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

####################################################################
# 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")

def wrap_into_ert(string, src, dst):
    " Wrap a something into an ERT"
    return string.replace(src, '\n\\begin_inset ERT\nstatus collapsed\n\\begin_layout Standard\n' 
      + dst + '\n\\end_layout\n\\end_inset\n')

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)

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

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."
        i = 0
        i = find_token(document.header, "\\use_hyperref", i)
        if i != -1:
            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:
            del document.header[i]
        i = find_token(document.header, "\\pdf_author", 0)
        if i != -1:
            del document.header[i]
        i = find_token(document.header, "\\pdf_subject", 0)
        if i != -1:
            del document.header[i]
        i = find_token(document.header, "\\pdf_keywords", 0)
        if i != -1:
            del document.header[i]
        i = find_token(document.header, "\\pdf_bookmarks", 0)
        if i != -1:
            del document.header[i]
        i = find_token(document.header, "\\pdf_bookmarksnumbered", i)
        if i != -1:
            del document.header[i]
        i = find_token(document.header, "\\pdf_bookmarksopen", i)
        if i != -1:
            del document.header[i]
        i = find_token(document.header, "\\pdf_bookmarksopenlevel", i)
        if i != -1:
            del document.header[i]
        i = find_token(document.header, "\\pdf_breaklinks", i)
        if i != -1:
            del document.header[i]
        i = find_token(document.header, "\\pdf_pdfborder", i)
        if i != -1:
            del document.header[i]
        i = find_token(document.header, "\\pdf_colorlinks", i)
        if i != -1:
            del document.header[i]
        i = find_token(document.header, "\\pdf_backref", i)
        if i != -1:
            del document.header[i]
        i = find_token(document.header, "\\pdf_pagebackref", i)
        if i != -1:
            del document.header[i]
        i = find_token(document.header, "\\pdf_pagemode", 0)
        if i != -1:
            del document.header[i]
        i = find_token(document.header, "\\pdf_quoted_options", 0)
        if i != -1:
            del document.header[i]


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, "lines", i)
        if i == -1:
            return
        j = find_token(document.body, "overhang", i+1)
        if j != i + 2 and j != -1:
            document.warning("Malformed LyX document: Couldn't find overhang parameter of wrap float.")
        if j == -1:
            return
        del document.body[i]
        del document.body[j-1]
        i = i + 1


def convert_latexcommand_index(document):
    "Convert from LatexCommand form to collapsable form."
    i = 0 
    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
        fullcontent = document.body[i + 2][6:].strip('"')
        document.body[i:i + 2] = ["\\begin_inset Index",
          "status collapsed",
          "\\begin_layout Standard"]
        # Put here the conversions needed from LaTeX string to LyXText.
        # Here we do a minimal conversion to prevent crashes and data loss.
        # Manual patch-up may be needed.
        # Umlauted characters (most common ones, can be extended):
        fullcontent = fullcontent.replace(r'\\\"a', u'ä').replace(r'\\\"o', u'ö').replace(r'\\\"u', u'ü')
        # Generic, \" -> ":
        fullcontent = wrap_into_ert(fullcontent, r'\"', '"')
        #fullcontent = fullcontent.replace(r'\"', '\n\\begin_inset ERT\nstatus collapsed\n\\begin_layout standard\n"\n\\end_layout\n\\end_inset\n')
        # Math:
        r = re.compile('^(.*?)(\$.*?\$)(.*)')
        g = fullcontent
        while r.match(g):
          m = r.match(g)
          s = m.group(1)
          f = m.group(2).replace('\\\\', '\\')
          g = m.group(3)
          if s:
            # this is non-math!
            s = wrap_into_ert(s, r'\\', '\\backslash')
            s = wrap_into_ert(s, '{', '{')
            s = wrap_into_ert(s, '}', '}')
            document.body.insert(i + 3, s)
            i += 1
          document.body.insert(i + 3, "\\begin_inset Formula " + f)
          document.body.insert(i + 4, "\\end_inset")
          i += 2
        # Generic, \\ -> \backslash:
        g = wrap_into_ert(g, r'\\', '\\backslash{}')
        g = wrap_into_ert(g, '{', '{')
        g = wrap_into_ert(g, '}', '}')
        document.body.insert(i + 3, g)
        document.body[i + 4] = "\\end_layout"
        i = i + 5


