lyx_mirror/lib/lyx2lyx/lyx_1_4.py

# This file is part of lyx2lyx
# Copyright (C) 2002 Dekel Tsur <dekel@lyx.org>
# Copyright (C) 2002-2004 José Matos <jamatos@lyx.org>
# Copyright (C) 2004-2005 Georg Baum <Georg.Baum@post.rwth-aachen.de>
#
# 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., 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301  USA

"""Convert files to the file format generated by lyx 1.4"""

import os.path
import re
from os import F_OK, access

from lyx_0_12 import update_latexaccents
from parser_tools import (
    check_token,
    find_beginning_of,
    find_end_of,
    find_re,
    find_token,
    find_token_exact,
    find_tokens,
    find_tokens_backwards,
    find_tokens_exact,
    get_value,
    is_nonempty_line,
)

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


def get_layout(line, default_layout):
    "Get layout, if empty return the default layout."
    tokens = line.split()
    if len(tokens) > 1:
        return tokens[1]
    return default_layout


def get_paragraph(lines, i, format):
    "Finds the paragraph that contains line i."

    if format < 225:
        begin_layout = "\\layout"
    else:
        begin_layout = "\\begin_layout"
    while i != -1:
        i = find_tokens_backwards(lines, ["\\end_inset", begin_layout], i)
        if i == -1:
            return -1
        if check_token(lines[i], begin_layout):
            return i
        i = find_beginning_of_inset(lines, i)
    return -1


def find_beginning_of_inset(lines, i):
    "Find beginning of inset, where lines[i] is included."
    return find_beginning_of(lines, i, "\\begin_inset", "\\end_inset")


def get_next_paragraph(lines, i, format):
    "Finds the paragraph after the paragraph that contains line i."

    if format < 225:
        tokens = ["\\begin_inset", "\\layout", "\\end_float", "\\the_end"]
    elif format < 236:
        tokens = ["\\begin_inset", "\\begin_layout", "\\end_float", "\\end_document"]
    else:
        tokens = [
            "\\begin_inset",
            "\\begin_layout",
            "\\end_float",
            "\\end_body",
            "\\end_document",
        ]
    while i != -1:
        i = find_tokens(lines, tokens, i)
        if not check_token(lines[i], "\\begin_inset"):
            return i
        i = find_end_of_inset(lines, i)
    return -1


def find_end_of_inset(lines, i):
    r"Finds the matching \end_inset"
    return find_end_of(lines, i, "\\begin_inset", "\\end_inset")


def del_token(lines, token, start, end):
    """del_token(lines, token, start, end) -> int

    Find the lower line in lines where token is the first element and
    delete that line.

    Returns the number of lines remaining."""

    k = find_token_exact(lines, token, start, end)
    if k == -1:
        return end
    else:
        del lines[k]
        return end - 1


# End of helper functions
####################################################################


def remove_color_default(document):
    r"Remove \color default"
    i = 0
    while True:
        i = find_token(document.body, "\\color default", i)
        if i == -1:
            return
        document.body[i] = document.body[i].replace("\\color default", "\\color inherit")


def add_end_header(document):
    r"Add \end_header"
    document.header.append("\\end_header")


def rm_end_header(document):
    r"Remove \end_header"
    i = find_token(document.header, "\\end_header", 0)
    if i == -1:
        return
    del document.header[i]


def convert_amsmath(document):
    "Convert \\use_amsmath"
    i = find_token(document.header, "\\use_amsmath", 0)
    if i == -1:
        document.warning("Malformed LyX document: Missing '\\use_amsmath'.")
        return
    tokens = document.header[i].split()
    if len(tokens) != 2:
        document.warning(
            "Malformed LyX document: Could not parse line '%s'." % document.header[i]
        )
        use_amsmath = "0"
    else:
        use_amsmath = tokens[1]
    # old: 0 == off, 1 == on
    # new: 0 == off, 1 == auto, 2 == on
    # translate off -> auto, since old format 'off' means auto in reality
    if use_amsmath == "0":
        document.header[i] = "\\use_amsmath 1"
    else:
        document.header[i] = "\\use_amsmath 2"


def revert_amsmath(document):
    "Revert \\use_amsmath"
    i = find_token(document.header, "\\use_amsmath", 0)
    if i == -1:
        document.warning("Malformed LyX document: Missing '\\use_amsmath'.")
        return
    tokens = document.header[i].split()
    if len(tokens) != 2:
        document.warning(
            "Malformed LyX document: Could not parse line '%s'." % document.header[i]
        )
        use_amsmath = "0"
    else:
        use_amsmath = tokens[1]
    # old: 0 == off, 1 == on
    # new: 0 == off, 1 == auto, 2 == on
    # translate auto -> off, since old format 'off' means auto in reality
    if use_amsmath == "2":
        document.header[i] = "\\use_amsmath 1"
    else:
        document.header[i] = "\\use_amsmath 0"


def convert_spaces(document):
    r"\SpecialChar ~ -> \InsetSpace ~"
    for i in range(len(document.body)):
        document.body[i] = document.body[i].replace("\\SpecialChar ~", "\\InsetSpace ~")


def revert_spaces(document):
    r"\InsetSpace ~ -> \SpecialChar ~"
    regexp = re.compile(r"(.*)(\\InsetSpace\s+)(\S+)")
    i = 0
    while True:
        i = find_re(document.body, regexp, i)
        if i == -1:
            break
        space = regexp.match(document.body[i]).group(3)
        prepend = regexp.match(document.body[i]).group(1)
        if space == "~":
            document.body[i] = regexp.sub(prepend + "\\SpecialChar ~", document.body[i])
            i = i + 1
        else:
            document.body[i] = regexp.sub(prepend, document.body[i])
            document.body[i + 1 : i + 1] = ""
            if space == "\\space":
                space = "\\ "
            i = insert_ert(
                document.body,
                i + 1,
                "Collapsed",
                space,
                document.format - 1,
                document.default_layout,
            )


def rename_spaces(document):
    """\\InsetSpace \\, -> \\InsetSpace \thinspace{}
    \\InsetSpace \\space -> \\InsetSpace \\space{}"""
    for i in range(len(document.body)):
        document.body[i] = document.body[i].replace(
            "\\InsetSpace \\space", "\\InsetSpace \\space{}"
        )
        document.body[i] = document.body[i].replace(
            "\\InsetSpace \\,", "\\InsetSpace \\thinspace{}"
        )


def revert_space_names(document):
    """\\InsetSpace \thinspace{} -> \\InsetSpace \\,
    \\InsetSpace \\space{} -> \\InsetSpace \\space"""
    for i in range(len(document.body)):
        document.body[i] = document.body[i].replace(
            "\\InsetSpace \\space{}", "\\InsetSpace \\space"
        )
        document.body[i] = document.body[i].replace(
            "\\InsetSpace \\thinspace{}", "\\InsetSpace \\,"
        )


def lyx_support_escape(lab):
    "Equivalent to pre-unicode lyx::support::escape()"
    hexdigit = [
        "0",
        "1",
        "2",
        "3",
        "4",
        "5",
        "6",
        "7",
        "8",
        "9",
        "A",
        "B",
        "C",
        "D",
        "E",
        "F",
    ]
    enc = ""
    for c in lab:
        o = ord(c)
        if o >= 128 or c == "=" or c == "%":
            enc = enc + "="
            enc = enc + hexdigit[o >> 4]
            enc = enc + hexdigit[o & 15]
        else:
            enc = enc + c
    return enc


def revert_eqref(document):
    "\\begin_inset LatexCommand \\eqref -> ERT"
    regexp = re.compile(r"^\\begin_inset\s+LatexCommand\s+\\eqref")
    i = 0
    while True:
        i = find_re(document.body, regexp, i)
        if i == -1:
            break
        eqref = lyx_support_escape(regexp.sub("", document.body[i]))
        document.body[i : i + 1] = [
            "\\begin_inset ERT",
            "status Collapsed",
            "",
            "\\layout %s" % document.default_layout,
            "",
            "\\backslash ",
            "eqref" + eqref,
        ]
        i = i + 7


def convert_bibtex(document):
    "Convert BibTeX changes."
    for i in range(len(document.body)):
        document.body[i] = document.body[i].replace(
            "\\begin_inset LatexCommand \\BibTeX", "\\begin_inset LatexCommand \\bibtex"
        )


def revert_bibtex(document):
    "Revert BibTeX changes."
    for i in range(len(document.body)):
        document.body[i] = document.body[i].replace(
            "\\begin_inset LatexCommand \\bibtex", "\\begin_inset LatexCommand \\BibTeX"
        )


def remove_insetparent(document):
    r"Remove \lyxparent"
    i = 0
    while True:
        i = find_token(document.body, "\\begin_inset LatexCommand \\lyxparent", i)
        if i == -1:
            break
        del document.body[i : i + 3]


def convert_external(document):
    "Convert inset External."
    external_rexp = re.compile(r'\\begin_inset External ([^,]*),"([^"]*)",')
    external_header = "\\begin_inset External"
    i = 0
    while True:
        i = find_token(document.body, external_header, i)
        if i == -1:
            break
        look = external_rexp.search(document.body[i])
        args = ["", ""]
        if look:
            args[0] = look.group(1)
            args[1] = look.group(2)
        # FIXME: if the previous search fails then warn

        if args[0] == "RasterImage":
            # Convert a RasterImage External Inset to a Graphics Inset.
            top = "\\begin_inset Graphics"
            if args[1]:
                filename = "\tfilename " + args[1]
            document.body[i : i + 1] = [top, filename]
            i = i + 1
        else:
            # Convert the old External Inset format to the new.
            top = external_header
            template = "\ttemplate " + args[0]
            if args[1]:
                filename = "\tfilename " + args[1]
                document.body[i : i + 1] = [top, template, filename]
                i = i + 2
            else:
                document.body[i : i + 1] = [top, template]
                i = i + 1


def revert_external_1(document):
    "Revert inset External."
    external_header = "\\begin_inset External"
    i = 0
    while True:
        i = find_token(document.body, external_header, i)
        if i == -1:
            break

