#! /usr/bin/env python
#

import sys
import bibtexparser
from   bibtexparser.bparser import BibTexParser
import bibtexparser.customization 

areas = { u'cv':(u'Computer Vision', u'Images/Vision-100.png'),
          u'ai':(u'Artificial Intelligence', u'Images/AI-100.png'),
          u'ml':(u'Machine Learning', u'Images/ML-80.png'),
          u'forensics':(u'Forensics', u'Images/Forensics-80.png'),
          u'access':(u'Accessibility', u'Images/Access-80.png'),
          u'cg':(u'Computer Graphics', u'Images/Graphics-80.png'),
          u'asl':(u'Sign Language', u'Images/ASL-100.png'),
          u'wavelets':(u'Wavelets', u'Images/Wavelet-100.png'),
          u'sport': (u'Sport', u'Images/Sport-80.png'),
          u'uncert': (u'Uncertanties', u'Images/Uncert-100.png'),
          u'virtcrd': (u'Virtual Crowds', u'Images/Crowd-80.png'),
          u'med': (u'Biomedical', u'Images/Med-64.png'),
          u'biblio': (u'Bibliometrics', u'Images/Book-80.png')
          }

areas_keys = areas.keys()


def stripnline(record):
    if u"author" in record:
        record[u"author"] = record[u"author"].replace(u'\n', u' ').strip()

    if "title" in record:
        record[u"title"] = record[u"title"].replace(u'\n', u' ').strip()
        
    return record


def customizations(record):
    """Use some functions delivered by the library

    :param record: a record
    :returns: -- customized record
    """
    record = bibtexparser.customization.type(record)
#    record = author(record)
#    record = editor(record)
#    record = journal(record)
    record = bibtexparser.customization.keyword(record)
#    record = link(record)
#    record = page_double_hyphen(record)
#    record = doi(record)
    record = stripnline(record)
    return record


#
#  input is the variable with all the text of the bibtex file
#  output is a list of dictionaries
#
def parseBibtex( bibtex_fname ):

    with open(bibtex_fname) as bibtex_file:
        parser = BibTexParser()
        parser.customization = customizations
        bibtex_database = bibtexparser.load(bibtex_file, parser=parser)


    # fix encoding
    #//    source = parser.checkEncoding(source)

    # fix latex special characters, among other things
#    source = parser.preprocess(source) #.encode('utf-8'))

    # get list of dictionaries
    return bibtex_database.entries



def toBibtex (ref,
            valid_fields=[u'title', u'editor', u'author', u'journal', u'booktitle',
                          u'publisher', u'school', u'issn', u'volume', u'number',  
                          u'pages', u'year', u'doi', u'pdf'],
              omit_fields=[]):

    omit   = [each.lower() for each in omit_fields]
    fields = [each.lower() for each in valid_fields]

#    ttable = [('&', r'\&'),
#              ('~', r'\~'),
#              ('—',r'---'),
#              ('–', r'--'),
#              ]

    bib_key = ref[u'ID']
    ref_type = ref[u'ENTRYTYPE'].replace(u'Reference','')
    bibtex = u'\n@' + ref_type + u'{' + bib_key + u',\n' # '%s{%s,\n" %  (ref_type, bib_key)

    ref_keys = ref.keys()

    for k in fields:
        if k in ref_keys and k not in omit:
            if type(ref[k]) == list:
                nv = ref[k][0]
            else:
                nv = ref[k]

#            for a,b in ttable:
#                nv = nv.replace(a,b)

            bibtex = bibtex + u'  '
            bibtex = bibtex + k
            bibtex = bibtex + u' = {'   
            #            bibtex = bibtex + unicode(nv.encode('utf8'))
            #            bibtex = bibtex + unicode(nv, encoding='utf8')
            bibtex = bibtex + nv
            bibtex = bibtex + u'},\n'
            
