pyx/pdfwriter.py

   1 #!/usr/bin/env python
   2 # -*- coding: ISO-8859-1 -*-
   3 #
   4 #
   5 # Copyright (C) 2005 Jörg Lehmann <joergl@users.sourceforge.net>
   6 # Copyright (C) 2005 André Wobst <wobsta@users.sourceforge.net>
   7 #
   8 # This file is part of PyX (http://pyx.sourceforge.net/).
   9 #
  10 # PyX is free software; you can redistribute it and/or modify
  11 # it under the terms of the GNU General Public License as published by
  12 # the Free Software Foundation; either version 2 of the License, or
  13 # (at your option) any later version.
  14 #
  15 # PyX is distributed in the hope that it will be useful,
  16 # but WITHOUT ANY WARRANTY; without even the implied warranty of
  17 # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  18 # GNU General Public License for more details.
  19 #
  20 # You should have received a copy of the GNU General Public License
  21 # along with PyX; if not, write to the Free Software
  22 # Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA
  23
  24 import copy, warnings, time
  25 try:
  26     import zlib
  27     haszlib = 1
  28 except:
  29     haszlib = 0
  30
  31 import unit, style, type1font, version
  32
  33 try:
  34     enumerate([])
  35 except NameError:
  36     # fallback implementation for Python 2.2 and below
  37     def enumerate(list):
  38         return zip(xrange(len(list)), list)
  39
  40
  41 class PDFregistry:
  42
  43     def __init__(self):
  44         self.types = {}
  45         # we need to keep the original order of the resources (for PDFcontentlength)
  46         self.resources = []
  47
  48     def add(self, resource):
  49         """ register resource, merging it with an already registered resource of the same type and id"""
  50         resources = self.types.setdefault(resource.type, {})
  51         if resources.has_key(resource.id):
  52             resources[resource.id].merge(resource)
  53         else:
  54             self.resources.append(resource)
  55             resources[resource.id] = resource
  56
  57     def getrefno(self, resource):
  58         return self.types[resource.type][resource.id].refno
  59
  60     def mergeregistry(self, registry):
  61         for resource in registry.resources:
  62             self.add(resource)
  63
  64     def write(self, file, writer, catalog):
  65         # first we set all refnos
  66         refno = 1
  67
  68         # we recursively inserted the resources such that the topmost resources in
  69         # the dependency tree of the resources come last. Hence, we need to
  70         # reverse the resources list before writing the output
  71         self.resources.reverse()
  72         for resource in self.resources:
  73             resource.refno = refno
  74             refno += 1
  75
  76         # second, all objects are written, keeping the positions in the output file
  77         fileposes = []
  78         for resource in self.resources:
  79             fileposes.append(file.tell())
  80             file.write("%i 0 obj\n" % resource.refno)
  81             resource.outputPDF(file, writer, self)
  82             file.write("endobj\n")
  83
  84         # xref
  85         xrefpos = file.tell()
  86         file.write("xref\n"
  87                    "0 %d\n"
  88                    "0000000000 65535 f \n" % refno)
  89
  90         for filepos in fileposes:
  91             file.write("%010i 00000 n \n" % filepos)
  92
  93         # trailer
  94         file.write("trailer\n"
  95                    "<<\n"
  96                    "/Size %i\n" % refno)
  97         file.write("/Root %i 0 R\n" % self.getrefno(catalog))
  98         file.write("/Info %i 0 R\n" % self.getrefno(catalog.PDFinfo))
  99         file.write(">>\n"
 100                    "startxref\n"
 101                    "%i\n" % xrefpos)
 102         file.write("%%EOF\n")
 103
 104
 105 class PDFobject:
 106
 107     def __init__(self, type, _id=None, pageresource=None, pageprocset=None):
 108         """create a PDFobject
 109           - type has to be a string describing the type of the object
 110           - _id is a unique identification used for the object if it is not None.
 111             Otherwise id(self) is used
 112           - If pageresource is not None, it has to be a string describing the name
 113             of the resource to be included in the resource dictionary of the pages
 114             including the PDFobject.
 115           - If pageprocset is not None, it has to be a string describing the name
 116             to be used in the ProcSet list of the pages including the PDFObject.
