lyx_mirror/lib/lyx2lyx/lyx_1_6.py

# This file is part of lyx2lyx
# -*- coding: utf-8 -*-
# Copyright (C) 2007-2008 The LyX Team <lyx-devel@lists.lyx.org>
#
# This program is free software; you can redistribute it and/or
# modify it under the terms of the GNU General Public License
# as published by the Free Software Foundation; either version 2
# of the License, or (at your option) any later version.
#
# This program is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with this program; if not, write to the Free Software
# Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA  02111-1307, USA.

""" Convert files to the file format generated by lyx 1.6"""

import re
import unicodedata
import sys, os

from parser_tools import find_token, find_end_of, find_tokens, get_value, get_value_string

####################################################################
# Private helper functions

def find_end_of_inset(lines, i):
    " Find end of inset, where lines[i] is included."
    return find_end_of(lines, i, "\\begin_inset", "\\end_inset")

# WARNING!
# DO NOT do this:
#   document.body[i] = wrap_insert_ert(...)
# wrap_into_ert may returns a multiline string, which should NOT appear
# in document.body. Insetad, do something like this:
#   subst = wrap_inset_ert(...)
#   subst = subst.split('\n')
#   document.body[i:i+1] = subst
#   i+= len(subst) - 1
# where the last statement resets the counter to accord with the added
# lines.
def wrap_into_ert(string, src, dst):
    '''Within string, replace occurrences of src with dst, wrapped into ERT
       E.g.: wrap_into_ert('sch\"on', "\\", "\\backslash") is:
       sch<ERT>\\backslash</ERT>"on'''
    return string.replace(src, '\n\\begin_inset ERT\nstatus collapsed\n\\begin_layout Standard\n'
      + dst + '\n\\end_layout\n\\end_inset\n')

def put_cmd_in_ert(string):
    string = string.replace('\\', "\\backslash\n")
    string = "\\begin_inset ERT\nstatus collapsed\n\\begin_layout Standard\n" \
      + string + "\n\\end_layout\n\\end_inset"
    return string

def add_to_preamble(document, text):
    """ Add text to the preamble if it is not already there.
    Only the first line is checked!"""

    if find_token(document.preamble, text[0], 0) != -1:
        return

    document.preamble.extend(text)

# Convert a LyX length into a LaTeX length
def convert_len(len):
    units = {"text%":"\\backslash\ntextwidth", "col%":"\\backslash\ncolumnwidth",
             "page%":"\\backslash\npagewidth", "line%":"\\backslash\nlinewidth",
             "theight%":"\\backslash\ntextheight", "pheight%":"\\backslash\npageheight"}

    # Convert LyX units to LaTeX units
    for unit in units.keys():
        if len.find(unit) != -1:
            len = '%f' % (len2value(len) / 100)
            len = len.strip('0') + units[unit]
            break

    return len

# Return the value of len without the unit in numerical form.
def len2value(len):
    result = re.search('([+-]?[0-9.]+)', len)
    if result:
        return float(result.group(1))
    # No number means 1.0
    return 1.0

# Unfortunately, this doesn't really work, since Standard isn't always default.
# But it's as good as we can do right now.
def find_default_layout(document, start, end):
    l = find_token(document.body, "\\begin_layout Standard", start, end)
    if l == -1:
        l = find_token(document.body, "\\begin_layout PlainLayout", start, end)
    if l == -1:
        l = find_token(document.body, "\\begin_layout Plain Layout", start, end)
    return l

def get_option(document, m, option, default):
    l = document.body[m].find(option)
    val = default
    if l != -1:
        val = document.body[m][l:].split('"')[1]
    return val

def remove_option(document, m, option):
    l = document.body[m].find(option)
    if l != -1:
        val = document.body[m][l:].split('"')[1]
        document.body[m] = document.body[m][:l-1] + document.body[m][l+len(option + '="' + val + '"'):]
    return l

def set_option(document, m, option, value):
    l = document.body[m].find(option)
    if l != -1:
        oldval = document.body[m][l:].split('"')[1]
        l = l + len(option + '="')
        document.body[m] = document.body[m][:l] + value + document.body[m][l+len(oldval):]
    else:
        document.body[m] = document.body[m][:-1] + ' ' + option + '="' + value + '">'
    return l


def read_unicodesymbols():
    " Read the unicodesymbols list of unicode characters and corresponding commands."
    pathname = os.path.abspath(os.path.dirname(sys.argv[0]))
    fp = open(os.path.join(pathname.strip('lyx2lyx'), 'unicodesymbols'))
    spec_chars = []
    # Two backslashes, followed by some non-word character, and then a character
    # in brackets. The idea is to check for constructs like: \"{u}, which is how
    # they are written in the unicodesymbols file; but they can also be written
    # as: \"u.
    r = re.compile(r'\\\\(\W)\{(\w)\}')
    for line in fp.readlines():
        if line[0] != '#' and line.strip() != "":
            line=line.replace(' "',' ') # remove all quotation marks with spaces before
            line=line.replace('" ',' ') # remove all quotation marks with spaces after
            line=line.replace(r'\"','"') # replace \" by " (for characters with diaeresis)
            try:
                [ucs4,command,dead] = line.split(None,2)
                if command[0:1] != "\\":
                    continue
                spec_chars.append([command, unichr(eval(ucs4))])
            except:
                continue
            m = r.match(command)
            if m != None:
                command = "\\\\"
                # If the character is a double-quote, then we need to escape it, too,
                # since it is done that way in the LyX file.
                if m.group(1) == "\"":
                    command += "\\"
                command += m.group(1) + m.group(2)
                spec_chars.append([command, unichr(eval(ucs4))])
    fp.close()
    return spec_chars


def extract_argument(line):
    'Extracts a LaTeX argument from the start of line. Returns (arg, rest).'

    if not line:
        return (None, "")

    bracere = re.compile("(\s*)(.*)")
    n = bracere.match(line)
    whitespace = n.group(1)
    stuff = n.group(2)
    brace = stuff[:1]
    if brace != "[" and brace != "{":
        return (None, line)

    # find closing brace
    remain = stuff[1:]
    pos = 0
    num = 1
    term = "}"
    if brace == "[":
        term = "]"
    skip = False
    for c in remain:
        if skip:
            skip = False
        elif c == "\\":
            skip = True
        elif c == brace:
            num += 1
        elif c == term:
            num -= 1
        if c == 0:
            break
        pos += 1
    if num != 0:
        # We never found the matching brace
        # So, to be on the safe side, let's just return everything
        # which will then get wrapped as ERT
        return (line, "")
    return (line[:pos + 1], line[pos + 1:])


def latex2ert(line):
    '''Converts LaTeX commands into ERT. line may well be a multi-line
       string when it is returned.'''
    if not line:
        return line

    retval = ""
    ## FIXME Escaped \ ??
    # This regex looks for a LaTeX command---i.e., something of the form
    # "\alPhaStuFF", or "\X", where X is any character---where the command
    # may also be preceded by an additional backslash, which is how it would
    # appear (e.g.) in an InsetIndex.
    labelre = re.compile(r'(.*?)\\?(\\(?:[a-zA-Z]+|.))(.*)')

    m = labelre.match(line)
    while m != None:
        retval += m.group(1)
        cmd = m.group(2)
        end = m.group(3)

        while True:
            (arg, rest) = extract_argument(end)
            if arg == None:
                break
            cmd += arg
            end = rest
        # If we wanted to put labels into an InsetLabel, for example, then we
        # would just need to test here for cmd == "label" and then take some
        # appropriate action, i.e., to use arg to get the content and then 
        # wrap it appropriately.
        cmd = put_cmd_in_ert(cmd)
        retval += "\n" + cmd + "\n"
        line = end
        m = labelre.match(line)
    retval += line
    return retval


def latex2lyx(data):
    '''Takes a string, possibly multi-line, and returns the result of 
    converting LaTeX constructs into LyX constructs. Returns a list of
    lines, suitable for insertion into document.body.'''

    retval = []

    # Convert LaTeX to Unicode
    reps = read_unicodesymbols()
    # Commands of this sort need to be checked to make sure they are
    # followed by a non-alpha character, lest we replace too much.
    hardone = re.compile(r'^\\\\[a-zA-Z]+$')

    for rep in reps:
        if hardone.match(rep[0]):
            pos = 0
            while True:
                pos = data.find(rep[0], pos)
                if pos == -1:
                    break
                nextpos = pos + len(rep[0])
                if nextpos < len(data) and data[nextpos].isalpha():
                    # not the end of that command
                    pos = nextpos
                    continue
                data = data[:pos] + rep[1] + data[nextpos:]
                pos = nextpos
        else:
            data = data.replace(rep[0], rep[1])

    # Generic, \" -> ":
    data = wrap_into_ert(data, r'\"', '"')

    # Math:
    mathre = re.compile('^(.*?)(\$.*?\$)(.*)')
    lines = data.split('\n')
    for line in lines:
        #document.warning("LINE: " + line)
        #document.warning(str(i) + ":" + document.body[i])
        #document.warning("LAST: " + document.body[-1])
        g = line
        m = mathre.match(g)
        while m != None:
            s = m.group(1)
            f = m.group(2).replace('\\\\', '\\')
            g = m.group(3)
            if s:
                # this is non-math!
                s = latex2ert(s)
                subst = s.split('\n')
                retval += subst
            retval.append("\\begin_inset Formula " + f)
            retval.append("\\end_inset")
            m = mathre.match(g)
        # Handle whatever is left, which is just text
        g = latex2ert(g)
        subst = g.split('\n')
        retval += subst
    return retval


def lyx2latex(lines):
    'Convert some LyX stuff into corresponding LaTeX stuff, as best we can.'
    # clean up multiline stuff
    content = ""
    ert_end = 0
    reps = read_unicodesymbols()
  
    for curline in range(len(lines)):
      line = lines[curline]
      if line.startswith("\\begin_inset ERT"):
          # We don't want to replace things inside ERT, so figure out
          # where the end of the inset is.
          ert_end = find_end_of_inset(lines, curline + 1)
          continue
      elif line.startswith("\\begin_inset Formula"):
          line = line[20:]
      elif line.startswith("\\begin_inset Quotes"):
          # For now, we do a very basic reversion. Someone who understands
          # quotes is welcome to fix it up.
          qtype = line[20:].strip()
          # lang = qtype[0]
          side = qtype[1]
          dbls = qtype[2]
          if side == "l":
              if dbls == "d":
                  line = "``"
              else:
                  line = "`"
          else:
              if dbls == "d":
                  line = "''"
              else:
                  line = "'"
      elif line.isspace() or \
            line.startswith("\\begin_layout") or \
            line.startswith("\\end_layout") or \
            line.startswith("\\begin_inset") or \
            line.startswith("\\end_inset") or \
            line.startswith("\\lang") or \
            line.strip() == "status collapsed" or \
            line.strip() == "status open":
          #skip all that stuff
          continue
  
      # a lossless reversion is not possible
      # try at least to handle some common insets and settings
      # do not replace inside ERTs
      if ert_end >= curline:
          line = line.replace(r'\backslash', r'\\')
      else:
          # Do the LyX text --> LaTeX conversion
          for rep in reps:
            line = line.replace(rep[1], rep[0])
          line = line.replace(r'\backslash', r'\textbackslash{}')
          line = line.replace(r'\series bold', r'\bfseries{}').replace(r'\series default', r'\mdseries{}')
          line = line.replace(r'\shape italic', r'\itshape{}').replace(r'\shape smallcaps', r'\scshape{}')
          line = line.replace(r'\shape slanted', r'\slshape{}').replace(r'\shape default', r'\upshape{}')
          line = line.replace(r'\emph on', r'\em{}').replace(r'\emph default', r'\em{}')
          line = line.replace(r'\noun on', r'\scshape{}').replace(r'\noun default', r'\upshape{}')
          line = line.replace(r'\bar under', r'\underbar{').replace(r'\bar default', r'}')
          line = line.replace(r'\family sans', r'\sffamily{}').replace(r'\family default', r'\normalfont{}')
          line = line.replace(r'\family typewriter', r'\ttfamily{}').replace(r'\family roman', r'\rmfamily{}')
          line = line.replace(r'\InsetSpace ', r'').replace(r'\SpecialChar ', r'')
      content += line
    return content


