lib/lyx2lyx/lyxconvert_218.py

   1 # This file is part of lyx2lyx
   2 # Copyright (C) 2002 Dekel Tsur <dekel@lyx.org>
   3 #
   4 # This program is free software; you can redistribute it and/or
   5 # modify it under the terms of the GNU General Public License
   6 # as published by the Free Software Foundation; either version 2
   7 # of the License, or (at your option) any later version.
   8 #
   9 # This program is distributed in the hope that it will be useful,
  10 # but WITHOUT ANY WARRANTY; without even the implied warranty of
  11 # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  12 # GNU General Public License for more details.
  13 #
  14 # You should have received a copy of the GNU General Public License
  15 # along with this program; if not, write to the Free Software
  16 # Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA  02111-1307, USA.
  17
  18
  19 import sys,string,re
  20 from parser_tools import *
  21
  22 floats = {
  23     "footnote": ["\\begin_inset Foot",
  24                  "collapsed true"],
  25     "margin":   ["\\begin_inset Marginal",
  26                  "collapsed true"],
  27     "fig":      ["\\begin_inset Float figure",
  28                  "wide false",
  29                  "collapsed false"],
  30     "tab":      ["\\begin_inset Float table",
  31                  "wide false",
  32                  "collapsed false"],
  33     "alg":      ["\\begin_inset Float algorithm",
  34                  "wide false",
  35                  "collapsed false"],
  36     "wide-fig": ["\\begin_inset Float figure",
  37                  "wide true",
  38                  "collapsed false"],
  39     "wide-tab": ["\\begin_inset Float table",
  40                  "wide true",
  41                  "collapsed false"]
  42 }
  43
  44 font_tokens = ["\\family", "\\series", "\\shape", "\\size", "\\emph",
  45                "\\bar", "\\noun", "\\color", "\\lang", "\\latex"]
  46
  47 #
  48 # Change \begin_float .. \end_float into \begin_inset Float .. \end_inset
  49 #
  50
  51 pextra_type3_rexp = re.compile(r".*\\pextra_type\s+3")
  52 pextra_rexp = re.compile(r"\\pextra_type\s+(\S+)"+\
  53                          r"(\s+\\pextra_alignment\s+(\S+))?"+\
  54                          r"(\s+\\pextra_hfill\s+(\S+))?"+\
  55                          r"(\s+\\pextra_start_minipage\s+(\S+))?"+\
  56                          r"(\s+(\\pextra_widthp?)\s+(\S*))?")
  57
  58 def get_width(mo):
  59     if mo.group(9) == "\\pextra_widthp":
  60         return mo.group(10)+"col%"
  61     elif mo.group(10):
  62         return mo.group(10)
  63     else:
  64         return "100col%"
  65
  66 def remove_oldfloat(lines, language):
  67     i = 0
  68     while 1:
  69         i = find_token(lines, "\\begin_float", i)
  70         if i == -1:
  71             break
  72         # There are no nested floats, so finding the end of the float is simple
  73         j = find_token(lines, "\\end_float", i+1)
  74
  75         floattype = string.split(lines[i])[1]
  76         if not floats.has_key(floattype):
  77             sys.stderr.write("Error! Unknown float type "+floattype+"\n")
  78             floattype = "fig"
  79
  80         # skip \end_deeper tokens
  81         i2 = i+1
  82         while check_token(lines[i2], "\\end_deeper"):
  83             i2 = i2+1
  84         if i2 > i+1:
  85             j2 = get_next_paragraph(lines, j+1)
  86             lines[j2:j2] = ["\\end_deeper "]*(i2-(i+1))
  87
  88         new = floats[floattype]+[""]
  89
  90         # Check if the float is floatingfigure
  91         k = find_re(lines, pextra_type3_rexp, i, j)
  92         if k != -1:
  93             mo = pextra_rexp.search(lines[k])
  94             width = get_width(mo)
  95             lines[k] = re.sub(pextra_rexp, "", lines[k])
  96             new = ["\\begin_inset Wrap figure",
  97                    'width "%s"' % width,
  98                    "collapsed false",
  99                    ""]
 100
 101         new = new+lines[i2:j]+["\\end_inset ", ""]
 102
 103         # After a float, all font attributes are reseted.
 104         # We need to output '\foo default' for every attribute foo
 105         # whose value is not default before the float.