        template = document.body[i + 1].split()
        template.reverse()
        del document.body[i + 1]

        filename = document.body[i + 1].split()
        filename.reverse()
        del document.body[i + 1]

        params = document.body[i + 1].split()
        params.reverse()
        if document.body[i + 1]:
            del document.body[i + 1]

        document.body[i] = (
            document.body[i]
            + " "
            + template[0]
            + ', "'
            + filename[0]
            + '", " '
            + " ".join(params[1:])
            + '"'
        )
        i = i + 1


def revert_external_2(document):
    "Revert inset External. (part II)"
    draft_token = "\tdraft"
    i = 0
    while True:
        i = find_token(document.body, "\\begin_inset External", i)
        if i == -1:
            break
        j = find_end_of_inset(document.body, i + 1)
        if j == -1:
            # this should not happen
            break
        k = find_token(document.body, draft_token, i + 1, j - 1)
        if k != -1 and len(draft_token) == len(document.body[k]):
            del document.body[k]
        i = j + 1


def convert_comment(document):
    "Convert \\layout comment"
    i = 0
    comment = "\\layout Comment"
    while True:
        i = find_token(document.body, comment, i)
        if i == -1:
            return

        document.body[i : i + 1] = [
            "\\layout %s" % document.default_layout,
            "",
            "",
            "\\begin_inset Comment",
            "collapsed true",
            "",
            "\\layout %s" % document.default_layout,
        ]
        i = i + 7

        while True:
            old_i = i
            i = find_token(document.body, "\\layout", i)
            if i == -1:
                i = len(document.body) - 1
                document.body[i:i] = ["\\end_inset", "", ""]
                return

            j = find_token(document.body, "\\begin_deeper", old_i, i)
            if j == -1:
                j = i + 1
            k = find_token(document.body, "\\begin_inset", old_i, i)
            if k == -1:
                k = i + 1

            if j < i and j < k:
                i = j
                del document.body[i]
                i = find_end_of(document.body, i, "\\begin_deeper", "\\end_deeper")
                if i == -1:
                    # This case should not happen
                    # but if this happens deal with it greacefully adding
                    # the missing \end_deeper.
                    i = len(document.body) - 1
                    document.body[i:i] = ["\\end_deeper", ""]
                    return
                else:
                    del document.body[i]
                    continue

            if k < i:
                i = k
                i = find_end_of(document.body, i, "\\begin_inset", "\\end_inset")
                if i == -1:
                    # This case should not happen
                    # but if this happens deal with it greacefully adding
                    # the missing \end_inset.
                    i = len(document.body) - 1
                    document.body[i:i] = ["\\end_inset", "", "", "\\end_inset", "", ""]
                    return
                else:
                    i = i + 1
                    continue

            if document.body[i].find(comment) == -1:
                document.body[i:i] = ["\\end_inset"]
                i = i + 1
                break
            document.body[i : i + 1] = ["\\layout %s" % document.default_layout]
            i = i + 1


def revert_comment(document):
    "Revert comments"
    i = 0
    while True:
        i = find_tokens(document.body, ["\\begin_inset Comment", "\\begin_inset Greyedout"], i)

        if i == -1:
            return
        document.body[i] = "\\begin_inset Note"
        i = i + 1


def add_end_layout(document):
    r"Add \end_layout"
    i = find_token(document.body, "\\layout", 0)

    if i == -1:
        return

    i = i + 1
    struct_stack = ["\\layout"]

    while True:
        i = find_tokens(
            document.body,
            [
                "\\begin_inset",
                "\\end_inset",
                "\\layout",
                "\\begin_deeper",
                "\\end_deeper",
                "\\the_end",
            ],
            i,
        )

        if i != -1:
            token = document.body[i].split()[0]
        else:
            document.warning("Truncated document.")
            i = len(document.body)
            document.body.insert(i, "\\the_end")
            token = ""

        if token == "\\begin_inset":
            struct_stack.append(token)
            i = i + 1
            continue

        if token == "\\end_inset":
            tail = struct_stack.pop()
            if tail == "\\layout":
                document.body.insert(i, "")
                document.body.insert(i, "\\end_layout")
                i = i + 2
                # Check if it is the correct tag
                struct_stack.pop()
            i = i + 1
            continue

        if token == "\\layout":
            tail = struct_stack.pop()
            if tail == token:
                document.body.insert(i, "")
                document.body.insert(i, "\\end_layout")
                i = i + 3
            else:
                struct_stack.append(tail)
                i = i + 1
            struct_stack.append(token)
            continue

        if token == "\\begin_deeper":
            document.body.insert(i, "")
            document.body.insert(i, "\\end_layout")
            i = i + 3
            # consecutive begin_deeper only insert one end_layout
            while document.body[i].startswith("\\begin_deeper"):
                i += 1
            struct_stack.append(token)
            continue

        if token == "\\end_deeper":
            if struct_stack[-1] == "\\layout":
                document.body.insert(i, "\\end_layout")
                i = i + 1
                struct_stack.pop()
            i = i + 1
            continue

        # case \end_document
        document.body.insert(i, "")
        document.body.insert(i, "\\end_layout")
        return


def rm_end_layout(document):
    r"Remove \end_layout"
    i = 0
    while True:
        i = find_token(document.body, "\\end_layout", i)

        if i == -1:
            return

        del document.body[i]


def insert_tracking_changes(document):
    "Handle change tracking keywords."
    i = find_token(document.header, "\\tracking_changes", 0)
    if i == -1:
        document.header.append("\\tracking_changes 0")


def rm_tracking_changes(document):
    "Remove change tracking keywords."
    i = find_token(document.header, "\\author", 0)
    if i != -1:
        del document.header[i]

    i = find_token(document.header, "\\tracking_changes", 0)
    if i == -1:
        return
    del document.header[i]


def rm_body_changes(document):
    "Remove body changes."
    i = 0
    while True:
        i = find_token(document.body, "\\change_", i)
        if i == -1:
            return

        del document.body[i]


def layout2begin_layout(document):
    r"\layout -> \begin_layout"
    i = 0
    while True:
        i = find_token(document.body, "\\layout", i)
        if i == -1:
            return

        document.body[i] = document.body[i].replace("\\layout", "\\begin_layout")
        i = i + 1


def begin_layout2layout(document):
    r"\begin_layout -> \layout"
    i = 0
    while True:
        i = find_token(document.body, "\\begin_layout", i)
        if i == -1:
            return

        document.body[i] = document.body[i].replace("\\begin_layout", "\\layout")
        i = i + 1


def convert_valignment_middle(body, start, end):
    'valignment="center" -> valignment="middle"'
    for i in range(start, end):
        if re.search('^<(column|cell) .*valignment="center".*>$', body[i]):
            body[i] = body[i].replace('valignment="center"', 'valignment="middle"')


def convert_table_valignment_middle(document):
    "Convert table  valignment, center -> middle"
    regexp = re.compile(r"^\\begin_inset\s+Tabular")
    i = 0
    while True:
        i = find_re(document.body, regexp, i)
        if i == -1:
            return
        j = find_end_of_inset(document.body, i + 1)
        if j == -1:
            # this should not happen
            convert_valignment_middle(document.body, i + 1, len(document.body))
            return
        convert_valignment_middle(document.body, i + 1, j)
        i = j + 1


def revert_table_valignment_middle(body, start, end):
    "valignment, middle -> center"
    for i in range(start, end):
        if re.search('^<(column|cell) .*valignment="middle".*>$', body[i]):
            body[i] = body[i].replace('valignment="middle"', 'valignment="center"')


def revert_valignment_middle(document):
    "Convert table  valignment, middle -> center"
    regexp = re.compile(r"^\\begin_inset\s+Tabular")
    i = 0
    while True:
        i = find_re(document.body, regexp, i)
        if i == -1:
            return
        j = find_end_of_inset(document.body, i + 1)
        if j == -1:
            # this should not happen
            revert_table_valignment_middle(document.body, i + 1, len(document.body))
            return
        revert_table_valignment_middle(document.body, i + 1, j)
        i = j + 1


def convert_end_document(document):
    "\\the_end -> \\end_document"
    i = find_token(document.body, "\\the_end", 0)
    if i == -1:
        document.body.append("\\end_document")
        return
    document.body[i] = "\\end_document"


def revert_end_document(document):
    "\\end_document -> \\the_end"
    i = find_token(document.body, "\\end_document", 0)
    if i == -1:
        document.body.append("\\the_end")
        return
    document.body[i] = "\\the_end"


def convert_breaks(document):
    r"""
    Convert line and page breaks
     Old:
    \layout Standard
    \line_top \line_bottom \pagebreak_top \pagebreak_bottom \added_space_top xxx \added_space_bottom yyy
    0

     New:
    \begin layout Standard

    \newpage

    \lyxline
    \begin_inset ERT
    \begin layout Standard
    \backslash
    vspace{-1\backslash
    parskip}
    \end_layout
    \end_inset

    \begin_inset VSpace xxx
    \end_inset

    0

    \begin_inset VSpace xxx
    \end_inset
    \lyxline

    \newpage

    \end_layout
    """
    par_params = (
        "added_space_bottom",
        "added_space_top",
        "align",
        "labelwidthstring",
        "line_bottom",
        "line_top",
        "noindent",
        "pagebreak_bottom",
        "pagebreak_top",
        "paragraph_spacing",
        "start_of_appendix",
    )
    font_attributes = [
        "\\family",
        "\\series",
        "\\shape",
        "\\emph",
        "\\numeric",
        "\\bar",
        "\\noun",
        "\\color",
        "\\lang",
    ]
    attribute_values = [
        "default",
        "default",
        "default",
        "default",
        "default",
        "default",
        "default",
        "none",
        document.language,
    ]
    i = 0
    while True:
        i = find_token(document.body, "\\begin_layout", i)
        if i == -1:
            return
        layout = get_layout(document.body[i], document.default_layout)
        i = i + 1