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

def convert_ltcaption(document):
    i = 0
    while True:
        i = find_token(document.body, "\\begin_inset Tabular", i)
        if i == -1:
            return
        j = find_end_of_inset(document.body, i + 1)
        if j == -1:
            document.warning("Malformed LyX document: Could not find end of tabular.")
            continue

        nrows = int(document.body[i+1].split('"')[3])
        ncols = int(document.body[i+1].split('"')[5])

        m = i + 1
        for k in range(nrows):
            m = find_token(document.body, "<row", m)
            r = m
            caption = 'false'
            for k in range(ncols):
                m = find_token(document.body, "<cell", m)
                if (k == 0):
                    mend = find_token(document.body, "</cell>", m + 1)
                    # first look for caption insets
                    mcap = find_token(document.body, "\\begin_inset Caption", m + 1, mend)
                    # then look for ERT captions
                    if mcap == -1:
                        mcap = find_token(document.body, "caption", m + 1, mend)
                        if mcap > -1:
                            mcap = find_token(document.body, "\\backslash", mcap - 1, mcap)
                    if mcap > -1:
                        caption = 'true'
                if caption == 'true':
                    if (k == 0):
                        set_option(document, r, 'caption', 'true')
                        set_option(document, m, 'multicolumn', '1')
                        set_option(document, m, 'bottomline', 'false')
                        set_option(document, m, 'topline', 'false')
                        set_option(document, m, 'rightline', 'false')
                        set_option(document, m, 'leftline', 'false')
                        #j = find_end_of_inset(document.body, j + 1)
                    else:
                        set_option(document, m, 'multicolumn', '2')
                m = m + 1
            m = m + 1

        i = j + 1


#FIXME Use of wrap_into_ert can confuse lyx2lyx
def revert_ltcaption(document):
    i = 0
    while True:
        i = find_token(document.body, "\\begin_inset Tabular", i)
        if i == -1:
            return
        j = find_end_of_inset(document.body, i + 1)
        if j == -1:
            document.warning("Malformed LyX document: Could not find end of tabular.")
            continue

        m = i + 1
        nrows = int(document.body[i+1].split('"')[3])
        ncols = int(document.body[i+1].split('"')[5])

        for k in range(nrows):
            m = find_token(document.body, "<row", m)
            caption = get_option(document, m, 'caption', 'false')
            if caption == 'true':
                remove_option(document, m, 'caption')
                for k in range(ncols):
                    m = find_token(document.body, "<cell", m)
                    remove_option(document, m, 'multicolumn')
                    if k == 0:
                        m = find_token(document.body, "\\begin_inset Caption", m)
                        if m == -1:
                            return
                        m = find_end_of_inset(document.body, m + 1)
                        document.body[m] += wrap_into_ert("","","\\backslash\n\\backslash\n%")
                    m = m + 1
            m = m + 1
        i = j + 1


def convert_tablines(document):
    i = 0
    while True:
        i = find_token(document.body, "\\begin_inset Tabular", i)
        if i == -1:
            # LyX 1.3 inserted an extra space between \begin_inset
            # and Tabular so let us try if this is the case and fix it.
            i = find_token(document.body, "\\begin_inset  Tabular", i)
            if i == -1:
                return
            else:
                document.body[i] = "\\begin_inset Tabular"
        j = find_end_of_inset(document.body, i + 1)
        if j == -1:
            document.warning("Malformed LyX document: Could not find end of tabular.")
            continue

        m = i + 1
        nrows = int(document.body[i+1].split('"')[3])
        ncols = int(document.body[i+1].split('"')[5])

        col_info = []
        for k in range(ncols):
            m = find_token(document.body, "<column", m)
            left = get_option(document, m, 'leftline', 'false')
            right = get_option(document, m, 'rightline', 'false')
            col_info.append([left, right])
            remove_option(document, m, 'leftline')
            remove_option(document, m, 'rightline')
            m = m + 1

        row_info = []
        for k in range(nrows):
            m = find_token(document.body, "<row", m)
            top = get_option(document, m, 'topline', 'false')
            bottom = get_option(document, m, 'bottomline', 'false')
            row_info.append([top, bottom])
            remove_option(document, m, 'topline')
            remove_option(document, m, 'bottomline')
            m = m + 1

        m = i + 1
        mc_info = []
        for k in range(nrows*ncols):
            m = find_token(document.body, "<cell", m)
            mc_info.append(get_option(document, m, 'multicolumn', '0'))
            m = m + 1
        m = i + 1
        for l in range(nrows):
            for k in range(ncols):
                m = find_token(document.body, '<cell', m)
                if mc_info[l*ncols + k] == '0':
                    r = set_option(document, m, 'topline', row_info[l][0])
                    r = set_option(document, m, 'bottomline', row_info[l][1])
                    r = set_option(document, m, 'leftline', col_info[k][0])
                    r = set_option(document, m, 'rightline', col_info[k][1])
                elif mc_info[l*ncols + k] == '1':
                    s = k + 1
                    while s < ncols and mc_info[l*ncols + s] == '2':
                        s = s + 1
                    if s < ncols and mc_info[l*ncols + s] != '1':
                        r = set_option(document, m, 'rightline', col_info[k][1])
                    if k > 0 and mc_info[l*ncols + k - 1] == '0':
                        r = set_option(document, m, 'leftline', col_info[k][0])
                m = m + 1
        i = j + 1


def revert_tablines(document):
    i = 0
    while True:
        i = find_token(document.body, "\\begin_inset Tabular", i)
        if i == -1:
            return
        j = find_end_of_inset(document.body, i + 1)
        if j == -1:
            document.warning("Malformed LyX document: Could not find end of tabular.")
            continue

        m = i + 1
        nrows = int(document.body[i+1].split('"')[3])
        ncols = int(document.body[i+1].split('"')[5])

        lines = []
        for k in range(nrows*ncols):
            m = find_token(document.body, "<cell", m)
            top = get_option(document, m, 'topline', 'false')
            bottom = get_option(document, m, 'bottomline', 'false')
            left = get_option(document, m, 'leftline', 'false')
            right = get_option(document, m, 'rightline', 'false')
            lines.append([top, bottom, left, right])
            m = m + 1

        # we will want to ignore longtable captions
        m = i + 1
        caption_info = []
        for k in range(nrows):
            m = find_token(document.body, "<row", m)
            caption = get_option(document, m, 'caption', 'false')
            caption_info.append([caption])
            m = m + 1

        m = i + 1
        col_info = []
        for k in range(ncols):
            m = find_token(document.body, "<column", m)
            left = 'true'
            for l in range(nrows):
                left = lines[l*ncols + k][2]
                if left == 'false' and caption_info[l] == 'false':
                    break
            set_option(document, m, 'leftline', left)
            right = 'true'
            for l in range(nrows):
                right = lines[l*ncols + k][3]
                if right == 'false' and caption_info[l] == 'false':
                    break
            set_option(document, m, 'rightline', right)
            m = m + 1

        row_info = []
        for k in range(nrows):
            m = find_token(document.body, "<row", m)
            top = 'true'
            for l in range(ncols):
                top = lines[k*ncols + l][0]
                if top == 'false':
                    break
            if caption_info[k] == 'false':
                top = 'false'
            set_option(document, m, 'topline', top)
            bottom = 'true'
            for l in range(ncols):
                bottom = lines[k*ncols + l][1]
                if bottom == 'false':
                    break
            if caption_info[k] == 'false':
                bottom = 'false'
            set_option(document, m, 'bottomline', bottom)
            m = m + 1

        i = j + 1


def fix_wrong_tables(document):
    i = 0
    while True:
        i = find_token(document.body, "\\begin_inset Tabular", i)
        if i == -1:
            return
        j = find_end_of_inset(document.body, i + 1)
        if j == -1:
            document.warning("Malformed LyX document: Could not find end of tabular.")
            continue

        m = i + 1
        nrows = int(document.body[i+1].split('"')[3])
        ncols = int(document.body[i+1].split('"')[5])

        for l in range(nrows):
            prev_multicolumn = 0
            for k in range(ncols):
                m = find_token(document.body, '<cell', m)

                if document.body[m].find('multicolumn') != -1:
                    multicol_cont = int(document.body[m].split('"')[1])

                    if multicol_cont == 2 and (k == 0 or prev_multicolumn == 0):
                        document.body[m] = document.body[m][:5] + document.body[m][21:]
                        prev_multicolumn = 0
                    else:
                        prev_multicolumn = multicol_cont
                else:
                    prev_multicolumn = 0

        i = j + 1


def close_begin_deeper(document):
    i = 0
    depth = 0
    while True:
        i = find_tokens(document.body, ["\\begin_deeper", "\\end_deeper"], i)

        if i == -1:
            break

        if document.body[i][:13] == "\\begin_deeper":
            depth += 1
        else:
            depth -= 1

        i += 1

    document.body[-2:-2] = ['\\end_deeper' for i in range(depth)]


def long_charstyle_names(document):
    i = 0
    while True:
        i = find_token(document.body, "\\begin_inset CharStyle", i)
        if i == -1:
            return
        document.body[i] = document.body[i].replace("CharStyle ", "CharStyle CharStyle:")
        i += 1

def revert_long_charstyle_names(document):
    i = 0
    while True:
        i = find_token(document.body, "\\begin_inset CharStyle", i)
        if i == -1:
            return
        document.body[i] = document.body[i].replace("CharStyle CharStyle:", "CharStyle")
        i += 1


def axe_show_label(document):
    i = 0
    while True:
        i = find_token(document.body, "\\begin_inset CharStyle", i)
        if i == -1:
            return
        if document.body[i + 1].find("show_label") != -1:
            if document.body[i + 1].find("true") != -1:
                document.body[i + 1] = "status open"
                del document.body[ i + 2]
            else:
                if document.body[i + 1].find("false") != -1:
                    document.body[i + 1] = "status collapsed"
                    del document.body[ i + 2]
                else:
                    document.warning("Malformed LyX document: show_label neither false nor true.")
        else:
            document.warning("Malformed LyX document: show_label missing in CharStyle.")

        i += 1


def revert_show_label(document):
    i = 0
    while True:
        i = find_token(document.body, "\\begin_inset CharStyle", i)
        if i == -1:
            return
        if document.body[i + 1].find("status open") != -1:
            document.body.insert(i + 1, "show_label true")
        else:
            if document.body[i + 1].find("status collapsed") != -1:
                document.body.insert(i + 1, "show_label false")
            else:
                document.warning("Malformed LyX document: no legal status line in CharStyle.")
        i += 1

def revert_begin_modules(document):
    i = 0
    while True:
        i = find_token(document.header, "\\begin_modules", i)
        if i == -1:
            return
        j = find_end_of(document.header, i, "\\begin_modules", "\\end_modules")
        if j == -1:
            # this should not happen
            break
        document.header[i : j + 1] = []

def convert_flex(document):
    "Convert CharStyle to Flex"
    i = 0
    while True:
        i = find_token(document.body, "\\begin_inset CharStyle", i)
        if i == -1:
            return
        document.body[i] = document.body[i].replace('\\begin_inset CharStyle', '\\begin_inset Flex')

def revert_flex(document):
    "Convert Flex to CharStyle"
    i = 0
    while True:
        i = find_token(document.body, "\\begin_inset Flex", i)
        if i == -1:
            return
        document.body[i] = document.body[i].replace('\\begin_inset Flex', '\\begin_inset CharStyle')


