X-Git-Url: https://git.lyx.org/gitweb/?a=blobdiff_plain;f=lib%2Flyx2lyx%2Flyx_2_0.py;h=1771e185f45cde2dfb555aacb9dbf18ef4a58694;hb=96bcf145d92dc80bbc7ca3413be52e66d7d95327;hp=58745669a8407de98422f05fc5252289c1ff9ff1;hpb=4f04086c9748f6c93f4be3f6de37e2e6d7389654;p=lyx.git diff --git a/lib/lyx2lyx/lyx_2_0.py b/lib/lyx2lyx/lyx_2_0.py index 58745669a8..1771e185f4 100644 --- a/lib/lyx2lyx/lyx_2_0.py +++ b/lib/lyx2lyx/lyx_2_0.py @@ -23,85 +23,26 @@ import re, string import unicodedata import sys, os -from parser_tools import find_token, find_end_of, find_tokens, get_value, get_value_string +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, get_option_value + +from lyx2lyx_tools import add_to_preamble, insert_to_preamble, \ + put_cmd_in_ert, lyx2latex, latex_length, revert_flex_inset, \ + revert_font_attrs, hex2ratio, str2bool #################################################################### # Private helper functions -def remove_option(document, m, option): - l = document.body[m].find(option) - if l != -1: - val = document.body[m][l:].split('"')[1] - document.body[m] = document.body[m][:l - 1] + document.body[m][l+len(option + '="' + val + '"'):] - return l - -def find_end_of_inset(lines, i): - " Find end of inset, where lines[i] is included." - return find_end_of(lines, i, "\\begin_inset", "\\end_inset") - - -# Note that text can be either a list of lines or a single line. -def add_to_preamble(document, text): - """ Add text to the preamble if it is not already there. - Only the first line is checked!""" - - if not type(text) is list: - # split on \n just in case - # it'll give us the one element list we want - # if there's no \n, too - text = text.split('\n') - - if find_token(document.preamble, text[0], 0) != -1: - return - - document.preamble.extend(text) - - -def insert_to_preamble(index, document, text): - """ Insert text to the preamble at a given line""" - - document.preamble.insert(index, text) - - -def read_unicodesymbols(): - " Read the unicodesymbols list of unicode characters and corresponding commands." - pathname = os.path.abspath(os.path.dirname(sys.argv[0])) - fp = open(os.path.join(pathname.strip('lyx2lyx'), 'unicodesymbols')) - spec_chars = [] - # Two backslashes, followed by some non-word character, and then a character - # in brackets. The idea is to check for constructs like: \"{u}, which is how - # they are written in the unicodesymbols file; but they can also be written - # as: \"u or even \" u. - r = re.compile(r'\\\\(\W)\{(\w)\}') - for line in fp.readlines(): - if line[0] != '#' and line.strip() != "": - line=line.replace(' "',' ') # remove all quotation marks with spaces before - line=line.replace('" ',' ') # remove all quotation marks with spaces after - line=line.replace(r'\"','"') # replace \" by " (for characters with diaeresis) - try: - [ucs4,command,dead] = line.split(None,2) - if command[0:1] != "\\": - continue - spec_chars.append([command, unichr(eval(ucs4))]) - except: - continue - m = r.match(command) - if m != None: - command = "\\\\" - # If the character is a double-quote, then we need to escape it, too, - # since it is done that way in the LyX file. - if m.group(1) == "\"": - command += "\\" - commandbl = command - command += m.group(1) + m.group(2) - commandbl += m.group(1) + ' ' + m.group(2) - spec_chars.append([command, unichr(eval(ucs4))]) - spec_chars.append([commandbl, unichr(eval(ucs4))]) - fp.close() - return spec_chars - - -unicode_reps = read_unicodesymbols() +def remove_option(lines, m, option): + ''' removes option from line m. returns whether we did anything ''' + l = lines[m].find(option) + if l == -1: + return False + val = lines[m][l:].split('"')[1] + lines[m] = lines[m][:l - 1] + lines[m][l+len(option + '="' + val + '"'):] + return True # DO NOT USE THIS ROUTINE ANY MORE. Better yet, replace the uses that @@ -115,304 +56,6 @@ def old_put_cmd_in_ert(string): return string -# This routine wraps some content in an ERT inset. -# -# NOTE: The function accepts either a single string or a LIST of strings as -# argument. But it returns a LIST of strings, split on \n, so that it does -# not have embedded newlines. -# -# This is how lyx2lyx represents a LyX document: as a list of strings, -# each representing a line of a LyX file. Embedded newlines confuse -# lyx2lyx very much. -# -# A call to this routine will often go something like this: -# i = find_token('\\begin_inset FunkyInset', ...) -# ... -# j = find_end_of_inset(document.body, i) -# content = ...extract content from insets -# # that could be as simple as: -# # content = lyx2latex(document[i:j + 1]) -# ert = put_cmd_in_ert(content) -# document.body[i:j] = ert -# Now, before we continue, we need to reset i appropriately. Normally, -# this would be: -# i += len(ert) -# That puts us right after the ERT we just inserted. -# -def put_cmd_in_ert(arg): - ret = ["\\begin_inset ERT", "status collapsed", "\\begin_layout Plain Layout", ""] - # Despite the warnings just given, it will be faster for us to work - # with a single string internally. That way, we only go through the - # unicode_reps loop once. - if type(arg) is list: - s = "\n".join(arg) - else: - s = arg - for rep in unicode_reps: - s = s.replace(rep[1], rep[0].replace('\\\\', '\\')) - s = s.replace('\\', "\\backslash\n") - ret += s.splitlines() - ret += ["\\end_layout", "\\end_inset"] - return ret - - -def lyx2latex(document, lines): - 'Convert some LyX stuff into corresponding LaTeX stuff, as best we can.' - # clean up multiline stuff - content = "" - ert_end = 0 - note_end = 0 - hspace = "" - - for curline in range(len(lines)): - line = lines[curline] - if line.startswith("\\begin_inset Note Note"): - # We want to skip LyX notes, so remember where the inset ends - note_end = find_end_of_inset(lines, curline + 1) - continue - elif note_end >= curline: - # Skip LyX notes - continue - elif line.startswith("\\begin_inset ERT"): - # We don't want to replace things inside ERT, so figure out - # where the end of the inset is. - ert_end = find_end_of_inset(lines, curline + 1) - continue - elif line.startswith("\\begin_inset Formula"): - line = line[20:] - elif line.startswith("\\begin_inset Quotes"): - # For now, we do a very basic reversion. Someone who understands - # quotes is welcome to fix it up. - qtype = line[20:].strip() - # lang = qtype[0] - side = qtype[1] - dbls = qtype[2] - if side == "l": - if dbls == "d": - line = "``" - else: - line = "`" - else: - if dbls == "d": - line = "''" - else: - line = "'" - elif line.startswith("\\begin_inset space"): - line = line[18:].strip() - if line.startswith("\\hspace"): - # Account for both \hspace and \hspace* - hspace = line[:-2] - continue - elif line == "\\space{}": - line = "\\ " - elif line == "\\thinspace{}": - line = "\\," - elif hspace != "": - # The LyX length is in line[8:], after the \length keyword - length = latex_length(line[8:])[1] - line = hspace + "{" + length + "}" - hspace = "" - elif line.isspace() or \ - line.startswith("\\begin_layout") or \ - line.startswith("\\end_layout") or \ - line.startswith("\\begin_inset") or \ - line.startswith("\\end_inset") or \ - line.startswith("\\lang") or \ - line.strip() == "status collapsed" or \ - line.strip() == "status open": - #skip all that stuff - continue - - # this needs to be added to the preamble because of cases like - # \textmu, \textbackslash, etc. - add_to_preamble(document, ['% added by lyx2lyx for converted index entries', - '\\@ifundefined{textmu}', - ' {\\usepackage{textcomp}}{}']) - # a lossless reversion is not possible - # try at least to handle some common insets and settings - if ert_end >= curline: - line = line.replace(r'\backslash', '\\') - else: - # No need to add "{}" after single-nonletter macros - line = line.replace('&', '\\&') - line = line.replace('#', '\\#') - line = line.replace('^', '\\textasciicircum{}') - line = line.replace('%', '\\%') - line = line.replace('_', '\\_') - line = line.replace('$', '\\$') - - # Do the LyX text --> LaTeX conversion - for rep in unicode_reps: - line = line.replace(rep[1], rep[0] + "{}") - line = line.replace(r'\backslash', r'\textbackslash{}') - line = line.replace(r'\series bold', r'\bfseries{}').replace(r'\series default', r'\mdseries{}') - line = line.replace(r'\shape italic', r'\itshape{}').replace(r'\shape smallcaps', r'\scshape{}') - line = line.replace(r'\shape slanted', r'\slshape{}').replace(r'\shape default', r'\upshape{}') - line = line.replace(r'\emph on', r'\em{}').replace(r'\emph default', r'\em{}') - line = line.replace(r'\noun on', r'\scshape{}').replace(r'\noun default', r'\upshape{}') - line = line.replace(r'\bar under', r'\underbar{').replace(r'\bar default', r'}') - line = line.replace(r'\family sans', r'\sffamily{}').replace(r'\family default', r'\normalfont{}') - line = line.replace(r'\family typewriter', r'\ttfamily{}').replace(r'\family roman', r'\rmfamily{}') - line = line.replace(r'\InsetSpace ', r'').replace(r'\SpecialChar ', r'') - content += line - return content - - -def latex_length(slen): - ''' - Convert lengths to their LaTeX representation. Returns (bool, length), - where the bool tells us if it was a percentage, and the length is the - LaTeX representation. - ''' - i = 0 - percent = False - # the slen has the form - # ValueUnit+ValueUnit-ValueUnit or - # ValueUnit+-ValueUnit - # the + and - (glue lengths) are optional - # the + always precedes the - - - # Convert relative lengths to LaTeX units - units = {"text%":"\\textwidth", "col%":"\\columnwidth", - "page%":"\\paperwidth", "line%":"\\linewidth", - "theight%":"\\textheight", "pheight%":"\\paperheight"} - for unit in units.keys(): - i = slen.find(unit) - if i == -1: - continue - percent = True - minus = slen.rfind("-", 1, i) - plus = slen.rfind("+", 0, i) - latex_unit = units[unit] - if plus == -1 and minus == -1: - value = slen[:i] - value = str(float(value)/100) - end = slen[i + len(unit):] - slen = value + latex_unit + end - if plus > minus: - value = slen[plus + 1:i] - value = str(float(value)/100) - begin = slen[:plus + 1] - end = slen[i+len(unit):] - slen = begin + value + latex_unit + end - if plus < minus: - value = slen[minus + 1:i] - value = str(float(value)/100) - begin = slen[:minus + 1] - slen = begin + value + latex_unit - - # replace + and -, but only if the - is not the first character - slen = slen[0] + slen[1:].replace("+", " plus ").replace("-", " minus ") - # handle the case where "+-1mm" was used, because LaTeX only understands - # "plus 1mm minus 1mm" - if slen.find("plus minus"): - lastvaluepos = slen.rfind(" ") - lastvalue = slen[lastvaluepos:] - slen = slen.replace(" ", lastvalue + " ") - return (percent, slen) - - -def revert_flex_inset(document, name, LaTeXname, position): - " Convert flex insets to TeX code " - i = position - while True: - i = find_token(document.body, '\\begin_inset Flex ' + name, i) - if i == -1: - return - z = find_end_of_inset(document.body, i) - if z == -1: - document.warning("Malformed LyX document: Can't find end of Flex " + name + " inset.") - return - # remove the \end_inset - document.body[z - 2:z + 1] = put_cmd_in_ert("}") - # we need to reset character layouts if necessary - j = find_token(document.body, '\\emph on', i, z) - k = find_token(document.body, '\\noun on', i, z) - l = find_token(document.body, '\\series', i, z) - m = find_token(document.body, '\\family', i, z) - n = find_token(document.body, '\\shape', i, z) - o = find_token(document.body, '\\color', i, z) - p = find_token(document.body, '\\size', i, z) - q = find_token(document.body, '\\bar under', i, z) - r = find_token(document.body, '\\uuline on', i, z) - s = find_token(document.body, '\\uwave on', i, z) - t = find_token(document.body, '\\strikeout on', i, z) - if j != -1: - document.body.insert(z - 2, "\\emph default") - if k != -1: - document.body.insert(z - 2, "\\noun default") - if l != -1: - document.body.insert(z - 2, "\\series default") - if m != -1: - document.body.insert(z - 2, "\\family default") - if n != -1: - document.body.insert(z - 2, "\\shape default") - if o != -1: - document.body.insert(z - 2, "\\color inherit") - if p != -1: - document.body.insert(z - 2, "\\size default") - if q != -1: - document.body.insert(z - 2, "\\bar default") - if r != -1: - document.body.insert(z - 2, "\\uuline default") - if s != -1: - document.body.insert(z - 2, "\\uwave default") - if t != -1: - document.body.insert(z - 2, "\\strikeout default") - document.body[i:i + 4] = put_cmd_in_ert(LaTeXname + "{") - i += 1 - - -def revert_font_attrs(document, name, LaTeXname): - " Reverts font changes to TeX code " - i = 0 - changed = False - while True: - i = find_token(document.body, name + ' on', i) - if i == -1: - return changed - j = find_token(document.body, name + ' default', i) - k = find_token(document.body, name + ' on', i + 1) - # if there is no default set, the style ends with the layout - # assure hereby that we found the correct layout end - if j != -1 and (j < k or k == -1): - document.body[j:j + 1] = put_cmd_in_ert("}") - else: - j = find_token(document.body, '\\end_layout', i) - document.body[j:j] = put_cmd_in_ert("}") - document.body[i:i + 1] = put_cmd_in_ert(LaTeXname + "{") - changed = True - i += 1 - - -def revert_layout_command(document, name, LaTeXname, position): - " Reverts a command from a layout to TeX code " - i = position - while True: - i = find_token(document.body, '\\begin_layout ' + name, i) - if i == -1: - return - k = -1 - # find the next layout - j = i + 1 - while k == -1: - j = find_token(document.body, '\\begin_layout', j) - l = len(document.body) - # if nothing was found it was the last layout of the document - if j == -1: - document.body[l - 4:l - 4] = put_cmd_in_ert("}") - k = 0 - # exclude plain layout because this can be TeX code or another inset - elif document.body[j] != '\\begin_layout Plain Layout': - document.body[j - 2:j - 2] = put_cmd_in_ert("}") - k = 0 - else: - j += 1 - document.body[i] = '\\begin_layout Standard' - document.body[i + 1:i + 1] = put_cmd_in_ert(LaTeXname + "{") - i += 1 - - ############################################################################### ### ### Conversion and reversion routines @@ -524,7 +167,7 @@ def revert_phantom_types(document, ptype, cmd): document.warning("Can't find layout for inset at line " + str(i)) i = end continue - bend = find_token(document.body, "\\end_layout", blay, end) + bend = find_end_of_layout(document.body, blay) if bend == -1: document.warning("Malformed LyX document: Could not find end of Phantom inset's layout.") i = end @@ -552,109 +195,117 @@ def revert_vphantom(document): def revert_xetex(document): " Reverts documents that use XeTeX " + i = find_token(document.header, '\\use_xetex', 0) if i == -1: document.warning("Malformed LyX document: Missing \\use_xetex.") return - if get_value(document.header, "\\use_xetex", i) == 'false': + if not str2bool(get_value(document.header, "\\use_xetex", i)): del document.header[i] return del document.header[i] + # 1.) set doc encoding to utf8-plain i = find_token(document.header, "\\inputencoding", 0) if i == -1: document.warning("Malformed LyX document: Missing \\inputencoding.") - document.header[i] = "\\inputencoding utf8-plain" + else: + document.header[i] = "\\inputencoding utf8-plain" + # 2.) check font settings - l = find_token(document.header, "\\font_roman", 0) - if l == -1: - document.warning("Malformed LyX document: Missing \\font_roman.") - line = document.header[l] - l = re.compile(r'\\font_roman (.*)$') - m = l.match(line) - roman = m.group(1) - l = find_token(document.header, "\\font_sans", 0) - if l == -1: - document.warning("Malformed LyX document: Missing \\font_sans.") - line = document.header[l] - l = re.compile(r'\\font_sans (.*)$') - m = l.match(line) - sans = m.group(1) - l = find_token(document.header, "\\font_typewriter", 0) - if l == -1: - document.warning("Malformed LyX document: Missing \\font_typewriter.") - line = document.header[l] - l = re.compile(r'\\font_typewriter (.*)$') - m = l.match(line) - typewriter = m.group(1) - osf = get_value(document.header, '\\font_osf', 0) == "true" - sf_scale = float(get_value(document.header, '\\font_sf_scale', 0)) - tt_scale = float(get_value(document.header, '\\font_tt_scale', 0)) - # 3.) set preamble stuff - pretext = '%% This document must be processed with xelatex!\n' - pretext += '\\usepackage{fontspec}\n' - if roman != "default": - pretext += '\\setmainfont[Mapping=tex-text]{' + roman + '}\n' - if sans != "default": - pretext += '\\setsansfont[' - if sf_scale != 100: - pretext += 'Scale=' + str(sf_scale / 100) + ',' - pretext += 'Mapping=tex-text]{' + sans + '}\n' - if typewriter != "default": - pretext += '\\setmonofont' - if tt_scale != 100: - pretext += '[Scale=' + str(tt_scale / 100) + ']' - pretext += '{' + typewriter + '}\n' - if osf: - pretext += '\\defaultfontfeatures{Numbers=OldStyle}\n' - pretext += '\usepackage{xunicode}\n' - pretext += '\usepackage{xltxtra}\n' - insert_to_preamble(0, document, pretext) - # 4.) reset font settings + # defaults + roman = sans = typew = default + osf = False + sf_scale = tt_scale = 100.0 + i = find_token(document.header, "\\font_roman", 0) if i == -1: document.warning("Malformed LyX document: Missing \\font_roman.") - document.header[i] = "\\font_roman default" + else: + roman = get_value(document.header, "\\font_roman", i) + document.header[i] = "\\font_roman default" + i = find_token(document.header, "\\font_sans", 0) if i == -1: document.warning("Malformed LyX document: Missing \\font_sans.") - document.header[i] = "\\font_sans default" + else: + sans = get_value(document.header, "\\font_sans", i) + document.header[i] = "\\font_sans default" + i = find_token(document.header, "\\font_typewriter", 0) if i == -1: document.warning("Malformed LyX document: Missing \\font_typewriter.") - document.header[i] = "\\font_typewriter default" + else: + typew = get_value(document.header, "\\font_typewriter", i) + document.header[i] = "\\font_typewriter default" + i = find_token(document.header, "\\font_osf", 0) if i == -1: document.warning("Malformed LyX document: Missing \\font_osf.") - document.header[i] = "\\font_osf false" + else: + osf = str2bool(get_value(document.header, "\\font_osf", i)) + document.header[i] = "\\font_osf false" + i = find_token(document.header, "\\font_sc", 0) if i == -1: document.warning("Malformed LyX document: Missing \\font_sc.") - document.header[i] = "\\font_sc false" + else: + # we do not need this value. + document.header[i] = "\\font_sc false" + i = find_token(document.header, "\\font_sf_scale", 0) if i == -1: document.warning("Malformed LyX document: Missing \\font_sf_scale.") - document.header[i] = "\\font_sf_scale 100" + else: + val = get_value(document.header, '\\font_sf_scale', i) + try: + # float() can throw + sf_scale = float(val) + except: + document.warning("Invalid font_sf_scale value: " + val) + document.header[i] = "\\font_sf_scale 100" + i = find_token(document.header, "\\font_tt_scale", 0) if i == -1: document.warning("Malformed LyX document: Missing \\font_tt_scale.") - document.header[i] = "\\font_tt_scale 100" + else: + val = get_value(document.header, '\\font_tt_scale', i) + try: + # float() can throw + tt_scale = float(val) + except: + document.warning("Invalid font_tt_scale value: " + val) + document.header[i] = "\\font_tt_scale 100" + + # 3.) set preamble stuff + pretext = ['%% This document must be processed with xelatex!'] + pretext.append('\\usepackage{fontspec}') + if roman != "default": + pretext.append('\\setmainfont[Mapping=tex-text]{' + roman + '}') + if sans != "default": + sf = '\\setsansfont[' + if sf_scale != 100.0: + sf += 'Scale=' + str(sf_scale / 100.0) + ',' + sf += 'Mapping=tex-text]{' + sans + '}' + pretext.append(sf) + if typewriter != "default": + tw = '\\setmonofont' + if tt_scale != 100.0: + tw += '[Scale=' + str(tt_scale / 100.0) + ']' + tw += '{' + typewriter + '}' + pretext.append(tw) + if osf: + pretext.append('\\defaultfontfeatures{Numbers=OldStyle}') + pretext.append('\usepackage{xunicode}') + pretext.append('\usepackage{xltxtra}') + insert_to_preamble(document, pretext) def