        # Merge all paragraph parameters into a single line
        # We cannot check for '\\' only because paragraphs may start e.g.
        # with '\\backslash'
        while (
            document.body[i + 1][:1] == "\\"
            and document.body[i + 1][1:].split()[0] in par_params
        ):
            document.body[i] = document.body[i + 1] + " " + document.body[i]
            del document.body[i + 1]

        line_top = document.body[i].find("\\line_top")
        line_bot = document.body[i].find("\\line_bottom")
        pb_top = document.body[i].find("\\pagebreak_top")
        pb_bot = document.body[i].find("\\pagebreak_bottom")
        vspace_top = document.body[i].find("\\added_space_top")
        vspace_bot = document.body[i].find("\\added_space_bottom")

        if (
            line_top == -1
            and line_bot == -1
            and pb_bot == -1
            and pb_top == -1
            and vspace_top == -1
            and vspace_bot == -1
        ):
            continue

        # Do we have a nonstandard paragraph? We need to create new paragraphs
        # if yes to avoid putting lyxline etc. inside of special environments.
        # This is wrong for itemize and enumerate environments, but it is
        # impossible to convert these correctly.
        # We want to avoid new paragraphs if possible becauase we want to
        # inherit font sizes.
        nonstandard = 0
        if (
            not document.is_default_layout(layout)
            or document.body[i].find("\\align") != -1
            or document.body[i].find("\\labelwidthstring") != -1
            or document.body[i].find("\\noindent") != -1
        ):
            nonstandard = 1

        # get the font size of the beginning of this paragraph, since we need
        # it for the lyxline inset
        j = i + 1
        while not is_nonempty_line(document.body[j]):
            j = j + 1
        size_top = ""
        if document.body[j].find("\\size") != -1:
            size_top = document.body[j].split()[1]

        for tag in (
            "\\line_top",
            "\\line_bottom",
            "\\pagebreak_top",
            "\\pagebreak_bottom",
        ):
            document.body[i] = document.body[i].replace(tag, "")

        if vspace_top != -1:
            # the position could be change because of the removal of other
            # paragraph properties above
            vspace_top = document.body[i].find("\\added_space_top")
            tmp_list = document.body[i][vspace_top:].split()
            vspace_top_value = tmp_list[1]
            document.body[i] = document.body[i][:vspace_top] + " ".join(tmp_list[2:])

        if vspace_bot != -1:
            # the position could be change because of the removal of other
            # paragraph properties above
            vspace_bot = document.body[i].find("\\added_space_bottom")
            tmp_list = document.body[i][vspace_bot:].split()
            vspace_bot_value = tmp_list[1]
            document.body[i] = document.body[i][:vspace_bot] + " ".join(tmp_list[2:])

        document.body[i] = document.body[i].strip()
        i = i + 1

        # Create an empty paragraph or paragraph fragment for line and
        # page break that belong above the paragraph
        if pb_top != -1 or line_top != -1 or vspace_top != -1:
            paragraph_above = list()
            if nonstandard:
                # We need to create an extra paragraph for nonstandard environments
                paragraph_above = ["\\begin_layout %s" % document.default_layout, ""]

            if pb_top != -1:
                paragraph_above.extend(["\\newpage ", ""])

            if vspace_top != -1:
                paragraph_above.extend(
                    ["\\begin_inset VSpace " + vspace_top_value, "\\end_inset", "", ""]
                )

            if line_top != -1:
                if size_top != "":
                    paragraph_above.extend(["\\size " + size_top + " "])
                # We need an additional vertical space of -\parskip.
                # We can't use the vspace inset because it does not know \parskip.
                paragraph_above.extend(["\\lyxline ", "", ""])
                insert_ert(
                    paragraph_above,
                    len(paragraph_above) - 1,
                    "Collapsed",
                    "\\vspace{-1\\parskip}\n",
                    document.format + 1,
                    document.default_layout,
                )
                paragraph_above.extend([""])

            if nonstandard:
                paragraph_above.extend(["\\end_layout ", ""])
                # insert new paragraph above the current paragraph
                document.body[i - 2 : i - 2] = paragraph_above
            else:
                # insert new lines at the beginning of the current paragraph
                document.body[i:i] = paragraph_above

            i = i + len(paragraph_above)

        # Ensure that nested style are converted later.
        k = find_end_of(document.body, i, "\\begin_layout", "\\end_layout")

        if k == -1:
            return

        if pb_bot != -1 or line_bot != -1 or vspace_bot != -1:
            # get the font size of the end of this paragraph
            size_bot = size_top
            j = i + 1
            while j < k:
                if document.body[j].find("\\size") != -1:
                    size_bot = document.body[j].split()[1]
                    j = j + 1
                elif document.body[j].find("\\begin_inset") != -1:
                    # skip insets
                    j = find_end_of_inset(document.body, j)
                else:
                    j = j + 1

            paragraph_below = list()
            if nonstandard:
                # We need to create an extra paragraph for nonstandard environments
                paragraph_below = [
                    "",
                    "\\begin_layout %s" % document.default_layout,
                    "",
                ]
            else:
                for a in range(len(font_attributes)):
                    if find_token(document.body, font_attributes[a], i, k) != -1:
                        paragraph_below.extend([font_attributes[a] + " " + attribute_values[a]])

            if line_bot != -1:
                if nonstandard and size_bot != "":
                    paragraph_below.extend(["\\size " + size_bot + " "])
                paragraph_below.extend(["\\lyxline ", ""])
                if size_bot != "":
                    paragraph_below.extend(["\\size default "])

            if vspace_bot != -1:
                paragraph_below.extend(
                    ["\\begin_inset VSpace " + vspace_bot_value, "\\end_inset", "", ""]
                )

            if pb_bot != -1:
                paragraph_below.extend(["\\newpage ", ""])

            if nonstandard:
                paragraph_below.extend(["\\end_layout "])
                # insert new paragraph below the current paragraph
                document.body[k + 1 : k + 1] = paragraph_below
            else:
                # insert new lines at the end of the current paragraph
                document.body[k:k] = paragraph_below


def convert_note(document):
    "Convert Notes."
    i = 0
    while True:
        i = find_tokens(
            document.body,
            ["\\begin_inset Note", "\\begin_inset Comment", "\\begin_inset Greyedout"],
            i,
        )
        if i == -1:
            break

        document.body[i] = document.body[i][0:13] + "Note " + document.body[i][13:]
        i = i + 1


def revert_note(document):
    "Revert Notes."
    note_header = "\\begin_inset Note "
    i = 0
    while True:
        i = find_token(document.body, note_header, i)
        if i == -1:
            break

        document.body[i] = "\\begin_inset " + document.body[i][len(note_header) :]
        i = i + 1


def convert_box(document):
    "Convert Boxes."
    i = 0
    while True:
        i = find_tokens(
            document.body,
            [
                "\\begin_inset Boxed",
                "\\begin_inset Doublebox",
                "\\begin_inset Frameless",
                "\\begin_inset ovalbox",
                "\\begin_inset Ovalbox",
                "\\begin_inset Shadowbox",
            ],
            i,
        )
        if i == -1:
            break

        document.body[i] = document.body[i][0:13] + "Box " + document.body[i][13:]
        i = i + 1


def revert_box(document):
    "Revert Boxes."
    box_header = "\\begin_inset Box "
    i = 0
    while True:
        i = find_token(document.body, box_header, i)
        if i == -1:
            break

        document.body[i] = "\\begin_inset " + document.body[i][len(box_header) :]
        i = i + 1


def convert_collapsible(document):
    "Convert collapsed insets."
    i = 0
    while True:
        i = find_tokens_exact(
            document.body,
            [
                "\\begin_inset Box",
                "\\begin_inset Branch",
                "\\begin_inset CharStyle",
                "\\begin_inset Float",
                "\\begin_inset Foot",
                "\\begin_inset Marginal",
                "\\begin_inset Note",
                "\\begin_inset OptArg",
                "\\begin_inset Wrap",
            ],
            i,
        )
        if i == -1:
            break

        # Seach for a line starting 'collapsed'
        # If, however, we find a line starting '\begin_layout'
        # (_always_ present) then break with a warning message
        i = i + 1
        while True:
            if document.body[i] == "collapsed false":
                document.body[i] = "status open"
                break
            elif document.body[i] == "collapsed true":
                document.body[i] = "status collapsed"
                break
            elif document.body[i][:13] == "\\begin_layout":
                document.warning("Malformed LyX document: Missing 'collapsed'.")
                break
            i = i + 1

        i = i + 1


def revert_collapsible(document):
    "Revert collapsed insets."
    i = 0
    while True:
        i = find_tokens_exact(
            document.body,
            [
                "\\begin_inset Box",
                "\\begin_inset Branch",
                "\\begin_inset CharStyle",
                "\\begin_inset Float",
                "\\begin_inset Foot",
                "\\begin_inset Marginal",
                "\\begin_inset Note",
                "\\begin_inset OptArg",
                "\\begin_inset Wrap",
            ],
            i,
        )
        if i == -1:
            break

        # Seach for a line starting 'status'
        # If, however, we find a line starting '\begin_layout'
        # (_always_ present) then break with a warning message
        i = i + 1
        while True:
            if document.body[i] == "status open":
                document.body[i] = "collapsed false"
                break
            elif document.body[i] == "status collapsed" or document.body[i] == "status inlined":
                document.body[i] = "collapsed true"
                break
            elif document.body[i][:13] == "\\begin_layout":
                document.warning("Malformed LyX document: Missing 'status'.")
                break
            i = i + 1

        i = i + 1


def convert_ert(document):
    "Convert ERT."
    i = 0
    while True:
        i = find_token(document.body, "\\begin_inset ERT", i)
        if i == -1:
            break

        # Seach for a line starting 'status'
        # If, however, we find a line starting '\begin_layout'
        # (_always_ present) then break with a warning message
        i = i + 1
        while True:
            if document.body[i] == "status Open":
                document.body[i] = "status open"
                break
            elif document.body[i] == "status Collapsed":
                document.body[i] = "status collapsed"
                break
            elif document.body[i] == "status Inlined":
                document.body[i] = "status inlined"
                break
            elif document.body[i][:13] == "\\begin_layout":
                document.warning("Malformed LyX document: Missing 'status'.")
                break
            i = i + 1

        i = i + 1


def revert_ert(document):
    "Revert ERT."
    i = 0
    while True:
        i = find_token(document.body, "\\begin_inset ERT", i)
        if i == -1:
            break

