# -*- coding: utf-8 -*-
# This file is part of lyx2lyx
# -*- coding: utf-8 -*-
-# Copyright (C) 2010 The LyX team
+# Copyright (C) 2011 The LyX team
#
# This program is free software; you can redistribute it and/or
# modify it under the terms of the GNU General Public License
from parser_tools import find_token, find_end_of, find_tokens, \
find_token_exact, find_end_of_inset, find_end_of_layout, \
find_token_backwards, is_in_inset, get_value, get_quoted_value, \
- del_token, check_token
+ del_token, check_token, get_option_value
from lyx2lyx_tools import add_to_preamble, insert_to_preamble, \
put_cmd_in_ert, lyx2latex, latex_length, revert_flex_inset, \
# 2.) check font settings
# defaults
- roman = sans = typew = default
+ roman = sans = typew = "default"
osf = False
sf_scale = tt_scale = 100.0
sf += 'Scale=' + str(sf_scale / 100.0) + ','
sf += 'Mapping=tex-text]{' + sans + '}'
pretext.append(sf)
- if typewriter != "default":
+ if typew != "default":
tw = '\\setmonofont'
if tt_scale != 100.0:
tw += '[Scale=' + str(tt_scale / 100.0) + ']'
- tw += '{' + typewriter + '}'
+ tw += '{' + typew + '}'
pretext.append(tw)
if osf:
pretext.append('\\defaultfontfeatures{Numbers=OldStyle}')
name = m.group(2)
email = m.group(3)
document.header[i] = "\\author %i %s %s" % (anum, name, email)
- # FIXME Should this really be incremented if we didn't match?
anum += 1
i += 1
del document.header[i]
+def convert_mhchem(document):
+ "Set mhchem to off for versions older than 1.6.x"
+ if document.start < 277:
+ # LyX 1.5.x and older did never load mhchem.
+ # Therefore we must switch it off: Documents that use mhchem have
+ # a manual \usepackage anyway, and documents not using mhchem but
+ # custom macros with the same names as mhchem commands might get
+ # corrupted if mhchem is automatically loaded.
+ mhchem = 0 # off
+ else:
+ # LyX 1.6.x did always load mhchem automatically.
+ mhchem = 1 # auto
+ i = find_token(document.header, "\\use_esint", 0)
+ if i == -1:
+ # pre-1.5.x document
+ i = find_token(document.header, "\\use_amsmath", 0)
+ if i == -1:
+ document.warning("Malformed LyX document: Could not find amsmath os esint setting.")
+ return
+ document.header.insert(i + 1, "\\use_mhchem %d" % mhchem)
+
+
def revert_mhchem(document):
"Revert mhchem loading to preamble code"
def revert_multirow(document):
" Revert multirow cells in tables to TeX-code"
- i = 0
- multirow = False
+
+ # first, let's find out if we need to do anything
+ # cell type 3 is multirow begin cell
+ i = find_token(document.body, '<cell multirow="3"', 0)
+ if i == -1:
+ return
+
+ add_to_preamble(document, ["\\usepackage{multirow}"])
+
+ begin_table = 0
while True:
- # cell type 3 is multirow begin cell
- i = find_token(document.body, '<cell multirow="3"', i)
- if i == -1:
- break
- # a multirow cell was found
- multirow = True
- # remove the multirow tag, set the valignment to top
- # and remove the bottom line
- # FIXME Are we sure these always have space around them?
- document.body[i] = document.body[i].replace(' multirow="3" ', ' ')
- document.body[i] = document.body[i].replace('valignment="middle"', 'valignment="top"')
- document.body[i] = document.body[i].replace(' bottomline="true" ', ' ')
- # write ERT to create the multirow cell
- # use 2 rows and 2cm as default with because the multirow span
- # and the column width is only hardly accessible
- cend = find_token(document.body, "</cell>", i)
- if cend == -1:
- document.warning("Malformed LyX document: Could not find end of tabular cell.")
- i += 1
- continue
- blay = find_token(document.body, "\\begin_layout", i, cend)
- if blay == -1:
- document.warning("Can't find layout for cell!")
- i = j
- continue
- bend = find_end_of_layout(document.body, blay)
- if blay == -1:
- document.warning("Can't find end of layout for cell!")
- i = cend
+ # find begin/end of table
+ begin_table = find_token(document.body, '<lyxtabular version=', begin_table)
+ if begin_table == -1:
+ break
+ end_table = find_end_of(document.body, begin_table, '<lyxtabular', '</lyxtabular>')
+ if end_table == -1:
+ document.warning("Malformed LyX document: Could not find end of table.")