#  Discard PDF options for hyperref
def revert_pdf_options(document):
        "Revert PDF options for hyperref."
        # store the PDF options and delete the entries from the Lyx file
        i = 0
        hyperref = False
        title = ""
        author = ""
        subject = ""
        keywords = ""
        bookmarks = ""
        bookmarksnumbered = ""
        bookmarksopen = ""
        bookmarksopenlevel = ""
        breaklinks = ""
        pdfborder = ""
        colorlinks = ""
        backref = ""
        pagebackref = ""
        pagemode = ""
        otheroptions = ""
        i = find_token(document.header, "\\use_hyperref", i)
        if i != -1:
            hyperref = get_value(document.header, "\\use_hyperref", i) == 'true'
            del document.header[i]
        i = find_token(document.header, "\\pdf_store_options", i)
        if i != -1:
            del document.header[i]
        i = find_token(document.header, "\\pdf_title", 0)
        if i != -1:
            title = get_value_string(document.header, '\\pdf_title', 0, 0, True)
            title = ' pdftitle={' + title + '}'
            del document.header[i]
        i = find_token(document.header, "\\pdf_author", 0)
        if i != -1:
            author = get_value_string(document.header, '\\pdf_author', 0, 0, True)
            if title == "":
                author = ' pdfauthor={' + author + '}'
            else:
                author = ',\n pdfauthor={' + author + '}'
            del document.header[i]
        i = find_token(document.header, "\\pdf_subject", 0)
        if i != -1:
            subject = get_value_string(document.header, '\\pdf_subject', 0, 0, True)
            if title == "" and author == "":
                subject = ' pdfsubject={' + subject + '}'
            else:
                subject = ',\n pdfsubject={' + subject + '}'
            del document.header[i]
        i = find_token(document.header, "\\pdf_keywords", 0)
        if i != -1:
            keywords = get_value_string(document.header, '\\pdf_keywords', 0, 0, True)
            if title == "" and author == "" and subject == "":
                keywords = ' pdfkeywords={' + keywords + '}'
            else:
                keywords = ',\n pdfkeywords={' + keywords + '}'
            del document.header[i]
        i = find_token(document.header, "\\pdf_bookmarks", 0)
        if i != -1:
            bookmarks = get_value_string(document.header, '\\pdf_bookmarks', 0)
            bookmarks = ',\n bookmarks=' + bookmarks
            del document.header[i]
        i = find_token(document.header, "\\pdf_bookmarksnumbered", i)
        if i != -1:
            bookmarksnumbered = get_value_string(document.header, '\\pdf_bookmarksnumbered', 0)
            bookmarksnumbered = ',\n bookmarksnumbered=' + bookmarksnumbered
            del document.header[i]
        i = find_token(document.header, "\\pdf_bookmarksopen", i)
        if i != -1:
            bookmarksopen = get_value_string(document.header, '\\pdf_bookmarksopen', 0)
            bookmarksopen = ',\n bookmarksopen=' + bookmarksopen
            del document.header[i]
        i = find_token(document.header, "\\pdf_bookmarksopenlevel", i)
        if i != -1:
            bookmarksopenlevel = get_value_string(document.header, '\\pdf_bookmarksopenlevel', 0, 0, True)
            bookmarksopenlevel = ',\n bookmarksopenlevel=' + bookmarksopenlevel
            del document.header[i]
        i = find_token(document.header, "\\pdf_breaklinks", i)
        if i != -1:
            breaklinks = get_value_string(document.header, '\\pdf_breaklinks', 0)
            breaklinks = ',\n breaklinks=' + breaklinks
            del document.header[i]
        i = find_token(document.header, "\\pdf_pdfborder", i)
        if i != -1:
            pdfborder = get_value_string(document.header, '\\pdf_pdfborder', 0)
            if pdfborder == 'true':
                pdfborder = ',\n pdfborder={0 0 0}'
            else:
                pdfborder = ',\n pdfborder={0 0 1}'
            del document.header[i]
        i = find_token(document.header, "\\pdf_colorlinks", i)
        if i != -1:
            colorlinks = get_value_string(document.header, '\\pdf_colorlinks', 0)
            colorlinks = ',\n colorlinks=' + colorlinks
            del document.header[i]
        i = find_token(document.header, "\\pdf_backref", i)
        if i != -1:
            backref = get_value_string(document.header, '\\pdf_backref', 0)
            backref = ',\n backref=' + backref
            del document.header[i]
        i = find_token(document.header, "\\pdf_pagebackref", i)
        if i != -1:
            pagebackref = get_value_string(document.header, '\\pdf_pagebackref', 0)
            pagebackref = ',\n pagebackref=' + pagebackref
            del document.header[i]
        i = find_token(document.header, "\\pdf_pagemode", 0)
        if i != -1:
            pagemode = get_value_string(document.header, '\\pdf_pagemode', 0)
            pagemode = ',\n pdfpagemode=' + pagemode
            del document.header[i]
        i = find_token(document.header, "\\pdf_quoted_options", 0)
        if i != -1:
            otheroptions = get_value_string(document.header, '\\pdf_quoted_options', 0, 0, True)
            if title == "" and author == "" and subject == "" and keywords == "":
                otheroptions = ' ' + otheroptions
            else:
                otheroptions = ',\n ' + otheroptions
            del document.header[i]

        # write to the preamble when hyperref was used
        if hyperref == True:
            # preamble write preparations
            # bookmark numbers are only output when they are turned on
            if bookmarksopen == ',\n bookmarksopen=true':
                bookmarksopen = bookmarksopen + bookmarksopenlevel
            if bookmarks == ',\n bookmarks=true':
                bookmarks = bookmarks + bookmarksnumbered + bookmarksopen
            else:
                bookmarks = bookmarks
            # hypersetup is only output when there are things to be set up
            setupstart = '\\hypersetup{%\n'
            setupend = ' }\n'
            if otheroptions == "" and title == "" and  author == ""\
               and  subject == "" and keywords == "":
                setupstart = ""
                setupend = ""
            # write the preamble
            add_to_preamble(document,
                                ['% Commands inserted by lyx2lyx for PDF properties',
                                 '\\usepackage[unicode=true'
                                 + bookmarks
                                 + breaklinks
                                 + pdfborder
                                 + backref
                                 + pagebackref
                                 + colorlinks
                                 + pagemode
                                 + ']\n'
                                 ' {hyperref}\n'
                                 + setupstart
                                 + title
                                 + author
                                 + subject
                                 + keywords
                                 + otheroptions
                                 + setupend])


def remove_inzip_options(document):
    "Remove inzipName and embed options from the Graphics inset"
    i = 0
    while 1:
        i = find_token(document.body, "\\begin_inset Graphics", i)
        if i == -1:
            return
        j = find_end_of_inset(document.body, i + 1)
        if j == -1:
            # should not happen
            document.warning("Malformed LyX document: Could not find end of graphics inset.")
        # If there's a inzip param, just remove that
        k = find_token(document.body, "\tinzipName", i + 1, j)
        if k != -1:
            del document.body[k]
            # embed option must follow the inzipName option
            del document.body[k+1]
        i = i + 1


def convert_inset_command(document):
    """
        Convert:
            \begin_inset LatexCommand cmd
        to
            \begin_inset CommandInset InsetType
            LatexCommand cmd
    """
    i = 0
    while 1:
        i = find_token(document.body, "\\begin_inset LatexCommand", i)
        if i == -1:
            return
        line = document.body[i]
        r = re.compile(r'\\begin_inset LatexCommand (.*)$')
        m = r.match(line)
        cmdName = m.group(1)
        insetName = ""
        #this is adapted from factory.cpp
        if cmdName[0:4].lower() == "cite":
            insetName = "citation"
        elif cmdName == "url" or cmdName == "htmlurl":
            insetName = "url"
        elif cmdName[-3:] == "ref":
            insetName = "ref"
        elif cmdName == "tableofcontents":
            insetName = "toc"
        elif cmdName == "printnomenclature":
            insetName = "nomencl_print"
        elif cmdName == "printindex":
            insetName = "index_print"
        else:
            insetName = cmdName
        insertion = ["\\begin_inset CommandInset " + insetName, "LatexCommand " + cmdName]
        document.body[i : i+1] = insertion


def revert_inset_command(document):
    """
        Convert:
            \begin_inset CommandInset InsetType
            LatexCommand cmd
        to
            \begin_inset LatexCommand cmd
        Some insets may end up being converted to insets earlier versions of LyX
        will not be able to recognize. Not sure what to do about that.
    """
    i = 0
    while 1:
        i = find_token(document.body, "\\begin_inset CommandInset", i)
        if i == -1:
            return
        nextline = document.body[i+1]
        r = re.compile(r'LatexCommand\s+(.*)$')
        m = r.match(nextline)
        if not m:
            document.warning("Malformed LyX document: Missing LatexCommand in " + document.body[i] + ".")
            continue
        cmdName = m.group(1)
        insertion = ["\\begin_inset LatexCommand " + cmdName]
        document.body[i : i+2] = insertion


def convert_wrapfig_options(document):
    "Convert optional options for wrap floats (wrapfig)."
    # adds the tokens "lines", "placement", and "overhang"
    i = 0
    while True:
        i = find_token(document.body, "\\begin_inset Wrap figure", i)
        if i == -1:
            return
        document.body.insert(i + 1, "lines 0")
        j = find_token(document.body, "placement", i)
        # placement can be already set or not; if not, set it
        if j == i+2:
            document.body.insert(i + 3, "overhang 0col%")
        else:
           document.body.insert(i + 2, "placement o")
           document.body.insert(i + 3, "overhang 0col%")
        i = i + 1


def revert_wrapfig_options(document):
    "Revert optional options for wrap floats (wrapfig)."
    i = 0
    while True:
        i = find_token(document.body, "\\begin_inset Wrap figure", i)
        if i == -1:
            return
        j = find_end_of_inset(document.body, i)
        if j == -1:
            document.warning("Can't find end of Wrap inset at line " + str(i))
            i += 1
            continue
        k = find_default_layout(document, i, j)
        if k == -1:
            document.warning("Can't find default layout for Wrap figure!")
            i = j
            continue
        # Options should be between i and k now
        l = find_token(document.body, "lines", i, k)
        if l == -1:
            document.warning("Can't find lines option for Wrap figure!")
            i = k
            continue
        m = find_token(document.body, "overhang", i + 1, k)
        if m == -1:
            document.warning("Malformed LyX document: Couldn't find overhang parameter of wrap float!")
            i = k
            continue
        # Do these in reverse order
        del document.body[m]
        del document.body[l]
        i = k


def convert_latexcommand_index(document):
    "Convert from LatexCommand form to collapsable form."
    i = 0
    r1 = re.compile('name "(.*)"')
    while True:
        i = find_token(document.body, "\\begin_inset CommandInset index", i)
        if i == -1:
            return
        if document.body[i + 1] != "LatexCommand index": # Might also be index_print
            return
        m = r1.match(document.body[i + 2])
        if m == None:
            document.warning("Unable to match: " + document.body[i+2])
            i += 1
            continue
        fullcontent = m.group(1)
        #document.warning(fullcontent)
        document.body[i:i + 3] = ["\\begin_inset Index",
          "status collapsed",
          "\\begin_layout Standard"]
        i += 3
        # We are now on the blank line preceding "\end_inset"
        # We will write the content here, into the inset.

        linelist = latex2lyx(fullcontent)
        document.body[i+1:i+1] = linelist
        i += len(linelist)

        document.body.insert(i + 1, "\\end_layout")
        i += 1


def revert_latexcommand_index(document):
    "Revert from collapsable form to LatexCommand form."
    i = 0
    while True:
        i = find_token(document.body, "\\begin_inset Index", i)
        if i == -1:
          return
        j = find_end_of_inset(document.body, i + 1)
        if j == -1:
          return

        content = lyx2latex(document.body[i:j])
        # escape quotes
        content = content.replace('"', r'\"')
        document.body[i:j] = ["\\begin_inset CommandInset index", "LatexCommand index",
            "name " + '"' + content + '"', ""]
        i += 5


def revert_wraptable(document):
    "Revert wrap table to wrap figure."
    i = 0
    while True:
        i = find_token(document.body, "\\begin_inset Wrap table", i)
        if i == -1:
            return
        document.body[i] = document.body[i].replace('\\begin_inset Wrap table', '\\begin_inset Wrap figure')
        i = i + 1


def revert_vietnamese(document):
    "Set language Vietnamese to English"
    # Set document language from Vietnamese to English
    i = 0
    if document.language == "vietnamese":
        document.language = "english"
        i = find_token(document.header, "\\language", 0)
        if i != -1:
            document.header[i] = "\\language english"
    j = 0
    while True:
        j = find_token(document.body, "\\lang vietnamese", j)
        if j == -1:
            return
        document.body[j] = document.body[j].replace("\\lang vietnamese", "\\lang english")
        j = j + 1


def convert_japanese_cjk(document):
    "Set language japanese to japanese-cjk"
    # Set document language from japanese-plain to japanese
    i = 0
    if document.language == "japanese":
        document.language = "japanese-cjk"
        i = find_token(document.header, "\\language", 0)
        if i != -1:
            document.header[i] = "\\language japanese-cjk"
    j = 0
    while True:
        j = find_token(document.body, "\\lang japanese", j)
        if j == -1:
            return
        document.body[j] = document.body[j].replace("\\lang japanese", "\\lang japanese-cjk")
        j = j + 1


def revert_japanese(document):
    "Set language japanese-plain to japanese"
    # Set document language from japanese-plain to japanese
    i = 0
    if document.language == "japanese-plain":
        document.language = "japanese"
        i = find_token(document.header, "\\language", 0)
        if i != -1:
            document.header[i] = "\\language japanese"
    j = 0
    while True:
        j = find_token(document.body, "\\lang japanese-plain", j)
        if j == -1:
            return
        document.body[j] = document.body[j].replace("\\lang japanese-plain", "\\lang japanese")
        j = j + 1


