self.logger.error("error in %s on line %d" % (fname, lineno))
_trace = _trace + "error in %s on line %d" % (fname, lineno)
with open(self.LOGFILE, "a") as outfile:
t = time.asctime(time.localtime(time.time()))
json.dump(
{
'link': response.url,
'error': 'parsing poetry failed',
'trace': _trace,
'time': t
},
outfile,
indent=4)
def parse_poetry_page(self, response):
"""
Parse the poetry page
1. First check if the page contains a poetry or not.
2. If Poetry not found, call parse_url_list because it may contain list of poems/collections/books.
3. If Poetry found, then extract the poem and save in the database.
"""
self.logger.debug("parse_poetry_page: IN.")
self.crawler.stats.inc_value('kangaroo/poetry_page_visit')
flag_poetry_found = True
##
# Find out if the page contains a Poetry
try:
p = response.xpath(
"//div[@id='mw-content-text']/div[@class='poem']//p").extract(
)
if len(p) is 0:
# in some pages, the poetry is not in div[@class='poem']
# e.g. http://www.kavitakosh.org/kk/%E0%A4%AE%E0%A4%A8_%E0%A4%B2%E0%A4%BE%E0%A4%97%E0%A5%8D%E0%A4%AF%E0%A5%8B_%E0%A4%AE%E0%A5%87%E0%A4%B0%E0%A5%8B_%E0%A4%AF%E0%A4%BE%E0%A4%B0_%E0%A4%AB%E0%A4%BC%E0%A4%95%E0%A5%80%E0%A4%B0%E0%A5%80_%E0%A4%AE%E0%A5%87%E0%A4%82_/_%E0%A4%95%E0%A4%AC%E0%A5%80%E0%A4%B0
p = response.xpath("//div[@id='mw-content-text']//p").extract()
if len(p):
# Now check for the length of the text under the <p>
# Because it may contains empty <p>, or one line text stating poetry is not available etc.
# Here we assume that the length of Poetry should be greater than ARTICLE_MIN_LEN.
p_t = response.xpath(
"//div[@id='mw-content-text']//p/text()").extract()
p_t = "".join(x.encode('utf-8') for x in p_t)
if len(p_t) <= ARTICLE_MIN_LEN:
flag_poetry_found = Flase
else:
flag_poetry_found = Flase
else:
# Check the length of the article
p_t = response.xpath(
"//div[@id='mw-content-text']/div[@class='poem']//p/text()"
).extract()
p_t = "".join(x.encode('utf-8') for x in p_t)
if len(p_t) <= ARTICLE_MIN_LEN:
flag_poetry_found = Flase
except:
self.logger.error("parse_poetry_page: xpath error.")
flag_poetry_found = False
##
# If poetry not found...
if (flag_poetry_found is False):
# It may contains list of poems/collections/books
self.logger.info(
'parse_poetry_page: no poetry found on this page.')
return self.parse_url_list(response)
##
# If Poetry found...
# This is poetry page, extract the poetry
try:
h1 = response.xpath("//h1[@id='firstHeading']//text()").extract(
)[0].encode('utf-8')
h1_list = h1.split('/')
title = '/'.join(h1_list[:-1])
poet = h1_list[-1]
# Process and create Poetry
poem = " ".join(x.encode('utf-8') for x in p)
data = {}
data['poem'] = poem
data['url'] = response.url.encode('utf-8')
data['title'] = title
data['author'] = poet.strip()
data['language'] = 'hi' # Content of this site are in hindi
# Store these information in DB
save_to_db_poem(data)
self.crawler.stats.inc_value('kangaroo/poetry_found')
except:
self.logger.error("parse_poetry_page: %s", sys.exc_info()[0])
_trace = ''
for frame in traceback.extract_tb(sys.exc_info()[2]):
fname, lineno, fn, text = frame
self.logger.error("error in %s on line %d" % (fname, lineno))
_trace = _trace + "error in %s on line %d" % (fname, lineno)
with open(self.LOGFILE, "a") as outfile:
t = time.asctime(time.localtime(time.time()))
json.dump(
{
'link': response.url,
'error': 'parsing poetry failed',
'trace': _trace,
'time': t
},
outfile,
indent=4)