    # remove trailing command
    bibtex = bibtex[0:-2] + u"}\n"

    return bibtex



def toHTML (ref,
            numb=0,
            valid_fields=[u'title', u'editor', u'author', u'journal', u'booktitle',
                          u'publisher', u'school', u'issn', u'volume', u'number',  
                          u'pages', u'year', u'doi', u'pdf', u'abstract'],
            omit_fields=[]):

    omit = [each.lower() for each in omit_fields]
    fields = [each.lower() for each in valid_fields]

    ref_type = ref[u'ENTRYTYPE'].replace('Reference','').lower()

    bib_key = ref[u'ID']
    bid = 'b' + bib_key
    ref_keys = ref.keys()

    html = u'<table width="100%" class="pub">\n'
    html += u'<col align="center" width=230px />\n'
    html += u'<col align="left" />\n'
    html += u'<col align="left" />\n'
    html += u'<tr>\n'
    html += u'<td class="keywords" align="center">\n'
    #    html += u'<td class="keywords" align="center" padding="20px">\n'
    if u'keyword' in ref_keys:
#        for k,v in ref[u'keyword'].iteritems():
#            print k, v
        for k in ref[u'keyword']:
            lk = k.lower()
            if lk in areas_keys:
                html += u'<img class="keyw" src="%s" alt="%s" title="%s" align="middle" />\n' % (areas[lk][1], areas[lk][0], areas[lk][0])
    else:
        html += u'No keywords!\n'
    
    html += u'</td>\n'

        
    html += u'<td><div class="%s" name="%s" ><a name="%s">\n' % (ref_type, bib_key, bib_key)
    #    if numb>0:
    #        html += u'%d. \n' % numb


    for k in fields:
        if k in ref_keys and k not in omit:
            if type(ref[k]) == list:
                v = ref[k][0]
            else:
                v = ref[k]
            if numb>0 and k=='title':
                html += u'<span class="%s">%d. %s </span>\n' % (k, numb, v)
            elif k=='issn':
                html += u'<span class="%s">ISSN: %s, </span>\n' % (k, v)
            elif k=='volume':
                html += u'<span class="%s">Vol. %s, </span>\n' % (k, v)
            elif k=='number':
                html += u'<span class="%s">Num. %s, </span>\n' % (k, v)
            elif k=='pages':
                html += u'<span class="%s">pgs. %s, </span>\n' % (k, v)
            elif k=='doi':
                html += u'<span class="%s"> <a href="http://dx.doi.org/%s">doi:%s</a> </span>\n' % (k, v, v)
            elif k == 'pdf':
                html += u'<span class="pdf"> <a href="papers/%s">PDF.</a> </span>\n' % v
            elif k == 'abstract':
                html += u'<span class="linkabs"><a href="javascript:unhide(\'%s\');">Show/Hide Abstract.</a></span>\n' % (bid + u'-abs')
                html += u'<div class="hidden" id="%s"><div class="abstract"><b>Abstract:</b> %s\n</div></div>\n' % (bid + u'-abs', v)
            else:
                html += u'<span class="%s"> %s </span>\n' % (k, v)
    html += u'<span class="linkbib"><a href="javascript:unhide(\'%s\');">Show/Hide Bibtex.</a></span>\n' % bid
    html += u'<div class="hidden" id="%s"><pre>%s\n</pre></div>\n' % (bid, toBibtex(ref))


    html += u'\n  </a></div></td>\n'


    html += u'  </tr>\n'
    html += u'</table>\n\n'

    return html

def referencesByYear(bibs):
    #    byYears = {first:0, last: 3000}
    byYears = {}

    for ref in bibs:
        if u'year' in ref.keys():   
            year = int(ref[u'year'])
            if year not in byYears.keys():
                byYears[year] = [ ref ]
            else:
                byYears[year].append(ref)

    return byYears


def preamble():
    a = """<html xmlns="http://www.w3.org/1999/xhtml" xml:lang="en" lang="en"><head>
    