+ begin_table += 1
+ continue
+ # does this table have multirow?
+ i = find_token(document.body, '<cell multirow="3"', begin_table, end_table)
+ if i == -1:
+ begin_table = end_table
+ continue
+
+ # store the number of rows and columns
+ numrows = get_option_value(document.body[begin_table], "rows")
+ numcols = get_option_value(document.body[begin_table], "columns")
+ try:
+ numrows = int(numrows)
+ numcols = int(numcols)
+ except:
+ document.warning(numrows)
+ document.warning("Unable to determine rows and columns!")
+ begin_table = end_table
continue
- # do the later one first, so as not to mess up the numbering
- # we are wrapping the whole cell in this ert
- # so before the end of the layout...
- document.body[bend:bend] = put_cmd_in_ert("}")
- # ...and after the beginning
- document.body[blay+1:blay+1] = put_cmd_in_ert("\\multirow{2}{2cm}{")
-
- while True:
- # cell type 4 is multirow part cell
- k = find_token(document.body, '<cell multirow="4"', cend)
- if k == -1:
+ mrstarts = []
+ multirows = []
+ # collect info on rows and columns of this table.
+ begin_row = begin_table
+ for row in range(numrows):
+ begin_row = find_token(document.body, '<row>', begin_row, end_table)
+ if begin_row == -1:
+ document.warning("Can't find row " + str(row + 1))
break
- # remove the multirow tag, set the valignment to top
- # and remove the top line
- # FIXME Are we sure these always have space around them?
- document.body[k] = document.body[k].replace(' multirow="4" ', ' ')
- document.body[k] = document.body[k].replace('valignment="middle"', 'valignment="top"')
- document.body[k] = document.body[k].replace(' topline="true" ', ' ')
- k += 1
- # this will always be ok
- i = cend
-
- if multirow == True:
- add_to_preamble(document, ["\\usepackage{multirow}"])
+ end_row = find_end_of(document.body, begin_row, '<row>', '</row>')
+ if end_row == -1:
+ document.warning("Can't find end of row " + str(row + 1))
+ break
+ begin_cell = begin_row
+ multirows.append([])
+ for column in range(numcols):
+ begin_cell = find_token(document.body, '<cell ', begin_cell, end_row)
+ if begin_cell == -1:
+ document.warning("Can't find column " + str(column + 1) + \
+ "in row " + str(row + 1))
+ break
+ # NOTE
+ # this will fail if someone puts "</cell>" in a cell, but
+ # that seems fairly unlikely.
+ end_cell = find_end_of(document.body, begin_cell, '<cell', '</cell>')
+ if end_cell == -1:
+ document.warning("Can't find end of column " + str(column + 1) + \
+ "in row " + str(row + 1))
+ break
+ multirows[row].append([begin_cell, end_cell, 0])
+ if document.body[begin_cell].find('multirow="3"') != -1:
+ multirows[row][column][2] = 3 # begin multirow
+ mrstarts.append([row, column])
+ elif document.body[begin_cell].find('multirow="4"') != -1:
+ multirows[row][column][2] = 4 # in multirow
+ begin_cell = end_cell
+ begin_row = end_row
+ # end of table info collection
+
+ # work from the back to avoid messing up numbering
+ mrstarts.reverse()
+ for m in mrstarts:
+ row = m[0]
+ col = m[1]
+ # get column width
+ col_width = get_option_value(document.body[begin_table + 2 + col], "width")
+ # "0pt" means that no width is specified
+ if not col_width or col_width == "0pt":
+ col_width = "*"
+ # determine the number of cells that are part of the multirow
+ nummrs = 1
+ for r in range(row + 1, numrows):
+ if multirows[r][col][2] != 4:
+ break
+ nummrs += 1
+ # take the opportunity to revert this line
+ lineno = multirows[r][col][0]
+ document.body[lineno] = document.body[lineno].\
+ replace(' multirow="4" ', ' ').\
+ replace('valignment="middle"', 'valignment="top"').\
+ replace(' topline="true" ', ' ')
+ # remove bottom line of previous multirow-part cell
+ lineno = multirows[r-1][col][0]
+ document.body[lineno] = document.body[lineno].replace(' bottomline="true" ', ' ')
+ # revert beginning cell
+ bcell = multirows[row][col][0]
+ ecell = multirows[row][col][1]
+ document.body[bcell] = document.body[bcell].\
+ replace(' multirow="3" ', ' ').\
+ replace('valignment="middle"', 'valignment="top"')
+ blay = find_token(document.body, "\\begin_layout", bcell, ecell)
+ if blay == -1:
+ document.warning("Can't find layout for cell!")