 106         # The check here is not accurate, but it doesn't matter
 107         # as extra '\foo default' commands are ignored.
 108         # In fact, it might be safer to output '\foo default' for all
 109         # font attributes.
 110         k = get_paragraph(lines, i)
 111         flag = 0
 112         for token in font_tokens:
 113             if find_token(lines, token, k, i) != -1:
 114                 if not flag:
 115                     # This is not necessary, but we want the output to be
 116                     # as similar as posible to the lyx format
 117                     flag = 1
 118                     new.append("")
 119                 if token == "\\lang":
 120                     new.append(token+" "+language)
 121                 else:
 122                     new.append(token+" default ")
 123
 124         lines[i:j+1] = new
 125         i = i+1
 126
 127 pextra_type2_rexp = re.compile(r".*\\pextra_type\s+2")
 128 pextra_type2_rexp2 = re.compile(r".*(\\layout|\\pextra_type\s+2)")
 129
 130 def remove_oldminipage(lines):
 131     i = 0
 132     flag = 0
 133     while 1:
 134         i = find_re(lines, pextra_type2_rexp, i)
 135         if i == -1:
 136             break
 137
 138         mo = pextra_rexp.search(lines[i])
 139         position = mo.group(3)
 140         hfill = mo.group(5)
 141         width = get_width(mo)
 142         lines[i] = re.sub(pextra_rexp, "", lines[i])
 143
 144         start = ["\\begin_inset Minipage",
 145                  "position " + position,
 146                  "inner_position 0",
 147                  'height "0pt"',
 148                  'width "%s"' % width,
 149                  "collapsed false"
 150                  ]
 151         if flag:
 152             flag = 0
 153             if hfill:
 154                 start = ["","\hfill",""]+start
 155         else:
 156             start = ["\\layout Standard"] + start
 157
 158         j = find_token_backwards(lines,"\\layout", i-1)
 159         j0 = j
 160         mid = lines[j:i]
 161
 162         j = find_tokens(lines, ["\\layout", "\\end_float"], i+1)
 163         # j can be -1, but this is still ok
 164         mid = mid+lines[i+1:j]
 165
 166         count = 0
 167         while 1:
 168             # collect more paragraphs to the minipage
 169             count = count+1
 170             if j == -1 or not check_token(lines[j], "\\layout"):
 171                 break
 172             i = find_re(lines, pextra_type2_rexp2, j+1)
 173             if i == -1:
 174                 break
 175             mo = pextra_rexp.search(lines[i])
 176             if not mo:
 177                 break
 178             if mo.group(7) == "1":
 179                 flag = 1
 180                 break
 181             j = find_token_backwards(lines,"\\layout", i-1)
 182             mid = mid+lines[j:i]
 183             j = find_tokens(lines, ["\\layout", "\\end_float"], i+1)
 184             mid = mid+lines[i+1:j]
 185
 186         end = ["\\end_inset "]
 187
 188         lines[j0:j] = start+mid+end
 189         i = i+1
 190
 191 def is_empty(lines):
 192     return filter(is_nonempty_line, lines) == []
 193
 194 font_rexp =  re.compile(r"\\(family|series|shape|size|emph|numeric|bar|noun)")
 195 ert_rexp = re.compile(r"\\begin_inset|.*\\SpecialChar")
 196 spchar_rexp = re.compile(r"(.*)(\\SpecialChar.*)")
 197 ert_begin = ["\\begin_inset ERT",
 198              "status Collapsed",
 199              "",
 200              "\\layout Standard"]
 201
 202 def remove_oldert(lines):
 203     i = 0
 204     while 1:
 205         i = find_tokens(lines, ["\\latex latex", "\\layout LaTeX"], i)
 206         if i == -1:
 207             break
 208         j = i+1
 209         while 1:
 210             j = find_tokens(lines, ["\\latex default", "\\begin_inset", "\\layout", "\\end_float", "\\the_end"],
 211                             j)
 212             if check_token(lines[j], "\\begin_inset"):
 213                 j = find_end_of_inset(lines, j)
 214             else:
 215                 break
 216
 217         if check_token(lines[j], "\\layout"):
 218             while j-1 >= 0 and check_token(lines[j-1], "\\begin_deeper"):
 219                 j = j-1
 220
 221         # We need to remove insets, special chars & font commands from ERT text
 222         new = []
 223         new2 = []
 224         if check_token(lines[i], "\\layout LaTeX"):
 225             new = ["\layout Standard", "", ""]
 226             # We have a problem with classes in which Standard is not the default layout!