        # Seach for a line starting 'status'
        # If, however, we find a line starting '\begin_layout'
        # (_always_ present) then break with a warning message
        i = i + 1
        while True:
            if document.body[i] == "status open":
                document.body[i] = "status Open"
                break
            elif document.body[i] == "status collapsed":
                document.body[i] = "status Collapsed"
                break
            elif document.body[i] == "status inlined":
                document.body[i] = "status Inlined"
                break
            elif document.body[i][:13] == "\\begin_layout":
                document.warning("Malformed LyX document : Missing 'status'.")
                break
            i = i + 1

        i = i + 1


def convert_minipage(document):
    """Convert minipages to the box inset.
    We try to use the same order of arguments as lyx does.
    """
    pos = ["t", "c", "b"]
    inner_pos = ["c", "t", "b", "s"]

    i = 0
    while True:
        i = find_token(document.body, "\\begin_inset Minipage", i)
        if i == -1:
            return

        document.body[i] = "\\begin_inset Box Frameless"
        i = i + 1

        # convert old to new position using the pos list
        if document.body[i][:8] == "position":
            document.body[i] = 'position "%s"' % pos[int(document.body[i][9])]
        else:
            document.body.insert(i, 'position "%s"' % pos[0])
        i = i + 1

        document.body.insert(i, 'hor_pos "c"')
        i = i + 1
        document.body.insert(i, "has_inner_box 1")
        i = i + 1

        # convert the inner_position
        if document.body[i][:14] == "inner_position":
            innerpos = inner_pos[int(document.body[i][15])]
            del document.body[i]
        else:
            innerpos = inner_pos[0]

        # We need this since the new file format has a height and width
        # in a different order.
        if document.body[i][:6] == "height":
            height = document.body[i][6:]
            # test for default value of 221 and convert it accordingly
            if height == ' "0pt"' or height == ' "0"':
                height = ' "1pt"'
            del document.body[i]
        else:
            height = ' "1pt"'

        if document.body[i][:5] == "width":
            width = document.body[i][5:]
            del document.body[i]
        else:
            width = ' "0"'

        if document.body[i][:9] == "collapsed":
            if document.body[i][9:] == "true":
                status = "collapsed"
            else:
                status = "open"
            del document.body[i]
        else:
            status = "collapsed"

        # Handle special default case:
        if height == ' "1pt"' and innerpos == "c":
            innerpos = "t"

        document.body.insert(i, 'inner_pos "' + innerpos + '"')
        i = i + 1
        document.body.insert(i, "use_parbox 0")
        i = i + 1
        document.body.insert(i, "width" + width)
        i = i + 1
        document.body.insert(i, 'special "none"')
        i = i + 1
        document.body.insert(i, "height" + height)
        i = i + 1
        document.body.insert(i, 'height_special "totalheight"')
        i = i + 1
        document.body.insert(i, "status " + status)
        i = i + 1


def convert_ertbackslash(body, i, ert, format, default_layout):
    r"""-------------------------------------------------------------------------------------------
    Convert backslashes and '\n' into valid ERT code, append the converted
    text to body[i] and return the (maybe incremented) line index i"""

    for c in ert:
        if c == "\\":
            body[i] = body[i] + "\\backslash "
            i = i + 1
            body.insert(i, "")
        elif c == "\n":
            if format <= 240:
                body[i + 1 : i + 1] = ["\\newline ", ""]
                i = i + 2
            else:
                body[i + 1 : i + 1] = [
                    "\\end_layout",
                    "",
                    "\\begin_layout %s" % default_layout,
                    "",
                ]
                i = i + 4
        else:
            body[i] = body[i] + c
    return i


def ert2latex(lines, format):
    r"""Converts lines in ERT code to LaTeX
    The surrounding \begin_layout ... \end_layout pair must not be included"""

    backslash = re.compile(r"\\backslash\s*$")
    newline = re.compile(r"\\newline\s*$")
    if format <= 224:
        begin_layout = re.compile(r"\\layout\s*\S+$")
    else:
        begin_layout = re.compile(r"\\begin_layout\s*\S+$")
    end_layout = re.compile(r"\\end_layout\s*$")
    ert = ""
    for i in range(len(lines)):
        line = backslash.sub("\\\\", lines[i])
        if format <= 240:
            if begin_layout.match(line):
                line = "\n\n"
            else:
                line = newline.sub("\n", line)
        else:
            if begin_layout.match(line):
                line = "\n"
        if format > 224 and end_layout.match(line):
            line = ""
        ert = ert + line
    return ert


def get_par_params(lines, i):
    """get all paragraph parameters. They can be all on one line or on several lines.
    lines[i] must be the first parameter line"""
    par_params = (
        "added_space_bottom",
        "added_space_top",
        "align",
        "labelwidthstring",
        "line_bottom",
        "line_top",
        "noindent",
        "pagebreak_bottom",
        "pagebreak_top",
        "paragraph_spacing",
        "start_of_appendix",
    )
    # We cannot check for '\\' only because paragraphs may start e.g.
    # with '\\backslash'
    params = ""
    while lines[i][:1] == "\\" and lines[i][1:].split()[0] in par_params:
        params = params + " " + lines[i].strip()
        i = i + 1
    return params.strip()


def lyxsize2latexsize(lyxsize):
    "Convert LyX font size to LaTeX fontsize."
    sizes = {
        "tiny": "tiny",
        "scriptsize": "scriptsize",
        "footnotesize": "footnotesize",
        "small": "small",
        "normal": "normalsize",
        "large": "large",
        "larger": "Large",
        "largest": "LARGE",
        "huge": "huge",
        "giant": "Huge",
    }
    if lyxsize in sizes:
        return "\\" + sizes[lyxsize]
    return ""


def revert_breaks(document):
    """Change vspace insets, page breaks and lyxlines to paragraph options
    (if possible) or ERT"""

    # Get default spaceamount
    i = find_token(document.header, "\\defskip", 0)
    if i == -1:
        defskipamount = "medskip"
    else:
        defskipamount = document.header[i].split()[1]

    keys = {"\\begin_inset": "vspace", "\\lyxline": "lyxline", "\\newpage": "newpage"}
    keywords_top = {
        "vspace": "\\added_space_top",
        "lyxline": "\\line_top",
        "newpage": "\\pagebreak_top",
    }
    keywords_bot = {
        "vspace": "\\added_space_bottom",
        "lyxline": "\\line_bottom",
        "newpage": "\\pagebreak_bottom",
    }
    tokens = ["\\begin_inset VSpace", "\\lyxline", "\\newpage"]

    # Convert the insets
    i = 0
    while True:
        i = find_tokens(document.body, tokens, i)
        if i == -1:
            return

        # Are we at the beginning of a paragraph?
        paragraph_start = 1
        this_par = get_paragraph(document.body, i, document.format - 1)
        start = this_par + 1
        params = get_par_params(document.body, start)
        size = "normal"
        # Paragraph parameters may be on one or more lines.
        # Find the start of the real paragraph text.
        while document.body[start][:1] == "\\" and document.body[start].split()[0] in params:
            start = start + 1
        for k in range(start, i):
            if document.body[k].find("\\size") != -1:
                # store font size
                size = document.body[k].split()[1]
            elif is_nonempty_line(document.body[k]):
                paragraph_start = 0
                break
        # Find the end of the real paragraph text.
        next_par = get_next_paragraph(document.body, i, document.format - 1)
        if next_par == -1:
            document.warning("Malformed LyX document: Missing next paragraph.")
            i = i + 1
            continue

        # first line of our insets
        inset_start = i
        # last line of our insets
        inset_end = inset_start
        # Are we at the end of a paragraph?
        paragraph_end = 1
        # start and end line numbers to delete if we convert this inset
        del_lines = list()
        # is this inset a lyxline above a paragraph?
        top = list()
        # raw inset information
        lines = list()
        # name of this inset
        insets = list()
        # font size of this inset
        sizes = list()

        # Detect subsequent lyxline, vspace and pagebreak insets created by convert_breaks()
        n = 0
        k = inset_start
        while k < next_par:
            if find_tokens(document.body, tokens, k) == k:
                # inset to convert
                lines.append(document.body[k].split())
                insets.append(keys[lines[n][0]])
                del_lines.append([k, k])
                top.append(0)
                sizes.append(size)
                n = n + 1
                inset_end = k
            elif document.body[k].find("\\size") != -1:
                # store font size
                size = document.body[k].split()[1]
            elif find_token(document.body, "\\begin_inset ERT", k) == k:
                ert_begin = find_token(document.body, "\\layout", k) + 1
                if ert_begin == 0:
                    document.warning("Malformed LyX document: Missing '\\layout'.")
                    continue
                ert_end = find_end_of_inset(document.body, k)
                if ert_end == -1:
                    document.warning("Malformed LyX document: Missing '\\end_inset'.")
                    continue
                ert = ert2latex(document.body[ert_begin:ert_end], document.format - 1)
                if n > 0 and insets[n - 1] == "lyxline" and ert == "\\vspace{-1\\parskip}\n":
                    # vspace ERT created by convert_breaks() for top lyxline
                    top[n - 1] = 1
                    del_lines[n - 1][1] = ert_end
                    inset_end = ert_end
                    k = ert_end
                else:
                    paragraph_end = 0
                    break
            elif (
                n > 0
                and insets[n - 1] == "vspace"
                and find_token(document.body, "\\end_inset", k) == k
            ):
                # ignore end of vspace inset
                del_lines[n - 1][1] = k
                inset_end = k
            elif is_nonempty_line(document.body[k]):
                paragraph_end = 0
                break
            k = k + 1

        # Determine space amount for vspace insets
        spaceamount = list()
        arguments = list()
        for k in range(n):
            if insets[k] == "vspace":
                spaceamount.append(lines[k][2])
                arguments.append(" " + spaceamount[k] + " ")
            else:
                spaceamount.append("")
                arguments.append(" ")

