X-Git-Url: https://git.lyx.org/gitweb/?a=blobdiff_plain;f=lib%2Flyx2lyx%2Flyx_2_0.py;h=2756418f88d09afb0671878977800b6a8f525584;hb=f6cd5ab863cc9a460efb7a1dc7eb48b87d867655;hp=7556609bc2a5487faa08f0f86d88f41765728865;hpb=cd544b25d3c56e68d60f0ad8b43c75d49f0a825c;p=lyx.git

diff --git a/lib/lyx2lyx/lyx_2_0.py b/lib/lyx2lyx/lyx_2_0.py
index 7556609bc2..2756418f88 100644
--- a/lib/lyx2lyx/lyx_2_0.py
+++ b/lib/lyx2lyx/lyx_2_0.py
@@ -1,7 +1,7 @@
 # -*- 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
@@ -15,7 +15,7 @@
 #
 # You should have received a copy of the GNU General Public License
 # along with this program; if not, write to the Free Software
-# Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA  02111-1307, USA.
+# Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301  USA
 
 """ Convert files to the file format generated by lyx 2.0"""
 
@@ -24,12 +24,13 @@ import unicodedata
 import sys, os
 
 from parser_tools import find_token, find_end_of, find_tokens, \
-  find_end_of_inset, find_end_of_layout, find_token_backwards, \
-  is_in_inset, get_value, get_quoted_value
+  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, revert_layout_command, hex2ratio, str2bool
+  revert_font_attrs, hex2ratio, str2bool
 
 ####################################################################
 # Private helper functions
@@ -44,17 +45,6 @@ def remove_option(lines, m, option):
     return True
 
 
-# DO NOT USE THIS ROUTINE ANY MORE. Better yet, replace the uses that
-# have been made of it with uses of put_cmd_in_ert.
-def old_put_cmd_in_ert(string):
-    for rep in unicode_reps:
-        string = string.replace(rep[1], rep[0].replace('\\\\', '\\'))
-    string = string.replace('\\', "\\backslash\n")
-    string = "\\begin_inset ERT\nstatus collapsed\n\\begin_layout Plain Layout\n" \
-      + string + "\n\\end_layout\n\\end_inset"
-    return string
-
-
 ###############################################################################
 ###
 ### Conversion and reversion routines
@@ -166,7 +156,7 @@ 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
@@ -213,7 +203,7 @@ def revert_xetex(document):
 
     # 2.) check font settings
     # defaults
-    roman = sans = typew = default
+    roman = sans = typew = "default"
     osf = False
     sf_scale = tt_scale = 100.0
     
@@ -287,26 +277,24 @@ def revert_xetex(document):
             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}')
     pretext.append('\usepackage{xunicode}')
     pretext.append('\usepackage{xltxtra}')
-    insert_to_preamble(0, document, pretext)
+    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 revert_backgroundcolor(document):
@@ -322,12 +310,11 @@ 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):
@@ -362,7 +349,7 @@ def revert_splitindex(document):
               preamble.append("\\newindex[" + iname + "]{" + ishortcut + "}")
         del document.header[i:k + 1]
     if preamble:
-        insert_to_preamble(0, document, preamble)
+        insert_to_preamble(document, preamble)
         
     # deal with index insets
     # these need to have the argument removed
@@ -486,14 +473,14 @@ 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):
@@ -501,28 +488,24 @@ 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):
@@ -544,13 +527,9 @@ 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):
@@ -568,8 +547,7 @@ def revert_nomencl_cwidth(document):
         continue
       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
 
 
@@ -632,8 +610,7 @@ 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]
 
 
@@ -650,8 +627,7 @@ def revert_percent_skip_lengths(document):
     # handle percent lengths
     percent, length = latex_length(length)
     if percent:
-        add_to_preamble(document, ["% this command was inserted by lyx2lyx"])
-        add_to_preamble(document, ["\\setlength{\\parskip}{" + length + "}"])
+        insert_to_preamble(document, ["\\setlength{\\parskip}{" + length + "}"])
         # set defskip to medskip as default
         document.header[i] = "\\defskip medskip"
 
@@ -691,7 +667,7 @@ 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)
@@ -722,7 +698,7 @@ 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)
@@ -763,7 +739,6 @@ def convert_author_id(document):
             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
         