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
        del document.body[j - 1]
        del document.body[j - 2] # \end_layout
        document.body[i] =  "\\begin_inset CommandInset index"
        document.body[i + 1] =  "LatexCommand index"
        # clean up multiline stuff
        content = ""
        for k in range(i + 3, j - 2):
          line = document.body[k]
          if line.startswith("\\begin_inset ERT"):
            line = line[16:]
          if line.startswith("\\begin_inset Formula"):
            line = line[20:]
          if line.startswith("\\begin_layout Standard"):
            line = line[22:]
          if line.startswith("\\end_layout"):
            line = line[11:]
          if line.startswith("\\end_inset"):
            line = line[10:]
          if line.startswith("status collapsed"):
            line = line[16:]
          line = line.replace(u'ä', r'\\\"a').replace(u'ö', r'\\\"o').replace(u'ü', r'\\\"u')
          content = content + line;
        document.body[i + 3] = "name " + '"' + content + '"'
        for k in range(i + 4, j - 2):
          del document.body[i + 4]
        document.body.insert(i + 4, "")
        del document.body[i + 2] # \begin_layout standard
        i = 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 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_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 " + i + " of body!")
      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 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
  r1 = re.compile('LatexCommand (.+)')
  r2 = re.compile('filename (.+)')
  r3 = re.compile('options (.*)')
  while True:
    i = find_token(document.body, "\\begin_inset CommandInset include", i)
    if i == -1:
      return
    previewline = document.body[i + 1]
    m = r1.match(document.body[i + 2])
    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)
    m = r2.match(document.body[i + 3])
    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)
    options = ""
    numlines = 4
    if (cmd == "lstinputlisting"):
      m = r3.match(document.body[i + 4])
      if m != None:
        options = m.group(1)
        numlines = 5
    newline = "\\begin_inset Include \\" + cmd + "{" + fn + "}"
    if options:
      newline += ("[" + options + "]")
    insertion = [newline, previewline]
    document.body[i : i + numlines] = 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
        document.body[i] = document.body[i].replace("\\begin_inset Note", "\\begin_inset Box")
        document.body.insert(i + 1, 'position "t"\nhor_pos "c"\nhas_inner_box 0\ninner_pos "t"\n' \
        'use_parbox 0\nwidth "100col%"\nspecial "none"\nheight "1in"\n' \
        'height_special "totalheight"')
        i = i + 1


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_token(document.body, "\\begin_layout Standard", i + 1, j)
        if l == -1:
            document.warning("Malformed LyX document: Missing `\\begin_layout Standard' 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")
            document.body.insert(l + 1, "\\begin_inset Note Shaded\n" + status + "\n\\begin_layout Standard\n")
            document.body.insert(m + 1, "\\end_layout\n\\end_inset")
        i = i + 1


def revert_slash(document):
    'Revert \\SpecialChar \\slash{} to ERT'
    for i in range(len(document.body)):
        document.body[i] = document.body[i].replace('\\SpecialChar \\slash{}', \
        '\\begin_inset ERT\nstatus collapsed\n\n' \
        '\\begin_layout Standard\n\n\n\\backslash\n' \
        'slash{}\n\\end_layout\n\n\\end_inset\n\n')