        # Can we convert to top paragraph parameters?
        before = 0
        if (
            (
                n == 3
                and insets[0] == "newpage"
                and insets[1] == "vspace"
                and insets[2] == "lyxline"
                and top[2]
            )
            or (
                n == 2
                and (
                    (insets[0] == "newpage" and insets[1] == "vspace")
                    or (insets[0] == "newpage" and insets[1] == "lyxline" and top[1])
                    or (insets[0] == "vspace" and insets[1] == "lyxline" and top[1])
                )
            )
            or (n == 1 and insets[0] == "lyxline" and top[0])
        ):
            # These insets have been created before a paragraph by
            # convert_breaks()
            before = 1

        # Can we convert to bottom paragraph parameters?
        after = 0
        if (
            (
                n == 3
                and insets[0] == "lyxline"
                and not top[0]
                and insets[1] == "vspace"
                and insets[2] == "newpage"
            )
            or (
                n == 2
                and (
                    (insets[0] == "lyxline" and not top[0] and insets[1] == "vspace")
                    or (insets[0] == "lyxline" and not top[0] and insets[1] == "newpage")
                    or (insets[0] == "vspace" and insets[1] == "newpage")
                )
            )
            or (n == 1 and insets[0] == "lyxline" and not top[0])
        ):
            # These insets have been created after a paragraph by
            # convert_breaks()
            after = 1

        if paragraph_start and paragraph_end:
            # We are in a paragraph of our own.
            # We must not delete this paragraph if it has parameters
            if params == "":
                # First try to merge with the previous paragraph.
                # We try the previous paragraph first because we would
                # otherwise need ERT for two subsequent vspaces.
                prev_par = get_paragraph(document.body, this_par - 1, document.format - 1) + 1
                if prev_par > 0 and not before:
                    prev_params = get_par_params(document.body, prev_par + 1)
                    ert = 0
                    # determine font size
                    prev_size = "normal"
                    k = prev_par + 1
                    while (
                        document.body[k][:1] == "\\"
                        and document.body[k].split()[0] in prev_params
                    ):
                        k = k + 1
                    while k < this_par:
                        if document.body[k].find("\\size") != -1:
                            prev_size = document.body[k].split()[1]
                            break
                        elif document.body[k].find("\\begin_inset") != -1:
                            # skip insets
                            k = find_end_of_inset(document.body, k)
                        elif is_nonempty_line(document.body[k]):
                            break
                        k = k + 1
                    for k in range(n):
                        if keywords_bot[insets[k]] in prev_params or (
                            insets[k] == "lyxline" and sizes[k] != prev_size
                        ):
                            ert = 1
                            break
                    if not ert:
                        for k in range(n):
                            document.body.insert(
                                prev_par + 1, keywords_bot[insets[k]] + arguments[k]
                            )
                        del document.body[this_par + n : next_par - 1 + n]
                        i = this_par + n
                        continue
                # Then try next paragraph
                if next_par > 0 and not after:
                    next_params = get_par_params(document.body, next_par + 1)
                    ert = 0
                    while (
                        document.body[k][:1] == "\\"
                        and document.body[k].split()[0] in next_params
                    ):
                        k = k + 1
                    # determine font size
                    next_size = "normal"
                    k = next_par + 1
                    while k < this_par:
                        if document.body[k].find("\\size") != -1:
                            next_size = document.body[k].split()[1]
                            break
                        elif is_nonempty_line(document.body[k]):
                            break
                        k = k + 1
                    for k in range(n):
                        if keywords_top[insets[k]] in next_params or (
                            insets[k] == "lyxline" and sizes[k] != next_size
                        ):
                            ert = 1
                            break
                    if not ert:
                        for k in range(n):
                            document.body.insert(
                                next_par + 1, keywords_top[insets[k]] + arguments[k]
                            )
                        del document.body[this_par : next_par - 1]
                        i = this_par
                        continue
        elif paragraph_start or paragraph_end:
            # Convert to paragraph formatting if we are at the beginning or end
            # of a paragraph and the resulting paragraph would not be empty
            # The order is important: del and insert invalidate some indices
            if paragraph_start:
                keywords = keywords_top
            else:
                keywords = keywords_bot
            ert = 0
            for k in range(n):
                if keywords[insets[k]] in params:
                    ert = 1
                    break
            if not ert:
                for k in range(n):
                    document.body.insert(this_par + 1, keywords[insets[k]] + arguments[k])
                    for j in range(k, n):
                        del_lines[j][0] = del_lines[j][0] + 1
                        del_lines[j][1] = del_lines[j][1] + 1
                    del document.body[del_lines[k][0] : del_lines[k][1] + 1]
                    deleted = del_lines[k][1] - del_lines[k][0] + 1
                    for j in range(k + 1, n):
                        del_lines[j][0] = del_lines[j][0] - deleted
                        del_lines[j][1] = del_lines[j][1] - deleted
                i = this_par
                continue

        # Convert the first inset to ERT.
        # The others are converted in the next loop runs (if they exist)
        if insets[0] == "vspace":
            document.body[i : i + 1] = [
                "\\begin_inset ERT",
                "status Collapsed",
                "",
                "\\layout %s" % document.default_layout,
                "",
                "\\backslash ",
            ]
            i = i + 6
            if spaceamount[0][-1] == "*":
                spaceamount[0] = spaceamount[0][:-1]
                keep = 1
            else:
                keep = 0

            # Replace defskip by the actual value
            if spaceamount[0] == "defskip":
                spaceamount[0] = defskipamount

            # LaTeX does not know \\smallskip* etc
            if keep:
                if spaceamount[0] == "smallskip":
                    spaceamount[0] = "\\smallskipamount"
                elif spaceamount[0] == "medskip":
                    spaceamount[0] = "\\medskipamount"
                elif spaceamount[0] == "bigskip":
                    spaceamount[0] = "\\bigskipamount"
                elif spaceamount[0] == "vfill":
                    spaceamount[0] = "\\fill"

            # Finally output the LaTeX code
            if (
                spaceamount[0] == "smallskip"
                or spaceamount[0] == "medskip"
                or spaceamount[0] == "bigskip"
                or spaceamount[0] == "vfill"
            ):
                document.body.insert(i, spaceamount[0] + "{}")
            else:
                if keep:
                    document.body.insert(i, "vspace*{")
                else:
                    document.body.insert(i, "vspace{")
                i = convert_ertbackslash(
                    document.body,
                    i,
                    spaceamount[0],
                    document.format - 1,
                    document.default_layout,
                )
                document.body[i] = document.body[i] + "}"
            i = i + 1
        elif insets[0] == "lyxline":
            document.body[i] = ""
            latexsize = lyxsize2latexsize(size)
            if latexsize == "":
                document.warning(
                    "Could not convert LyX fontsize '%s' to LaTeX font size." % size
                )
                latexsize = "\\normalsize"
            i = insert_ert(
                document.body,
                i,
                "Collapsed",
                "\\lyxline{%s}" % latexsize,
                document.format - 1,
                document.default_layout,
            )
            # We use \providecommand so that we don't get an error if native
            # lyxlines are used (LyX writes first its own preamble and then
            # the user specified one)
            add_to_preamble(
                document,
                [
                    "% Commands inserted by lyx2lyx for lyxlines",
                    "\\providecommand{\\lyxline}[1]{",
                    "  {#1 \\vspace{1ex} \\hrule width \\columnwidth \\vspace{1ex}}" "}",
                ],
            )
        elif insets[0] == "newpage":
            document.body[i] = ""
            i = insert_ert(
                document.body,
                i,
                "Collapsed",
                "\\newpage{}",
                document.format - 1,
                document.default_layout,
            )


# Convert a LyX length into a LaTeX length
def convert_len(len, special):
    units = {
        "text%": "\\textwidth",
        "col%": "\\columnwidth",
        "page%": "\\pagewidth",
        "line%": "\\linewidth",
        "theight%": "\\textheight",
        "pheight%": "\\pageheight",
    }

    # Convert special lengths
    if special != "none":
        len = "%f\\" % len2value(len) + special

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

    return len


def convert_ertlen(body, i, len, special, format, default_layout):
    """Convert a LyX length into valid ERT code and append it to body[i]
    Return the (maybe incremented) line index i
    Convert backslashes and insert the converted length into body."""
    return convert_ertbackslash(body, i, convert_len(len, special), format, default_layout)


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


def insert_ert(body, i, status, text, format, default_layout):
    """Convert text to ERT and insert it at body[i]
    Return the index of the line after the inserted ERT"""

    body[i:i] = ["\\begin_inset ERT", "status " + status, ""]
    i = i + 3
    if format <= 224:
        body[i:i] = ["\\layout %s" % default_layout, ""]
    else:
        body[i:i] = ["\\begin_layout %s" % default_layout, ""]
    i = i + 1  # i points now to the just created empty line
    i = convert_ertbackslash(body, i, text, format, default_layout) + 1
    if format > 224:
        body[i:i] = ["\\end_layout"]
        i = i + 1
    body[i:i] = ["", "\\end_inset", ""]
    i = i + 3
    return i


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 convert_frameless_box(document):
    "Convert frameless box."
    pos = ["t", "c", "b"]
    inner_pos = ["c", "t", "b", "s"]
    i = 0
    while True:
        i = find_token(document.body, "\\begin_inset Frameless", 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
        del document.body[i]
        j = j - 1

        # Gather parameters
        params = {
            "position": 0,
            "hor_pos": "c",
            "has_inner_box": "1",
            "inner_pos": 1,
            "use_parbox": "0",
            "width": "100col%",
            "special": "none",
            "height": "1in",
            "height_special": "totalheight",
            "collapsed": "false",
        }
        for key in list(params.keys()):
            value = get_value(document.body, key, i, j).replace('"', "")
            if value != "":
                if key == "position":
                    # convert new to old position: 'position "t"' -> 0
                    value = find_token(pos, value, 0)
                    if value != -1:
                        params[key] = value
                elif key == "inner_pos":
                    # convert inner position
                    value = find_token(inner_pos, value, 0)
                    if value != -1:
                        params[key] = value
                else:
                    params[key] = value
                j = del_token(document.body, key, i, j)
        i = i + 1

        # Convert to minipage or ERT?
        # Note that the inner_position and height parameters of a minipage
        # inset are ignored and not accessible for the user, although they
        # are present in the file format and correctly read in and written.
        # Therefore we convert to ERT if they do not have their LaTeX
        # defaults. These are:
        # - the value of "position" for "inner_pos"
        # - "\totalheight"          for "height"
        if (
            params["use_parbox"] != "0"
            or params["has_inner_box"] != "1"
            or params["special"] != "none"
            or params["height_special"] != "totalheight"
            or len2value(params["height"]) != 1.0
        ):
            # Here we know that this box is not supported in file format 224.
            # Therefore we need to convert it to ERT. We can't simply convert
            # the beginning and end of the box to ERT, because the
            # box inset may contain layouts that are different from the
            # surrounding layout. After the conversion the contents of the
            # box inset is on the same level as the surrounding text, and
            # paragraph layouts and align parameters can get mixed up.

