RhodeCode

"""

  BibTeX to Doxygen converter

  Usage: python bib2dox.py bibfile.bib > bibfile.dox

  This code is the modification of the BibTeX to XML converter

  **********************************************************************

  Decoder for bibliographic data, BibTeX

  Usage: python bibtex2xml.py bibfile.bib > bibfile.xml

  v.8

  (c)2002-06-23 Vidar Bronken Gundersen

  http://bibtexml.sf.net/

  Reuse approved as long as this notification is kept.

  Licence: GPL.

  Contributions/thanks to:

  Egon Willighagen, http://sf.net/projects/jreferences/

  Richard Mahoney (for providing a test case)

  Editted by Sara Sprenkle to be more robust and handle more bibtex features.

  (c) 2003-01-15

  1.  Changed bibtex: tags to bibxml: tags.

  2.  Use xmlns:bibxml="http://bibtexml.sf.net/"

  3.  Allow spaces between @type and first {

  4.  "author" fields with multiple authors split by " and "

      are put in separate xml "bibxml:author" tags.

  5.  Option for Titles: words are capitalized

      only if first letter in title or capitalized inside braces

  6.  Removes braces from within field values

  7.  Ignores comments in bibtex file (including @comment{ or % )

  8.  Replaces some special latex tags, e.g., replaces ~ with ' '

  9.  Handles bibtex @string abbreviations

        --> includes bibtex's default abbreviations for months

        --> does concatenation of abbr # " more " and " more " # abbr

  10. Handles @type( ... ) or @type{ ... }

  11. The keywords field is split on , or ; and put into separate xml

      "bibxml:keywords" tags

  12. Ignores @preamble

  Known Limitations

  1.  Does not transform Latex encoding like math mode and special

      latex symbols.

  2.  Does not parse author fields into first and last names.

      E.g., It does not do anything special to an author whose name is

      in the form LAST_NAME, FIRST_NAME

      In "author" tag, will show up as

      <bibxml:author>LAST_NAME, FIRST_NAME</bibxml:author>

  3.  Does not handle "crossref" fields other than to print

      <bibxml:crossref>...</bibxml:crossref>

  4.  Does not inform user of the input's format errors.  You just won't

      be able to transform the file later with XSL

  You will have to manually edit the XML output if you need to handle

  these (and unknown) limitations.

"""

import string, re

# set of valid name characters

valid_name_chars = '[\w\-:]'

#

# define global regular expression variables

#

author_rex = re.compile('\s+and\s+')

rembraces_rex = re.compile('[{}]')

capitalize_rex = re.compile('({[^}]*})')

# used by bibtexkeywords(data)

keywords_rex = re.compile('[,;]')

# used by concat_line(line)

concatsplit_rex = re.compile('\s*#\s*')

# split on {, }, or " in verify_out_of_braces

delimiter_rex = re.compile('([{}"])',re.I)

field_rex = re.compile('\s*(\w*)\s*=\s*(.*)')

data_rex = re.compile('\s*(\w*)\s*=\s*([^,]*),?')

url_rex = re.compile('\\\url\{([^}]*)\}')

#

# styles for html formatting

#

divstyle = 'margin-top: -4ex; margin-left: 8em;'

#

# return the string parameter without braces

#

def transformurls(str):

    return url_rex.sub(r'<a href="\1">\1</a>', str)

#

# return the string parameter without braces

#

def removebraces(str):

    return rembraces_rex.sub('', str)

#

# latex-specific replacements

# (do this after braces were removed)

#

def latexreplacements(line):

    line = string.replace(line, '~', ' ')

    line = string.replace(line, '\\\'a', 'á')

    line = string.replace(line, '\\"a', 'ä')

    line = string.replace(line, '\\\'e', 'é')

    line = string.replace(line, '\\"e', 'ë')

    line = string.replace(line, '\\\'i', 'í')

    line = string.replace(line, '\\"i', 'ï')

    line = string.replace(line, '\\\'o', 'ó')

    line = string.replace(line, '\\"o', 'ö')

    line = string.replace(line, '\\\'u', 'ú')

    line = string.replace(line, '\\"u', 'ü')

    line = string.replace(line, '\\H o', 'õ')

    line = string.replace(line, '\\H u', 'ü')   # ũ does not exist

    line = string.replace(line, '\\\'A', 'Á')

    line = string.replace(line, '\\"A', 'Ä')

    line = string.replace(line, '\\\'E', 'É')

    line = string.replace(line, '\\"E', 'Ë')

    line = string.replace(line, '\\\'I', 'Í')

    line = string.replace(line, '\\"I', 'Ï')

    line = string.replace(line, '\\\'O', 'Ó')

    line = string.replace(line, '\\"O', 'Ö')

    line = string.replace(line, '\\\'U', 'Ú')

    line = string.replace(line, '\\"U', 'Ü')

    line = string.replace(line, '\\H O', 'Õ')

    line = string.replace(line, '\\H U', 'Ü')   # Ũ does not exist

    return line

#

# copy characters form a string decoding html expressions (&xyz;)

#

def copychars(str, ifrom, count):

    result = ''

    i = ifrom

    c = 0

    html_spec = False

    while (i < len(str)) and (c < count):

        if str[i] == '&':

            html_spec = True;

            if i+1 < len(str):

                result += str[i+1]

            c += 1

            i += 2

        else:

            if not html_spec:

                if ((str[i] >= 'A') and (str[i] <= 'Z')) or \

                   ((str[i] >= 'a') and (str[i] <= 'z')):

                    result += str[i]

                    c += 1

            elif str[i] == ';':

                html_spec = False;

            i += 1

    return result

# 

# Handle a list of authors (separated by 'and').

# It gives back an array of the follwing values:

#  - num: the number of authors,

#  - list: the list of the author names,

#  - text: the bibtex text (separated by commas and/or 'and')

#  - abbrev: abbreviation that can be used for indicate the

#    bibliography entries

#

def bibtexauthor(data):

    result = {}

    bibtex = ''

    result['list'] = author_rex.split(data)

    result['num'] = len(result['list'])

    for i, author in enumerate(result['list']):

        # general transformations

        author = latexreplacements(removebraces(author.strip()))

        # transform "Xyz, A. B." to "A. B. Xyz"

        pos = author.find(',')

        if pos != -1:

            author = author[pos+1:].strip() + ' ' + author[:pos].strip()

        result['list'][i] = author

        bibtex += author + '#'

    bibtex = bibtex[:-1]

    if result['num'] > 1:

        ix = bibtex.rfind('#')

        if result['num'] == 2:

            bibtex = bibtex[:ix] + ' and ' + bibtex[ix+1:]

        else:

            bibtex = bibtex[:ix] + ', and ' + bibtex[ix+1:]

    bibtex = bibtex.replace('#', ', ')

    result['text'] = bibtex