+ continue
+ bend = find_end_of_layout(document.body, blay)
+ if bend == -1:
+ document.warning("Can't find end of layout for cell!")
+ continue
+ # do the later one first, so as not to mess up the numbering
+ # we are wrapping the whole cell in this ert
+ # so before the end of the layout...
+ document.body[bend:bend] = put_cmd_in_ert("}")
+ # ...and after the beginning
+ document.body[blay + 1:blay + 1] = \
+ put_cmd_in_ert("\\multirow{" + str(nummrs) + "}{" + col_width + "}{")
+
+ begin_table = end_table
def convert_math_output(document):
i = z
continue
document.body.insert(k + 1, "use_makebox 0")
- i = z + 1
+ i = blay + 1 # not z + 1 (box insets may be nested)
def revert_IEEEtran(document):
def convert_mathdots(document):
" Load mathdots automatically "
- i = find_token(document.header, "\\use_esint" , 0)
+ i = find_token(document.header, "\\use_mhchem" , 0)
+ if i == -1:
+ i = find_token(document.header, "\\use_esint" , 0)
if i != -1:
document.header.insert(i + 1, "\\use_mathdots 1")
# only need to do it once!
return
+chapters = ("amsbook", "book", "docbook-book", "elsart", "extbook", "extreport",
+ "jbook", "jreport", "jsbook", "literate-book", "literate-report", "memoir",
+ "mwbk", "mwrep", "recipebook", "report", "scrbook", "scrreprt", "svmono",
+ "svmult", "tbook", "treport", "tufte-book")
def convert_bibtex_clearpage(document):
" insert a clear(double)page bibliographystyle if bibtotoc option is used "
+ if document.textclass not in chapters:
+ return
+
i = find_token(document.header, '\\papersides', 0)
sides = 0
if i == -1:
document.warning("Can't find end of layout at line " + str(beg))
beg += 1
continue
+
# we are now going to replace newline insets within this layout
# by new instances of this layout. so we have repeated layouts
# instead of newlines.
+
+ # if the paragraph has any customization, however, we do not want to
+ # do the replacement.
+ if document.body[beg + 1].startswith("\\"):
+ beg = end + 1
+ continue
+
ns = beg
while True:
ns = find_token(document.body, "\\begin_inset Newline newline", ns, end)
# do ns properly, too
newlines = (ne - ns) - len(subst)
ns += newlines + 2
- end += newlines + 1
+ end += newlines + 2
+
# ok, we now want to find out if the next layout is the
# same as this one. if so, we will insert an extra copy of it
didit = False
end = newend
beg = end + 1
+
+def revert_multirowOffset(document):
+ " Revert multirow cells with offset in tables to TeX-code"
+ # this routine is the same as the revert_multirow routine except that
+ # it checks additionally for the offset
+
+ # first, let's find out if we need to do anything
+ i = find_token(document.body, '<cell multirow="3" mroffset=', 0)
+ if i == -1:
+ return
+
+ add_to_preamble(document, ["\\usepackage{multirow}"])
+
+ rgx = re.compile(r'mroffset="[^"]+?"')
+ begin_table = 0
+
+ while True:
+ # find begin/end of table
+ begin_table = find_token(document.body, '<lyxtabular version=', begin_table)
+ if begin_table == -1:
+ break
+ end_table = find_end_of(document.body, begin_table, '<lyxtabular', '</lyxtabular>')
+ if end_table == -1:
+ document.warning("Malformed LyX document: Could not find end of table.")
+ begin_table += 1
+ continue
+ # does this table have multirow?
+ i = find_token(document.body, '<cell multirow="3"', begin_table, end_table)
+ if i == -1:
+ begin_table = end_table
+ continue
+
+ # store the number of rows and columns
+ numrows = get_option_value(document.body[begin_table], "rows")
+ numcols = get_option_value(document.body[begin_table], "columns")
+ try:
+ numrows = int(numrows)
+ numcols = int(numcols)
+ except:
+ document.warning(numrows)
+ document.warning("Unable to determine rows and columns!")
+ begin_table = end_table
+ continue
+
+ mrstarts = []
+ multirows = []
+ # collect info on rows and columns of this table.
