X-Git-Url: https://git.lyx.org/gitweb/?a=blobdiff_plain;f=lib%2Flyx2lyx%2Flyx_2_0.py;h=1771e185f45cde2dfb555aacb9dbf18ef4a58694;hb=96bcf145d92dc80bbc7ca3413be52e66d7d95327;hp=ff45486eb49a356b9d486464441ea638f284e03f;hpb=44279587e559b95088b33dca73bede7d4fce6328;p=lyx.git

diff --git a/lib/lyx2lyx/lyx_2_0.py b/lib/lyx2lyx/lyx_2_0.py
index ff45486eb4..1771e185f4 100644
--- a/lib/lyx2lyx/lyx_2_0.py
+++ b/lib/lyx2lyx/lyx_2_0.py
@@ -26,7 +26,7 @@ import sys, os
 from parser_tools import find_token, find_end_of, find_tokens, \
   find_token_exact, find_end_of_inset, find_end_of_layout, \
   find_token_backwards, is_in_inset, get_value, get_quoted_value, \
-  del_token, check_token
+  del_token, check_token, get_option_value
   
 from lyx2lyx_tools import add_to_preamble, insert_to_preamble, \
   put_cmd_in_ert, lyx2latex, latex_length, revert_flex_inset, \
@@ -539,7 +539,7 @@ def revert_nomencl_width(document):
         break
       j = find_end_of_inset(document.body, i)
       if not del_token(document.body, "set_width", i, j):
-            document.warning("Can't find set_width option for nomencl_print!")
+        document.warning("Can't find set_width option for nomencl_print!")
       i = j
 
 
@@ -750,7 +750,6 @@ 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
         
@@ -817,6 +816,28 @@ def revert_suppress_date(document):
     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"
 
@@ -940,64 +961,131 @@ def revert_includeall(document):
 
 def revert_multirow(document):
     " Revert multirow cells in tables to TeX-code"
-    i = 0
-    multirow = False
+
+    # first, let's find out if we need to do anything
+    # cell type 3 is multirow begin cell
+    i = find_token(document.body, '<cell multirow="3"', 0)
+    if i == -1:
+      return
+
+    add_to_preamble(document, ["\\usepackage{multirow}"])
+
+    begin_table = 0
     while True:
-      # cell type 3 is multirow begin cell
-      i = find_token(document.body, '<cell multirow="3"', i)
-      if i == -1:
-          break
-      # a multirow cell was found
-      multirow = True
-      # remove the multirow tag, set the valignment to top
-      # and remove the bottom line
-      # FIXME Are we sure these always have space around them?
-      document.body[i] = document.body[i].replace(' multirow="3" ', ' ')
-      document.body[i] = document.body[i].replace('valignment="middle"', 'valignment="top"')
-      document.body[i] = document.body[i].replace(' bottomline="true" ', ' ')
-      # write ERT to create the multirow cell
-      # use 2 rows and 2cm as default with because the multirow span
-      # and the column width is only hardly accessible
-      cend = find_token(document.body, "</cell>", i)
-      if cend == -1:
-          document.warning("Malformed LyX document: Could not find end of tabular cell.")
-          i += 1
-          continue
-      blay = find_token(document.body, "\\begin_layout", i, cend)
-      if blay == -1:
-          document.warning("Can't find layout for cell!")
-          i = j
-          continue
-      bend = find_end_of_layout(document.body, blay)
-      if blay == -1:
-          document.warning("Can't find end of layout for cell!")
-          i = cend
+        # find begin/end of table
+        begin_table = find_token(document.body, '<lyxtabular version=', begin_table)
+        if begin_table == -1:
+            break
+        end_table = find_end_of(document.body, begin_table, '<lyxtabular', '</lyxtabular>')
+        if end_table == -1:
+            document.warning("Malformed LyX document: Could not find end of table.")
+            begin_table += 1
+            continue
+        # does this table have multirow?
+        i = find_token(document.body, '<cell multirow="3"', begin_table, end_table)
+        if i == -1:
+            begin_table = end_table
+            continue
+        
+        # store the number of rows and columns
+        numrows = get_option_value(document.body[begin_table], "rows")
+        numcols = get_option_value(document.body[begin_table], "columns")
+        try:
+          numrows = int(numrows)
+          numcols = int(numcols)
+        except:
+          document.warning(numrows)
+          document.warning("Unable to determine rows and columns!")
+          begin_table = end_table
           continue
 