revert_outputformat(document): " Remove default output format param " - i = find_token(document.header, '\\default_output_format', 0) - if i == -1: + + if not del_token(document.header, '\\default_output_format', 0): document.warning("Malformed LyX document: Missing \\default_output_format.") - return - del document.header[i] - - -def hex2ratio(s): - val = string.atoi(s, 16) - if val != 0: - val += 1 - return str(val / 256.0) def revert_backgroundcolor(document): @@ -670,12 +321,11 @@ def revert_backgroundcolor(document): red = hex2ratio(colorcode[1:3]) green = hex2ratio(colorcode[3:5]) blue = hex2ratio(colorcode[5:7]) - insert_to_preamble(0, document, - '% Commands inserted by lyx2lyx to set the background color\n' - + '\\@ifundefined{definecolor}{\\usepackage{color}}{}\n' - + '\\definecolor{page_backgroundcolor}{rgb}{' - + red + ',' + green + ',' + blue + '}\n' - + '\\pagecolor{page_backgroundcolor}\n') + insert_to_preamble(document, \ + ['% To set the background color', + '\\@ifundefined{definecolor}{\\usepackage{color}}{}', + '\\definecolor{page_backgroundcolor}{rgb}{' + red + ',' + green + ',' + blue + '}', + '\\pagecolor{page_backgroundcolor}']) def revert_splitindex(document): @@ -684,12 +334,11 @@ def revert_splitindex(document): if i == -1: document.warning("Malformed LyX document: Missing \\use_indices.") return - indices = get_value(document.header, "\\use_indices", i) - preamble = "" - useindices = (indices == "true") - if useindices: - preamble += "\\usepackage{splitidx}\n" + useindices = str2bool(get_value(document.header, "\\use_indices", i)) del document.header[i] + preamble = [] + if useindices: + preamble.append("\\usepackage{splitidx})") # deal with index declarations in the preamble i = 0 @@ -708,10 +357,10 @@ def revert_splitindex(document): iname = m.group(1) ishortcut = get_value(document.header, '\\shortcut', i, k) if ishortcut != "": - preamble += "\\newindex[" + iname + "]{" + ishortcut + "}\n" + preamble.append("\\newindex[" + iname + "]{" + ishortcut + "}") del document.header[i:k + 1] - if preamble != "": - insert_to_preamble(0, document, preamble) + if preamble: + insert_to_preamble(document, preamble) # deal with index insets # these need to have the argument removed @@ -746,7 +395,7 @@ def revert_splitindex(document): if i == -1: return k = find_end_of_inset(document.body, i) - ptype = get_value(document.body, 'type', i, k).strip('"') + ptype = get_quoted_value(document.body, 'type', i, k) if ptype == "idx": j = find_token(document.body, "type", i, k) del document.body[j] @@ -788,8 +437,7 @@ def revert_subindex(document): if i == -1: document.warning("Malformed LyX document: Missing \\use_indices.") return - indices = get_value(document.header, "\\use_indices", i) - useindices = (indices == "true") + useindices = str2bool(get_value(document.header, "\\use_indices", i)) i = 0 while True: i = find_token(document.body, "\\begin_inset CommandInset index_print", i) @@ -800,7 +448,7 @@ def revert_subindex(document): if ctype != "printsubindex": i = k + 1 continue - ptype = get_value(document.body, 'type', i, k).strip('"') + ptype = get_quoted_value(document.body, 'type', i, k) if not useindices: del document.body[i:k + 1] else: @@ -815,8 +463,7 @@ def revert_printindexall(document): if i == -1: document.warning("Malformed LyX document: Missing \\use_indices.") return - indices = get_value(document.header, "\\use_indices", i) - useindices = (indices == "true") + useindices = str2bool(get_value(document.header, "\\use_indices", i)) i = 0 while True: i = find_token(document.body, "\\begin_inset CommandInset index_print", i) @@ -837,14 +484,14 @@ def revert_printindexall(document): def revert_strikeout(document): " Reverts \\strikeout font attribute " - changed = revert_font_attrs(document, "\\uuline", "\\uuline") - changed = revert_font_attrs(document, "\\uwave", "\\uwave") or changed - changed = revert_font_attrs(document, "\\strikeout", "\\sout") or changed + changed = revert_font_attrs(document.body, "\\uuline", "\\uuline") + changed = revert_font_attrs(document.body, "\\uwave", "\\uwave") or changed + changed = revert_font_attrs(document.body, "\\strikeout", "\\sout") or changed if changed == True: - insert_to_preamble(0, document, - '% Commands inserted by lyx2lyx for proper underlining\n' - + '\\PassOptionsToPackage{normalem}{ulem}\n' - + '\\usepackage{ulem}\n') + insert_to_preamble(document, \ + ['% for proper underlining', + '\\PassOptionsToPackage{normalem}{ulem}', + '\\usepackage{ulem}']) def revert_ulinelatex(document): @@ -852,28 +499,24 @@ def revert_ulinelatex(document): i = find_token(document.body, '\\bar under', 0) if i == -1: return - insert_to_preamble(0, document, - '% Commands inserted by lyx2lyx for proper underlining\n' - + '\\PassOptionsToPackage{normalem}{ulem}\n' - + '\\usepackage{ulem}\n' - + '\\let\\cite@rig\\cite\n' - + '\\newcommand{\\b@xcite}[2][\\%]{\\def\\def@pt{\\%}\\def\\pas@pt{#1}\n' - + ' \\mbox{\\ifx\\def@pt\\pas@pt\\cite@rig{#2}\\else\\cite@rig[#1]{#2}\\fi}}\n' - + '\\renewcommand{\\underbar}[1]{{\\let\\cite\\b@xcite\\uline{#1}}}\n') + insert_to_preamble(document,\ + ['% for proper underlining', + '\\PassOptionsToPackage{normalem}{ulem}', + '\\usepackage{ulem}', + '\\let\\cite@rig\\cite', + '\\newcommand{\\b@xcite}[2][\\%]{\\def\\def@pt{\\%}\\def\\pas@pt{#1}', + ' \\mbox{\\ifx\\def@pt\\pas@pt\\cite@rig{#2}\\else\\cite@rig[#1]{#2}\\fi}}', + '\\renewcommand{\\underbar}[1]{{\\let\\cite\\b@xcite\\uline{#1}}}']) def revert_custom_processors(document): " Remove bibtex_command and index_command params " - i = find_token(document.header, '\\bibtex_command', 0) - if i == -1: + + if not del_token(document.header, '\\bibtex_command', 0): document.warning("Malformed LyX document: Missing \\bibtex_command.") - else: - del document.header[i] - i = find_token(document.header, '\\index_command', 0) - if i == -1: + + if not del_token(document.header, '\\index_command', 0): document.warning("Malformed LyX document: Missing \\index_command.") - else: - del document.header[i] def convert_nomencl_width(document): @@ -895,13 +538,9 @@ def revert_nomencl_width(document): if i == -1: break j = find_end_of_inset(document.body, i) - l = find_token(document.body, "set_width", i, j) - if l == -1: - document.warning("Can't find set_width option for nomencl_print!") - i = j - continue - del document.body[l] - i = j - 1 + if not del_token(document.body, "set_width", i, j): + document.warning("Can't find set_width option for nomencl_print!") + i = j def revert_nomencl_cwidth(document): @@ -917,10 +556,9 @@ def revert_nomencl_cwidth(document): document.warning("Can't find width option for nomencl_print!") i = j continue - width = get_value(document.body, "width", i, j).strip('"') + width = get_quoted_value(document.body, "width", i, j) del document.body[l] - add_to_preamble(document, ["% this command was inserted by lyx2lyx"]) - add_to_preamble(document, ["\\setlength{\\nomlabelwidth}{" + width + "}"]) + insert_to_preamble(document, ["\\setlength{\\nomlabelwidth}{" + width + "}"]) i = j - 1 @@ -983,8 +621,7 @@ def revert_paragraph_indentation(document): if length != "default": # handle percent lengths length = latex_length(length)[1] - add_to_preamble(document, ["% this command was inserted by lyx2lyx"]) - add_to_preamble(document, ["\\setlength{\\parindent}{" + length + "}"]) + insert_to_preamble(document, ["\\setlength{\\parindent}{" + length + "}"]) del document.header[i] @@ -1000,9 +637,8 @@ def revert_percent_skip_lengths(document): return # handle percent lengths percent, length = latex_length(length) - if percent == "True": - add_to_preamble(document, ["% this command was inserted by lyx2lyx"]) - add_to_preamble(document, ["\\setlength{\\parskip}{" + length + "}"]) + if percent: + insert_to_preamble(document, ["\\setlength{\\parskip}{" + length + "}"]) # set defskip to medskip as default document.header[i] = "\\defskip medskip" @@ -1029,7 +665,7 @@ def revert_percent_vspace_lengths(document): # handle percent lengths percent, length = latex_length(length) # revert the VSpace inset to ERT - if percent == "True": + if percent: if protected: subst = put_cmd_in_ert("\\vspace*{" + length + "}") else: @@ -1042,7 +678,7 @@ def revert_percent_hspace_lengths(document): " Revert relative HSpace lengths to ERT " i = 0 while True: - i = find_token(document.body, "\\begin_inset space \\hspace", i) + i = find_token_exact(document.body, "\\begin_inset space \\hspace", i) if i == -1: break j = find_end_of_inset(document.body, i) @@ -1062,7 +698,7 @@ def revert_percent_hspace_lengths(document): # ...and if it used a percent length percent, length = latex_length(length) # revert the HSpace inset to ERT - if percent == "True": + if percent: subst = put_cmd_in_ert("\\hspace" + protected + "{" + length + "}") document.body[i:j + 1] = subst # if we did a substitution, this will still be ok @@ -1073,7 +709,7 @@ def revert_hspace_glue_lengths(document): " Revert HSpace glue lengths to ERT " i = 0 while True: - i = find_token(document.body, "\\begin_inset space \\hspace", i) + i = find_token_exact(document.body, "\\begin_inset space \\hspace", i) if i == -1: break j = find_end_of_inset(document.body, i) @@ -1102,133 +738,148 @@ def revert_hspace_glue_lengths(document): def convert_author_id(document): " Add the author_id to the \\author definition and make sure 0 is not used" i = 0 - j = 1 + anum = 1 + re_author = re.compile(r'(\\author) (\".*\")\s*(.*)$') + while True: i = find_token(document.header, "\\author", i) if i == -1: break - - r = re.compile(r'(\\author) (\".*\")\s?(.