def revert_nobreakdash(document):
    'Revert \\SpecialChar \\nobreakdash- to ERT'
    found = 0
    for i in range(len(document.body)):
        line = document.body[i]
        r = re.compile(r'\\SpecialChar \\nobreakdash-')
        m = r.match(line)
        if m:
            found = 1
        document.body[i] = document.body[i].replace('\\SpecialChar \\nobreakdash-', \
        '\\begin_inset ERT\nstatus collapsed\n\n' \
        '\\begin_layout Standard\n\n\n\\backslash\n' \
        'nobreakdash-\n\\end_layout\n\n\\end_inset\n\n')
    if not found:
        return
    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"


def revert_nocite_key(body, start, end):
    'key "..." -> \nocite{...}'
    for i in range(start, end):
        if (body[i][0:5] == 'key "'):
            body[i] = body[i].replace('key "', "\\backslash\nnocite{")
            body[i] = body[i].replace('"', "}")
        else:
            body[i] = ""


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
        i = i + 1
        if (document.body[i] == "LatexCommand nocite"):
            j = find_end_of_inset(document.body, i + 1)
            if j == -1:
                #this should not happen
                document.warning("End of CommandInset citation not found in revert_nocite!")
                revert_nocite_key(document.body, i + 1, len(document.body))
                return
            revert_nocite_key(document.body, i + 1, j)
            document.body[i-1] = "\\begin_inset ERT"
            document.body[i] = "status collapsed\n\n" \
            "\\begin_layout Standard"
            document.body.insert(j, "\\end_layout\n");
            i = j


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)
            for k in range(i, j):
                if (document.body[k] == 'btprint "btPrintAll"'):
                    del document.body[k]
                    document.body.insert(i, "\\begin_inset ERT\n" \
                    "status collapsed\n\n\\begin_layout Standard\n\n" \
                    "\\backslash\nnocite{*}\n" \
                    "\\end_layout\n\\end_inset\n")
            i = j


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:
        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)
        floattype = m.group(1)
        if floattype == "figure" or floattype == "table":
            i = i + 1
            continue
        j = find_end_of_inset(document.body, i)
        if j == -1:
            document.warning("Malformed lyx document: Missing '\\end_inset'.")
            i = i + 1
            continue
        if get_value(document.body, 'sideways', i, j) != "false":
            l = find_token(document.body, "\\begin_layout Standard", i + 1, j)
            if l == -1:
                document.warning("Malformed LyX document: Missing `\\begin_layout Standard' in Float inset.")
                return
            document.body[j] = '\\begin_layout Standard\n\\begin_inset ERT\nstatus collapsed\n\n' \
            '\\begin_layout Standard\n\n\n\\backslash\n' \
            'end{sideways' + floattype + '}\n\\end_layout\n\n\\end_inset\n'
            del document.body[i+1:l-1]
            document.body[i] = '\\begin_inset ERT\nstatus collapsed\n\n' \
            '\\begin_layout Standard\n\n\n\\backslash\n' \
            'begin{sideways' + floattype + '}\n\\end_layout\n\n\\end_inset\n\n\\end_layout\n\n'
            if floattype == "algorithm":
                add_to_preamble(document,
                                ['% Commands inserted by lyx2lyx for sideways algorithm float',
                                 '\\usepackage{rotfloat}\n'
                                 '\\floatstyle{ruled}\n'
                                 '\\newfloat{algorithm}{tbp}{loa}\n'
                                 '\\floatname{algorithm}{Algorithm}\n'])
            else:
                document.warning("Cannot create preamble definition for custom float" + floattype + ".")
            i = i + 1
            continue
        i = i + 1