<link href="publications.css" type="text/css" rel="stylesheet" />

<meta content="text/html; charset=utf-8" http-equiv="Content-Type" />

<script type="text/javascript">
 function unhide(divID) {
 var item = document.getElementById(divID);
 if (item) {
 item.className=(item.className=='hidden')?'unhidden':'hidden';
 }
 }
</script>
  
<title>Publications</title>

</head><body>
<div class="header1"><br />
<table class="header1" cellpadding="0" cellspacing="0">
  <colgroup span="1" width="80%" /><colgroup span="1" width="20%" /><tbody>
    <tr>
      <td>
      <h1>Publications</h1>
      </td>
      <td><!--#include virtual="navigation.html"-->&nbsp; </td>
    </tr>
  </tbody>
</table>
</div>

<p>
The full CV, at the CNPq CV repository, is available
<a href="http://lattes.cnpq.br/3575273076267374 ">here</a>
(in portuguese).
</p>"""
    return a

def posamble():
    a = """<h2>&nbsp;</h2>

<div>
<address>Siome Klein Goldenstein:
[myfirstname](at) ic unicamp br </address>
<!-- hhmts start -->Lastmodified:&nbsp;
<!--#flastmod file="publications.shtml" -->
<!-- hhmts end -->
</div>

</body></html>"""
    return a


def filterType( entry, valid_types ):
    vtypes = [each.lower() for each in valid_types]
    tp = entry[u'ENTRYTYPE']
    tp = tp.lower()
    tp = tp.replace(u'reference','')

    return tp in vtypes
    


def generateWebpage (bibs,
                     pord=[ ([u'article'], u'Journal Publications', u'byYear'),
                            ([u'patent'], u'Patents', None),
                            ([u'book', u'inbook'], u'Books and Book Chapters', None),
                            ([u'inproceedings'], u'Conferences and Workshops', u'byYear'),
                            ([u'incollection'], u'Magazines', None),
                            ([u'phdthesis', u'mastersthesis'], u'Thesis', None)
                          ]
                     ):

    webpg = preamble() + u'\n\n'

    webpg += u'<br /> <br /> <p style="text-align: center;"> '
    for ent in pord:
      webpg += u'&nbsp;&nbsp;&nbsp;&nbsp;'
      webpg += u'    <a href="#' + unicode(ent[1]) + u'"> ' + unicode(ent[1]) + u'</a> &nbsp;&nbsp;&nbsp;&nbsp;'
    webpg += u' </p"> <br /> <br />'


    for ent in pord:
        fbib = filter(lambda x: filterType(x, ent[0]), bibs)
        print >>sys.stderr, u'%s: has %d elements out of %d total' % (ent[1],len(fbib),len(bibs))
        sys.stderr.flush()
        
        webpg += u'<h2> <a name="%s" > %s </a></h2>\n' % (ent[1],ent[1])
#        webpg += u'<ol reversed="reversed">'

        numb = len(fbib)
        byear = referencesByYear(fbib)
        years = byear.keys()
        years.sort(reverse=True)

        for y in years:
            if ent[2] == u'byYear':
                webpg += u'<h3> <span> %d </span></h3>\n' % y

            for r in byear[y]:
                webpg += u'\n %s \n' % toHTML(r, numb)
                numb -=1

    webpg += posamble()

    return webpg


if __name__ == "__main__":
    if len(sys.argv) != 3:
        print '\n' +  sys.argv[0] + ': requires name of the bibtex file and output file!\n'
        sys.exit()
    fname = sys.argv[1]
#    bibfile = codecs.open(fname,encoding='latin_1').read()
    bibs = parseBibtex( fname )
    s = generateWebpage( bibs )

    with open(sys.argv[2],'w') as fout:
        fout.write(s.encode('utf8'))
#    sys.stdout.write(s)
#    print s.encode('latin_1')
#    print (s)