            # A possible solution for this problem:
            # Convert the box to a minipage and redefine the minipage
            # environment in ERT so that the original box is simulated.
            # For minipages we could do this in a way that the width and
            # position can still be set from LyX, but this did not work well.
            # This is not possible for parboxes either, so we convert the
            # original box to ERT, put the minipage inset inside the box
            # and redefine the minipage environment to be empty.

            # Commands that are independant of a particular box can go to
            # the preamble.
            # We need to define lyxtolyxrealminipage with 3 optional
            # arguments although LyX 1.3 uses only the first one.
            # Otherwise we will get LaTeX errors if this document is
            # converted to format 225 or above again (LyX 1.4 uses all
            # optional arguments).
            add_to_preamble(
                document,
                [
                    "% Commands inserted by lyx2lyx for frameless boxes",
                    "% Save the original minipage environment",
                    "\\let\\lyxtolyxrealminipage\\minipage",
                    "\\let\\endlyxtolyxrealminipage\\endminipage",
                    "% Define an empty lyxtolyximinipage environment",
                    "% with 3 optional arguments",
                    "\\newenvironment{lyxtolyxiiiminipage}[4]{}{}",
                    "\\newenvironment{lyxtolyxiiminipage}[2][\\lyxtolyxargi]%",
                    "  {\\begin{lyxtolyxiiiminipage}{\\lyxtolyxargi}{\\lyxtolyxargii}{#1}{#2}}%",
                    "  {\\end{lyxtolyxiiiminipage}}",
                    "\\newenvironment{lyxtolyximinipage}[1][\\totalheight]%",
                    "  {\\def\\lyxtolyxargii{{#1}}\\begin{lyxtolyxiiminipage}}%",
                    "  {\\end{lyxtolyxiiminipage}}",
                    "\\newenvironment{lyxtolyxminipage}[1][c]%",
                    "  {\\def\\lyxtolyxargi{{#1}}\\begin{lyxtolyximinipage}}",
                    "  {\\end{lyxtolyximinipage}}",
                ],
            )

            if params["use_parbox"] != "0":
                ert = "\\parbox"
            else:
                ert = "\\begin{lyxtolyxrealminipage}"

            # convert optional arguments only if not latex default
            if (
                pos[params["position"]] != "c"
                or inner_pos[params["inner_pos"]] != pos[params["position"]]
                or params["height_special"] != "totalheight"
                or len2value(params["height"]) != 1.0
            ):
                ert = ert + "[" + pos[params["position"]] + "]"
            if (
                inner_pos[params["inner_pos"]] != pos[params["position"]]
                or params["height_special"] != "totalheight"
                or len2value(params["height"]) != 1.0
            ):
                ert = ert + "[" + convert_len(params["height"], params["height_special"]) + "]"
            if inner_pos[params["inner_pos"]] != pos[params["position"]]:
                ert = ert + "[" + inner_pos[params["inner_pos"]] + "]"

            ert = ert + "{" + convert_len(params["width"], params["special"]) + "}"

            if params["use_parbox"] != "0":
                ert = ert + "{"
            ert = ert + "\\let\\minipage\\lyxtolyxminipage%\n"
            ert = ert + "\\let\\endminipage\\endlyxtolyxminipage%\n"

            old_i = i
            i = insert_ert(
                document.body,
                i,
                "Collapsed",
                ert,
                document.format - 1,
                document.default_layout,
            )
            j = j + i - old_i - 1

            document.body[i:i] = [
                "\\begin_inset Minipage",
                "position %d" % params["position"],
                "inner_position 1",
                'height "1in"',
                'width "' + params["width"] + '"',
                "collapsed " + params["collapsed"],
            ]
            i = i + 6
            j = j + 6

            # Restore the original minipage environment since we may have
            # minipages inside this box.
            # Start a new paragraph because the following may be nonstandard
            document.body[i:i] = ["\\layout %s" % document.default_layout, "", ""]
            i = i + 2
            j = j + 3
            ert = "\\let\\minipage\\lyxtolyxrealminipage%\n"
            ert = ert + "\\let\\endminipage\\lyxtolyxrealendminipage%"
            old_i = i
            i = insert_ert(
                document.body,
                i,
                "Collapsed",
                ert,
                document.format - 1,
                document.default_layout,
            )
            j = j + i - old_i - 1

            # Redefine the minipage end before the inset end.
            # Start a new paragraph because the previous may be nonstandard
            document.body[j:j] = ["\\layout %s" % document.default_layout, "", ""]
            j = j + 2
            ert = "\\let\\endminipage\\endlyxtolyxminipage"
            j = insert_ert(
                document.body,
                j,
                "Collapsed",
                ert,
                document.format - 1,
                document.default_layout,
            )
            j = j + 1
            document.body.insert(j, "")
            j = j + 1

            # LyX writes '%\n' after each box. Therefore we need to end our
            # ERT with '%\n', too, since this may swallow a following space.
            if params["use_parbox"] != "0":
                ert = "}%\n"
            else:
                ert = "\\end{lyxtolyxrealminipage}%\n"
            j = insert_ert(
                document.body,
                j,
                "Collapsed",
                ert,
                document.format - 1,
                document.default_layout,
            )

            # We don't need to restore the original minipage after the inset
            # end because the scope of the redefinition is the original box.

        else:
            # Convert to minipage
            document.body[i:i] = [
                "\\begin_inset Minipage",
                "position %d" % params["position"],
                "inner_position %d" % params["inner_pos"],
                'height "' + params["height"] + '"',
                'width "' + params["width"] + '"',
                "collapsed " + params["collapsed"],
            ]
            i = i + 6


def remove_branches(document):
    "Remove branches."
    i = 0
    while True:
        i = find_token(document.header, "\\branch", i)
        if i == -1:
            break
        document.warning("Removing branch %s." % document.header[i].split()[1])
        j = find_token(document.header, "\\end_branch", i)
        if j == -1:
            document.warning("Malformed LyX document: Missing '\\end_branch'.")
            break
        del document.header[i : j + 1]

    i = 0
    while True:
        i = find_token(document.body, "\\begin_inset Branch", 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
        del document.body[i]
        del document.body[j - 1]
        # Seach for a line starting 'collapsed'
        # If, however, we find a line starting '\layout'
        # (_always_ present) then break with a warning message
        collapsed_found = 0
        while True:
            if document.body[i][:9] == "collapsed":
                del document.body[i]
                collapsed_found = 1
                continue
            elif document.body[i][:7] == "\\layout":
                if collapsed_found == 0:
                    document.warning("Malformed LyX document: Missing 'collapsed'.")
                # Delete this new paragraph, since it would not appear in
                # .tex output. This avoids also empty paragraphs.
                del document.body[i]
                break
            i = i + 1


def convert_jurabib(document):
    "Convert jurabib."
    i = find_token(document.header, "\\use_numerical_citations", 0)
    if i == -1:
        document.warning("Malformed lyx document: Missing '\\use_numerical_citations'.")
        return
    document.header.insert(i + 1, "\\use_jurabib 0")


def revert_jurabib(document):
    "Revert jurabib."
    i = find_token(document.header, "\\use_jurabib", 0)
    if i == -1:
        document.warning("Malformed lyx document: Missing '\\use_jurabib'.")
        return
    if get_value(document.header, "\\use_jurabib", 0) != "0":
        document.warning("Conversion of '\\use_jurabib = 1' not yet implemented.")
        # Don't remove '\\use_jurabib' so that people will get warnings by lyx
        return
    del document.header[i]


def convert_bibtopic(document):
    "Convert bibtopic."
    i = find_token(document.header, "\\use_jurabib", 0)
    if i == -1:
        document.warning("Malformed lyx document: Missing '\\use_jurabib'.")
        return
    document.header.insert(i + 1, "\\use_bibtopic 0")


def revert_bibtopic(document):
    "Revert bibtopic."
    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) != "0":
        document.warning("Conversion of '\\use_bibtopic = 1' not yet implemented.")
        # Don't remove '\\use_jurabib' so that people will get warnings by lyx
    del document.header[i]


def convert_float(document):
    "Convert sideway floats."
    i = 0
    while True:
        i = find_token_exact(document.body, "\\begin_inset Float", i)
        if i == -1:
            return
        # Seach for a line starting 'wide'
        # If, however, we find a line starting '\begin_layout'
        # (_always_ present) then break with a warning message
        i = i + 1
        while True:
            if document.body[i][:4] == "wide":
                document.body.insert(i + 1, "sideways false")
                break
            elif document.body[i][:13] == "\\begin_layout":
                document.warning("Malformed lyx document: Missing 'wide'.")
                break
            i = i + 1
        i = i + 1