+ begin_row = begin_table
+ for row in range(numrows):
+ begin_row = find_token(document.body, '<row>', begin_row, end_table)
+ if begin_row == -1:
+ document.warning("Can't find row " + str(row + 1))
+ break
+ end_row = find_end_of(document.body, begin_row, '<row>', '</row>')
+ if end_row == -1:
+ document.warning("Can't find end of row " + str(row + 1))
+ break
+ begin_cell = begin_row
+ multirows.append([])
+ for column in range(numcols):
+ begin_cell = find_token(document.body, '<cell ', begin_cell, end_row)
+ if begin_cell == -1:
+ document.warning("Can't find column " + str(column + 1) + \
+ "in row " + str(row + 1))
+ break
+ # NOTE
+ # this will fail if someone puts "</cell>" in a cell, but
+ # that seems fairly unlikely.
+ end_cell = find_end_of(document.body, begin_cell, '<cell', '</cell>')
+ if end_cell == -1:
+ document.warning("Can't find end of column " + str(column + 1) + \
+ "in row " + str(row + 1))
+ break
+ multirows[row].append([begin_cell, end_cell, 0])
+ if document.body[begin_cell].find('multirow="3" mroffset=') != -1:
+ multirows[row][column][2] = 3 # begin multirow
+ mrstarts.append([row, column])
+ elif document.body[begin_cell].find('multirow="4"') != -1:
+ multirows[row][column][2] = 4 # in multirow
+ begin_cell = end_cell
+ begin_row = end_row
+ # end of table info collection
+
+ # work from the back to avoid messing up numbering
+ mrstarts.reverse()
+ for m in mrstarts:
+ row = m[0]
+ col = m[1]
+ # get column width
+ col_width = get_option_value(document.body[begin_table + 2 + col], "width")
+ # "0pt" means that no width is specified
+ if not col_width or col_width == "0pt":
+ col_width = "*"
+ # determine the number of cells that are part of the multirow
+ nummrs = 1
+ for r in range(row + 1, numrows):
+ if multirows[r][col][2] != 4:
+ break
+ nummrs += 1
+ # take the opportunity to revert this line
+ lineno = multirows[r][col][0]
+ document.body[lineno] = document.body[lineno].\
+ replace(' multirow="4" ', ' ').\
+ replace('valignment="middle"', 'valignment="top"').\
+ replace(' topline="true" ', ' ')
+ # remove bottom line of previous multirow-part cell
+ lineno = multirows[r-1][col][0]
+ document.body[lineno] = document.body[lineno].replace(' bottomline="true" ', ' ')
+ # revert beginning cell
+ bcell = multirows[row][col][0]
+ ecell = multirows[row][col][1]
+ offset = get_option_value(document.body[bcell], "mroffset")
+ document.body[bcell] = document.body[bcell].\
+ replace(' multirow="3" ', ' ').\
+ replace('valignment="middle"', 'valignment="top"')
+ # remove mroffset option
+ document.body[bcell] = rgx.sub('', document.body[bcell])
+
+ blay = find_token(document.body, "\\begin_layout", bcell, ecell)
+ if blay == -1:
+ document.warning("Can't find layout for cell!")
+ continue
+ bend = find_end_of_layout(document.body, blay)
+ if bend == -1:
+ document.warning("Can't find end of layout for cell!")
+ continue
+ # do the later one first, so as not to mess up the numbering
+ # we are wrapping the whole cell in this ert
+ # so before the end of the layout...
+ document.body[bend:bend] = put_cmd_in_ert("}")
+ # ...and after the beginning
+ document.body[blay + 1:blay + 1] = \
+ put_cmd_in_ert("\\multirow{" + str(nummrs) + "}{" + col_width + "}[" \
+ + offset + "]{")
+
+ # on to the next table
+ begin_table = end_table
+
+
+def revert_script(document):
+ " Convert subscript/superscript inset to TeX code "
+ i = 0
+ foundsubscript = False
+ while 1:
+ i = find_token(document.body, '\\begin_inset script', i)
+ if i == -1:
+ break
+ z = find_end_of_inset(document.body, i)
+ if z == -1:
+ document.warning("Malformed LyX document: Can't find end of script inset.")
+ i += 1
+ continue
+ blay = find_token(document.body, "\\begin_layout", i, z)
+ if blay == -1:
+ document.warning("Malformed LyX document: Can't find layout in script inset.")
+ i = z
+ continue
+
+ if check_token(document.body[i], "\\begin_inset script subscript"):
+ subst = '\\textsubscript{'
+ foundsubscript = True
+ elif check_token(document.body[i], "\\begin_inset script superscript"):
+ subst = '\\textsuperscript{'
+ else:
+ document.warning("Malformed LyX document: Unknown type of script inset.")