def revert_japanese_cjk(document):
    "Set language japanese-cjk to japanese"
    # Set document language from japanese-plain to japanese
    i = 0
    if document.language == "japanese-cjk":
        document.language = "japanese"
        i = find_token(document.header, "\\language", 0)
        if i != -1:
            document.header[i] = "\\language japanese"
    j = 0
    while True:
        j = find_token(document.body, "\\lang japanese-cjk", j)
        if j == -1:
            return
        document.body[j] = document.body[j].replace("\\lang japanese-cjk", "\\lang japanese")
        j = j + 1


def revert_japanese_encoding(document):
    "Set input encoding form EUC-JP-plain to EUC-JP etc."
    # Set input encoding form EUC-JP-plain to EUC-JP etc.
    i = 0
    i = find_token(document.header, "\\inputencoding EUC-JP-plain", 0)
    if i != -1:
        document.header[i] = "\\inputencoding EUC-JP"
    j = 0
    j = find_token(document.header, "\\inputencoding JIS-plain", 0)
    if j != -1:
        document.header[j] = "\\inputencoding JIS"
    k = 0
    k = find_token(document.header, "\\inputencoding SJIS-plain", 0)
    if k != -1: # convert to UTF8 since there is currently no SJIS encoding
        document.header[k] = "\\inputencoding UTF8"


def revert_inset_info(document):
    'Replace info inset with its content'
    i = 0
    while 1:
        i = find_token(document.body, '\\begin_inset Info', i)
        if i == -1:
            return
        j = find_end_of_inset(document.body, i + 1)
        if j == -1:
            # should not happen
            document.warning("Malformed LyX document: Could not find end of Info inset.")
        type = 'unknown'
        arg = ''
        for k in range(i, j+1):
            if document.body[k].startswith("arg"):
                arg = document.body[k][3:].strip().strip('"')
            if document.body[k].startswith("type"):
                type = document.body[k][4:].strip().strip('"')
        # I think there is a newline after \\end_inset, which should be removed.
        if document.body[j + 1].strip() == "":
            document.body[i : (j + 2)] = [type + ':' + arg]
        else:
            document.body[i : (j + 1)] = [type + ':' + arg]


def convert_pdf_options(document):
    # Set the pdfusetitle tag, delete the pdf_store_options,
    # set quotes for bookmarksopenlevel"
    has_hr = get_value(document.header, "\\use_hyperref", 0, default = "0")
    if has_hr == "1":
        k = find_token(document.header, "\\use_hyperref", 0)
        document.header.insert(k + 1, "\\pdf_pdfusetitle true")
    k = find_token(document.header, "\\pdf_store_options", 0)
    if k != -1:
        del document.header[k]
    i = find_token(document.header, "\\pdf_bookmarksopenlevel", k)
    if i == -1: return
    document.header[i] = document.header[i].replace('"', '')


def revert_pdf_options_2(document):
    # reset the pdfusetitle tag, set quotes for bookmarksopenlevel"
    k = find_token(document.header, "\\use_hyperref", 0)
    i = find_token(document.header, "\\pdf_pdfusetitle", k)
    if i != -1:
        del document.header[i]
    i = find_token(document.header, "\\pdf_bookmarksopenlevel", k)
    if i == -1: return
    values = document.header[i].split()
    values[1] = ' "' + values[1] + '"'
    document.header[i] = ''.join(values)


def convert_htmlurl(document):
    'Convert "htmlurl" to "href" insets for docbook'
    if document.backend != "docbook":
      return
    i = 0
    while True:
      i = find_token(document.body, "\\begin_inset CommandInset url", i)
      if i == -1:
        return
      document.body[i] = "\\begin_inset CommandInset href"
      document.body[i + 1] = "LatexCommand href"
      i = i + 1


def convert_url(document):
    'Convert url insets to url charstyles'
    if document.backend == "docbook":
      return
    i = 0
    while True:
      i = find_token(document.body, "\\begin_inset CommandInset url", i)
      if i == -1:
        break
      n = find_token(document.body, "name", i)
      if n == i + 2:
        # place the URL name in typewriter before the new URL insert
        # grab the name 'bla' from the e.g. the line 'name "bla"',
        # therefore start with the 6th character
        name = document.body[n][6:-1]
        newname = [name + " "]
        document.body[i:i] = newname
        i = i + 1
      j = find_token(document.body, "target", i)
      if j == -1:
        document.warning("Malformed LyX document: Can't find target for url inset")
        i = j
        continue
      target = document.body[j][8:-1]
      k = find_token(document.body, "\\end_inset", j)
      if k == -1:
        document.warning("Malformed LyX document: Can't find end of url inset")
        i = k
        continue
      newstuff = ["\\begin_inset Flex URL",
        "status collapsed", "",
        "\\begin_layout Standard",
        "",
        target,
        "\\end_layout",
        ""]
      document.body[i:k] = newstuff
      i = k

def convert_ams_classes(document):
  tc = document.textclass
  if (tc != "amsart" and tc != "amsart-plain" and
      tc != "amsart-seq" and tc != "amsbook"):
    return
  if tc == "amsart-plain":
    document.textclass = "amsart"
    document.set_textclass()
    document.add_module("Theorems (Starred)")
    return
  if tc == "amsart-seq":
    document.textclass = "amsart"
    document.set_textclass()
  document.add_module("Theorems (AMS)")

  #Now we want to see if any of the environments in the extended theorems
  #module were used in this document. If so, we'll add that module, too.
  layouts = ["Criterion", "Algorithm", "Axiom", "Condition", "Note",  \
    "Notation", "Summary", "Acknowledgement", "Conclusion", "Fact", \
    "Assumption"]

  r = re.compile(r'^\\begin_layout (.*?)\*?\s*$')
  i = 0
  while True:
    i = find_token(document.body, "\\begin_layout", i)
    if i == -1:
      return
    m = r.match(document.body[i])
    if m == None:
      document.warning("Weirdly formed \\begin_layout at line %d of body!" % i)
      i += 1
      continue
    m = m.group(1)
    if layouts.count(m) != 0:
      document.add_module("Theorems (AMS-Extended)")
      return
    i += 1

def revert_href(document):
    'Reverts hyperlink insets (href) to url insets (url)'
    i = 0
    while True:
      i = find_token(document.body, "\\begin_inset CommandInset href", i)
      if i == -1:
          return
      document.body[i : i + 2] = \
        ["\\begin_inset CommandInset url", "LatexCommand url"]
      i = i + 2

def revert_url(document):
    'Reverts Flex URL insets to old-style URL insets'
    i = 0
    while True:
        i = find_token(document.body, "\\begin_inset Flex URL", i)
        if i == -1:
            return
        j = find_end_of_inset(document.body, i)
        if j == -1:
            document.warning("Can't find end of inset in revert_url!")
            return
        k = find_default_layout(document, i, j)
        if k == -1:
            document.warning("Can't find default layout in revert_url!")
            i = j
            continue
        l = find_end_of(document.body, k, "\\begin_layout", "\\end_layout")
        if l == -1 or l >= j:
            document.warning("Can't find end of default layout in revert_url!")
            i = j
            continue
        # OK, so the inset's data is between lines k and l.
        data =  " ".join(document.body[k+1:l])
        data = data.strip()
        newinset = ["\\begin_inset LatexCommand url", "target \"" + data + "\"",\
                    "", "\\end_inset"]
        document.body[i:j+1] = newinset
        i = i + len(newinset)


def convert_include(document):
  'Converts include insets to new format.'
  i = 0
  r = re.compile(r'\\begin_inset Include\s+\\([^{]+){([^}]*)}(?:\[(.*)\])?')
  while True:
    i = find_token(document.body, "\\begin_inset Include", i)
    if i == -1:
      return
    line = document.body[i]
    previewline = document.body[i + 1]
    m = r.match(line)
    if m == None:
      document.warning("Unable to match line " + str(i) + " of body!")
      i += 1
      continue
    cmd = m.group(1)
    fn  = m.group(2)
    opt = m.group(3)
    insertion = ["\\begin_inset CommandInset include",
       "LatexCommand " + cmd, previewline,
       "filename \"" + fn + "\""]
    newlines = 2
    if opt:
      insertion.append("lstparams " + '"' + opt + '"')
      newlines += 1
    document.body[i : i + 2] = insertion
    i += newlines


def revert_include(document):
  'Reverts include insets to old format.'
  i = 0
  r0 = re.compile('preview.*')
  r1 = re.compile('LatexCommand (.+)')
  r2 = re.compile('filename "(.+)"')
  r3 = re.compile('lstparams "(.*)"')
  while True:
    i = find_token(document.body, "\\begin_inset CommandInset include", i)
    if i == -1:
      return
    nextline = i + 1
    if r0.match(document.body[nextline]):
      previewline = document.body[nextline]
      nextline += 1
    else:
      previewline = ""
    m = r1.match(document.body[nextline])
    if m == None:
      document.warning("Malformed LyX document: No LatexCommand line for `" +
        document.body[i] + "' on line " + str(i) + ".")
      i += 1
      continue
    cmd = m.group(1)
    nextline += 1
    m = r2.match(document.body[nextline])
    if m == None:
      document.warning("Malformed LyX document: No filename line for `" + \
        document.body[i] + "' on line " + str(i) + ".")
      i += 2
      continue
    fn = m.group(1)
    nextline += 1
    options = ""
    if (cmd == "lstinputlisting"):
      m = r3.match(document.body[nextline])
      if m != None:
        options = m.group(1)
        numlines = 5
        nextline += 1
    newline = "\\begin_inset Include \\" + cmd + "{" + fn + "}"
    if options:
      newline += ("[" + options + "]")
    insertion = [newline]
    if previewline != "":
      insertion.append(previewline)
    document.body[i : nextline] = insertion
    i += 2


def revert_albanian(document):
    "Set language Albanian to English"
    i = 0
    if document.language == "albanian":
        document.language = "english"
        i = find_token(document.header, "\\language", 0)
        if i != -1:
            document.header[i] = "\\language english"
    j = 0
    while True:
        j = find_token(document.body, "\\lang albanian", j)
        if j == -1:
            return
        document.body[j] = document.body[j].replace("\\lang albanian", "\\lang english")
        j = j + 1


def revert_lowersorbian(document):
    "Set language lower Sorbian to English"
    i = 0
    if document.language == "lowersorbian":
        document.language = "english"
        i = find_token(document.header, "\\language", 0)
        if i != -1:
            document.header[i] = "\\language english"
    j = 0
    while True:
        j = find_token(document.body, "\\lang lowersorbian", j)
        if j == -1:
            return
        document.body[j] = document.body[j].replace("\\lang lowersorbian", "\\lang english")
        j = j + 1


def revert_uppersorbian(document):
    "Set language uppersorbian to usorbian as this was used in LyX 1.5"
    i = 0
    if document.language == "uppersorbian":
        document.language = "usorbian"
        i = find_token(document.header, "\\language", 0)
        if i != -1:
            document.header[i] = "\\language usorbian"
    j = 0
    while True:
        j = find_token(document.body, "\\lang uppersorbian", j)
        if j == -1:
            return
        document.body[j] = document.body[j].replace("\\lang uppersorbian", "\\lang usorbian")
        j = j + 1


def convert_usorbian(document):
    "Set language usorbian to uppersorbian"
    i = 0
    if document.language == "usorbian":
        document.language = "uppersorbian"
        i = find_token(document.header, "\\language", 0)
        if i != -1:
            document.header[i] = "\\language uppersorbian"
    j = 0
    while True:
        j = find_token(document.body, "\\lang usorbian", j)
        if j == -1:
            return
        document.body[j] = document.body[j].replace("\\lang usorbian", "\\lang uppersorbian")
        j = j + 1


def revert_macro_optional_params(document):
    "Convert macro definitions with optional parameters into ERTs"
    # Stub to convert macro definitions with one or more optional parameters
    # into uninterpreted ERT insets


def revert_hyperlinktype(document):
    'Reverts hyperlink type'
    i = 0
    j = 0
    while True:
      i = find_token(document.body, "target", i)
      if i == -1:
          return
      j = find_token(document.body, "type", i)
      if j == -1:
          return
      if j == i + 1:
          del document.body[j]
      i = i + 1


def revert_pagebreak(document):
    'Reverts pagebreak to ERT'
    i = 0
    while True:
      i = find_token(document.body, "\\pagebreak", i)
      if i == -1:
          return
      document.body[i] = '\\begin_inset ERT\nstatus collapsed\n\n' \
      '\\begin_layout Standard\n\n\n\\backslash\n' \
      'pagebreak{}\n\\end_layout\n\n\\end_inset\n\n'
      i = i + 1