 227
 228         k = i+1
 229         while 1:
 230             k2 = find_re(lines, ert_rexp, k, j)
 231             inset = specialchar = 0
 232             if k2 == -1:
 233                 k2 = j
 234             elif check_token(lines[k2], "\\begin_inset"):
 235                 inset = 1
 236             else:
 237                 specialchar = 1
 238                 mo = spchar_rexp.match(lines[k2])
 239                 lines[k2] = mo.group(1)
 240                 specialchar_str = mo.group(2)
 241                 k2 = k2+1
 242
 243             tmp = []
 244             for line in lines[k:k2]:
 245                 if font_rexp.match(line):
 246                     if new2 == []:
 247                         # This is not necessary, but we want the output to be
 248                         # as similar as posible to the lyx format
 249                         new2 = [""]
 250                     new2.append(line)
 251                 elif not check_token(line, "\\latex"):
 252                     tmp.append(line)
 253
 254             if is_empty(tmp):
 255                 if filter(lambda x:x != "", tmp) != []:
 256                     if new == []:
 257                         # This is not necessary, but we want the output to be
 258                         # as similar as posible to the lyx format
 259                         lines[i-1] = lines[i-1]+" "
 260                     else:
 261                         new = new+[" "]
 262             else:
 263                 new = new+ert_begin+tmp+["\\end_inset ", ""]
 264
 265             if inset:
 266                 k3 = find_end_of_inset(lines, k2)
 267                 new = new+[""]+lines[k2:k3+1]+[""] # Put an empty line after \end_inset
 268                 k = k3+1
 269                 # Skip the empty line after \end_inset
 270                 if not is_nonempty_line(lines[k]):
 271                     k = k+1
 272                     new.append("")
 273             elif specialchar:
 274                 if new == []:
 275                     # This is not necessary, but we want the output to be
 276                     # as similar as posible to the lyx format
 277                     lines[i-1] = lines[i-1]+specialchar_str
 278                     new = [""]
 279                 else:
 280                     new = new+[specialchar_str, ""]
 281                 k = k2
 282             else:
 283                 break
 284
 285         new = new+new2
 286         if not check_token(lines[j], "\\latex "):
 287             new = new+[""]+[lines[j]]
 288         lines[i:j+1] = new
 289         i = i+1
 290
 291     i = 0
 292     while 1:
 293         i = find_token(lines, "\\latex ", i)
 294         if i == -1:
 295             break
 296         del lines[i]
 297
 298
 299 def remove_oldertinset(lines):
 300     i = 0
 301     while 1:
 302         i = find_token(lines, "\\begin_inset ERT", i)
 303         if i == -1:
 304             break
 305         j = find_end_of_inset(lines, i)
 306         k = find_token(lines, "\\layout", i+1)
 307         l = get_paragraph(lines, i)
 308         if lines[k] == lines[l]: # same layout
 309             k = k+1
 310         new = lines[k:j]
 311         lines[i:j+1] = new
 312         i = i+1
 313
 314 def is_ert_paragraph(lines, i):
 315     i = find_nonempty_line(lines, i+1)
 316     if not check_token(lines[i], "\\begin_inset ERT"):
 317         return 0
 318     j = find_end_of_inset(lines, i)
 319     k = find_nonempty_line(lines, j+1)
 320     return check_token(lines[k], "\\layout")
 321
 322 def combine_ert(lines):
 323     i = 0
 324     while 1:
 325         i = find_token(lines, "\\begin_inset ERT", i)
 326         if i == -1:
 327             break
 328         j = find_token_backwards(lines,"\\layout", i-1)
 329         count = 0
 330         text = []
 331         while is_ert_paragraph(lines, j):
 332
 333             count = count+1
 334             i2 = find_token(lines, "\\layout", j+1)
 335             k = find_token(lines, "\\end_inset", i2+1)
 336             text = text+lines[i2:k]
 337             j = find_token(lines, "\\layout", k+1)
 338             if j == -1:
 339                 break
 340
 341         if count >= 2:
 342             j = find_token(lines, "\\layout", i+1)
 343             lines[j:k] = text
 344
 345         i = i+1
 346
 347 oldunits = ["pt", "cm", "in", "text%", "col%"]
 348
 349 def get_length(lines, name, start, end):
 350     i = find_token(lines, name, start, end)
 351     if i == -1:
 352         return ""
 353     x = string.split(lines[i])
 354     return x[2]+oldunits[int(x[1])]
 355
 356 def write_attribute(x, token, value):
 357     if value != "":
 358         x.append("\t"+token+" "+value)
 359
 360 def remove_figinset(lines):
 361     i = 0
 362     while 1:
 363         i = find_token(lines, "\\begin_inset Figure", i)
 364         if i == -1:
 365             break
 366         j = find_end_of_inset(lines, i)
 367
 368         lyxwidth = string.split(lines[i])[3]+"pt"
 369         lyxheight = string.split(lines[i])[4]+"pt"
 370
 371         filename = get_value(lines, "file", i+1, j)
 372
 373         width = get_length(lines, "width", i+1, j)
 374         # what does width=5 mean ?