 117         """
 118         self.type = type
 119         if _id is None:
 120             self.id = id(self)
 121         else:
 122             self.id = _id
 123         self.pageresource = pageresource
 124         self.pageprocset = pageprocset
 125         self.refno = None
 126
 127     def merge(self, other):
 128         pass
 129
 130     def outputPDF(self, file, writer, registry):
 131         raise NotImplementedError("outputPDF method has to be provided by PDFobject subclass")
 132
 133
 134 class PDFcatalog(PDFobject):
 135
 136     def __init__(self, document, registry):
 137         PDFobject.__init__(self, "catalog")
 138         self.PDFpages = PDFpages(document, registry)
 139         registry.add(self.PDFpages)
 140         self.PDFinfo = PDFinfo()
 141         registry.add(self.PDFinfo)
 142
 143     def outputPDF(self, file, writer, registry):
 144         file.write("<<\n"
 145                    "/Type /Catalog\n"
 146                    "/Pages %i 0 R\n" % registry.getrefno(self.PDFpages))
 147         if writer.fullscreen:
 148             file.write("/PageMode /FullScreen\n")
 149         file.write(">>\n")
 150
 151
 152 class PDFinfo(PDFobject):
 153
 154     def __init__(self):
 155         PDFobject.__init__(self, "info")
 156
 157     def outputPDF(self, file, writer, registry):
 158         if time.timezone < 0:
 159             # divmod on positive numbers, otherwise the minutes have a different sign from the hours
 160             timezone = "-%02i'%02i'" % divmod(-time.timezone/60, 60)
 161         elif time.timezone > 0:
 162             timezone = "+%02i'%02i'" % divmod(time.timezone/60, 60)
 163         else:
 164             timezone = "Z00'00'"
 165
 166         def pdfstring(s):
 167             r = ""
 168             for c in s:
 169                 if 32 <= ord(c) <= 127 and c not in "()[]<>\\":
 170                     r += c
 171                 else:
 172                     r += "\\%03o" % ord(c)
 173             return r
 174
 175         file.write("<<\n")
 176         if writer.title:
 177             file.write("/Title (%s)\n" % pdfstring(writer.title))
 178         if writer.author:
 179             file.write("/Author (%s)\n" % pdfstring(writer.author))
 180         if writer.subject:
 181             file.write("/Subject (%s)\n" % pdfstring(writer.subject))
 182         if writer.keywords:
 183             file.write("/Keywords (%s)\n" % pdfstring(writer.keywords))
 184         file.write("/Creator (PyX %s)\n" % version.version)
 185         file.write("/CreationDate (D:%s%s)\n" % (time.strftime("%Y%m%d%H%M"), timezone))
 186         file.write(">>\n")
 187
 188
 189 class PDFpages(PDFobject):
 190
 191     def __init__(self, document, registry):
 192         PDFobject.__init__(self, "pages")
 193         self.PDFpagelist = []
 194         for pageno, page in enumerate(document.pages):
 195             page = PDFpage(page, pageno, self, registry)
 196             self.PDFpagelist.append(page)
 197         for i in range(len(self.PDFpagelist), 0, -1):
 198             registry.add(self.PDFpagelist[i-1])
 199
 200     def outputPDF(self, file, writer, registry):
 201         file.write("<<\n"
 202                    "/Type /Pages\n"
 203                    "/Kids [%s]\n"
 204                    "/Count %i\n"
 205                    ">>\n" % (" ".join(["%i 0 R" % registry.getrefno(page)
 206                                        for page in self.PDFpagelist]),
 207                              len(self.PDFpagelist)))
 208
 209
 210 class PDFpage(PDFobject):
 211
 212     def __init__(self, page, pageno, PDFpages, registry):
 213         PDFobject.__init__(self, "page", pageno)
 214         self.PDFpages = PDFpages
 215         self.page = page
 216
 217         # every page uses its own registry in order to find out which
 218         # resources are used within the page. However, the
 219         # pageregistry is also merged in the global registry
 220         self.pageregistry = PDFregistry()
 221
 222         self.bbox = page.bbox()
 223         self.pagetrafo = page.pagetrafo(self.bbox)
 224         if self.pagetrafo:
 225             self.transformedbbox = self.bbox.transformed(self.pagetrafo)
 226         else:
 227             self.transformedbbox = self.bbox
 228         self.PDFcontent = PDFcontent(page.canvas, self.pagetrafo, self.pageregistry)
 229         self.pageregistry.add(self.PDFcontent)
 230         self.page.canvas.registerPDF(self.pageregistry)
 231         registry.mergeregistry(self.pageregistry)