+ i = z
+ continue
+ bend = find_end_of_layout(document.body, blay)
+ if bend == -1 or bend > z:
+ document.warning("Malformed LyX document: Can't find end of layout in script inset.")
+ i = z
+ continue
+ # remove the \end_layout \end_inset pair
+ document.body[bend:z + 1] = put_cmd_in_ert("}")
+ document.body[i:blay + 1] = put_cmd_in_ert(subst)
+ i += 1
+ # these classes provide a \textsubscript command:
+ # FIXME: Would be nice if we could use the information of the .layout file here
+ classes = ["memoir", "scrartcl", "scrbook", "scrlttr2", "scrreprt"]
+ if foundsubscript and find_token_exact(classes, document.textclass, 0) == -1:
+ add_to_preamble(document, ['\\usepackage{subscript}'])
+
+
+def convert_use_xetex(document):
+ " convert \\use_xetex to \\use_non_tex_fonts "
+ i = 0
+ i = find_token(document.header, "\\use_xetex", 0)
+ if i == -1:
+ return
+
+ val = get_value(document.header, "\\use_xetex", 0)
+ document.header[i] = "\\use_non_tex_fonts " + val
+
+
+def revert_use_xetex(document):
+ " revert \\use_non_tex_fonts to \\use_xetex "
+ i = 0
+ i = find_token(document.header, "\\use_non_tex_fonts", 0)
+ if i == -1:
+ document.warning("Malformed document. No \\use_non_tex_fonts param!")
+ return
+
+ val = get_value(document.header, "\\use_non_tex_fonts", 0)
+ document.header[i] = "\\use_xetex " + val
+
+
+def revert_labeling(document):
+ koma = ("scrartcl", "scrarticle-beamer", "scrbook", "scrlettr",
+ "scrlttr2", "scrreprt")
+ if document.textclass in koma:
+ return
+ i = 0
+ while True:
+ i = find_token_exact(document.body, "\\begin_layout Labeling", i)
+ if i == -1:
+ return
+ document.body[i] = "\\begin_layout List"
+
+
+def revert_langpack(document):
+ " revert \\language_package parameter "
+ i = 0
+ i = find_token(document.header, "\\language_package", 0)
+ if i == -1:
+ document.warning("Malformed document. No \\language_package param!")
+ return
+
+ del document.header[i]
+
+
+def convert_langpack(document):
+ " Add \\language_package parameter "
+ i = find_token(document.header, "\language" , 0)
+ if i == -1:
+ document.warning("Malformed document. No \\language defined!")
+ return
+
+ document.header.insert(i + 1, "\\language_package default")
+
+
+def revert_tabularwidth(document):
+ i = 0
+ while True:
+ i = find_token(document.body, "\\begin_inset Tabular", i)
+ if i == -1:
+ return
+ j = find_end_of_inset(document.body, i)
+ if j == -1:
+ document.warning("Unable to find end of Tabular inset at line " + str(i))
+ i += 1
+ continue
+ i += 1
+ features = find_token(document.body, "<features", i, j)
+ if features == -1:
+ document.warning("Can't find any features in Tabular inset at line " + str(i))
+ i = j
+ continue
+ if document.body[features].find('alignment="tabularwidth"') != -1:
+ remove_option(document.body, features, 'tabularwidth')
+
+def revert_html_css_as_file(document):
+ if not del_token(document.header, '\\html_css_as_file', 0):
+ document.warning("Malformed LyX document: Missing \\html_css_as_file.")
+
+
##
# Conversion hub
#
[368, []],
[369, [convert_author_id]],
[370, []],
- [371, []],
+ [371, [convert_mhchem]],
[372, []],
[373, [merge_gbrief]],
[374, []],
[403, [convert_flexnames]],
[404, [convert_prettyref]],
[405, []],
- [406, [convert_passthru]]
+ [406, [convert_passthru]],
+ [407, []],
+ [408, []],
+ [409, [convert_use_xetex]],
+ [410, []],
+ [411, [convert_langpack]],
+ [412, []],
+ [413, []],
]
-revert = [[405, [revert_passthru]],
+revert = [[412, [revert_html_css_as_file]],
+ [411, [revert_tabularwidth]],
+ [410, [revert_langpack]],
+ [409, [revert_labeling]],
+ [408, [revert_use_xetex]],
+ [407, [revert_script]],
+ [406, [revert_multirowOffset]],
+ [405, [revert_passthru]],
[404, []],
[403, [revert_refstyle]],
[402, [revert_flexnames]],