-      # do the later one first, so as not to mess up the numbering
-      # we are wrapping the whole cell in this ert
-      # so before the end of the layout...
-      document.body[bend:bend] = put_cmd_in_ert("}")
-      # ...and after the beginning
-      document.body[blay+1:blay+1] = put_cmd_in_ert("\\multirow{2}{2cm}{")
-
-      while True:
-          # cell type 4 is multirow part cell
-          k = find_token(document.body, '<cell multirow="4"', cend)
-          if k == -1:
+        mrstarts = []
+        multirows = []
+        # collect info on rows and columns of this table.
+        begin_row = begin_table
+        for row in range(numrows):
+            begin_row = find_token(document.body, '<row>', begin_row, end_table)
+            if begin_row == -1:
+              document.warning("Can't find row " + str(row + 1))
               break
-          # remove the multirow tag, set the valignment to top
-          # and remove the top line
-          # FIXME Are we sure these always have space around them?
-          document.body[k] = document.body[k].replace(' multirow="4" ', ' ')
-          document.body[k] = document.body[k].replace('valignment="middle"', 'valignment="top"')
-          document.body[k] = document.body[k].replace(' topline="true" ', ' ')
-          k += 1
-      # this will always be ok
-      i = cend
-
-    if multirow == True:
-        add_to_preamble(document, ["\\usepackage{multirow}"])
+            end_row = find_end_of(document.body, begin_row, '<row>', '</row>')
+            if end_row == -1:
+              document.warning("Can't find end of row " + str(row + 1))
+              break
+            begin_cell = begin_row
+            multirows.append([])
+            for column in range(numcols):            
+                begin_cell = find_token(document.body, '<cell ', begin_cell, end_row)
+                if begin_cell == -1:
+                  document.warning("Can't find column " + str(column + 1) + \
+                    "in row " + str(row + 1))
+                  break
+                # NOTE 
+                # this will fail if someone puts "</cell>" in a cell, but
+                # that seems fairly unlikely.
+                end_cell = find_end_of(document.body, begin_cell, '<cell', '</cell>')
+                if end_cell == -1:
+                  document.warning("Can't find end of column " + str(column + 1) + \
+                    "in row " + str(row + 1))
+                  break
+                multirows[row].append([begin_cell, end_cell, 0])
+                if document.body[begin_cell].find('multirow="3"') != -1:
+                  multirows[row][column][2] = 3 # begin multirow
+                  mrstarts.append([row, column])
+                elif document.body[begin_cell].find('multirow="4"') != -1:
+                  multirows[row][column][2] = 4 # in multirow
+                begin_cell = end_cell
+            begin_row = end_row
+        # end of table info collection
+
+        # work from the back to avoid messing up numbering
+        mrstarts.reverse()
+        for m in mrstarts:
+            row = m[0]
+            col = m[1]
+            # get column width
+            col_width = get_option_value(document.body[begin_table + 2 + col], "width")
+            # "0pt" means that no width is specified
+            if not col_width or col_width == "0pt":
+              col_width = "*"
+            # determine the number of cells that are part of the multirow
+            nummrs = 1
+            for r in range(row + 1, numrows):
+                if multirows[r][col][2] != 4:
+                  break
+                nummrs += 1
+                # take the opportunity to revert this line
+                lineno = multirows[r][col][0]
+                document.body[lineno] = document.body[lineno].\
+                  replace(' multirow="4" ', ' ').\
+                  replace('valignment="middle"', 'valignment="top"').\
+                  replace(' topline="true" ', ' ')
+                # remove bottom line of previous multirow-part cell
+                lineno = multirows[r-1][col][0]
+                document.body[lineno] = document.body[lineno].replace(' bottomline="true" ', ' ')
+            # revert beginning cell
+            bcell = multirows[row][col][0]
+            ecell = multirows[row][col][1]
+            document.body[bcell] = document.body[bcell].\
+              replace(' multirow="3" ', ' ').\
+              replace('valignment="middle"', 'valignment="top"')
+            blay = find_token(document.body, "\\begin_layout", bcell, ecell)
+            if blay == -1:
+              document.warning("Can't find layout for cell!")
+              continue
+            bend = find_end_of_layout(document.body, blay)
+            if bend == -1:
+              document.warning("Can't find end of layout for cell!")
+              continue
+            # do the later one first, so as not to mess up the numbering
+            # we are wrapping the whole cell in this ert
+            # so before the end of the layout...
+            document.body[bend:bend] = put_cmd_in_ert("}")
+            # ...and after the beginning
+            document.body[blay + 1:blay + 1] = \
+              put_cmd_in_ert("\\multirow{" + str(nummrs) + "}{" + col_width + "}{")
+
+        begin_table = end_table
 
 
 def convert_math_output(document):
@@ -1483,7 +1571,7 @@ 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):
@@ -1739,7 +1827,9 @@ def revert_flexnames(document):
 
 def convert_mathdots(document):
     " Load mathdots automatically "
-    i = find_token(document.header, "\\use_esint" , 0)
+    i = find_token(document.header, "\\use_mhchem" , 0)
+    if i == -1:
+      i = find_token(document.header, "\\use_esint" , 0)
     if i != -1:
       document.header.insert(i + 1, "\\use_mathdots 1")
 