 232
 233     def outputPDF(self, file, writer, registry):
 234         file.write("<<\n"
 235                    "/Type /Page\n"
 236                    "/Parent %i 0 R\n" % registry.getrefno(self.PDFpages))
 237         paperformat = self.page.paperformat
 238         if paperformat:
 239             file.write("/MediaBox [0 0 %f %f]\n" % (unit.topt(paperformat.width), unit.topt(paperformat.height)))
 240         else:
 241             file.write("/MediaBox [%f %f %f %f]\n" % self.transformedbbox.highrestuple_pt())
 242         if self.transformedbbox and writer.writebbox:
 243             file.write("/CropBox [%f %f %f %f]\n" % self.transformedbbox.highrestuple_pt())
 244         procset = ["PDF"]
 245         resources = {}
 246         for type in self.pageregistry.types.keys():
 247             for resource in self.pageregistry.types[type].values():
 248                 if resource.pageprocset is not None and resource.pageprocset not in procset:
 249                     procset.append(resource.pageprocset)
 250                 if resource.pageresource is not None:
 251                     resources.setdefault(resource.pageresource, []).append(resource)
 252         file.write("/Resources <<\n"
 253                    "/ProcSet [ %s ]\n" % " ".join(["/%s" % p for p in procset]))
 254         for pageresource, resources in resources.items():
 255             file.write("/%s <<\n%s\n>>\n" % (pageresource, "\n".join(["/%s %i 0 R" % (resource.name, registry.getrefno(resource))
 256                                                                       for resource in resources])))
 257         file.write(">>\n")
 258         file.write("/Contents %i 0 R\n"
 259                    ">>\n" % registry.getrefno(self.PDFcontent))
 260
 261
 262 class PDFcontent(PDFobject):
 263
 264     def __init__(self, canvas, pagetrafo, registry):
 265         PDFobject.__init__(self, "content")
 266         self.canvas = canvas
 267         self.pagetrafo = pagetrafo
 268         self.contentlength = PDFcontentlength((self.type, self.id))
 269         registry.add(self.contentlength)
 270
 271     def outputPDF(self, file, writer, registry):
 272         file.write("<<\n"
 273                    "/Length %i 0 R\n" % registry.getrefno(self.contentlength))
 274         if writer.compress:
 275             file.write("/Filter /FlateDecode\n")
 276         file.write(">>\n"
 277                    "stream\n")
 278         beginstreampos = file.tell()
 279
 280         if writer.compress:
 281             stream = compressedstream(file, writer.compresslevel)
 282         else:
 283             stream = file
 284
 285         # XXX this should maybe be handled by the page since removing
 286         # this code would allow us to (nearly, since we also need to
 287         # set more info in the content dict) reuse PDFcontent for
 288         # patterns
 289         acontext = context()
 290         # apply a possible global transformation
 291         if self.pagetrafo:
 292             self.pagetrafo.outputPDF(stream, writer, acontext)
 293         style.linewidth.normal.outputPDF(stream, writer, acontext)
 294
 295         self.canvas.outputPDF(stream, writer, acontext)
 296         if writer.compress:
 297             stream.flush()
 298
 299         self.contentlength.contentlength = file.tell() - beginstreampos
 300         if writer.compress:
 301             file.write("\n")
 302         file.write("endstream\n")
 303
 304
 305 class PDFcontentlength(PDFobject):
 306
 307     def __init__(self, contentid):
 308         PDFobject.__init__(self, "_contentlength", contentid)
 309         self.contentlength = None
 310
 311     def outputPDF(self, file, writer, registry):
 312         # initially we do not know about the content length
 313         # -> it has to be written into the instance later on
 314         file.write("%d\n" % self.contentlength)
 315
 316
 317 class PDFfont(PDFobject):
 318
 319     def __init__(self, font, chars, registry):
 320         PDFobject.__init__(self, "font", font.name, "Font", "Text")
 321
 322         self.fontdescriptor = PDFfontdescriptor(font, chars, registry)
 323         registry.add(self.fontdescriptor)
 324
 325         if font.encoding:
 326             self.encoding = PDFencoding(font.encoding)
 327             registry.add(self.encoding)
 328         else:
 329             self.encoding = None
 330
 331         self.name = font.name
 332         self.basefontname = font.basefontname