*)$') - m = r.match(document.header[i]) - if m != None: + m = re_author.match(document.header[i]) + if m: name = m.group(2) - - email = '' - if m.lastindex == 3: - email = m.group(3) - document.header[i] = "\\author %i %s %s" % (j, name, email) - j = j + 1 - i = i + 1 + email = m.group(3) + document.header[i] = "\\author %i %s %s" % (anum, name, email) + anum += 1 + i += 1 - k = 0 + i = 0 while True: - k = find_token(document.body, "\\change_", k) - if k == -1: + i = find_token(document.body, "\\change_", i) + if i == -1: break - - change = document.body[k].split(' '); + change = document.body[i].split(' '); if len(change) == 3: type = change[0] author_id = int(change[1]) time = change[2] - document.body[k] = "%s %i %s" % (type, author_id + 1, time) - k = k + 1 + document.body[i] = "%s %i %s" % (type, author_id + 1, time) + i += 1 + def revert_author_id(document): " Remove the author_id from the \\author definition " i = 0 - j = 0 + anum = 0 + rx = re.compile(r'(\\author)\s+(\d+)\s+(\".*\")\s*(.*)$') idmap = dict() + while True: i = find_token(document.header, "\\author", i) if i == -1: break - - r = re.compile(r'(\\author) (\d+) (\".*\")\s?(.*)$') - m = r.match(document.header[i]) - if m != None: + m = rx.match(document.header[i]) + if m: author_id = int(m.group(2)) - idmap[author_id] = j + idmap[author_id] = anum name = m.group(3) - - email = '' - if m.lastindex == 4: - email = m.group(4) + email = m.group(4) document.header[i] = "\\author %s %s" % (name, email) - i = i + 1 - j = j + 1 + i += 1 + # FIXME Should this be incremented if we didn't match? + anum += 1 - k = 0 + i = 0 while True: - k = find_token(document.body, "\\change_", k) - if k == -1: + i = find_token(document.body, "\\change_", i) + if i == -1: break - - change = document.body[k].split(' '); + change = document.body[i].split(' '); if len(change) == 3: type = change[0] author_id = int(change[1]) time = change[2] - document.body[k] = "%s %i %s" % (type, idmap[author_id], time) - k = k + 1 + document.body[i] = "%s %i %s" % (type, idmap[author_id], time) + i += 1 def revert_suppress_date(document): " Revert suppressing of default document date to preamble code " - i = 0 - while True: - i = find_token(document.header, "\\suppress_date", i) - if i == -1: - break - # remove the preamble line and write to the preamble - # when suppress_date was true - date = get_value(document.header, "\\suppress_date", i) - if date == "true": - add_to_preamble(document, ["% this command was inserted by lyx2lyx"]) - add_to_preamble(document, ["\\date{}"]) - del document.header[i] - i = i + 1 + i = find_token(document.header, "\\suppress_date", 0) + if i == -1: + return + # remove the preamble line and write to the preamble + # when suppress_date was true + date = str2bool(get_value(document.header, "\\suppress_date", i)) + if date: + add_to_preamble(document, ["\\date{}"]) + 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" - i = 0 - j = 0 - k = 0 + mhchem = "off" - i = find_token(document.header, "\\use_mhchem 1", 0) - if i != -1: + i = find_token(document.header, "\\use_mhchem", 0) + if i == -1: + document.warning("Malformed LyX document: Could not find mhchem setting.") mhchem = "auto" else: - i = find_token(document.header, "\\use_mhchem 2", 0) - if i != -1: + val = get_value(document.header, "\\use_mhchem", i) + if val == "1": + mhchem = "auto" + elif val == "2": mhchem = "on" + del document.header[i] + + if mhchem == "off": + # don't load case + return + if mhchem == "auto": - j = find_token(document.body, "\\cf{", 0) - if j != -1: - mhchem = "on" - else: - j = find_token(document.body, "\\ce{", 0) - if j != -1: - mhchem = "on" + i = 0 + while True: + i = find_token(document.body, "\\begin_inset Formula", i) + if i == -1: + break + line = document.body[i] + if line.find("\\ce{") != -1 or line.find("\\cf{") != -1: + mhchem = "on" + break + i += 1 + if mhchem == "on": - add_to_preamble(document, ["% this command was inserted by lyx2lyx"]) - add_to_preamble(document, ["\\PassOptionsToPackage{version=3}{mhchem}"]) - add_to_preamble(document, ["\\usepackage{mhchem}"]) - k = find_token(document.header, "\\use_mhchem", 0) - if k == -1: - document.warning("Malformed LyX document: Could not find mhchem setting.") - return - del document.header[k] + pre = ["\\PassOptionsToPackage{version=3}{mhchem}", + "\\usepackage{mhchem}"] + insert_to_preamble(document, pre) def revert_fontenc(document): " Remove fontencoding param " - i = find_token(document.header, '\\fontencoding', 0) - if i == -1: + if not del_token(document.header, '\\fontencoding', 0): document.warning("Malformed LyX document: Missing \\fontencoding.") - return - del document.header[i] def merge_gbrief(document): @@ -1286,12 +937,8 @@ def revert_gbrief(document): def revert_html_options(document): " Remove html options " - i = find_token(document.header, '\\html_use_mathml', 0) - if i != -1: - del document.header[i] - i = find_token(document.header, '\\html_be_strict', 0) - if i != -1: - del document.header[i] + del_token(document.header, '\\html_use_mathml', 0) + del_token(document.header, '\\html_be_strict', 0) def revert_includeonly(document): @@ -1302,58 +949,143 @@ def revert_includeonly(document): return j = find_end_of(document.header, i, "\\begin_includeonly", "\\end_includeonly") if j == -1: - # this should not happen + document.warning("Unable to find end of includeonly section!!") break document.header[i : j + 1] = [] def revert_includeall(document): " Remove maintain_unincluded_children param " - i = find_token(document.header, '\\maintain_unincluded_children', 0) - if i != -1: - del document.header[i] + del_token(document.header, '\\maintain_unincluded_children', 0) 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, '", i) - if i == -1: - document.warning("Malformed LyX document: Could not find end of tabular cell.") - break - subst = [old_put_cmd_in_ert("}")] - document.body[i - 3:i - 3] = subst - # cell type 4 is multirow part cell - i = find_token(document.body, '') + 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, '', 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, '', '') + 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, '" in a cell, but + # that seems fairly unlikely. + end_cell = find_end_of(document.body, begin_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): @@ -1365,8 +1097,8 @@ def convert_math_output(document): m = rgx.match(document.header[i]) newval = "0" # MathML if m: - val = m.group(1) - if val != "true": + val = str2bool(m.group(1)) + if not val: newval = "2" # Images else: document.warning("Can't match " + document.header[i]) @@ -1394,45 +1126,89 @@ def revert_math_output(document): def revert_inset_preview(document): " Dissolves the preview inset " i = 0 - j = 0 - k = 0 while True: i = find_token(document.body, "\\begin_inset Preview", i) if i == -1: return - j = find_end_of_inset(document.body, i) - if j == -1: + iend = find_end_of_inset(document.body, i) + if iend == -1: document.warning("Malformed LyX document: Could not find end of Preview inset.") - return - #If the layout is Standard we need to remove it, otherwise there - #will be paragraph breaks that shouldn't be there. - k = find_token(document.body, "\\begin_layout Standard", i) - if k == i + 2: - del document.body[i:i + 3] - del document.body[j - 5:j - 2] - i -= 6 - else: + i += 1 + continue + + # This has several issues. + # We need to do something about the layouts inside InsetPreview. + # If we just leave the first one, then we have something like: + # \begin_layout Standard + # ... + # \begin_layout Standard + # and we get a "no \end_layout" error. So something has to be done. + # Ideally, we would check if it is the same as the layout we are in. + # If so, we just remove it; if not, we end the active one. But it is + # not easy to know what layout we are in, due to depth changes, etc, + # and it is not clear to me how much work it is worth doing. In most + # cases, the layout will probably be the same. + # + # For the same reason, we have to remove the \end_layout tag at the + # end of the last layout in the inset. Again, that will sometimes be + # wrong, but it will usually be right. To know what to do, we would + # again have to know what layout the inset is in. + + blay = find_token(document.body, "\\begin_layout", i, iend) + if blay == -1: + document.warning("Can't find layout for preview inset!") + # always do the later one first... + del document.body[iend] del document.body[i] - del document.body[j - 1] - i -= 2 + # deletions mean we do not need to reset i + continue + + # This is where we would check what layout we are in. + # The check for Standard is definitely wrong. + # + # lay = document.body[blay].split(None, 1)[1] + # if lay != oldlayout: + # # record a boolean to tell us what to do later.... + # # better to do it later, since (a) it won't mess up + # # the numbering and (b) we only modify at the end. + + # we want to delete the last \\end_layout in this inset, too. + # note that this may not be the \\end_layout that goes with blay!! + bend = find_end_of_layout(document.body, blay) + while True: + tmp = find_token(document.body, "\\end_layout", bend + 1, iend) + if tmp == -1: + break + bend = tmp + if bend == blay: + document.warning("Unable to find last layout in preview inset!") + del document.body[iend] + del document.body[i] + # deletions mean we do not need to reset i + continue + # always do the later one first... + del document.body[iend] + del document.body[bend] + del document.body[i:blay + 1] + # we do not need to reset i def revert_equalspacing_xymatrix(document): " Revert a Formula with xymatrix@! to an ERT inset " i = 0 - j = 0 has_preamble = False has_equal_spacing = False + while True: - found = -1 i = find_token(document.body, "\\begin_inset Formula", i) if i == -1: break j = find_end_of_inset(document.body, i) if j == -1: document.warning("Malformed LyX document: Could not find end of Formula inset.") - break - + i += 1 + continue + for curline in range(i,j): found = document.body[curline].find("\\xymatrix@!") if found != -1: @@ -1444,7 +1220,7 @@ def revert_equalspacing_xymatrix(document): content += document.body[i + 1:j] subst = put_cmd_in_ert(content) document.body[i:j + 1] = subst - i += len(subst) + i += len(subst) - (j - i) + 1 else: for curline in range(i,j): l = document.body[curline].find("\\xymatrix") @@ -1452,134 +1228,96 @@ def revert_equalspacing_xymatrix(document): has_preamble = True; break; i = j + 1 + if has_equal_spacing and not has_preamble: add_to_preamble(document, ['\\usepackage[all]{xy}']) def revert_notefontcolor(document): " Reverts greyed-out note font color to preamble code " - i = 0 - colorcode = "" - while True: - i = find_token(document.header, "\\notefontcolor", i) - if i == -1: - return - colorcode = get_value(document.header, '\\notefontcolor', 0) - del document.header[i] - # the color code is in the form #rrggbb where every character denotes a hex number - # convert the string to an int - red = string.atoi(colorcode[1:3],16) - # we want the output "0.5" for the value "127" therefore increment here - if red != 0: - red = red + 1 - redout = float(red) / 256 - green = string.atoi(colorcode[3:5],16) - if green != 0: - green = green + 1 - greenout = float(green) / 256 - blue = string.atoi(colorcode[5:7],16) - if blue != 0: - blue = blue + 1 - blueout = float(blue) / 256 - # write the preamble - insert_to_preamble(0, document, - '% Commands inserted by lyx2lyx to set the font color\n' - '% for greyed-out notes\n' - + '\\@ifundefined{definecolor}{\\usepackage{color}}{}\n' - + '\\definecolor{note_fontcolor}{rgb}{' - + str(redout) + ', ' + str(greenout) - + ', ' + str(blueout) + '}\n' - + '\\renewenvironment{lyxgreyedout}\n' - + ' {\\textcolor{note_fontcolor}\\bgroup}{\\egroup}\n') + + i = find_token(document.header, "\\notefontcolor", 0) + if i == -1: + return + + colorcode = get_value(document.header, '\\notefontcolor', i) + del document.header[i] + + # are there any grey notes? + if find_token(document.body, "\\begin_inset Note Greyedout", 0) == -1: + # no need to do anything else, and \renewcommand will throw + # an error since lyxgreyedout will not exist. + return + + # the color code is in the form #rrggbb where every character denotes a hex number + red = hex2ratio(colorcode[1:3]) + green = hex2ratio(colorcode[3:5]) + blue = hex2ratio(colorcode[5:7]) + # write the preamble + insert_to_preamble(document, + [ '% for greyed-out notes', + '\\@ifundefined{definecolor}{\\usepackage{color}}{}' + '\\definecolor{note_fontcolor}{rgb}{%s,%s,%s}' % (red, green, blue), + '\\renewenvironment{lyxgreyedout}', + ' {\\textcolor{note_fontcolor}\\bgroup}{\\egroup}']) def revert_turkmen(document): "Set language Turkmen to English" - i = 0 + if document.language == "turkmen": document.language = "english" i = find_token(document.header, "\\language", 0) if i != -1: document.header[i] = "\\language english" + j = 0 while True: j = find_token(document.body, "\\lang turkmen", j) if j == -1: return document.body[j] = document.body[j].replace("\\lang turkmen", "\\lang english") - j = j + 1 + j += 1 def revert_fontcolor(document): " Reverts font color to preamble code " - i = 0 - colorcode = "" - while True: - i = find_token(document.header, "\\fontcolor", i) - if i == -1: - return - colorcode = get_value(document.header, '\\fontcolor', 0) - del document.header[i] - # don't clutter the preamble if backgroundcolor is not set - if colorcode == "#000000": - continue - # the color code is in the form #rrggbb where every character denotes a hex number - # convert the string to an int - red = string.atoi(colorcode[1:3],16) - # we want the output "0.5" for the value "127" therefore add here - if red != 0: - red = red + 1 - redout = float(red) / 256 - green = string.atoi(colorcode[3:5],16) - if green != 0: - green = green + 1 - greenout = float(green) / 256 - blue = string.atoi(colorcode[5:7],16) - if blue != 0: - blue = blue + 1 - blueout = float(blue) / 256 - # write the preamble - insert_to_preamble(0, document, - '% Commands inserted by lyx2lyx to set the font color\n' - + '\\@ifundefined{definecolor}{\\usepackage{color}}{}\n' - + '\\definecolor{document_fontcolor}{rgb}{' - + str(redout) + ', ' + str(greenout) - + ', ' + str(blueout) + '}\n' - + '\\color{document_fontcolor}\n') + i = find_token(document.header, "\\fontcolor", 0) + if i == -1: + return + colorcode = get_value(document.header, '\\fontcolor', i) + del document.header[i] + # don't clutter the preamble if font color is not set + if colorcode == "#000000": + return + # the color code is in the form #rrggbb where every character denotes a hex number + red = hex2ratio(colorcode[1:3]) + green = hex2ratio(colorcode[3:5]) + blue = hex2ratio(colorcode[5:7]) + # write the preamble + insert_to_preamble(document, + ['% Set the font color', + '\\@ifundefined{definecolor}{\\usepackage{color}}{}', + '\\definecolor{document_fontcolor}{rgb}{%s,%s,%s}' % (red, green, blue), + '\\color{document_fontcolor}']) + def revert_shadedboxcolor(document): " Reverts shaded box color to preamble code " - i = 0 - colorcode = "" - while True: - i = find_token(document.header, "\\boxbgcolor", i) - if i == -1: - return - colorcode = get_value(document.header, '\\boxbgcolor', 0) - del document.header[i] - # the color code is in the form #rrggbb where every character denotes a hex number - # convert the string to an int - red = string.atoi(colorcode[1:3],16) - # we want the output "0.5" for the value "127" therefore increment here - if red != 0: - red = red + 1 - redout = float(red) / 256 - green = string.atoi(colorcode[3:5],16) - if green != 0: - green = green + 1 - greenout = float(green) / 256 - blue = string.atoi(colorcode[5:7],16) - if blue != 0: - blue = blue + 1 - blueout = float(blue) / 256 - # write the preamble - insert_to_preamble(0, document, - '% Commands inserted by lyx2lyx to set the color\n' - '% of boxes with shaded background\n' - + '\\@ifundefined{definecolor}{\\usepackage{color}}{}\n' - + '\\definecolor{shadecolor}{rgb}{' - + str(redout) + ', ' + str(greenout) - + ', ' + str(blueout) + '}\n') + i = find_token(document.header, "\\boxbgcolor", 0) + if i == -1: + return + colorcode = get_value(document.header, '\\boxbgcolor', i) + del document.header[i] + # the color code is in the form #rrggbb + red = hex2ratio(colorcode[1:3]) + green = hex2ratio(colorcode[3:5]) + blue = hex2ratio(colorcode[5:7]) + # write the preamble + insert_to_preamble(document, + ['% Set the color of boxes with shaded background', + '\\@ifundefined{definecolor}{\\usepackage{color}}{}', + "\\definecolor{shadecolor}{rgb}{%s,%s,%s}" % (red, green, blue)]) def revert_lyx_version(document): @@ -1598,20 +1336,17 @@ def revert_lyx_version(document): return j = find_end_of_inset(document.body, i + 1) if j == -1: - # should not happen document.warning("Malformed LyX document: Could not find end of Info inset.") + i += 1 + continue + # We expect: # \begin_inset Info # type "lyxinfo" # arg "version" # \end_inset - # but we shall try to be forgiving. - arg = typ = "" - for k in range(i, j): - if document.body[k].startswith("arg"): - arg = document.body[k][3:].strip().strip('"') - if document.body[k].startswith("type"): - typ = document.body[k][4:].strip().strip('"') + typ = get_quoted_value(document.body, "type", i, j) + arg = get_quoted_value(document.body, "arg", i, j) if arg != "version" or typ != "lyxinfo": i = j + 1 continue @@ -1630,19 +1365,12 @@ def revert_lyx_version(document): def revert_math_scale(document): " Remove math scaling and LaTeX options " - i = find_token(document.header, '\\html_math_img_scale', 0) - if i != -1: - del document.header[i] - i = find_token(document.header, '\\html_latex_start', 0) - if i != -1: - del document.header[i] - i = find_token(document.header, '\\html_latex_end', 0) - if i != -1: - del document.header[i] + del_token(document.header, '\\html_math_img_scale', 0) + del_token(document.header, '\\html_latex_start', 0) + del_token(document.header, '\\html_latex_end', 0) def revert_pagesizes(document): - i = 0 " Revert page sizes to default " i = find_token(document.header, '\\papersize', 0) if i != -1: @@ -1656,7 +1384,6 @@ def revert_pagesizes(document): def revert_DIN_C_pagesizes(document): - i = 0 " Revert DIN C page sizes to default " i = find_token(document.header, '\\papersize', 0) if i != -1: @@ -1675,7 +1402,7 @@ def convert_html_quotes(document): line = document.header[i] l = re.compile(r'\\html_latex_start\s+"(.*)"') m = l.match(line) - if m != None: + if m: document.header[i] = "\\html_latex_start " + m.group(1) i = find_token(document.header, '\\html_latex_end', 0) @@ -1683,7 +1410,7 @@ def convert_html_quotes(document): line = document.header[i] l = re.compile(r'\\html_latex_end\s+"(.*)"') m = l.match(line) - if m != None: + if m: document.header[i] = "\\html_latex_end " + m.group(1) @@ -1695,126 +1422,58 @@ def revert_html_quotes(document): line = document.header[i] l = re.compile(r'\\html_latex_start\s+(.*)') m = l.match(line) - document.header[i] = "\\html_latex_start \"" + m.group(1) + "\"" + if not m: + document.warning("Weird html_latex_start line: " + line) + del document.header[i] + else: + document.header[i] = "\\html_latex_start \"" + m.group(1) + "\"" i = find_token(document.header, '\\html_latex_end', 0) if i != -1: line = document.header[i] l = re.compile(r'\\html_latex_end\s+(.