def revert_float(document):
    "Revert sideways floats."
    i = 0
    while True:
        i = find_token_exact(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":
            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] = (
                "\\layout Standard\n\\begin_inset ERT\nstatus Collapsed\n\n"
                "\\layout Standard\n\n\n\\backslash\n"
                "end{sideways" + floattype + "}\n\n\\end_inset\n"
            )
            del document.body[i + 1 : l - 1]
            document.body[i] = (
                "\\begin_inset ERT\nstatus Collapsed\n\n"
                "\\layout Standard\n\n\n\\backslash\n"
                "begin{sideways" + floattype + "}\n\n\\end_inset\n\n"
            )
            add_to_preamble(document, ["\\usepackage{rotfloat}\n"])
            i = i + 1
            continue
        del_token(document.body, "sideways", i, j)
        i = i + 1


def convert_graphics(document):
    """Add extension to documentnames of insetgraphics if necessary."""
    i = 0
    while True:
        i = find_token(document.body, "\\begin_inset Graphics", i)
        if i == -1:
            return

        j = find_token_exact(document.body, "documentname", i)
        if j == -1:
            return
        i = i + 1
        filename = document.body[j].split()[1]
        if document.dir == "" and not os.path.isabs(filename):
            # We don't know the directory and cannot check the document.
            # We could use a heuristic and take the current directory,
            # and we could try to find out if documentname has an extension,
            # but that would be just guesses and could be wrong.
            document.warning(
                """Warning: Cannot determine whether document
         %s
         needs an extension when reading from standard input.
         You may need to correct the document manually or run
         lyx2lyx again with the .lyx document as commandline argument."""
                % filename
            )
            continue
        absname = os.path.normpath(os.path.join(document.dir, filename))
        # This needs to be the same algorithm as in pre 233 insetgraphics
        if access(absname, F_OK):
            continue
        if access(absname + ".ps", F_OK):
            document.body[j] = document.body[j].replace(filename, filename + ".ps")
            continue
        if access(absname + ".eps", F_OK):
            document.body[j] = document.body[j].replace(filename, filename + ".eps")


def convert_names(document):
    """Convert in the docbook backend from firstname and surname style
    to charstyles.
    """
    if document.backend != "docbook":
        return

    i = 0

    while True:
        i = find_token(document.body, "\\begin_layout Author", i)
        if i == -1:
            return

        i = i + 1
        while document.body[i] == "":
            i = i + 1

        if (
            document.body[i][:11] != "\\end_layout"
            or document.body[i + 2][:13] != "\\begin_deeper"
        ):
            i = i + 1
            continue

        k = i
        i = find_end_of(document.body, i + 3, "\\begin_deeper", "\\end_deeper")
        if i == -1:
            # something is really wrong, abort
            document.warning("Missing \\end_deeper, after style Author.")
            document.warning("Aborted attempt to parse FirstName and Surname.")
            return
        firstname, surname = "", ""

        name = document.body[k:i]

        j = find_token(name, "\\begin_layout FirstName", 0)
        if j != -1:
            j = j + 1
            while name[j] != "\\end_layout":
                firstname = firstname + name[j]
                j = j + 1

        j = find_token(name, "\\begin_layout Surname", 0)
        if j != -1:
            j = j + 1
            while name[j] != "\\end_layout":
                surname = surname + name[j]
                j = j + 1

        # delete name
        del document.body[k + 2 : i + 1]

        document.body[k - 1 : k - 1] = [
            "",
            "",
            "\\begin_inset CharStyle Firstname",
            "status inlined",
            "",
            "\\begin_layout %s" % document.default_layout,
            "",
            "%s" % firstname,
            r"\end_layout",
            "",
            r"\end_inset",
            "",
            "",
            "\\begin_inset CharStyle Surname",
            "status inlined",
            "",
            "\\begin_layout %s" % document.default_layout,
            "",
            "%s" % surname,
            "\\end_layout",
            "",
            "\\end_inset",
            "",
        ]


def revert_names(document):
    """Revert in the docbook backend from firstname and surname char style
    to styles.
    """
    if document.backend != "docbook":
        return


def convert_cite_engine(document):
    r"""\use_natbib 1                       \cite_engine <style>
    \use_numerical_citations 0     ->   where <style> is one of
    \use_jurabib 0                      "basic", "natbib_authoryear","""

    a = find_token(document.header, "\\use_natbib", 0)
    if a == -1:
        document.warning("Malformed lyx document: Missing '\\use_natbib'.")
        return

    b = find_token(document.header, "\\use_numerical_citations", 0)
    if b == -1 or b != a + 1:
        document.warning("Malformed lyx document: Missing '\\use_numerical_citations'.")
        return

    c = find_token(document.header, "\\use_jurabib", 0)
    if c == -1 or c != b + 1:
        document.warning("Malformed lyx document: Missing '\\use_jurabib'.")
        return

    use_natbib = int(document.header[a].split()[1])
    use_numerical_citations = int(document.header[b].split()[1])
    use_jurabib = int(document.header[c].split()[1])

    cite_engine = "basic"
    if use_natbib:
        if use_numerical_citations:
            cite_engine = "natbib_numerical"
        else:
            cite_engine = "natbib_authoryear"
    elif use_jurabib:
        cite_engine = "jurabib"

    del document.header[a : c + 1]
    document.header.insert(a, "\\cite_engine " + cite_engine)


def revert_cite_engine(document):
    "Revert the cite engine."
    i = find_token(document.header, "\\cite_engine", 0)
    if i == -1:
        document.warning("Malformed lyx document: Missing '\\cite_engine'.")
        return

    cite_engine = document.header[i].split()[1]

    use_natbib = "0"
    use_numerical = "0"
    use_jurabib = "0"
    if cite_engine == "natbib_numerical":
        use_natbib = "1"
        use_numerical = "1"
    elif cite_engine == "natbib_authoryear":
        use_natbib = "1"
    elif cite_engine == "jurabib":
        use_jurabib = "1"

    del document.header[i]
    document.header.insert(i, "\\use_jurabib " + use_jurabib)
    document.header.insert(i, "\\use_numerical_citations " + use_numerical)
    document.header.insert(i, "\\use_natbib " + use_natbib)


def convert_paperpackage(document):
    "Convert paper package."
    i = find_token(document.header, "\\paperpackage", 0)
    if i == -1:
        return

    packages = {
        "default": "none",
        "a4": "none",
        "a4wide": "a4",
        "widemarginsa4": "a4wide",
    }
    if len(document.header[i].split()) > 1:
        paperpackage = document.header[i].split()[1]
        document.header[i] = document.header[i].replace(paperpackage, packages[paperpackage])
    else:
        document.header[i] = document.header[i] + " widemarginsa4"


def revert_paperpackage(document):
    "Revert paper package."
    i = find_token(document.header, "\\paperpackage", 0)
    if i == -1:
        return

    packages = {
        "none": "a4",
        "a4": "a4wide",
        "a4wide": "widemarginsa4",
        "widemarginsa4": "",
        "default": "default",
    }
    if len(document.header[i].split()) > 1:
        paperpackage = document.header[i].split()[1]
    else:
        paperpackage = "default"
    document.header[i] = document.header[i].replace(paperpackage, packages[paperpackage])


def convert_bullets(document):
    "Convert bullets."
    i = 0
    while True:
        i = find_token(document.header, "\\bullet", i)
        if i == -1:
            return
        if document.header[i][:12] == "\\bulletLaTeX":
            document.header[i] = document.header[i] + " " + document.header[i + 1].strip()
            n = 3
        else:
            document.header[i] = (
                document.header[i]
                + " "
                + document.header[i + 1].strip()
                + " "
                + document.header[i + 2].strip()
                + " "
                + document.header[i + 3].strip()
            )
            n = 5
        del document.header[i + 1 : i + n]
        i = i + 1


def revert_bullets(document):
    "Revert bullets."
    i = 0
    while True:
        i = find_token(document.header, "\\bullet", i)
        if i == -1:
            return
        if document.header[i][:12] == "\\bulletLaTeX":
            n = document.header[i].find('"')
            if n == -1:
                document.warning("Malformed header.")
                return
            else:
                document.header[i : i + 1] = [
                    document.header[i][: n - 1],
                    "\t" + document.header[i][n:],
                    "\\end_bullet",
                ]
            i = i + 3
        else:
            frag = document.header[i].split()
            if len(frag) != 5:
                document.warning("Malformed header.")
                return
            else:
                document.header[i : i + 1] = [
                    frag[0] + " " + frag[1],
                    "\t" + frag[2],
                    "\t" + frag[3],
                    "\t" + frag[4],
                    "\\end_bullet",
                ]
                i = i + 5


def add_begin_header(document):
    r"Add \begin_header and \begin_document."
    i = find_token(document.header, "\\lyxformat", 0)
    document.header.insert(i + 1, "\\begin_header")
    document.header.insert(i + 1, "\\begin_document")


def remove_begin_header(document):
    r"Remove \begin_header and \begin_document."
    i = find_token(document.header, "\\begin_document", 0)
    if i != -1:
        del document.header[i]
    i = find_token(document.header, "\\begin_header", 0)
    if i != -1:
        del document.header[i]


def add_begin_body(document):
    r"Add and \begin_document and \end_document"
    document.body.insert(0, "\\begin_body")
    document.body.insert(1, "")
    i = find_token(document.body, "\\end_document", 0)
    document.body.insert(i, "\\end_body")


def remove_begin_body(document):
    r"Remove \begin_body and \end_body"
    i = find_token(document.body, "\\begin_body", 0)
    if i != -1:
        del document.body[i]
        if not document.body[i]:
            del document.body[i]
    i = find_token(document.body, "\\end_body", 0)
    if i != -1:
        del document.body[i]


def normalize_papersize(document):
    r"Normalize \papersize"
    i = find_token(document.header, "\\papersize", 0)
    if i == -1:
        return

    tmp = document.header[i].split()
    if tmp[1] == "Default":
        document.header[i] = "\\papersize default"
        return
    if tmp[1] == "Custom":
        document.header[i] = "\\papersize custom"


def denormalize_papersize(document):
    r"Revert \papersize"
    i = find_token(document.header, "\\papersize", 0)
    if i == -1:
        return

    tmp = document.header[i].split()
    if tmp[1] == "custom":
        document.header[i] = "\\papersize Custom"


