#!/usr/bin/env python

import re, sys, urllib2, cookielib,sys
from urlparse import urlparse
from urllib import urlencode
from urllib2 import urlopen
import BeautifulSoup
from html5lib import treebuilders
import html5lib
import warnings

warnings.simplefilter("ignore",DeprecationWarning)



# error messages (taken from iop.py)
ERR_STR_PREFIX = 'status\terr\t'
ERR_STR_FETCH = 'Unable to fetch the page: '
ERR_STR_TRY_AGAIN = 'The server may be down.  Please try later.'
ERR_STR_NO_DOI = 'No document object identifier found on the page: '
ERR_STR_NO_BIBTEX = 'No BibTeX entry found for the DOI: '
ERR_STR_NO_URL = 'No URL found in the BibTeX entry for the DOI: '
ERR_STR_REPORT = 'Please report the error to plugins@citeulike.org.'

# read url from std input
url = sys.stdin.readline()

# get rid of the newline at the end
url = url.strip()

cookie_jar = cookielib.CookieJar()

handlers = []
if "--debug" in sys.argv:
	handlers.append( urllib2.HTTPHandler(debuglevel=True) )
handlers.append( urllib2.HTTPCookieProcessor(cookie_jar) )

opener=urllib2.build_opener(*handlers)
opener.addheaders = [('User-agent', 'lwp-request/5.810')]
urllib2.install_opener(opener)

# link.aps.org/doi is basically a DOI resolver, but not necessarily to
# aps.org
if re.search(r'/doi/',url):
	f = urlopen(url)
	# Open the URL, which automatically follows redirects.
	# I think this just does similar to HEAD (i.e, no data downloaded)
	# but not sure.
	url =  f.geturl()
	# if not an APS url we can handle, let another plugin handle it
	if not re.match(r'http://[^/]*(prola|aps)[^/]*/abstract', url):
		print "status\tredirect\t%s" %  url
		sys.exit(0)



# we'll be given a url of the form http://prola.aps.org/abstract/PRA/v18/i3/p787_1
#  http://prola.aps.org/abstract/journal/volume/issue/page
# the ris resource is located at http://prola.aps.org/export/___/v__/i__/p__?type=ris
# use type=bibtex for bibtex
# should we prefer one to the other?

host = 'http://'+ urlparse(url).netloc

address = url.split('abstract')[1]

bibtexurl = host +'/export' + address + '?type=bibtex'
try:
	f = urlopen(bibtexurl);
except:
	print ERR_STR_PREFIX + ERR_STR_FETCH + bibtexurl + '.  ' + ERR_STR_TRY_AGAIN
	sys.exit(1)
bibtex = f.read()
bibtex = bibtex.split('\n',2)[2]
bibtex = re.sub(r'\\ifmmode .*?\\else (.*?)\\fi\{\}',r'\1',bibtex)

#okay so that handles most of it  but we still need to get the actual title of the journal,
# and we need to find the abstract if it has one.
absurl = host +'/abstract' + address
try:
	f = urllib2.urlopen(absurl)
except:
	print ERR_STR_PREFIX + ERR_STR_FETCH + absurl + '.  ' + ERR_STR_TRY_AGAIN
	sys.exit(1)
content = f.read()

parser = html5lib.HTMLParser(tree=treebuilders.getTreeBuilder("beautifulsoup"))
soup = parser.parse(content)

abstract = ''

div = soup.find('div',attrs={'class':'aps-abstractbox'})
if div:
	abs = []
	for t in div.findAll(text=True):
		abs.append(t);
	abstract =  " ".join(abs).strip()

# We would much much rather extract the DOI from the bibtex feed, since it has a
# much more std structure, (that isn't as subject to change as the page conent.
# I don't ever expect that we should get to the else clause.  If the re is
# going to fail, we've probably failed to get the ris earlier.

match = re.search(r"""^  doi = \{(10\..*)\},$""", bibtex, re.MULTILINE)
if match:
	doi = match.group(1)
else:
	print ERR_STR_PREFIX + ERR_STR_NO_DOI + absurl + '.  ' + ERR_STR_REPORT
	sys.exit(1)

#We can look at the letter code in the address (/address/xxxxx/)
#to get the journal name.
#Journal  	URL
#Phys. Rev. A	http://pra.aps.org/ PRA
#Phys. Rev. B	http://prb.aps.org/ PRB
#Phys. Rev. C	http://prc.aps.org/ PRC
#Phys. Rev. D	http://prd.aps.org/ PRD
#Phys. Rev. E	http://pre.aps.org/ PRE
#Phys. Rev. Lett.	http://prl.aps.org/ PRL
#Phys. Rev. ST Phys. Educ. Res.	http://prst-per.aps.org/ PRSTPER
#Phys. Rev. ST Accel. Beams	http://prst-ab.aps.org/ PRSTAB
#Rev. Mod. Phys.	http://rmp.aps.org/ RMP
#PROLA Archive	http://prola.aps.org/ PR
#Physical Review (Series I) PRI

journal_key = address.split("/")[1]

journalmap = {'PRA' : 'Physical Review A',
              'PRB' : 'Physical Review B',
              'PRC' : 'Physical Review C',
              'PRD' : 'Physical Review D',
              'PRE' : 'Physical Review E',
              'PRL' : 'Physical Review Letters',
              'PRI' : 'Physical Review (Series I)',
              'PR' : 'Physical Review Online Archive (Prola)',
              'RMP' : 'Reviews of Modern Physics',
              'PRSTAB' : 'Physical Review Special Topics - Accelerators and Beams',
              'PRSTPER' : 'Phys. Rev. ST Phys. Educ. Res.'
             }
journal = journalmap[journal_key]

print 'begin_tsv'
print 'journal\t' + journal
if abstract:
	print 'abstract\t' + abstract
print "linkout\tPROLA\t\t%s\t\t" % (address[1:])
print "linkout\tDOI\t\t%s\t\t" % (doi)
print 'doi\t' + doi
print 'end_tsv'
print 'begin_bibtex'
print bibtex
print 'end_bibtex'
print 'status\tok'