def revert_linebreak(document):
    'Reverts linebreak to ERT'
    i = 0
    while True:
      i = find_token(document.body, "\\linebreak", i)
      if i == -1:
          return
      document.body[i] = '\\begin_inset ERT\nstatus collapsed\n\n' \
      '\\begin_layout Standard\n\n\n\\backslash\n' \
      'linebreak{}\n\\end_layout\n\n\\end_inset\n\n'
      i = i + 1


def revert_latin(document):
    "Set language Latin to English"
    i = 0
    if document.language == "latin":
        document.language = "english"
        i = find_token(document.header, "\\language", 0)
        if i != -1:
            document.header[i] = "\\language english"
    j = 0
    while True:
        j = find_token(document.body, "\\lang latin", j)
        if j == -1:
            return
        document.body[j] = document.body[j].replace("\\lang latin", "\\lang english")
        j = j + 1


def revert_samin(document):
    "Set language North Sami to English"
    i = 0
    if document.language == "samin":
        document.language = "english"
        i = find_token(document.header, "\\language", 0)
        if i != -1:
            document.header[i] = "\\language english"
    j = 0
    while True:
        j = find_token(document.body, "\\lang samin", j)
        if j == -1:
            return
        document.body[j] = document.body[j].replace("\\lang samin", "\\lang english")
        j = j + 1


def convert_serbocroatian(document):
    "Set language Serbocroatian to Croatian as this was really Croatian in LyX 1.5"
    i = 0
    if document.language == "serbocroatian":
        document.language = "croatian"
        i = find_token(document.header, "\\language", 0)
        if i != -1:
            document.header[i] = "\\language croatian"
    j = 0
    while True:
        j = find_token(document.body, "\\lang serbocroatian", j)
        if j == -1:
            return
        document.body[j] = document.body[j].replace("\\lang serbocroatian", "\\lang croatian")
        j = j + 1


def convert_framed_notes(document):
    "Convert framed notes to boxes. "
    i = 0
    while 1:
        i = find_tokens(document.body, ["\\begin_inset Note Framed", "\\begin_inset Note Shaded"], i)
        if i == -1:
            return
        subst = [document.body[i].replace("\\begin_inset Note", "\\begin_inset Box"),
                 'position "t"',
                 'hor_pos "c"',
                 'has_inner_box 0',
                 'inner_pos "t"', 
                 'use_parbox 0',
                 'width "100col%"',
                 'special "none"',
                 'height "1in"',
                 'height_special "totalheight"']
        document.body[i:i+1] = subst
        i = i + 9


def convert_module_names(document):
  modulemap = { 'Braille' : 'braille', 'Endnote' : 'endnotes', 'Foot to End' : 'foottoend',\
    'Hanging' : 'hanging', 'Linguistics' : 'linguistics', 'Logical Markup' : 'logicalmkup', \
    'Theorems (AMS-Extended)' : 'theorems-ams-extended', 'Theorems (AMS)' : 'theorems-ams', \
    'Theorems (Order By Chapter)' : 'theorems-chap', 'Theorems (Order By Section)' : 'theorems-sec', \
    'Theorems (Starred)' : 'theorems-starred', 'Theorems' : 'theorems-std' }
  modlist = document.get_module_list()
  if len(modlist) == 0:
    return
  newmodlist = []
  for mod in modlist:
    if modulemap.has_key(mod):
      newmodlist.append(modulemap[mod])
    else:
      document.warning("Can't find module %s in the module map!" % mod)
      newmodlist.append(mod)
  document.set_module_list(newmodlist)


def revert_module_names(document):
  modulemap = { 'braille' : 'Braille', 'endnotes' : 'Endnote', 'foottoend' : 'Foot to End',\
    'hanging' : 'Hanging', 'linguistics' : 'Linguistics', 'logicalmkup' : 'Logical Markup', \
    'theorems-ams-extended' : 'Theorems (AMS-Extended)', 'theorems-ams' : 'Theorems (AMS)', \
    'theorems-chap' : 'Theorems (Order By Chapter)', 'theorems-sec' : 'Theorems (Order By Section)', \
    'theorems-starred' : 'Theorems (Starred)', 'theorems-std' : 'Theorems'}
  modlist = document.get_module_list()
  if len(modlist) == 0:
    return
  newmodlist = []
  for mod in modlist:
    if modulemap.has_key(mod):
      newmodlist.append(modulemap[mod])
    else:
      document.warning("Can't find module %s in the module map!" % mod)
      newmodlist.append(mod)
  document.set_module_list(newmodlist)


def revert_colsep(document):
    i = find_token(document.header, "\\columnsep", 0)
    if i == -1:
        return
    colsepline = document.header[i]
    r = re.compile(r'\\columnsep (.*)')
    m = r.match(colsepline)
    if not m:
        document.warning("Malformed column separation line!")
        return
    colsep = m.group(1)
    del document.header[i]
    #it seems to be safe to add the package even if it is already used
    pretext = ["\\usepackage{geometry}", "\\geometry{columnsep=" + colsep + "}"]

    add_to_preamble(document, pretext)


def revert_framed_notes(document):
    "Revert framed boxes to notes. "
    i = 0
    while 1:
        i = find_tokens(document.body, ["\\begin_inset Box Framed", "\\begin_inset Box Shaded"], i)

        if i == -1:
            return
        j = find_end_of_inset(document.body, i + 1)
        if j == -1:
            # should not happen
            document.warning("Malformed LyX document: Could not find end of Box inset.")
        k = find_token(document.body, "status", i + 1, j)
        if k == -1:
            document.warning("Malformed LyX document: Missing `status' tag in Box inset.")
            return
        status = document.body[k]
        l = find_default_layout(document, i + 1, j)
        if l == -1:
            document.warning("Malformed LyX document: Missing `\\begin_layout' in Box inset.")
            return
        m = find_token(document.body, "\\end_layout", i + 1, j)
        if m == -1:
            document.warning("Malformed LyX document: Missing `\\end_layout' in Box inset.")
            return
        ibox = find_token(document.body, "has_inner_box 1", i + 1, k)
        pbox = find_token(document.body, "use_parbox 1", i + 1, k)
        if ibox == -1 and pbox == -1:
            document.body[i] = document.body[i].replace("\\begin_inset Box", "\\begin_inset Note")
            del document.body[i+1:k]
        else:
            document.body[i] = document.body[i].replace("\\begin_inset Box Shaded", "\\begin_inset Box Frameless")
            subst1 = [document.body[l],
                      "\\begin_inset Note Shaded",
                      status,
                      '\\begin_layout Standard']
            document.body[l:l + 1] = subst1
            subst2 = [document.body[m], "\\end_layout", "\\end_inset"]
            document.body[m:m + 1] = subst2
        i = i + 1


def revert_slash(document):
    'Revert \\SpecialChar \\slash{} to ERT'
    r = re.compile(r'\\SpecialChar \\slash{}')
    i = 0
    while i < len(document.body):
        m = r.match(document.body[i])
        if m:
          subst = ['\\begin_inset ERT',
                   'status collapsed', '',
                   '\\begin_layout Standard',
                   '', '', '\\backslash',
                   'slash{}',
                   '\\end_layout', '',
                   '\\end_inset', '']
          document.body[i: i+1] = subst
          i = i + len(subst)
        else:
          i = i + 1


def revert_nobreakdash(document):
    'Revert \\SpecialChar \\nobreakdash- to ERT'
    i = 0
    while i < len(document.body):
        line = document.body[i]
        r = re.compile(r'\\SpecialChar \\nobreakdash-')
        m = r.match(line)
        if m:
            subst = ['\\begin_inset ERT',
                    'status collapsed', '',
                    '\\begin_layout Standard', '', '',
                    '\\backslash',
                    'nobreakdash-',
                    '\\end_layout', '',
                    '\\end_inset', '']
            document.body[i:i+1] = subst
            i = i + len(subst)
            j = find_token(document.header, "\\use_amsmath", 0)
            if j == -1:
                document.warning("Malformed LyX document: Missing '\\use_amsmath'.")
                return
            document.header[j] = "\\use_amsmath 2"
        else:
            i = i + 1


#Returns number of lines added/removed
def revert_nocite_key(body, start, end):
    'key "..." -> \nocite{...}' 
    r = re.compile(r'^key "(.*)"')
    i = start
    j = end
    while i < j:
        m = r.match(body[i])
        if m:
            body[i:i+1] = ["\\backslash", "nocite{" + m.group(1) + "}"]
            j += 1     # because we added a line
            i += 2     # skip that line
        else:
            del body[i]
            j -= 1     # because we deleted a line
            # no need to change i, since it now points to the next line
    return j - end


def revert_nocite(document):
    "Revert LatexCommand nocite to ERT"
    i = 0
    while 1:
        i = find_token(document.body, "\\begin_inset CommandInset citation", i)
        if i == -1:
            return
        if (document.body[i+1] != "LatexCommand nocite"):
            # note that we already incremented i
            i = i + 1
            continue
        insetEnd = find_end_of_inset(document.body, i)
        if insetEnd == -1:
            #this should not happen
            document.warning("End of CommandInset citation not found in revert_nocite!")
            return

        paramLocation = i + 2 #start of the inset's parameters
        addedLines = 0
        document.body[i:i+2] = \
            ["\\begin_inset ERT", "status collapsed", "", "\\begin_layout Standard"]
        # that added two lines
        paramLocation += 2
        insetEnd += 2
        #print insetEnd, document.body[i: insetEnd + 1]
        insetEnd += revert_nocite_key(document.body, paramLocation, insetEnd)
        #print insetEnd, document.body[i: insetEnd + 1]
        document.body.insert(insetEnd, "\\end_layout")
        document.body.insert(insetEnd + 1, "")
        i = insetEnd + 1


def revert_btprintall(document):
    "Revert (non-bibtopic) btPrintAll option to ERT \nocite{*}"
    i = find_token(document.header, '\\use_bibtopic', 0)
    if i == -1:
        document.warning("Malformed lyx document: Missing '\\use_bibtopic'.")
        return
    if get_value(document.header, '\\use_bibtopic', 0) == "false":
        i = 0
        while i < len(document.body):
            i = find_token(document.body, "\\begin_inset CommandInset bibtex", i)
            if i == -1:
                return
            j = find_end_of_inset(document.body, i + 1)
            if j == -1:
                #this should not happen
                document.warning("End of CommandInset bibtex not found in revert_btprintall!")
                j = len(document.body)
            # this range isn't really right, but it should be OK, since we shouldn't
            # see more than one matching line in each inset
            addedlines = 0
            for k in range(i, j):
                if (document.body[k] == 'btprint "btPrintAll"'):
                    del document.body[k]
                    subst = ["\\begin_inset ERT",
                             "status collapsed", "",
                             "\\begin_layout Standard", "",
                             "\\backslash",
                             "nocite{*}",
                             "\\end_layout",
                             "\\end_inset"]
                    document.body[i:i] = subst
                    addlines = addedlines + len(subst) - 1
            i = j + addedlines


def revert_bahasam(document):
    "Set language Bahasa Malaysia to Bahasa Indonesia"
    i = 0
    if document.language == "bahasam":
        document.language = "bahasa"
        i = find_token(document.header, "\\language", 0)
        if i != -1:
            document.header[i] = "\\language bahasa"
    j = 0
    while True:
        j = find_token(document.body, "\\lang bahasam", j)
        if j == -1:
            return
        document.body[j] = document.body[j].replace("\\lang bahasam", "\\lang bahasa")
        j = j + 1


def revert_interlingua(document):
    "Set language Interlingua to English"
    i = 0
    if document.language == "interlingua":
        document.language = "english"
        i = find_token(document.header, "\\language", 0)
        if i != -1:
            document.header[i] = "\\language english"
    j = 0
    while True:
        j = find_token(document.body, "\\lang interlingua", j)
        if j == -1:
            return
        document.body[j] = document.body[j].replace("\\lang interlingua", "\\lang english")
        j = j + 1


def revert_serbianlatin(document):
    "Set language Serbian-Latin to Croatian"
    i = 0
    if document.language == "serbian-latin":
        document.language = "croatian"
        i = find_token(document.header, "\\language", 0)
        if i != -1:
            document.header[i] = "\\language croatian"
    j = 0
    while True:
        j = find_token(document.body, "\\lang serbian-latin", j)
        if j == -1:
            return
        document.body[j] = document.body[j].replace("\\lang serbian-latin", "\\lang croatian")
        j = j + 1