@@ -1952,6 +2042,361 @@ 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.
+        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, '<cell multirow="3" mroffset=', 0)
+    if i == -1:
+      return
+
+    add_to_preamble(document, ["\\usepackage{multirow}"])
+
+    rgx = re.compile(r'mroffset="[^"]+?"')
+    begin_table = 0
+
+    while True:
+        # find begin/end of table
+        begin_table = find_token(document.body, '<lyxtabular version=', begin_table)
+        if begin_table == -1:
+            break
+        end_table = find_end_of(document.body, begin_table, '<lyxtabular', '</lyxtabular>')
+        if end_table == -1:
+            document.warning("Malformed LyX document: Could not find end of table.")
+            begin_table += 1
+            continue
+        # does this table have multirow?
+        i = find_token(document.body, '<cell multirow="3"', begin_table, end_table)
+        if i == -1:
+            begin_table = end_table
+            continue
+        
+        # store the number of rows and columns
+        numrows = get_option_value(document.body[begin_table], "rows")
+        numcols = get_option_value(document.body[begin_table], "columns")
+        try:
+          numrows = int(numrows)
+          numcols = int(numcols)
+        except:
+          document.warning(numrows)
+          document.warning("Unable to determine rows and columns!")
+          begin_table = end_table
+          continue
+
+        mrstarts = []
+        multirows = []
+        # collect info on rows and columns of this table.
+        begin_row = begin_table
+        for row in range(numrows):
+            begin_row = find_token(document.body, '<row>', begin_row, end_table)
+            if begin_row == -1:
+              document.warning("Can't find row " + str(row + 1))
+              break
+            end_row = find_end_of(document.body, begin_row, '<row>', '</row>')
+            if end_row == -1:
+              document.warning("Can't find end of row " + str(row + 1))
+              break
+            begin_cell = begin_row
+            multirows.append([])
+            for column in range(numcols):            
+                begin_cell = find_token(document.body, '<cell ', begin_cell, end_row)
+                if begin_cell == -1:
+                  document.warning("Can't find column " + str(column + 1) + \
+                    "in row " + str(row + 1))
+                  break
+                # NOTE 
+                # this will fail if someone puts "</cell>" in a cell, but
+                # that seems fairly unlikely.
+                end_cell = find_end_of(document.body, begin_cell, '<cell', '</cell>')
+                if end_cell == -1:
+                  document.warning("Can't find end of column " + str(column + 1) + \
+                    "in row " + str(row + 1))
+                  break
+                multirows[row].append([begin_cell, end_cell, 0])
+                if document.body[begin_cell].find('multirow="3" mroffset=') != -1:
+                  multirows[row][column][2] = 3 # begin multirow
+                  mrstarts.append([row, column])
+                elif document.body[begin_cell].find('multirow="4"') != -1:
+                  multirows[row][column][2] = 4 # in multirow
+                begin_cell = end_cell
+            begin_row = end_row
+        # end of table info collection
+
+        # work from the back to avoid messing up numbering
+        mrstarts.reverse()
+        for m in mrstarts:
+            row = m[0]
+            col = m[1]
+            # get column width
+            col_width = get_option_value(document.body[begin_table + 2 + col], "width")
+            # "0pt" means that no width is specified
+            if not col_width or col_width == "0pt":
+              col_width = "*"
+            # determine the number of cells that are part of the multirow
+            nummrs = 1
+            for r in range(row + 1, numrows):
+                if multirows[r][col][2] != 4:
+                  break
+                nummrs += 1
+                # take the opportunity to revert this line
+                lineno = multirows[r][col][0]
+                document.body[lineno] = document.body[lineno].\
+                  replace(' multirow="4" ', ' ').\
+                  replace('valignment="middle"', 'valignment="top"').\
+                  replace(' topline="true" ', ' ')
+                # remove bottom line of previous multirow-part cell
+                lineno = multirows[r-1][col][0]
+                document.body[lineno] = document.body[lineno].replace(' bottomline="true" ', ' ')
+            # revert beginning cell