*)') m = l.match(line) - document.header[i] = "\\html_latex_end \"" + m.group(1) + "\"" + if not m: + document.warning("Weird html_latex_end line: " + line) + del document.header[i] + else: + document.header[i] = "\\html_latex_end \"" + m.group(1) + "\"" def revert_output_sync(document): " Remove forward search options " - i = find_token(document.header, '\\output_sync_macro', 0) - if i != -1: - del document.header[i] - i = find_token(document.header, '\\output_sync', 0) - if i != -1: - del document.header[i] - - -def convert_beamer_args(document): - " Convert ERT arguments in Beamer to InsetArguments " - - if document.textclass != "beamer" and document.textclass != "article-beamer": - return - - layouts = ("Block", "ExampleBlock", "AlertBlock") - for layout in layouts: - blay = 0 - while True: - blay = find_token(document.body, '\\begin_layout ' + layout, blay) - if blay == -1: - break - elay = find_end_of(document.body, blay, '\\begin_layout', '\\end_layout') - if elay == -1: - document.warning("Malformed LyX document: Can't find end of " + layout + " layout.") - blay += 1 - continue - bert = find_token(document.body, '\\begin_inset ERT', blay) - if bert == -1: - document.warning("Malformed Beamer LyX document: Can't find argument of " + layout + " layout.") - blay = elay + 1 - continue - eert = find_end_of_inset(document.body, bert) - if eert == -1: - document.warning("Malformed LyX document: Can't find end of ERT.") - blay = elay + 1 - continue - - # So the ERT inset begins at line k and goes to line l. We now wrap it in - # an argument inset. - # Do the end first, so as not to mess up the variables. - document.body[eert + 1:eert + 1] = ['', '\\end_layout', '', '\\end_inset', ''] - document.body[bert:bert] = ['\\begin_inset OptArg', 'status open', '', - '\\begin_layout Plain Layout'] - blay = elay + 9 - - -def revert_beamer_args(document): - " Revert Beamer arguments to ERT " - - if document.textclass != "beamer" and document.textclass != "article-beamer": - return - - layouts = ("Block", "ExampleBlock", "AlertBlock") - for layout in layouts: - blay = 0 - while True: - blay = find_token(document.body, '\\begin_layout ' + layout, blay) - if blay == -1: - break - elay = find_end_of(document.body, blay, '\\begin_layout', '\\end_layout') - if elay == -1: - document.warning("Malformed LyX document: Can't find end of " + layout + " layout.") - blay += 1 - continue - bopt = find_token(document.body, '\\begin_inset OptArg', blay) - if bopt == -1: - # it is legal not to have one of these - blay = elay + 1 - continue - eopt = find_end_of_inset(document.body, bopt) - if eopt == -1: - document.warning("Malformed LyX document: Can't find end of argument.") - blay = elay + 1 - continue - bplay = find_token(document.body, '\\begin_layout Plain Layout', blay) - if bplay == -1: - document.warning("Malformed LyX document: Can't find plain layout.") - blay = elay + 1 - continue - eplay = find_end_of(document.body, bplay, '\\begin_layout', '\\end_layout') - if eplay == -1: - document.warning("Malformed LyX document: Can't find end of plain layout.") - blay = elay + 1 - continue - # So the content of the argument inset goes from bplay + 1 to eplay - 1 - bcont = bplay + 1 - if bcont >= eplay: - # Hmm. - document.warning(str(bcont) + " " + str(eplay)) - blay = blay + 1 - continue - # we convert the content of the argument into pure LaTeX... - content = lyx2latex(document, document.body[bcont:eplay]) - strlist = put_cmd_in_ert(["{" + content + "}"]) - - # now replace the optional argument with the ERT - document.body[bopt:eopt + 1] = strlist - blay = blay + 1 + del_token(document.header, '\\output_sync_macro', 0) + del_token(document.header, '\\output_sync', 0) def revert_align_decimal(document): - l = 0 + i = 0 while True: - l = document.body[l].find('alignment=decimal') - if l == -1: - break - remove_option(document, l, 'decimal_point') - document.body[l].replace('decimal', 'center') + 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 + cell = find_token(document.body, " z or y == -1: - document.warning("Malformed LyX document: Can't find layout in box.") - return - # remove the \end_layout \end_inset pair - document.body[z - 2:z + 1] = put_cmd_in_ert("}") - # determine the alignment - k = find_token(document.body, 'hor_pos', j - 4) - align = document.body[k][9] - # determine the width - l = find_token(document.body, 'width "', j + 1) - length = document.body[l][7:] - # remove trailing '"' - length = length[:-1] - length = latex_length(length)[1] - subst = "\\makebox[" + length + "][" \ - + align + "]{" - document.body[i:y + 1] = put_cmd_in_ert(subst) + 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 box.") + i = z + continue + # by looking before the layout we make sure we're actually finding + # an option, not text. + j = find_token(document.body, 'use_makebox', i, blay) + if j == -1: + i = z + continue + + if not check_token(document.body[i], "\\begin_inset Box Frameless") \ + or get_value(document.body, 'use_makebox', j) != 1: + del document.body[j] + 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 box.") + i = z + continue + # determine the alignment + align = get_quoted_value(document.body, 'hor_pos', i, blay, "c") + # determine the width + length = get_quoted_value(document.body, 'width', i, blay, "50col%") + length = latex_length(length)[1] + # remove the \end_layout \end_inset pair + document.body[bend:z + 1] = put_cmd_in_ert("}") + subst = "\\makebox[" + length + "][" \ + + align + "]{" + document.body[i:blay + 1] = put_cmd_in_ert(subst) i += 1 -def revert_use_makebox(document): - " Deletes use_makebox option of boxes " - h = 0 - while 1: - # remove the option use_makebox - h = find_token(document.body, 'use_makebox', 0) - if h == -1: - return - del document.body[h] - h += 1 - - def convert_use_makebox(document): " Adds use_makebox option for boxes " i = 0 while 1: - # remove the option use_makebox i = find_token(document.body, '\\begin_inset Box', i) if i == -1: return - k = find_token(document.body, 'use_parbox', i) + # all of this is to make sure we actually find the use_parbox + # that is an option for this box, not some text elsewhere. + z = find_end_of_inset(document.body, i) + if z == -1: + document.warning("Can't find end of box inset!!") + i += 1 + continue + blay = find_token(document.body, "\\begin_layout", i, z) + if blay == -1: + document.warning("Can't find layout in box inset!!") + i = z + continue + # so now we are looking for use_parbox before the box's layout + k = find_token(document.body, 'use_parbox', i, blay) if k == -1: document.warning("Malformed LyX document: Can't find use_parbox statement in box.") - return + i = z + continue document.body.insert(k + 1, "use_makebox 0") - i = k + 1 + i = blay + 1 # not z + 1 (box insets may be nested) def revert_IEEEtran(document): " Convert IEEEtran layouts and styles to TeX code " if document.textclass != "IEEEtran": return - revert_flex_inset(document, "IEEE membership", "\\IEEEmembership", 0) - revert_flex_inset(document, "Lowercase", "\\MakeLowercase", 0) + revert_flex_inset(document.body, "IEEE membership", "\\IEEEmembership") + revert_flex_inset(document.body, "Lowercase", "\\MakeLowercase") layouts = ("Special Paper Notice", "After Title Text", "Publication ID", "Page headings", "Biography without photo") latexcmd = {"Special Paper Notice": "\\IEEEspecialpapernotice", @@ -1924,7 +1593,7 @@ def revert_IEEEtran(document): i = find_token(document.body, '\\begin_layout ' + layout, i) if i == -1: break - j = find_end_of(document.body, i, '\\begin_layout', '\\end_layout') + j = find_end_of_layout(document.body, i) if j == -1: document.warning("Malformed LyX document: Can't find end of " + layout + " layout.") i += 1 @@ -1932,10 +1601,11 @@ def revert_IEEEtran(document): if layout in obsoletedby: document.body[i] = "\\begin_layout " + obsoletedby[layout] i = j - else: - content = lyx2latex(document, document.body[i:j + 1]) - add_to_preamble(document, [latexcmd[layout] + "{" + content + "}"]) - del document.body[i:j + 1] + continue + content = lyx2latex(document, document.body[i:j + 1]) + add_to_preamble(document, [latexcmd[layout] + "{" + content + "}"]) + del document.body[i:j + 1] + # no need to reset i def convert_prettyref(document): @@ -1953,8 +1623,8 @@ def convert_prettyref(document): document.warning("Malformed LyX document: No end of InsetRef!") i += 1 continue - k = find_token(document.body, "LatexCommand prettyref", i) - if k != -1 and k < j: + k = find_token(document.body, "LatexCommand prettyref", i, j) + if k != -1: document.body[k] = "LatexCommand formatted" i = j + 1 document.header.insert(-1, "\\use_refstyle 0") @@ -1975,8 +1645,8 @@ def revert_refstyle(document): document.warning("Malformed LyX document: No end of InsetRef") i += 1 continue - k = find_token(document.body, "LatexCommand formatted", i) - if k != -1 and k < j: + k = find_token(document.body, "LatexCommand formatted", i, j) + if k != -1: document.body[k] = "LatexCommand prettyref" i = j + 1 i = find_token(document.header, "\\use_refstyle", 0) @@ -2001,23 +1671,16 @@ def revert_nameref(document): cmdloc = i i += 1 # Make sure it is actually in an inset! - # We could just check document.lines[i-1], but that relies - # upon something that might easily change. - # We'll look back a few lines. - stins = cmdloc - 10 - if stins < 0: - stins = 0 - stins = find_token(document.body, "\\begin_inset CommandInset ref", stins) - if stins == -1 or stins > cmdloc: - continue - endins = find_end_of_inset(document.body, stins) - if endins == -1: - document.warning("Can't find end of inset at line " + stins + "!!") - continue - if endins < cmdloc: - continue - refline = find_token(document.body, "reference", stins) - if refline == -1 or refline > endins: + # A normal line could begin with "LatexCommand nameref"! + val = is_in_inset(document.body, cmdloc, \ + "\\begin_inset CommandInset ref") + if not val: + continue + stins, endins = val + + # ok, so it is in an InsetRef + refline = find_token(document.body, "reference", stins, endins) + if refline == -1: document.warning("Can't find reference for inset at line " + stinst + "!!") continue m = rx.match(document.body[refline]) @@ -2026,12 +1689,11 @@ def revert_nameref(document): continue foundone = True ref = m.group(1) - newcontent = ['\\begin_inset ERT', 'status collapsed', '', \ - '\\begin_layout Plain Layout', '', '\\backslash', \ - cmd + '{' + ref + '}', '\\end_layout', '', '\\end_inset'] + newcontent = put_cmd_in_ert('\\' + cmd + '{' + ref + '}') document.body[stins:endins + 1] = newcontent + if foundone: - add_to_preamble(document, "\usepackage{nameref}") + add_to_preamble(document, ["\usepackage{nameref}"]) def remove_Nameref(document): @@ -2047,20 +1709,9 @@ def remove_Nameref(document): i += 1 # Make sure it is actually in an inset! - # We could just check document.lines[i-1], but that relies - # upon something that might easily change. - # We'll look back a few lines. - stins = cmdloc - 10 - if stins < 0: - stins = 0 - stins = find_token(document.body, "\\begin_inset CommandInset ref", stins) - if stins == -1 or stins > cmdloc: - continue - endins = find_end_of_inset(document.body, stins) - if endins == -1: - document.warning("Can't find end of inset at line " + stins + "!!") - continue - if endins < cmdloc: + val = is_in_inset(document.body, cmdloc, \ + "\\begin_inset CommandInset ref") + if not val: continue document.body[cmdloc] = "LatexCommand nameref" @@ -2068,16 +1719,10 @@ def remove_Nameref(document): def revert_mathrsfs(document): " Load mathrsfs if \mathrsfs us use in the document " i = 0 - end = len(document.body) - 1 - while True: - j = document.body[i].find("\\mathscr{") - if j != -1: - add_to_preamble(document, ["% this command was inserted by lyx2lyx"]) + for line in document.body: + if line.find("\\mathscr{") != -1: add_to_preamble(document, ["\\usepackage{mathrsfs}"]) - break - if i == end: - break - i += 1 + return def convert_flexnames(document): @@ -2095,66 +1740,66 @@ def convert_flexnames(document): i += 1 -flex_insets = [ - ["Alert", "CharStyle:Alert"], - ["Code", "CharStyle:Code"], - ["Concepts", "CharStyle:Concepts"], - ["E-Mail", "CharStyle:E-Mail"], - ["Emph", "CharStyle:Emph"], - ["Expression", "CharStyle:Expression"], - ["Initial", "CharStyle:Initial"], - ["Institute", "CharStyle:Institute"], - ["Meaning", "CharStyle:Meaning"], - ["Noun", "CharStyle:Noun"], - ["Strong", "CharStyle:Strong"], - ["Structure", "CharStyle:Structure"], - ["ArticleMode", "Custom:ArticleMode"], - ["Endnote", "Custom:Endnote"], - ["Glosse", "Custom:Glosse"], - ["PresentationMode", "Custom:PresentationMode"], - ["Tri-Glosse", "Custom:Tri-Glosse"] -] - -flex_elements = [ - ["Abbrev", "Element:Abbrev"], - ["CCC-Code", "Element:CCC-Code"], - ["Citation-number", "Element:Citation-number"], - ["City", "Element:City"], - ["Code", "Element:Code"], - ["CODEN", "Element:CODEN"], - ["Country", "Element:Country"], - ["Day", "Element:Day"], - ["Directory", "Element:Directory"], - ["Dscr", "Element:Dscr"], - ["Email", "Element:Email"], - ["Emph", "Element:Emph"], - ["Filename", "Element:Filename"], - ["Firstname", "Element:Firstname"], - ["Fname", "Element:Fname"], - ["GuiButton", "Element:GuiButton"], - ["GuiMenu", "Element:GuiMenu"], - ["GuiMenuItem", "Element:GuiMenuItem"], - ["ISSN", "Element:ISSN"], - ["Issue-day", "Element:Issue-day"], - ["Issue-months", "Element:Issue-months"], - ["Issue-number", "Element:Issue-number"], - ["KeyCap", "Element:KeyCap"], - ["KeyCombo", "Element:KeyCombo"], - ["Keyword", "Element:Keyword"], - ["Literal", "Element:Literal"], - ["MenuChoice", "Element:MenuChoice"], - ["Month", "Element:Month"], - ["Orgdiv", "Element:Orgdiv"], - ["Orgname", "Element:Orgname"], - ["Postcode", "Element:Postcode"], - ["SS-Code", "Element:SS-Code"], - ["SS-Title", "Element:SS-Title"], - ["State", "Element:State"], - ["Street", "Element:Street"], - ["Surname", "Element:Surname"], - ["Volume", "Element:Volume"], - ["Year", "Element:Year"] -] +flex_insets = { + "Alert" : "CharStyle:Alert", + "Code" : "CharStyle:Code", + "Concepts" : "CharStyle:Concepts", + "E-Mail" : "CharStyle:E-Mail", + "Emph" : "CharStyle:Emph", + "Expression" : "CharStyle:Expression", + "Initial" : "CharStyle:Initial", + "Institute" : "CharStyle:Institute", + "Meaning" : "CharStyle:Meaning", + "Noun" : "CharStyle:Noun", + "Strong" : "CharStyle:Strong", + "Structure" : "CharStyle:Structure", + "ArticleMode" : "Custom:ArticleMode", + "Endnote" : "Custom:Endnote", + "Glosse" : "Custom:Glosse", + "PresentationMode" : "Custom:PresentationMode", + "Tri-Glosse" : "Custom:Tri-Glosse" +} + +flex_elements = { + "Abbrev" : "Element:Abbrev", + "CCC-Code" : "Element:CCC-Code", + "Citation-number" : "Element:Citation-number", + "City" : "Element:City", + "Code" : "Element:Code", + "CODEN" : "Element:CODEN", + "Country" : "Element:Country", + "Day" : "Element:Day", + "Directory" : "Element:Directory", + "Dscr" : "Element:Dscr", + "Email" : "Element:Email", + "Emph" : "Element:Emph", + "Filename" : "Element:Filename", + "Firstname" : "Element:Firstname", + "Fname" : "Element:Fname", + "GuiButton" : "Element:GuiButton", + "GuiMenu" : "Element:GuiMenu", + "GuiMenuItem" : "Element:GuiMenuItem", + "ISSN" : "Element:ISSN", + "Issue-day" : "Element:Issue-day", + "Issue-months" : "Element:Issue-months", + "Issue-number" : "Element:Issue-number", + "KeyCap" : "Element:KeyCap", + "KeyCombo" : "Element:KeyCombo", + "Keyword" : "Element:Keyword", + "Literal" : "Element:Literal", + "MenuChoice" : "Element:MenuChoice", + "Month" : "Element:Month", + "Orgdiv" : "Element:Orgdiv", + "Orgname" : "Element:Orgname", + "Postcode" : "Element:Postcode", + "SS-Code" : "Element:SS-Code", + "SS-Title" : "Element:SS-Title", + "State" : "Element:State", + "Street" : "Element:Street", + "Surname" : "Element:Surname", + "Volume" : "Element:Volume", + "Year" : "Element:Year" +} def revert_flexnames(document): @@ -2174,100 +1819,116 @@ def revert_flexnames(document): document.warning("Illegal flex inset: " + document.body[i]) i += 1 continue - style = m.group(1) - for f in flexlist: - if f[0] == style: - document.body[i] = "\\begin_inset Flex " + f[1] - break - + if style in flexlist: + document.body[i] = "\\begin_inset Flex " + flexlist[style] i += 1 def convert_mathdots(document): " Load mathdots automatically " - while True: + 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") - break + if i != -1: + document.header.insert(i + 1, "\\use_mathdots 1") def revert_mathdots(document): " Load mathdots if used in the document " - i = 0 - ddots = re.compile(r'\\begin_inset Formula .*\\ddots', re.DOTALL) - vdots = re.compile(r'\\begin_inset Formula .*\\vdots', re.DOTALL) - iddots = re.compile(r'\\begin_inset Formula .*\\iddots', re.DOTALL) + mathdots = find_token(document.header, "\\use_mathdots" , 0) - no = find_token(document.header, "\\use_mathdots 0" , 0) - auto = find_token(document.header, "\\use_mathdots 1" , 0) - yes = find_token(document.header, "\\use_mathdots 2" , 0) - if mathdots != -1: + if mathdots == -1: + document.warning("No \\usemathdots line. Assuming auto.") + else: + val = get_value(document.header, "\\use_mathdots", mathdots) del document.header[mathdots] + try: + usedots = int(val) + except: + document.warning("Invalid \\use_mathdots value: " + val + ". Assuming auto.") + # probably usedots has not been changed, but be safe. + usedots = 1 + + if usedots == 0: + # do not load case + return + if usedots == 2: + # force load case + add_to_preamble(["\\usepackage{mathdots}"]) + return + + # so we are in the auto case. we want to load mathdots if \iddots is used. + i = 0 while True: i = find_token(document.body, '\\begin_inset Formula', i) if i == -1: return j = find_end_of_inset(document.body, i) if j == -1: - document.warning("Malformed LyX document: Can't find end of Formula inset.") - return - k = ddots.search("\n".join(document.body[i:j])) - l = vdots.search("\n".join(document.body[i:j])) - m = iddots.search("\n".join(document.body[i:j])) - if (yes == -1) and ((no != -1) or (not k and not l and not m) or (auto != -1 and not m)): + document.warning("Malformed LyX document: Can't find end of Formula inset at line " + str(i)) i += 1 continue - # use \@ifundefined to catch also the "auto" case - add_to_preamble(document, ["% this command was inserted by lyx2lyx"]) - add_to_preamble(document, ["\\@ifundefined{iddots}{\\usepackage{mathdots}}\n"]) - return + code = "\n".join(document.body[i:j]) + if code.find("\\iddots") != -1: + add_to_preamble(document, ["\\@ifundefined{iddots}{\\usepackage{mathdots}}"]) + return + i = j def convert_rule(document): - " Convert \\lyxline to CommandInset line " + " Convert \\lyxline to CommandInset line. " i = 0 + + inset = ['\\begin_inset CommandInset line', + 'LatexCommand rule', + 'offset "0.5ex"', + 'width "100line%"', + 'height "1pt"', '', + '\\end_inset', '', ''] + + # if paragraphs are indented, we may have to unindent to get the + # line to be full-width. + indent = get_value(document.header, "\\paragraph_separation", 0) + have_indent = (indent == "indent") + while True: i = find_token(document.body, "\\lyxline" , i) if i == -1: return - - j = find_token(document.body, "\\color" , i - 2) - if j == i - 2: - color = document.body[j] + '\n' - else: - color = '' - k = find_token(document.body, "\\begin_layout Standard" , i - 4) - # we need to handle the case that \lyxline is in a separate paragraph and that it is colored - # the result is then an extra empty paragraph which we get by adding an empty ERT inset - if k == i - 4 and j == i - 2 and document.body[i - 1] == '': - layout = '\\begin_inset ERT\nstatus collapsed\n\n\\begin_layout Plain Layout\n\n\n\\end_layout\n\n\\end_inset\n' \ - + '\\end_layout\n\n' \ - + '\\begin_layout Standard\n' - elif k == i - 2 and document.body[i - 1] == '': - layout = '' - else: - layout = '\\end_layout\n\n' \ - + '\\begin_layout Standard\n' - l = find_token(document.body, "\\begin_layout Standard" , i + 4) - if l == i + 4 and document.body[i + 1] == '': - layout2 = '' + + # we need to find out if this line follows other content + # in its paragraph. find its layout.... + lastlay = find_token_backwards(document.body, "\\begin_layout", i) + if lastlay == -1: + document.warning("Can't find layout for line at " + str(i)) + # do the best we can. + document.body[i:i+1] = inset + i += len(inset) + continue + + # ...and look for other content before it. + lineisfirst = True + for line in document.body[lastlay + 1:i]: + # is it empty or a paragraph option? + if not line or line[0] == '\\': + continue + lineisfirst = False + break + + if lineisfirst: + document.body[i:i+1] = inset + if indent: + # we need to unindent, lest the line be too long + document.body.insert(lastlay + 1, "\\noindent") + i += len(inset) else: - layout2 = '\\end_layout\n' \ - + '\n\\begin_layout Standard\n' - subst = layout \ - + '\\noindent\n\n' \ - + color \ - + '\\begin_inset CommandInset line\n' \ - + 'LatexCommand rule\n' \ - + 'offset "0.5ex"\n' \ - + 'width "100line%"\n' \ - + 'height "1pt"\n' \ - + '\n\\end_inset\n\n\n' \ - + layout2 - document.body[i] = subst - i += 1 + # so our line is in the