def revert_rotfloat(document):
    " Revert sideways custom floats. "
    i = 0
    while 1:
        # whitespace intended (exclude \\begin_inset FloatList)
        i = find_token(document.body, "\\begin_inset Float ", i)
        if i == -1:
            return
        line = document.body[i]
        r = re.compile(r'\\begin_inset Float (.*)$')
        m = r.match(line)
        if m == None:
            document.warning("Unable to match line " + str(i) + " of body!")
            i += 1
            continue
        floattype = m.group(1)
        if floattype == "figure" or floattype == "table":
            i += 1
            continue
        j = find_end_of_inset(document.body, i)
        if j == -1:
            document.warning("Malformed lyx document: Missing '\\end_inset' in revert_rotfloat.")
            i += 1
            continue
        addedLines = 0
        if get_value(document.body, 'sideways', i, j) == "false":
            i += 1
            continue
        l = find_default_layout(document, i + 1, j)
        if l == -1:
            document.warning("Malformed LyX document: Missing `\\begin_layout' in Float inset.")
            return
        subst = ['\\begin_layout Standard',
                  '\\begin_inset ERT',
                  'status collapsed', '',
                  '\\begin_layout Standard', '', '', 
                  '\\backslash', '',
                  'end{sideways' + floattype + '}',
                  '\\end_layout', '', '\\end_inset']
        document.body[j : j+1] = subst
        addedLines = len(subst) - 1
        del document.body[i+1 : l]
        addedLines -= (l-1) - (i+1) 
        subst = ['\\begin_inset ERT', 'status collapsed', '',
                  '\\begin_layout Standard', '', '', '\\backslash', 
                  'begin{sideways' + floattype + '}', 
                  '\\end_layout', '', '\\end_inset', '',
                  '\\end_layout', '']
        document.body[i : i+1] = subst
        addedLines += len(subst) - 1
        if floattype == "algorithm":
            add_to_preamble(document,
                            ['% Commands inserted by lyx2lyx for sideways algorithm float',
                              '\\usepackage{rotfloat}',
                              '\\floatstyle{ruled}',
                              '\\newfloat{algorithm}{tbp}{loa}',
                              '\\floatname{algorithm}{Algorithm}'])
        else:
            document.warning("Cannot create preamble definition for custom float" + floattype + ".")
        i += addedLines + 1


def revert_widesideways(document):
    " Revert wide sideways floats. "
    i = 0
    while 1:
        # whitespace intended (exclude \\begin_inset FloatList)
        i = find_token(document.body, '\\begin_inset Float ', i)
        if i == -1:
            return
        line = document.body[i]
        r = re.compile(r'\\begin_inset Float (.*)$')
        m = r.match(line)
        if m == None:
            document.warning("Unable to match line " + str(i) + " of body!")
            i += 1
            continue
        floattype = m.group(1)
        if floattype != "figure" and floattype != "table":
            i += 1
            continue
        j = find_end_of_inset(document.body, i)
        if j == -1:
            document.warning("Malformed lyx document: Missing '\\end_inset' in revert_widesideways.")
            i += 1
            continue
        if get_value(document.body, 'sideways', i, j) == "false" or \
           get_value(document.body, 'wide', i, j) == "false":
             i += 1
             continue
        l = find_default_layout(document, i + 1, j)
        if l == -1:
            document.warning("Malformed LyX document: Missing `\\begin_layout' in Float inset.")
            return
        subst = ['\\begin_layout Standard', '\\begin_inset ERT', 
                  'status collapsed', '', 
                  '\\begin_layout Standard', '', '', '\\backslash',
                  'end{sideways' + floattype + '*}', 
                  '\\end_layout', '', '\\end_inset']
        document.body[j : j+1] = subst
        addedLines = len(subst) - 1
        del document.body[i+1:l-1]
        addedLines -= (l-1) - (i+1)
        subst = ['\\begin_inset ERT', 'status collapsed', '',
                 '\\begin_layout Standard', '', '', '\\backslash',
                 'begin{sideways' + floattype + '*}', '\\end_layout', '',
                 '\\end_inset', '', '\\end_layout', '']
        document.body[i : i+1] = subst
        addedLines += len(subst) - 1
        add_to_preamble(document, ['\\usepackage{rotfloat}\n'])
        i += addedLines + 1


def revert_inset_embedding(document, type):
    ' Remove embed tag from certain type of insets'
    i = 0
    while 1:
        i = find_token(document.body, "\\begin_inset %s" % type, i)
        if i == -1:
            return
        j = find_end_of_inset(document.body, i)
        if j == -1:
            document.warning("Malformed lyx document: Missing '\\end_inset' in revert_inset_embedding.")
            i = i + 1
            continue
        k = find_token(document.body, "\tembed", i, j)
        if k == -1:
            k = find_token(document.body, "embed", i, j)
        if k != -1:
            del document.body[k]
        i = i + 1


def revert_external_embedding(document):
    ' Remove embed tag from external inset '
    revert_inset_embedding(document, 'External')


def convert_subfig(document):
    " Convert subfigures to subfloats. "
    i = 0
    while 1:
        i = find_token(document.body, '\\begin_inset Graphics', i)
        if i == -1:
            return
        endInset = find_end_of_inset(document.body, i)
        if endInset == -1:
            document.warning("Malformed lyx document: Missing '\\end_inset' in convert_subfig.")
            i += 1
            continue
        k = find_token(document.body, '\tsubcaption', i, endInset)
        if k == -1:
            i = endInset
            continue
        l = find_token(document.body, '\tsubcaptionText', i, endInset)
        if l == -1:
            document.warning("Malformed lyx document: Can't find subcaptionText!")
            i = endInset
            continue
        caption = document.body[l][16:].strip('"')
        del document.body[l]
        del document.body[k]
        addedLines = -2
        subst = ['\\begin_inset Float figure', 'wide false', 'sideways false', 
                 'status open', '', '\\begin_layout Plain Layout', '\\begin_inset Caption', 
                 '', '\\begin_layout Plain Layout'] + latex2lyx(caption) + \
                 [ '\\end_layout', '', '\\end_inset', '', 
                 '\\end_layout', '', '\\begin_layout Plain Layout']
        document.body[i : i] = subst
        addedLines += len(subst)
        endInset += addedLines
        subst = ['', '\\end_inset', '', '\\end_layout']
        document.body[endInset : endInset] = subst
        addedLines += len(subst)
        i += addedLines + 1


def revert_subfig(document):
    " Revert subfloats. "
    i = 0
    while 1:
        # whitespace intended (exclude \\begin_inset FloatList)
        i = find_tokens(document.body, ['\\begin_inset Float ', '\\begin_inset Wrap'], i)
        if i == -1:
            return
        j = 0
        addedLines = 0
        while j != -1:
            j = find_end_of_inset(document.body, i)
            if j == -1:
                document.warning("Malformed lyx document: Missing '\\end_inset' (float) at line " + str(i + len(document.header)) + ".\n\t" + document.body[i])
                # document.warning(document.body[i-1] + "\n" + document.body[i+1])
                i += 1
                continue # this will get us back to the outer loop, since j == -1
            # look for embedded float (= subfloat)
            # whitespace intended (exclude \\begin_inset FloatList)
            k = find_token(document.body, '\\begin_inset Float ', i + 1, j)
            if k == -1:
                break
            l = find_end_of_inset(document.body, k)
            if l == -1:
                document.warning("Malformed lyx document: Missing '\\end_inset' (embedded float).")
                i += 1
                j == -1
                continue # escape to the outer loop
            m = find_default_layout(document, k + 1, l)
            # caption?
            cap = find_token(document.body, '\\begin_inset Caption', k + 1, l)
            caption = ''
            shortcap = ''
            capend = cap
            if cap != -1:
                capend = find_end_of_inset(document.body, cap)
                if capend == -1:
                    document.warning("Malformed lyx document: Missing '\\end_inset' (caption).")
                    return
                # label?
                label = ''
                lbl = find_token(document.body, '\\begin_inset CommandInset label', cap, capend)
                if lbl != -1:
                    lblend = find_end_of_inset(document.body, lbl + 1)
                    if lblend == -1:
                        document.warning("Malformed lyx document: Missing '\\end_inset' (label).")
                        return
                    for line in document.body[lbl:lblend + 1]:
                        if line.startswith('name '):
                            label = line.split()[1].strip('"')
                            break
                else:
                    lbl = capend
                    lblend = capend
                    label = ''
                # opt arg?
                opt = find_token(document.body, '\\begin_inset OptArg', cap, capend)
                if opt != -1:
                    optend = find_end_of_inset(document.body, opt)
                    if optend == -1:
                        document.warning("Malformed lyx document: Missing '\\end_inset' (OptArg).")
                        return
                    optc = find_default_layout(document, opt, optend)
                    if optc == -1:
                        document.warning("Malformed LyX document: Missing `\\begin_layout' in Float inset.")
                        return
                    optcend = find_end_of(document.body, optc, "\\begin_layout", "\\end_layout")
                    for line in document.body[optc:optcend]:
                        if not line.startswith('\\'):
                            shortcap += line.strip()
                else:
                    opt = capend
                    optend = capend
                for line in document.body[cap:capend]:
                    if line in document.body[lbl:lblend]:
                        continue
                    elif line in document.body[opt:optend]:
                        continue
                    elif not line.startswith('\\'):
                        caption += line.strip()
                if len(label) > 0:
                    caption += "\\backslash\nlabel{" + label + "}"
            subst = '\\begin_layout Plain Layout\n\\begin_inset ERT\nstatus collapsed\n\n' \
                      '\\begin_layout Plain Layout\n\n}\n\\end_layout\n\n\\end_inset\n\n' \
                      '\\end_layout\n\n\\begin_layout Plain Layout\n'
            subst = subst.split('\n')
            document.body[l : l+1] = subst
            addedLines = len(subst) - 1
            # this is before l and so is unchanged by the multiline insertion
            if cap != capend:
                del document.body[cap:capend+1]
                addedLines -= (capend + 1 - cap)
            del document.body[k+1:m-1]
            addedLines -= (m - 1 - (k + 1))
            insertion = '\\begin_inset ERT\nstatus collapsed\n\n' \
                        '\\begin_layout Plain Layout\n\n\\backslash\n' \
                        'subfloat'
            if len(shortcap) > 0:
                insertion = insertion + "[" + shortcap + "]"
            if len(caption) > 0:
                insertion = insertion + "[" + caption + "]"
            insertion = insertion + '{%\n\\end_layout\n\n\\end_inset\n\n\\end_layout\n'
            insertion = insertion.split('\n')
            document.body[k : k + 1] = insertion
            addedLines += len(insertion) - 1
            add_to_preamble(document, ['\\usepackage{subfig}\n'])
        i += addedLines + 1


def revert_wrapplacement(document):
    " Revert placement options wrap floats (wrapfig). "
    i = 0
    while True:
        i = find_token(document.body, "\\begin_inset Wrap figure", i)
        if i == -1:
            return
        e = find_end_of_inset(document.body, i)
        j = find_token(document.body, "placement", i + 1, e)
        if j == -1:
            document.warning("Malformed LyX document: Couldn't find placement parameter of wrap float.")
            i += 1
            continue
        r = re.compile("placement (o|i|l|r)")
        m = r.match(document.body[j])
        if m == None:
            document.warning("Malformed LyX document: Placement option isn't O|I|R|L!")
        document.body[j] = "placement " + m.group(1).lower()
        i = j


def remove_extra_embedded_files(document):
    " Remove \extra_embedded_files from buffer params "
    i = find_token(document.header, '\\extra_embedded_files', 0)
    if i == -1:
        return
    document.header.pop(i)


def convert_spaceinset(document):
    " Convert '\\InsetSpace foo' to '\\begin_inset Space foo\n\\end_inset' "
    i = 0
    while i < len(document.body):
        m = re.match(r'(.*)\\InsetSpace (.*)', document.body[i])
        if m:
            before = m.group(1)
            after = m.group(2)
            subst = [before, "\\begin_inset Space " + after, "\\end_inset"]
            document.body[i: i+1] = subst
            i = i + len(subst)
        else:
            i = i + 1


def revert_spaceinset(document):
    " Revert '\\begin_inset Space foo\n\\end_inset' to '\\InsetSpace foo' "
    i = 0
    while True:
        i = find_token(document.body, "\\begin_inset Space", i)
        if i == -1:
            return
        j = find_end_of_inset(document.body, i)
        if j == -1:
            document.warning("Malformed LyX document: Could not find end of space inset.")
            continue
        document.body[i] = document.body[i].replace('\\begin_inset Space', '\\InsetSpace')
        del document.body[j]