@@ -785,7 +760,7 @@ def revert_author_id(document):
     " Remove the author_id from the \\author definition "
     i = 0
     anum = 0
-    rx = re.compile(r'(\\author)\s+(\d+)\s+(\".*\")\s*(.*)$')
+    rx = re.compile(r'(\\author)\s+(-?\d+)\s+(\".*\")\s*(.*)$')
     idmap = dict()
 
     while True:
@@ -826,11 +801,32 @@ def revert_suppress_date(document):
     # when suppress_date was true
     date = str2bool(get_value(document.header, "\\suppress_date", i))
     if date:
-        add_to_preamble(document, ["% this command was inserted by lyx2lyx"])
         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"
 
@@ -864,19 +860,15 @@ def revert_mhchem(document):
             i += 1
 
     if mhchem == "on":
-        pre = ["% lyx2lyx mhchem commands", 
-          "\\PassOptionsToPackage{version=3}{mhchem}", 
+        pre = ["\\PassOptionsToPackage{version=3}{mhchem}", 
           "\\usepackage{mhchem}"]
-        add_to_preamble(document, pre) 
+        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):
@@ -934,12 +926,8 @@ 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):
@@ -957,72 +945,135 @@ def revert_includeonly(document):
 
 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, '<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("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, 
-          ["% lyx2lyx multirow additions ", "\\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):
@@ -1167,7 +1218,7 @@ def revert_equalspacing_xymatrix(document):
           i = j + 1
   
     if has_equal_spacing and not has_preamble:
-        add_to_preamble(document, ['% lyx2lyx xymatrix addition', '\\usepackage[all]{xy}'])
+        add_to_preamble(document, ['\\usepackage[all]{xy}'])
 
 
 def revert_notefontcolor(document):
@@ -1177,22 +1228,22 @@ def revert_notefontcolor(document):
     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, and \renewcommand will throw an error
-        # since lyxgreyedout will not exist.
+        # no need to do anything else, and \renewcommand will throw 
+        # an error since lyxgreyedout will not exist.
         return
 
-    colorcode = get_value(document.header, '\\notefontcolor', i)
-    del document.header[i]
     # 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(0, document,
-      ['% Commands inserted by lyx2lyx to set the font color',
-        '% for greyed-out notes',
+    insert_to_preamble(document,
+      [ '%  for greyed-out notes',
         '\\@ifundefined{definecolor}{\\usepackage{color}}{}'
         '\\definecolor{note_fontcolor}{rgb}{%s,%s,%s}' % (red, green, blue),
         '\\renewenvironment{lyxgreyedout}',
@@ -1232,8 +1283,8 @@ def revert_fontcolor(document):
     green = hex2ratio(colorcode[3:5])
     blue = hex2ratio(colorcode[5:7])
     # write the preamble
-    insert_to_preamble(0, document,
-      ['% Commands inserted by lyx2lyx to set the font color',
+    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}'])
@@ -1251,8 +1302,8 @@ def revert_shadedboxcolor(document):
     green = hex2ratio(colorcode[3:5])
     blue = hex2ratio(colorcode[5:7])
     # write the preamble
-    insert_to_preamble(0, document,
-      ['% Commands inserted by lyx2lyx to set the color of boxes with shaded background',
+    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)])
 
@@ -1302,15 +1353,9 @@ 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):
@@ -1385,12 +1430,8 @@ def revert_html_quotes(document):
 
 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]
+  del_token(document.header, '\\output_sync_macro', 0)
+  del_token(document.header, '\\output_sync', 0)
 
 
 def revert_align_decimal(document):
@@ -1449,10 +1490,9 @@ def revert_makebox(document):
   " Convert \\makebox to TeX code "
   i = 0
   while 1:
-    # only revert frameless boxes without an inner box
-    i = find_token(document.body, '\\begin_inset Box Frameless', i)
+    i = find_token(document.body, '\\begin_inset Box', i)
     if i == -1:
-      return
+      break
     z = find_end_of_inset(document.body, i)
     if z == -1:
       document.warning("Malformed LyX document: Can't find end of box inset.")
@@ -1469,8 +1509,9 @@ def revert_makebox(document):
     if j == -1:
         i = z
         continue
-    val = get_value(document.body, 'use_makebox', j)
-    if val != "1":
+    
+    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
@@ -1518,15 +1559,15 @@ def convert_use_makebox(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):
   " 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",
@@ -1640,7 +1681,7 @@ def revert_nameref(document):
       document.body[stins:endins + 1] = newcontent
 