def strip_end_space(document):
    "Strip spaces at end of command line."
    for i in range(len(document.body)):
        if document.body[i][:1] == "\\":
            document.body[i] = document.body[i].strip()


def use_x_boolean(document):
    r"Use boolean values for \use_geometry, \use_bibtopic and \tracking_changes"
    bin2bool = {"0": "false", "1": "true"}
    for use in "\\use_geometry", "\\use_bibtopic", "\\tracking_changes":
        i = find_token(document.header, use, 0)
        if i == -1:
            continue
        decompose = document.header[i].split()
        document.header[i] = decompose[0] + " " + bin2bool[decompose[1]]


def use_x_binary(document):
    r"Use digit values for \use_geometry, \use_bibtopic and \tracking_changes"
    bool2bin = {"false": "0", "true": "1"}
    for use in "\\use_geometry", "\\use_bibtopic", "\\tracking_changes":
        i = find_token(document.header, use, 0)
        if i == -1:
            continue
        decompose = document.header[i].split()
        document.header[i] = decompose[0] + " " + bool2bin[decompose[1]]


def normalize_paragraph_params(document):
    "Place all the paragraph parameters in their own line."
    body = document.body

    allowed_parameters = (
        "\\paragraph_spacing",
        "\\noindent",
        "\\align",
        "\\labelwidthstring",
        "\\start_of_appendix",
        "\\leftindent",
    )

    i = 0
    while True:
        i = find_token(document.body, "\\begin_layout", i)
        if i == -1:
            return

        i = i + 1
        while True:
            if body[i].strip() and body[i].split()[0] not in allowed_parameters:
                break

            j = body[i].find("\\", 1)

            if j != -1:
                body[i : i + 1] = [body[i][:j].strip(), body[i][j:]]

            i = i + 1


def convert_output_changes(document):
    "Add output_changes parameter."
    i = find_token(document.header, "\\tracking_changes", 0)
    if i == -1:
        document.warning("Malformed lyx document: Missing '\\tracking_changes'.")
        return
    document.header.insert(i + 1, "\\output_changes true")


def revert_output_changes(document):
    "Remove output_changes parameter."
    i = find_token(document.header, "\\output_changes", 0)
    if i == -1:
        return
    del document.header[i]


def convert_ert_paragraphs(document):
    "Convert paragraph breaks and sanitize paragraphs."
    forbidden_settings = [
        # paragraph parameters
        "\\paragraph_spacing",
        "\\labelwidthstring",
        "\\start_of_appendix",
        "\\noindent",
        "\\leftindent",
        "\\align",
        # font settings
        "\\family",
        "\\series",
        "\\shape",
        "\\size",
        "\\emph",
        "\\numeric",
        "\\bar",
        "\\noun",
        "\\color",
        "\\lang",
    ]
    i = 0
    while True:
        i = find_token(document.body, "\\begin_inset ERT", 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

        # convert non-standard paragraphs to standard
        k = i
        while True:
            k = find_token(document.body, "\\begin_layout", k, j)
            if k == -1:
                break
            document.body[k] = "\\begin_layout %s" % document.default_layout
            k = k + 1

        # remove all paragraph parameters and font settings
        k = i
        while k < j:
            if document.body[k].strip() and document.body[k].split()[0] in forbidden_settings:
                del document.body[k]
                j = j - 1
            else:
                k = k + 1

        # insert an empty paragraph before each paragraph but the first
        k = i
        first_pagraph = 1
        while True:
            k = find_token(document.body, "\\begin_layout", k, j)
            if k == -1:
                break
            if first_pagraph:
                first_pagraph = 0
                k = k + 1
                continue
            document.body[k:k] = [
                "\\begin_layout %s" % document.default_layout,
                "",
                "\\end_layout",
                "",
            ]
            k = k + 5
            j = j + 4

        # convert \\newline to new paragraph
        k = i
        while True:
            k = find_token(document.body, "\\newline", k, j)
            if k == -1:
                break
            document.body[k : k + 1] = [
                "\\end_layout",
                "",
                "\\begin_layout %s" % document.default_layout,
            ]
            k = k + 3
            j = j + 2
            # We need an empty line if document.default_layout == ''
            if document.body[k] != "":
                document.body.insert(k, "")
                k = k + 1
                j = j + 1
        i = i + 1


def revert_ert_paragraphs(document):
    "Remove double paragraph breaks."
    i = 0
    while True:
        i = find_token(document.body, "\\begin_inset ERT", 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

        # replace paragraph breaks with \newline
        k = i
        while True:
            k = find_token(document.body, "\\end_layout", k, j)
            l = find_token(document.body, "\\begin_layout", k, j)
            if k == -1 or l == -1:
                break
            document.body[k : l + 1] = ["\\newline"]
            j = j - l + k
            k = k + 1

        # replace double \newlines with paragraph breaks
        k = i
        while True:
            k = find_token(document.body, "\\newline", k, j)
            if k == -1:
                break
            l = k + 1
            while document.body[l] == "":
                l = l + 1
            if document.body[l].strip() and document.body[l].split()[0] == "\\newline":
                document.body[k : l + 1] = [
                    "\\end_layout",
                    "",
                    "\\begin_layout %s" % document.default_layout,
                ]
                j = j - l + k + 2
                k = k + 3
                # We need an empty line if document.default_layout == ''
                if document.body[l + 1] != "":
                    document.body.insert(l + 1, "")
                    k = k + 1
                    j = j + 1
            else:
                k = k + 1
        i = i + 1


def convert_french(document):
    "Convert frenchb."
    regexp = re.compile(r"^\\language\s+frenchb")
    i = find_re(document.header, regexp, 0)
    if i != -1:
        document.header[i] = "\\language french"

    # Change language in the document body
    regexp = re.compile(r"^\\lang\s+frenchb")
    i = 0
    while True:
        i = find_re(document.body, regexp, i)
        if i == -1:
            break
        document.body[i] = "\\lang french"
        i = i + 1


def remove_paperpackage(document):
    "Remove paper package."
    i = find_token(document.header, "\\paperpackage", 0)

    if i == -1:
        return

    paperpackage = document.header[i].split()[1]

    del document.header[i]

    if paperpackage not in ("a4", "a4wide", "widemarginsa4"):
        return

    conv = {
        "a4": "\\usepackage{a4}",
        "a4wide": "\\usepackage{a4wide}",
        "widemarginsa4": "\\usepackage[widemargins]{a4}",
    }
    # for compatibility we ensure it is the first entry in preamble
    document.preamble[0:0] = [conv[paperpackage]]

    i = find_token(document.header, "\\papersize", 0)
    if i != -1:
        document.header[i] = "\\papersize default"


def remove_quotestimes(document):
    "Remove quotestimes."
    i = find_token(document.header, "\\quotes_times", 0)
    if i == -1:
        return
    del document.header[i]


def convert_sgml_paragraphs(document):
    "Convert SGML paragraphs."
    if document.backend != "docbook":
        return

    i = 0
    while True:
        i = find_token(document.body, "\\begin_layout SGML", i)

        if i == -1:
            return

        document.body[i] = "\\begin_layout Standard"
        j = find_token(document.body, "\\end_layout", i)

        document.body[j + 1 : j + 1] = ["", "\\end_inset", "", "", "\\end_layout"]
        document.body[i + 1 : i + 1] = [
            "\\begin_inset ERT",
            "status inlined",
            "",
            "\\begin_layout Standard",
            "",
        ]

        i = i + 10


##
# Conversion hub
#

supported_versions = ["1.4.%d" % i for i in range(3)] + ["1.4"]
convert = [
    [222, [insert_tracking_changes, add_end_header, convert_amsmath]],
    [223, [remove_color_default, convert_spaces, convert_bibtex, remove_insetparent]],
    [224, [convert_external, convert_comment]],
    [
        225,
        [
            add_end_layout,
            layout2begin_layout,
            convert_end_document,
            convert_table_valignment_middle,
            convert_breaks,
        ],
    ],
    [226, [convert_note]],
    [227, [convert_box]],
    [228, [convert_collapsible, convert_ert]],
    [229, [convert_minipage]],
    [230, [convert_jurabib]],
    [231, [convert_float]],
    [232, [convert_bibtopic]],
    [233, [convert_graphics, convert_names]],
    [234, [convert_cite_engine]],
    [235, [convert_paperpackage]],
    [
        236,
        [
            convert_bullets,
            add_begin_header,
            add_begin_body,
            normalize_papersize,
            strip_end_space,
        ],
    ],
    [237, [use_x_boolean]],
    [238, [update_latexaccents]],
    [239, [normalize_paragraph_params]],
    [240, [convert_output_changes]],
    [241, [convert_ert_paragraphs]],
    [242, [convert_french]],
    [243, [remove_paperpackage]],
    [244, [rename_spaces]],
    [245, [remove_quotestimes, convert_sgml_paragraphs]],
]

revert = [
    [244, []],
    [243, [revert_space_names]],
    [242, []],
    [241, []],
    [240, [revert_ert_paragraphs]],
    [239, [revert_output_changes]],
    [238, []],
    [237, []],
    [236, [use_x_binary]],
    [
        235,
        [denormalize_papersize, remove_begin_body, remove_begin_header, revert_bullets],
    ],
    [234, [revert_paperpackage]],
    [233, [revert_cite_engine]],
    [232, [revert_names]],
    [231, [revert_bibtopic]],
    [230, [revert_float]],
    [229, [revert_jurabib]],
    [228, []],
    [227, [revert_collapsible, revert_ert]],
    [226, [revert_box, revert_external_2]],
    [225, [revert_note]],
    [
        224,
        [
            rm_end_layout,
            begin_layout2layout,
            revert_end_document,
            revert_valignment_middle,
            revert_breaks,
            convert_frameless_box,
            remove_branches,
        ],
    ],
    [223, [revert_external_2, revert_comment, revert_eqref]],
    [222, [revert_spaces, revert_bibtex]],
    [221, [revert_amsmath, rm_end_header, rm_tracking_changes, rm_body_changes]],
]


if __name__ == "__main__":
    pass