def convert_hfill(document):
    " Convert hfill to space inset "
    i = 0
    while True:
        i = find_token(document.body, "\\hfill", i)
        if i == -1:
            return
        subst = document.body[i].replace('\\hfill', \
                  '\n\\begin_inset Space \\hfill{}\n\\end_inset')
        subst = subst.split('\n')
        document.body[i : i+1] = subst
        i += len(subst)


def revert_hfills(document):
    ' Revert \\hfill commands '
    hfill = re.compile(r'\\hfill')
    dotfill = re.compile(r'\\dotfill')
    hrulefill = re.compile(r'\\hrulefill')
    i = 0
    while True:
        i = find_token(document.body, "\\InsetSpace", i)
        if i == -1:
            return
        if hfill.search(document.body[i]):
            document.body[i] = \
              document.body[i].replace('\\InsetSpace \\hfill{}', '\\hfill')
            i += 1
            continue
        if dotfill.search(document.body[i]):
            subst = document.body[i].replace('\\InsetSpace \\dotfill{}', \
              '\\begin_inset ERT\nstatus collapsed\n\n' \
              '\\begin_layout Standard\n\n\n\\backslash\n' \
              'dotfill{}\n\\end_layout\n\n\\end_inset\n\n')
            subst = subst.split('\n')
            document.body[i : i+1] = subst
            i += len(subst)
            continue
        if hrulefill.search(document.body[i]):
            subst = document.body[i].replace('\\InsetSpace \\hrulefill{}', \
              '\\begin_inset ERT\nstatus collapsed\n\n' \
              '\\begin_layout Standard\n\n\n\\backslash\n' \
              'hrulefill{}\n\\end_layout\n\n\\end_inset\n\n')
            subst = subst.split('\n')
            document.body[i : i+1] = subst
            i += len(subst)
            continue
        i += 1

def revert_hspace(document):
    ' Revert \\InsetSpace \\hspace{} to ERT '
    i = 0
    hspace = re.compile(r'\\hspace{}')
    hstar  = re.compile(r'\\hspace\*{}')
    while True:
        i = find_token(document.body, "\\InsetSpace \\hspace", i)
        if i == -1:
            return
        length = get_value(document.body, '\\length', i+1)
        if length == '':
            document.warning("Malformed lyx document: Missing '\\length' in Space inset.")
            return
        del document.body[i+1]
        addedLines = -1
        if hstar.search(document.body[i]):
            subst = document.body[i].replace('\\InsetSpace \\hspace*{}', \
              '\\begin_inset ERT\nstatus collapsed\n\n' \
              '\\begin_layout Standard\n\n\n\\backslash\n' \
              'hspace*{' + length + '}\n\\end_layout\n\n\\end_inset\n\n')
            subst = subst.split('\n')
            document.body[i : i+1] = subst
            addedLines += len(subst) - 1
            i += addedLines + 1
            continue
        if hspace.search(document.body[i]):
            subst = document.body[i].replace('\\InsetSpace \\hspace{}', \
              '\\begin_inset ERT\nstatus collapsed\n\n' \
              '\\begin_layout Standard\n\n\n\\backslash\n' \
              'hspace{' + length + '}\n\\end_layout\n\n\\end_inset\n\n')
            subst = subst.split('\n')
            document.body[i : i+1] = subst
            addedLines += len(subst) - 1
            i += addedLines + 1
            continue
        i += 1


def revert_protected_hfill(document):
    ' Revert \\begin_inset Space \\hspace*{\\fill} to ERT '
    i = 0
    while True:
        i = find_token(document.body, '\\begin_inset Space \\hspace*{\\fill}', i)
        if i == -1:
            return
        j = find_end_of_inset(document.body, i)
        if j == -1:
            document.warning("Malformed LyX document: Could not find end of space inset.")
            continue
        del document.body[j]
        subst = document.body[i].replace('\\begin_inset Space \\hspace*{\\fill}', \
          '\\begin_inset ERT\nstatus collapsed\n\n' \
          '\\begin_layout Standard\n\n\n\\backslash\n' \
          'hspace*{\n\\backslash\nfill}\n\\end_layout\n\n\\end_inset\n\n')
        subst = subst.split('\n')
        document.body[i : i+1] = subst
        i += len(subst)


def revert_leftarrowfill(document):
    ' Revert \\begin_inset Space \\leftarrowfill{} to ERT '
    i = 0
    while True:
        i = find_token(document.body, '\\begin_inset Space \\leftarrowfill{}', i)
        if i == -1:
            return
        j = find_end_of_inset(document.body, i)
        if j == -1:
            document.warning("Malformed LyX document: Could not find end of space inset.")
            continue
        del document.body[j]
        subst = document.body[i].replace('\\begin_inset Space \\leftarrowfill{}', \
          '\\begin_inset ERT\nstatus collapsed\n\n' \
          '\\begin_layout Standard\n\n\n\\backslash\n' \
          'leftarrowfill{}\n\\end_layout\n\n\\end_inset\n\n')
        subst = subst.split('\n')
        document.body[i : i+1] = subst
        i += len(subst)


def revert_rightarrowfill(document):
    ' Revert \\begin_inset Space \\rightarrowfill{} to ERT '
    i = 0
    while True:
        i = find_token(document.body, '\\begin_inset Space \\rightarrowfill{}', i)
        if i == -1:
            return
        j = find_end_of_inset(document.body, i)
        if j == -1:
            document.warning("Malformed LyX document: Could not find end of space inset.")
            continue
        del document.body[j]
        subst = document.body[i].replace('\\begin_inset Space \\rightarrowfill{}', \
          '\\begin_inset ERT\nstatus collapsed\n\n' \
          '\\begin_layout Standard\n\n\n\\backslash\n' \
          'rightarrowfill{}\n\\end_layout\n\n\\end_inset\n\n')
        subst = subst.split('\n')
        document.body[i : i+1] = subst
        i += len(subst)


def revert_upbracefill(document):
    ' Revert \\begin_inset Space \\upbracefill{} to ERT '
    i = 0
    while True:
        i = find_token(document.body, '\\begin_inset Space \\upbracefill{}', i)
        if i == -1:
            return
        j = find_end_of_inset(document.body, i)
        if j == -1:
            document.warning("Malformed LyX document: Could not find end of space inset.")
            continue
        del document.body[j]
        subst = document.body[i].replace('\\begin_inset Space \\upbracefill{}', \
          '\\begin_inset ERT\nstatus collapsed\n\n' \
          '\\begin_layout Standard\n\n\n\\backslash\n' \
          'upbracefill{}\n\\end_layout\n\n\\end_inset\n\n')
        subst = subst.split('\n')
        document.body[i : i+1] = subst
        i += len(subst)


def revert_downbracefill(document):
    ' Revert \\begin_inset Space \\downbracefill{} to ERT '
    i = 0
    while True:
        i = find_token(document.body, '\\begin_inset Space \\downbracefill{}', i)
        if i == -1:
            return
        j = find_end_of_inset(document.body, i)
        if j == -1:
            document.warning("Malformed LyX document: Could not find end of space inset.")
            continue
        del document.body[j]
        subst = document.body[i].replace('\\begin_inset Space \\downbracefill{}', \
          '\\begin_inset ERT\nstatus collapsed\n\n' \
          '\\begin_layout Standard\n\n\n\\backslash\n' \
          'downbracefill{}\n\\end_layout\n\n\\end_inset\n\n')
        subst = subst.split('\n')
        document.body[i : i+1] = subst
        i += len(subst)


def revert_local_layout(document):
    ' Revert local layout headers.'
    i = 0
    while True:
        i = find_token(document.header, "\\begin_local_layout", i)
        if i == -1:
            return
        j = find_end_of(document.header, i, "\\begin_local_layout", "\\end_local_layout")
        if j == -1:
            # this should not happen
            break
        document.header[i : j + 1] = []


def convert_pagebreaks(document):
    ' Convert inline Newpage insets to new format '
    i = 0
    while True:
        i = find_token(document.body, '\\newpage', i)
        if i == -1:
            break
        document.body[i:i+1] = ['\\begin_inset Newpage newpage',
                                '\\end_inset']
    i = 0
    while True:
        i = find_token(document.body, '\\pagebreak', i)
        if i == -1:
            break
        document.body[i:i+1] = ['\\begin_inset Newpage pagebreak',
                                '\\end_inset']
    i = 0
    while True:
        i = find_token(document.body, '\\clearpage', i)
        if i == -1:
            break
        document.body[i:i+1] = ['\\begin_inset Newpage clearpage',
                                '\\end_inset']
    i = 0
    while True:
        i = find_token(document.body, '\\cleardoublepage', i)
        if i == -1:
            break
        document.body[i:i+1] = ['\\begin_inset Newpage cleardoublepage',
                                '\\end_inset']


def revert_pagebreaks(document):
    ' Revert \\begin_inset Newpage to previous inline format '
    i = 0
    while True:
        i = find_token(document.body, '\\begin_inset Newpage', i)
        if i == -1:
            return
        j = find_end_of_inset(document.body, i)
        if j == -1:
            document.warning("Malformed LyX document: Could not find end of Newpage inset.")
            continue
        del document.body[j]
        document.body[i] = document.body[i].replace('\\begin_inset Newpage newpage', '\\newpage')
        document.body[i] = document.body[i].replace('\\begin_inset Newpage pagebreak', '\\pagebreak')
        document.body[i] = document.body[i].replace('\\begin_inset Newpage clearpage', '\\clearpage')
        document.body[i] = document.body[i].replace('\\begin_inset Newpage cleardoublepage', '\\cleardoublepage')


def convert_linebreaks(document):
    ' Convert inline Newline insets to new format '
    i = 0
    while True:
        i = find_token(document.body, '\\newline', i)
        if i == -1:
            break
        document.body[i:i+1] = ['\\begin_inset Newline newline',
                                '\\end_inset']
    i = 0
    while True:
        i = find_token(document.body, '\\linebreak', i)
        if i == -1:
            break
        document.body[i:i+1] = ['\\begin_inset Newline linebreak',
                                '\\end_inset']


def revert_linebreaks(document):
    ' Revert \\begin_inset Newline to previous inline format '
    i = 0
    while True:
        i = find_token(document.body, '\\begin_inset Newline', i)
        if i == -1:
            return
        j = find_end_of_inset(document.body, i)
        if j == -1:
            document.warning("Malformed LyX document: Could not find end of Newline inset.")
            continue
        del document.body[j]
        document.body[i] = document.body[i].replace('\\begin_inset Newline newline', '\\newline')
        document.body[i] = document.body[i].replace('\\begin_inset Newline linebreak', '\\linebreak')


def convert_japanese_plain(document):
    ' Set language japanese-plain to japanese '
    i = 0
    if document.language == "japanese-plain":
        document.language = "japanese"
        i = find_token(document.header, "\\language", 0)
        if i != -1:
            document.header[i] = "\\language japanese"
    j = 0
    while True:
        j = find_token(document.body, "\\lang japanese-plain", j)
        if j == -1:
            return
        document.body[j] = document.body[j].replace("\\lang japanese-plain", "\\lang japanese")
        j = j + 1


def revert_pdfpages(document):
    ' Revert pdfpages external inset to ERT '
    i = 0
    while 1:
        i = find_token(document.body, "\\begin_inset External", i)
        if i == -1:
            return
        j = find_end_of_inset(document.body, i)
        if j == -1:
            document.warning("Malformed lyx document: Missing '\\end_inset' in revert_pdfpages.")
            i = i + 1
            continue
        if get_value(document.body, 'template', i, j) == "PDFPages":
            filename = get_value(document.body, 'filename', i, j)
            extra = ''
            r = re.compile(r'\textra PDFLaTeX \"(.*)\"$')
            for k in range(i, j):
                m = r.match(document.body[k])
                if m:
                    extra = m.group(1)
            angle = get_value(document.body, 'rotateAngle', i, j)
            width = get_value(document.body, 'width', i, j)
            height = get_value(document.body, 'height', i, j)
            scale = get_value(document.body, 'scale', i, j)
            keepAspectRatio = find_token(document.body, "\tkeepAspectRatio", i, j)
            options = extra
            if angle != '':
                 if options != '':
                     options += ",angle=" + angle
                 else:
                     options += "angle=" + angle
            if width != '':
                 if options != '':
                     options += ",width=" + convert_len(width)
                 else:
                     options += "width=" + convert_len(width)
            if height != '':
                 if options != '':
                     options += ",height=" + convert_len(height)
                 else:
                     options += "height=" + convert_len(height)
            if scale != '':
                 if options != '':
                     options += ",scale=" + scale
                 else:
                     options += "scale=" + scale
            if keepAspectRatio != '':
                 if options != '':
                     options += ",keepaspectratio"
                 else:
                     options += "keepaspectratio"
            if options != '':
                     options = '[' + options + ']'
            del document.body[i+1:j+1]
            document.body[i:i+1] = ['\\begin_inset ERT',
                                'status collapsed',
                                '',
                                '\\begin_layout Standard',
                                '',
                                '\\backslash',
                                'includepdf' + options + '{' + filename + '}',
                                '\\end_layout',
                                '',
                                '\\end_inset']
            add_to_preamble(document, ['\\usepackage{pdfpages}\n'])
            i = i + 1
            continue
        i = i + 1