+            bcell = multirows[row][col][0]
+            ecell = multirows[row][col][1]
+            offset = get_option_value(document.body[bcell], "mroffset")
+            document.body[bcell] = document.body[bcell].\
+              replace(' multirow="3" ', ' ').\
+              replace('valignment="middle"', 'valignment="top"')
+            # remove mroffset option
+            document.body[bcell] = rgx.sub('', document.body[bcell])
+            
+            blay = find_token(document.body, "\\begin_layout", bcell, ecell)
+            if blay == -1:
+              document.warning("Can't find layout for cell!")
+              continue
+            bend = find_end_of_layout(document.body, blay)
+            if bend == -1:
+              document.warning("Can't find end of layout for cell!")
+              continue
+            # do the later one first, so as not to mess up the numbering
+            # we are wrapping the whole cell in this ert
+            # so before the end of the layout...
+            document.body[bend:bend] = put_cmd_in_ert("}")
+            # ...and after the beginning
+            document.body[blay + 1:blay + 1] = \
+              put_cmd_in_ert("\\multirow{" + str(nummrs) + "}{" + col_width + "}[" \
+                  + offset + "]{")
+
+        # on to the next table
+        begin_table = end_table
+
+
+def revert_script(document):
+    " Convert subscript/superscript inset to TeX code "
+    i = 0
+    foundsubscript = False
+    while 1:
+        i = find_token(document.body, '\\begin_inset script', i)
+        if i == -1:
+            break
+        z = find_end_of_inset(document.body, i)
+        if z == -1:
+            document.warning("Malformed LyX document: Can't find end of script inset.")
+            i += 1
+            continue
+        blay = find_token(document.body, "\\begin_layout", i, z)
+        if blay == -1:
+            document.warning("Malformed LyX document: Can't find layout in script inset.")
+            i = z
+            continue
+
+        if check_token(document.body[i], "\\begin_inset script subscript"):
+            subst = '\\textsubscript{'
+            foundsubscript = True
+        elif check_token(document.body[i], "\\begin_inset script superscript"):
+            subst = '\\textsuperscript{'
+        else:
+            document.warning("Malformed LyX document: Unknown type of script inset.")
+            i = z
+            continue
+        bend = find_end_of_layout(document.body, blay)
+        if bend == -1 or bend > z:
+            document.warning("Malformed LyX document: Can't find end of layout in script inset.")
+            i = z
+            continue
+        # remove the \end_layout \end_inset pair
+        document.body[bend:z + 1] = put_cmd_in_ert("}")
+        document.body[i:blay + 1] = put_cmd_in_ert(subst)
+        i += 1
+    # these classes provide a \textsubscript command:
+    # FIXME: Would be nice if we could use the information of the .layout file here
+    classes = ["memoir", "scrartcl", "scrbook", "scrlttr2", "scrreprt"]
+    if foundsubscript and find_token_exact(classes, document.textclass, 0) == -1:
+        add_to_preamble(document, ['\\usepackage{subscript}'])
+
+
+def convert_use_xetex(document):
+    " convert \\use_xetex to \\use_non_tex_fonts "
+    i = 0
+    i = find_token(document.header, "\\use_xetex", 0)
+    if i == -1:
+        return
+    
+    val = get_value(document.header, "\\use_xetex", 0)
+    document.header[i] = "\\use_non_tex_fonts " + val
+
+
+def revert_use_xetex(document):
+    " revert \\use_non_tex_fonts to \\use_xetex "
+    i = 0
+    i = find_token(document.header, "\\use_non_tex_fonts", 0)
+    if i == -1:
+        document.warning("Malformed document. No \\use_non_tex_fonts param!")
+        return
+
+    val = get_value(document.header, "\\use_non_tex_fonts", 0)
+    document.header[i] = "\\use_xetex " + val
+
+
+def revert_labeling(document):
+    koma = ("scrartcl", "scrarticle-beamer", "scrbook", "scrlettr",
+        "scrlttr2", "scrreprt")
+    if document.textclass in koma:
+        return
+    i = 0
+    while True:
+        i = find_token_exact(document.body, "\\begin_layout Labeling", i)
+        if i == -1:
+            return
+        document.body[i] = "\\begin_layout List"
+
+
 ##
 # Conversion hub
 #
@@ -1982,7 +2427,7 @@ convert = [[346, []],
            [368, []],
            [369, [convert_author_id]],
            [370, []],
-           [371, []],
+           [371, [convert_mhchem]],
            [372, []],
            [373, [merge_gbrief]],
            [374, []],
@@ -2015,10 +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]],