#!/usr/bin/env python
# getbib by kkumer@phy.hr 
# v. 1.5 2007-08-24 

# Gets BibTeX info for an article from SPIRES database

# Usage:
# getbib quant-ph/0012149 
# getbib xxxxxxx  (where xxxxxx is SPIRES key which can be
#     found by looking there at links on keywords, bibtex's etc.)
#     This is useful for older papers which are not on arXiv org!


##########################################

# User modifiable options:

# SPIRES server:
#server="www.slac.stanford.edu" 
server="www-library.desy.de" 
# path to the SPIRES interface:
path="/spires/find/hep/www"
path="/cgi-bin/spiface/find/hep/www"
# GGI commands (%s stands for ID that will be substituted)
#cgibibtex="?rawcmd=fin+eprint+%s&FORMAT=WWWBRIEFBIBTEX"
cgibibtex="?key=%s&FORMAT=WWWBRIEFBIBTEX"
cgisummary="?rawcmd=fin+eprint+%s&FORMAT=WWW"
cgikeywords="topics?key=%s"

##########################################


import httplib,sys,re,string,commands,time


# Checking input
if len(sys.argv)!=2:
    print ''
    print 'Usage: getbib archive/number'
    print ''
    sys.exit(1)
else:
    id=sys.argv[1]


class hep:
    def __init__(self,command,paper_id):
        h=httplib.HTTP(server) #connecting
        # sending request the file
        h.putrequest('GET', path+command % paper_id)
        # servers insist on the following header
        h.putheader('User-Agent', 'getbib/1.0beta by kkumer[at]phy.hr')
        h.endheaders()
        self.errcode,self.errmsg,self.headers=h.getreply()
        if self.errcode!=200:
            print "Errcode: "+str(self.errcode)
            print "Server error!"
            sys.exit(1)
        # Getting data (gzipped paper or "invalid id" or "wait for PS")
        f=h.getfile()
        self.data=f.read()
        f.close()

# if hep number given on command line:
#     get summary page and extract paper "key" ID from there
if re.search('[/.]',id):
    id=string.replace(id,'/','%2F')  
    page=hep(cgisummary,id)
    try:
        keyid=re.search(r'www\?key=(.*?)&',page.data).group(1)
    except:
        sys.stderr.write('ERROR: No page. Are you sure paper %s exists?\n' % string.replace(id,'%2F','/'))
        sys.exit(1)
# Otherwise expect that argument given is keyid itself
else:
    keyid=id

# Getting  BiBTeX entry
page=hep(cgibibtex,keyid)
try:
    bibtex=re.search('@Article.*\n}\n',page.data,re.DOTALL).group()
except:
    sys.stderr.write('ERROR: No BibTeX. Are you sure paper %s exists?\n' % keyid)
    sys.exit(1)

# Getting  keywords
page=hep(cgikeywords,keyid)
kwpat=re.compile(r'www\?dk=.*?>(.*?)</a>',re.DOTALL)
keywords_raw=kwpat.findall(page.data)
# cleaning up
keywords=[re.sub('\n',' ',str).strip() for str in keywords_raw]
# putting in BibTeX format
kwlist=['     keywords  = "%s",\n' % str for str in keywords]
# appending it
bibtex=re.sub('"\n}\n','",\n'+string.join(kwlist),bibtex)

# appending SPIRES key id
bibtex=bibtex+'     spireskey  = "%s",\n' % keyid

# appending pdf, group and comments field
bibtex=bibtex+'     pdf  = "%s",\n     group  = "",\n     comments  = ""\n}\n' % string.replace(id+'.pdf','%2F','-')
print bibtex