middle of a paragraph + # we need to add a new line, lest this line follow the + # other content on that line and run off the side of the page + document.body[i:i+1] = inset + document.body[i:i] = ["\\begin_inset Newline newline", "\\end_inset", ""] + i += len(inset) def revert_rule(document): @@ -2279,41 +1940,28 @@ def revert_rule(document): return # find end of inset j = find_token(document.body, "\\end_inset" , i) - # assure we found the end_inset of the current inset - if j > i + 6 or j == -1: + if j == -1: document.warning("Malformed LyX document: Can't find end of line inset.") return # determine the optional offset - k = find_token(document.body, 'offset', i, j) - if k != -1: - offset = document.body[k][8:-1] - else: - offset = "" + offset = get_quoted_value(document.body, 'offset', i, j) + if offset: + offset = '[' + offset + ']' # determine the width - l = find_token(document.body, 'width', i, j) - if l != -1: - width = document.body[l][7:-1] - else: - width = "100col%" + width = get_quoted_value(document.body, 'width', i, j, "100col%") + width = latex_length(width)[1] # determine the height - m = find_token(document.body, 'height', i, j) - if m != -1: - height = document.body[m][8:-1] - else: - height = "1pt" + height = get_quoted_value(document.body, 'height', i, j, "1pt") + height = latex_length(height)[1] # output the \rule command - if offset: - subst = "\\rule[" + offset + "]{" + width + "}{" + height + "}" - else: - subst = "\\rule{" + width + "}{" + height + "}" + subst = "\\rule[" + offset + "]{" + width + "}{" + height + "}" document.body[i:j + 1] = put_cmd_in_ert(subst) - i += 1 + i += len(subst) - (j - i) def revert_diagram(document): " Add the feyn package if \\Diagram is used in math " i = 0 - re_diagram = re.compile(r'\\begin_inset Formula .*\\Diagram', re.DOTALL) while True: i = find_token(document.body, '\\begin_inset Formula', i) if i == -1: @@ -2322,12 +1970,11 @@ def revert_diagram(document): if j == -1: document.warning("Malformed LyX document: Can't find end of Formula inset.") return - m = re_diagram.search("\n".join(document.body[i:j])) - if not m: - i += 1 + lines = "\n".join(document.body[i:j]) + if lines.find("\\Diagram") == -1: + i = j continue - add_to_preamble(document, ["% this command was inserted by lyx2lyx"]) - add_to_preamble(document, "\\usepackage{feyn}") + add_to_preamble(document, ["\\usepackage{feyn}"]) # only need to do it once! return @@ -2336,10 +1983,21 @@ def convert_bibtex_clearpage(document): " insert a clear(double)page bibliographystyle if bibtotoc option is used " i = find_token(document.header, '\\papersides', 0) + sides = 0 if i == -1: document.warning("Malformed LyX document: Can't find papersides definition.") - return - sides = int(document.header[i][12]) + document.warning("Assuming single sided.") + sides = 1 + else: + val = get_value(document.header, "\\papersides", i) + try: + sides = int(val) + except: + pass + if sides != 1 and sides != 2: + document.warning("Invalid papersides value: " + val) + document.warning("Assuming single sided.") + sides = 1 j = 0 while True: @@ -2354,47 +2012,389 @@ def convert_bibtex_clearpage(document): continue # only act if there is the option "bibtotoc" - m = find_token(document.body, 'options', j, k) - if m == -1: + val = get_value(document.body, 'options', j, k) + if not val: document.warning("Can't find options for bibliography inset at line " + str(j)) j = k continue - optline = document.body[m] - idx = optline.find("bibtotoc") - if idx == -1: + if val.find("bibtotoc") == -1: j = k continue # so we want to insert a new page right before the paragraph that - # this bibliography thing is in. we'll look for it backwards. - lay = j - 1 - while lay >= 0: - if document.body[lay].startswith("\\begin_layout"): - break - lay -= 1 - - if lay < 0: + # this bibliography thing is in. + lay = find_token_backwards(document.body, "\\begin_layout", j) + if lay == -1: document.warning("Can't find layout containing bibliography inset at line " + str(j)) j = k continue - subst1 = '\\begin_layout Standard\n' \ - + '\\begin_inset Newpage clearpage\n' \ - + '\\end_inset\n\n\n' \ - + '\\end_layout\n' - subst2 = '\\begin_layout Standard\n' \ - + '\\begin_inset Newpage cleardoublepage\n' \ - + '\\end_inset\n\n\n' \ - + '\\end_layout\n' if sides == 1: - document.body.insert(lay, subst1) - document.warning(subst1) + cmd = "clearpage" else: - document.body.insert(lay, subst2) - document.warning(subst2) + cmd = "cleardoublepage" + subst = ['\\begin_layout Standard', + '\\begin_inset Newpage ' + cmd, + '\\end_inset', '', '', + '\\end_layout', ''] + document.body[lay:lay] = subst + j = k + len(subst) + + +def check_passthru(document): + tc = document.textclass + ok = (tc == "literate-article" or tc == "literate-book" or tc == "literate-report") + if not ok: + mods = document.get_module_list() + for mod in mods: + if mod == "sweave" or mod == "noweb": + ok = True + break + return ok - j = k + +def convert_passthru(document): + " http://www.mail-archive.com/lyx-devel@lists.lyx.org/msg161298.html " + if not check_passthru: + return + + rx = re.compile("\\\\begin_layout \s*(\w+)") + beg = 0 + for lay in ["Chunk", "Scrap"]: + while True: + beg = find_token(document.body, "\\begin_layout " + lay, beg) + if beg == -1: + break + end = find_end_of_layout(document.body, beg) + if end == -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. + ns = beg + while True: + ns = find_token(document.body, "\\begin_inset Newline newline", ns, end) + if ns == -1: + break + ne = find_end_of_inset(document.body, ns) + if ne == -1 or ne > end: + document.warning("Can't find end of inset at line " + str(nb)) + ns += 1 + continue + if document.body[ne + 1] == "": + ne += 1 + subst = ["\\end_layout", "", "\\begin_layout " + lay] + document.body[ns:ne + 1] = subst + # now we need to adjust end, in particular, but might as well + # do ns properly, too + newlines = (ne - ns) - len(subst) + ns += newlines + 2 + end += newlines + 1 + # 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 + next = find_token(document.body, "\\begin_layout", end) + if next != -1: + m = rx.match(document.body[next]) + if m: + nextlay = m.group(1) + if nextlay == lay: + subst = ["\\begin_layout " + lay, "", "\\end_layout", ""] + document.body[next:next] = subst + didit = True + beg = end + 1 + if didit: + beg += 4 # for the extra layout + + +def revert_passthru(document): + " http://www.mail-archive.com/lyx-devel@lists.lyx.org/msg161298.html " + if not check_passthru: + return + rx = re.compile("\\\\begin_layout \s*(\w+)") + beg = 0 + for lay in ["Chunk", "Scrap"]: + while True: + beg = find_token(document.body, "\\begin_layout " + lay, beg) + if beg == -1: + break + end = find_end_of_layout(document.body, beg) + if end == -1: + document.warning("Can't find end of layout at line " + str(beg)) + beg += 1 + continue + + # we now want to find out if the next layout is the + # same as this one. but we will need to do this over and + # over again. + while True: + next = find_token(document.body, "\\begin_layout", end) + if next == -1: + break + m = rx.match(document.body[next]) + if not m: + break + nextlay = m.group(1) + if nextlay != lay: + break + # so it is the same layout again. we now want to know if it is empty. + # but first let's check and make sure there is no content between the + # two layouts. i'm not sure if that can happen or not. + for l in range(end + 1, next): + document.warning("c'" + document.body[l] + "'") + if document.body[l] != "": + document.warning("Found content between adjacent " + lay + " layouts!") + break + nextend = find_end_of_layout(document.body, next) + if nextend == -1: + document.warning("Can't find end of layout at line " + str(next)) + break + empty = True + for l in range(next + 1, nextend): + document.warning("e'" + document.body[l] + "'") + if document.body[l] != "": + empty = False + break + if empty: + # empty layouts just get removed + # should we check if it's before yet another such layout? + del document.body[next : nextend + 1] + # and we do not want to check again. we know the next layout + # should be another Chunk and should be left as is. + break + else: + # if it's not empty, then we want to insert a newline in place + # of the layout switch + subst = ["\\begin_inset Newline newline", "\\end_inset", ""] + document.body[end : next + 1] = subst + # and now we have to find the end of the new, larger layout + newend = find_end_of_layout(document.body, beg) + if newend == -1: + document.warning("Can't find end of new layout at line " + str(beg)) + break + 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, '') + 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, '', 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, '', '') + 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, '" in a cell, but + # that seems fairly unlikely. + end_cell = find_end_of(document.body, begin_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" ## @@ -2427,7 +2427,7 @@ convert = [[346, []], [368, []], [369, [convert_author_id]], [370, []], - [371, []], + [371, [convert_mhchem]], [372, []], [373, [merge_gbrief]], [374, []], @@ -2460,10 +2460,22 @@ convert = [[346, []], [401, []], [402, [convert_bibtex_clearpage]], [403, [convert_flexnames]], - [404, [convert_prettyref]] + [404, [convert_prettyref]], + [405, []], + [406, [convert_passthru]], + [407, []], + [408, []], + [409, [convert_use_xetex]], + [410, []] ] -revert = [[403, [revert_refstyle]], +revert = [[409, [revert_labeling]], + [408, [revert_use_xetex]], + [407, [revert_script]], + [406, [revert_multirowOffset]], + [405, [revert_passthru]], + [404, []], + [403, [revert_refstyle]], [402, [revert_flexnames]], [401, []], [400, [revert_diagram]], @@ -2475,7 +2487,7 @@ revert = [[403, [revert_refstyle]], [394, [revert_DIN_C_pagesizes]], [393, [revert_makebox]], [392, [revert_argument]], - [391, [revert_beamer_args]], + [391, []], [390, [revert_align_decimal, revert_IEEEtran]], [389, [revert_output_sync]], [388, [revert_html_quotes]],