   if foundone:
-    add_to_preamble(document, "\usepackage{nameref}")
+    add_to_preamble(document, ["\usepackage{nameref}"])
 
 
 def remove_Nameref(document):
@@ -1668,7 +1709,7 @@ def revert_mathrsfs(document):
     i = 0
     for line in document.body:
       if line.find("\\mathscr{") != -1:
-        add_to_preamble(document, ["% lyx2lyx mathrsfs addition", "\\usepackage{mathrsfs}"])
+        add_to_preamble(document, ["\\usepackage{mathrsfs}"])
         return
 
 
@@ -1774,9 +1815,18 @@ def revert_flexnames(document):
 
 def convert_mathdots(document):
     " Load mathdots automatically "
-    i = find_token(document.header, "\\use_esint" , 0)
-    if i != -1:
-      document.header.insert(i + 1, "\\use_mathdots 1")
+    i = find_token(document.header, "\\use_mhchem" , 0)
+    if i == -1:
+        i = find_token(document.header, "\\use_esint" , 0)
+    if i == -1:
+        document.warning("Malformed LyX document: Can't find \\use_mhchem.")
+        return;
+    j = find_token(document.preamble, "\\usepackage{mathdots}", 0)
+    if j == -1:
+        document.header.insert(i + 1, "\\use_mathdots 0")
+    else:
+        document.header.insert(i + 1, "\\use_mathdots 2")
+        del document.preamble[j]
 
 
 def revert_mathdots(document):
@@ -1784,7 +1834,7 @@ def revert_mathdots(document):
 
     mathdots = find_token(document.header, "\\use_mathdots" , 0)
     if mathdots == -1:
-      document.warning("No \\usemathdots line. Assuming auto.")
+      document.warning("No \\use_mathdots line. Assuming auto.")
     else:
       val = get_value(document.header, "\\use_mathdots", mathdots)
       del document.header[mathdots]
@@ -1800,7 +1850,7 @@ def revert_mathdots(document):
         return
       if usedots == 2:
         # force load case
-        add_to_preamble(["% lyx2lyx mathdots addition", "\\usepackage{mathdots}"])
+        add_to_preamble(document, ["\\usepackage{mathdots}"])
         return
     
     # so we are in the auto case. we want to load mathdots if \iddots is used.
@@ -1816,8 +1866,7 @@ def revert_mathdots(document):
         continue
       code = "\n".join(document.body[i:j])
       if code.find("\\iddots") != -1:
-        add_to_preamble(document, ["% lyx2lyx mathdots addition", 
-        "\\@ifundefined{iddots}{\\usepackage{mathdots}}"])
+        add_to_preamble(document, ["\\@ifundefined{iddots}{\\usepackage{mathdots}}"])
         return
       i = j
 
@@ -1920,14 +1969,21 @@ def revert_diagram(document):
     if lines.find("\\Diagram") == -1:
       i = j
       continue
-    add_to_preamble(document, ["% lyx2lyx feyn package insertion ", "\\usepackage{feyn}"])
+    add_to_preamble(document, ["\\usepackage{feyn}"])
     # 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:
@@ -1988,6 +2044,415 @@ def convert_bibtex_clearpage(document):
     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
+
+
+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.
+
+        # 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)
+          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 + 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
+        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, '<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("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
 #
@@ -2018,7 +2483,7 @@ convert = [[346, []],
            [368, []],
            [369, [convert_author_id]],
            [370, []],
-           [371, []],
+           [371, [convert_mhchem]],
            [372, []],
            [373, [merge_gbrief]],
            [374, []],
@@ -2051,10 +2516,28 @@ 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, []],
+           [411, [convert_langpack]],
+           [412, []],
+           [413, []]
 ]
 
-revert =  [[403, [revert_refstyle]],
+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]],
            [401, []],
            [400, [revert_diagram]],