def revert_mexican(document):
    ' Set language Spanish(Mexico) to Spanish '
    i = 0
    if document.language == "spanish-mexico":
        document.language = "spanish"
        i = find_token(document.header, "\\language", 0)
        if i != -1:
            document.header[i] = "\\language spanish"
    j = 0
    while True:
        j = find_token(document.body, "\\lang spanish-mexico", j)
        if j == -1:
            return
        document.body[j] = document.body[j].replace("\\lang spanish-mexico", "\\lang spanish")
        j = j + 1


def remove_embedding(document):
    ' Remove embed tag from all insets '
    revert_inset_embedding(document, 'Graphics')
    revert_inset_embedding(document, 'External')
    revert_inset_embedding(document, 'CommandInset include')
    revert_inset_embedding(document, 'CommandInset bibtex')


def revert_master(document):
    ' Remove master param '
    i = find_token(document.header, "\\master", 0)
    if i != -1:
        del document.header[i]


def revert_graphics_group(document):
    ' Revert group information from graphics insets '
    i = 0
    while 1:
        i = find_token(document.body, "\\begin_inset Graphics", i)
        if i == -1:
            return
        j = find_end_of_inset(document.body, i)
        if j == -1:
            document.warning("Malformed lyx document: Missing '\\end_inset' in revert_graphics_group.")
            i = i + 1
            continue
        k = find_token(document.body, "	groupId", i, j)
        if k == -1:
            i = i + 1
            continue
        del document.body[k]
        i = i + 1


def update_apa_styles(document):
    ' Replace obsolete styles '

    if document.textclass != "apa":
        return

    obsoletedby = { "Acknowledgments": "Acknowledgements",
                    "Section*":        "Section",
                    "Subsection*":     "Subsection",
                    "Subsubsection*":  "Subsubsection",
                    "Paragraph*":      "Paragraph",
                    "Subparagraph*":   "Subparagraph"}
    i = 0
    while 1:
        i = find_token(document.body, "\\begin_layout", i)
        if i == -1:
            return

        layout = document.body[i][14:]
        if layout in obsoletedby:
            document.body[i] = "\\begin_layout " + obsoletedby[layout]

        i += 1


def convert_paper_sizes(document):
    ' exchange size options legalpaper and executivepaper to correct order '
    # routine is needed to fix http://bugzilla.lyx.org/show_bug.cgi?id=4868
    i = 0
    j = 0
    i = find_token(document.header, "\\papersize executivepaper", 0)
    if i != -1:
        document.header[i] = "\\papersize legalpaper"
        return
    j = find_token(document.header, "\\papersize legalpaper", 0)
    if j != -1:
        document.header[j] = "\\papersize executivepaper"


def revert_paper_sizes(document):
    ' exchange size options legalpaper and executivepaper to correct order '
    i = 0
    j = 0
    i = find_token(document.header, "\\papersize executivepaper", 0)
    if i != -1:
        document.header[i] = "\\papersize legalpaper"
        return
    j = find_token(document.header, "\\papersize legalpaper", 0)
    if j != -1:
        document.header[j] = "\\papersize executivepaper"


def convert_InsetSpace(document):
    " Convert '\\begin_inset Space foo' to '\\begin_inset space foo'"
    i = 0
    while True:
        i = find_token(document.body, "\\begin_inset Space", i)
        if i == -1:
            return
        document.body[i] = document.body[i].replace('\\begin_inset Space', '\\begin_inset space')


def revert_InsetSpace(document):
    " Revert '\\begin_inset space foo' to '\\begin_inset Space foo'"
    i = 0
    while True:
        i = find_token(document.body, "\\begin_inset space", i)
        if i == -1:
            return
        document.body[i] = document.body[i].replace('\\begin_inset space', '\\begin_inset Space')


def convert_display_enum(document):
    " Convert 'display foo' to 'display false/true'"
    i = 0
    while True:
        i = find_token(document.body, "\tdisplay", i)
        if i == -1:
            return
        val = get_value(document.body, 'display', i)
        if val == "none":
            document.body[i] = document.body[i].replace('none', 'false')
        if val == "default":
            document.body[i] = document.body[i].replace('default', 'true')
        if val == "monochrome":
            document.body[i] = document.body[i].replace('monochrome', 'true')
        if val == "grayscale":
            document.body[i] = document.body[i].replace('grayscale', 'true')
        if val == "color":
            document.body[i] = document.body[i].replace('color', 'true')
        if val == "preview":
            document.body[i] = document.body[i].replace('preview', 'true')
        i += 1


def revert_display_enum(document):
    " Revert 'display false/true' to 'display none/color'"
    i = 0
    while True:
        i = find_token(document.body, "\tdisplay", i)
        if i == -1:
            return
        val = get_value(document.body, 'display', i)
        if val == "false":
            document.body[i] = document.body[i].replace('false', 'none')
        if val == "true":
            document.body[i] = document.body[i].replace('true', 'default')
        i += 1


def remove_fontsCJK(document):
    ' Remove font_cjk param '
    i = find_token(document.header, "\\font_cjk", 0)
    if i != -1:
        del document.header[i]


def convert_plain_layout(document):
    " Convert 'PlainLayout' to 'Plain Layout'" 
    i = 0
    while True:
        i = find_token(document.body, '\\begin_layout PlainLayout', i)
        if i == -1:
            return
        document.body[i] = document.body[i].replace('\\begin_layout PlainLayout', \
          '\\begin_layout Plain Layout')
        i += 1


def revert_plain_layout(document):
    " Convert 'PlainLayout' to 'Plain Layout'" 
    i = 0
    while True:
        i = find_token(document.body, '\\begin_layout Plain Layout', i)
        if i == -1:
            return
        document.body[i] = document.body[i].replace('\\begin_layout Plain Layout', \
          '\\begin_layout PlainLayout')
        i += 1


def revert_plainlayout(document):
    " Convert 'PlainLayout' to 'Plain Layout'" 
    i = 0
    while True:
        i = find_token(document.body, '\\begin_layout PlainLayout', i)
        if i == -1:
            return
        # This will be incorrect for some document classes, since Standard is not always
        # the default. But (a) it is probably the best we can do and (b) it will actually
        # work, in fact, since an unknown layout will be converted to default.
        document.body[i] = document.body[i].replace('\\begin_layout PlainLayout', \
          '\\begin_layout Standard')
        i += 1


def revert_polytonicgreek(document):
    "Set language polytonic Greek to Greek"
    i = 0
    if document.language == "polutonikogreek":
        document.language = "greek"
        i = find_token(document.header, "\\language", 0)
        if i != -1:
            document.header[i] = "\\language greek"
    j = 0
    while True:
        j = find_token(document.body, "\\lang polutonikogreek", j)
        if j == -1:
            return
        document.body[j] = document.body[j].replace("\\lang polutonikogreek", "\\lang greek")
        j = j + 1


def revert_removed_modules(document):
    i = 0
    while True:
        i = find_token(document.header, "\\begin_remove_modules", i)
        if i == -1:
            return
        j = find_end_of(document.header, i, "\\begin_remove_modules", "\\end_remove_modules")
        if j == -1:
            # this should not happen
            break
        document.header[i : j + 1] = []


##
# Conversion hub
#

supported_versions = ["1.6.0","1.6"]
convert = [[277, [fix_wrong_tables]],
           [278, [close_begin_deeper]],
           [279, [long_charstyle_names]],
           [280, [axe_show_label]],
           [281, []],
           [282, []],
           [283, [convert_flex]],
           [284, []],
           [285, []],
           [286, []],
           [287, [convert_wrapfig_options]],
           [288, [convert_inset_command]],
           [289, [convert_latexcommand_index]],
           [290, []],
           [291, []],
           [292, [convert_japanese_cjk]],
           [293, []],
           [294, [convert_pdf_options]],
           [295, [convert_htmlurl, convert_url]],
           [296, [convert_include]],
           [297, [convert_usorbian]],
           [298, []],
           [299, []],
           [300, []],
           [301, []],
           [302, []],
           [303, [convert_serbocroatian]],
           [304, [convert_framed_notes]],
           [305, []],
           [306, []],
           [307, []],
           [308, []],
           [309, []],
           [310, []],
           [311, [convert_ams_classes]],
           [312, []],
           [313, [convert_module_names]],
           [314, []],
           [315, []],
           [316, [convert_subfig]],
           [317, []],
           [318, []],
           [319, [convert_spaceinset, convert_hfill]],
           [320, []],
           [321, [convert_tablines]],
           [322, [convert_plain_layout]],
           [323, [convert_pagebreaks]],
           [324, [convert_linebreaks]],
           [325, [convert_japanese_plain]],
           [326, []],
           [327, []],
           [328, [remove_embedding, remove_extra_embedded_files, remove_inzip_options]],
           [329, []],
           [330, []],
           [331, [convert_ltcaption]],
           [332, []],
           [333, [update_apa_styles]],
           [334, [convert_paper_sizes]],
           [335, [convert_InsetSpace]],
           [336, []],
           [337, [convert_display_enum]],
           [338, []],
           [339, []],
          ]

revert =  [[338, [revert_removed_modules]],
           [337, [revert_polytonicgreek]],
           [336, [revert_display_enum]],
           [335, [remove_fontsCJK]],
           [334, [revert_InsetSpace]],
           [333, [revert_paper_sizes]],
           [332, []],
           [331, [revert_graphics_group]],
           [330, [revert_ltcaption]],
           [329, [revert_leftarrowfill, revert_rightarrowfill, revert_upbracefill, revert_downbracefill]],
           [328, [revert_master]],
           [327, []],
           [326, [revert_mexican]],
           [325, [revert_pdfpages]],
           [324, []],
           [323, [revert_linebreaks]],
           [322, [revert_pagebreaks]],
           [321, [revert_local_layout, revert_plain_layout]],
           [320, [revert_tablines]],
           [319, [revert_protected_hfill]],
           [318, [revert_spaceinset, revert_hfills, revert_hspace]],
           [317, [remove_extra_embedded_files]],
           [316, [revert_wrapplacement]],
           [315, [revert_subfig]],
           [314, [revert_colsep, revert_plainlayout]],
           [313, []],
           [312, [revert_module_names]],
           [311, [revert_rotfloat, revert_widesideways]],
           [310, [revert_external_embedding]],
           [309, [revert_btprintall]],
           [308, [revert_nocite]],
           [307, [revert_serbianlatin]],
           [306, [revert_slash, revert_nobreakdash]],
           [305, [revert_interlingua]],
           [304, [revert_bahasam]],
           [303, [revert_framed_notes]],
           [302, []],
           [301, [revert_latin, revert_samin]],
           [300, [revert_linebreak]],
           [299, [revert_pagebreak]],
           [298, [revert_hyperlinktype]],
           [297, [revert_macro_optional_params]],
           [296, [revert_albanian, revert_lowersorbian, revert_uppersorbian]],
           [295, [revert_include]],
           [294, [revert_href, revert_url]],
           [293, [revert_pdf_options_2]],
           [292, [revert_inset_info]],
           [291, [revert_japanese, revert_japanese_encoding, revert_japanese_cjk]],
           [290, [revert_vietnamese]],
           [289, [revert_wraptable]],
           [288, [revert_latexcommand_index]],
           [287, [revert_inset_command]],
           [286, [revert_wrapfig_options]],
           [285, [revert_pdf_options]],
           [284, [remove_inzip_options]],
           [283, []],
           [282, [revert_flex]],
           [281, []],
           [280, [revert_begin_modules]],
           [279, [revert_show_label]],
           [278, [revert_long_charstyle_names]],
           [277, []],
           [276, []]
          ]


if __name__ == "__main__":
    pass