 375         height = get_length(lines, "height", i+1, j)
 376         rotateAngle = get_value(lines, "angle", i+1, j)
 377         if width == "" and height == "":
 378             size_type = "0"
 379         else:
 380             size_type = "1"
 381
 382         flags = get_value(lines, "flags", i+1, j)
 383         x = int(flags)%4
 384         if x == 1:
 385             display = "monochrome"
 386         elif x == 2:
 387             display = "gray"
 388         else:
 389             display = "color"
 390
 391         subcaptionText = get_value(lines, "subcaption", i+1, j)
 392         if subcaptionText != "":
 393             subcaptionText = '"'+subcaptionText+'"'
 394         k = find_token(lines, "subfigure", i+1,j)
 395         if k == -1:
 396             subcaption = 0
 397         else:
 398             subcaption = 1
 399
 400         new = ["\\begin_inset Graphics FormatVersion 1"]
 401         write_attribute(new, "filename", filename)
 402         write_attribute(new, "display", display)
 403         if subcaption:
 404             new.append("\tsubcaption")
 405         write_attribute(new, "subcaptionText", subcaptionText)
 406         write_attribute(new, "size_type", size_type)
 407         write_attribute(new, "width", width)
 408         write_attribute(new, "height", height)
 409         if rotateAngle != "":
 410             new.append("\trotate")
 411             write_attribute(new, "rotateAngle", rotateAngle)
 412         write_attribute(new, "rotateOrigin", "leftBaseline")
 413         write_attribute(new, "lyxsize_type", "1")
 414         write_attribute(new, "lyxwidth", lyxwidth)
 415         write_attribute(new, "lyxheight", lyxheight)
 416         new = new + ["\end_inset"]
 417         lines[i:j+1] = new
 418
 419 attr_re = re.compile(r' \w*="(false|0|)"')
 420 line_re = re.compile(r'<(features|column|row|cell)')
 421
 422 def update_tabular(lines):
 423     i = 0
 424     while 1:
 425         i = find_token(lines, '\\begin_inset  Tabular', i)
 426         if i == -1:
 427             break
 428
 429         for k in get_tabular_lines(lines, i):
 430             if check_token(lines[k], "<lyxtabular"):
 431                 lines[k] = string.replace(lines[k], 'version="2"', 'version="3"')
 432             elif check_token(lines[k], "<column"):
 433                 lines[k] = string.replace(lines[k], 'width=""', 'width="0pt"')
 434
 435             if line_re.match(lines[k]):
 436                 lines[k] = re.sub(attr_re, "", lines[k])
 437
 438         i = i+1
 439
 440 def change_preamble(lines):
 441     i = find_token(lines, "\\use_amsmath", 0)
 442     if i == -1:
 443         return
 444     lines[i+1:i+1] = ["\\use_natbib 0",
 445                       "\use_numerical_citations 0"]
 446
 447 def convert(header, body):
 448     language = get_value(header, "\\language", 0)
 449     if language == "":
 450         language = "english"
 451
 452     change_preamble(header)
 453     update_tabular(body)
 454     remove_oldminipage(body)
 455     remove_oldfloat(body, language)
 456     remove_figinset(body)
 457     remove_oldertinset(body)
 458     remove_oldert(body)
 459     combine_ert(body)
 460
 461 if __name__ == "__main__":
 462     pass