def revert_widesideways(document):
    " Revert wide sideways floats. "
    i = 0
    while 1:
        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)
        floattype = m.group(1)
        if floattype != "figure" and floattype != "table":
            i = i + 1
            continue
        j = find_end_of_inset(document.body, i)
        if j == -1:
            document.warning("Malformed lyx document: Missing '\\end_inset'.")
            i = i + 1
            continue
        if get_value(document.body, 'sideways', i, j) != "false":
            if get_value(document.body, 'wide', i, j) != "false":
                l = find_token(document.body, "\\begin_layout Standard", i + 1, j)
                if l == -1:
                    document.warning("Malformed LyX document: Missing `\\begin_layout Standard' in Float inset.")
                    return
                document.body[j] = '\\begin_layout Standard\n\\begin_inset ERT\nstatus collapsed\n\n' \
                '\\begin_layout Standard\n\n\n\\backslash\n' \
                'end{sideways' + floattype + '*}\n\\end_layout\n\n\\end_inset\n'
                del document.body[i+1:l-1]
                document.body[i] = '\\begin_inset ERT\nstatus collapsed\n\n' \
                '\\begin_layout Standard\n\n\n\\backslash\n' \
                'begin{sideways' + floattype + '*}\n\\end_layout\n\n\\end_inset\n\n\\end_layout\n\n'
                add_to_preamble(document,
                                ['\\usepackage{rotfloat}\n'])
                i = i + 1
                continue
        i = i + 1


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
        j = find_end_of_inset(document.body, i)
        if j == -1:
            document.warning("Malformed lyx document: Missing '\\end_inset'.")
            i = i + 1
            continue
        k = find_token(document.body, '\tsubcaption', i, j)
        if k == -1:
            i = i + 1
            continue
        l = find_token(document.body, '\tsubcaptionText', i, j)
        caption = get_value(document.body, '\tsubcaptionText', i, j).strip('"')
        savestr = document.body[i]
        document.body[i] = '\\begin_inset Float figure\nwide false\nsideways false\n' \
        'status open\n\n\\begin_layout PlainLayout\n\\begin_inset Caption\n\n\\begin_layout PlainLayout\n' \
        + caption + '\n\\end_layout\n\n\\end_inset\n\n\\end_layout\n\n\\begin_layout PlainLayout\n' + savestr
        savestr = document.body[j]
        document.body[j] = '\n\\end_layout\n\n\\end_inset\n' + savestr
        del document.body[k]
        del document.body[l]


def revert_subfig(document):
    " Revert subfloats. "
    i = 0
    while 1:
        i = find_token(document.body, '\\begin_inset Float', i)
        if i == -1:
            return
        while 1:
            j = find_end_of_inset(document.body, i)
            if j == -1:
                document.warning("Malformed lyx document: Missing '\\end_inset' (float).")
                i = i + 1
                continue
            # look for embedded float (= subfloat)
            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 = i + 1
                continue
            m = find_token(document.body, "\\begin_layout PlainLayout", k + 1, l)
            # caption?
            cap = find_token(document.body, '\\begin_inset Caption', k + 1, l)
            caption = ''
            shortcap = ''
            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_token(document.body, "\\begin_layout PlainLayout", opt, optend)
                    if optc == -1:
                        document.warning("Malformed LyX document: Missing `\\begin_layout PlainLayout' 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 + "}"
            document.body[l] = '\\begin_layout PlainLayout\n\\begin_inset ERT\nstatus collapsed\n\n' \
            '\\begin_layout PlainLayout\n\n}\n\\end_layout\n\n\\end_inset\n\n\\end_layout\n\n\\begin_layout PlainLayout\n'
            del document.body[cap:capend+1]
            del document.body[k+1:m-1]
            insertion = '\\begin_inset ERT\nstatus collapsed\n\n' \
            '\\begin_layout PlainLayout\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'
            document.body[k] = insertion
            add_to_preamble(document,
                            ['\\usepackage{subfig}\n'])
        i = i + 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, []],
           [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]]
          ]

revert =  [[315, [revert_subfig]],
           [314, [revert_colsep]],
           [313, []],
           [312, [revert_module_names]],
           [311, [revert_rotfloat, revert_widesideways]],
           [310, []],
           [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]],
           [293, [revert_pdf_options_2]],
           [292, [revert_inset_info]],
           [291, [revert_japanese, revert_japanese_encoding]],
           [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