 333         self.metric = font.metric
 334
 335     def outputPDF(self, file, writer, registry):
 336         file.write("<<\n"
 337                    "/Type /Font\n"
 338                    "/Subtype /Type1\n")
 339         file.write("/Name /%s\n" % self.name)
 340         file.write("/BaseFont /%s\n" % self.basefontname)
 341         if self.fontdescriptor.fontfile is not None and self.fontdescriptor.fontfile.usedchars is not None:
 342             usedchars = self.fontdescriptor.fontfile.usedchars
 343             firstchar = min(usedchars.keys())
 344             lastchar = max(usedchars.keys())
 345             file.write("/FirstChar %d\n" % firstchar)
 346             file.write("/LastChar %d\n" % lastchar)
 347             file.write("/Widths\n"
 348                        "[")
 349             for i in range(firstchar, lastchar+1):
 350                 if i and not (i % 8):
 351                     file.write("\n")
 352                 else:
 353                     file.write(" ")
 354                 if usedchars.has_key(i):
 355                     file.write("%f" % self.metric.getwidth_ds(i))
 356                 else:
 357                     file.write("0")
 358             file.write(" ]\n")
 359         else:
 360             file.write("/FirstChar 0\n"
 361                        "/LastChar 255\n"
 362                        "/Widths\n"
 363                        "[")
 364             for i in range(256):
 365                 if i and not (i % 8):
 366                     file.write("\n")
 367                 else:
 368                     file.write(" ")
 369                 try:
 370                     width = self.metric.getwidth_ds(i)
 371                 except (IndexError, AttributeError):
 372                     width = 0
 373                 file.write("%f" % width)
 374             file.write(" ]\n")
 375         file.write("/FontDescriptor %d 0 R\n" % registry.getrefno(self.fontdescriptor))
 376         if self.encoding:
 377             file.write("/Encoding %d 0 R\n" % registry.getrefno(self.encoding))
 378         file.write(">>\n")
 379
 380
 381 class PDFfontdescriptor(PDFobject):
 382
 383     def __init__(self, font, chars, registry):
 384         PDFobject.__init__(self, "fontdescriptor", font.basefontname)
 385
 386         if font.filename is None:
 387             self.fontfile = None
 388         else:
 389             self.fontfile = PDFfontfile(font.basefontname, font.filename, font.encoding, chars)
 390             registry.add(self.fontfile)
 391
 392         self.name = font.basefontname
 393         self.fontinfo = font.metric.fontinfo()
 394
 395     def outputPDF(self, file, writer, registry):
 396         file.write("<<\n"
 397                    "/Type /FontDescriptor\n"
 398                    "/FontName /%s\n" % self.name)
 399         if self.fontfile is None:
 400             file.write("/Flags 32\n")
 401         else:
 402             file.write("/Flags %d\n" % self.fontfile.getflags())
 403         file.write("/FontBBox [%d %d %d %d]\n" % self.fontinfo.fontbbox)
 404         file.write("/ItalicAngle %d\n" % self.fontinfo.italicangle)
 405         file.write("/Ascent %d\n" % self.fontinfo.ascent)
 406         file.write("/Descent %d\n" % self.fontinfo.descent)
 407         file.write("/CapHeight %d\n" % self.fontinfo.capheight)
 408         file.write("/StemV %d\n" % self.fontinfo.vstem)
 409         if self.fontfile is not None:
 410             file.write("/FontFile %d 0 R\n" % registry.getrefno(self.fontfile))
 411         file.write(">>\n")
 412
 413
 414 class PDFfontfile(PDFobject):
 415
 416     def __init__(self, name, filename, encoding, chars):
 417         PDFobject.__init__(self, "fontfile", filename)
 418         self.name = name
 419         self.filename = filename
 420         if encoding is None:
 421             self.encodingfilename = None
 422         else:
 423             self.encodingfilename = encoding.filename
 424         self.usedchars = {}
 425         for char in chars:
 426             self.usedchars[char] = 1
 427
 428         self.strip = 1
 429         self.font = None
 430
 431     def merge(self, other):
 432         if self.encodingfilename == other.encodingfilename:
 433             self.usedchars.update(other.usedchars)
 434         else:
 435             # TODO: need to resolve the encoding when several encodings are in the play
 436             self.strip = 0
 437
 438     def mkfontfile(self):
 439         import font.t1font
 440         self.font = font.t1font.T1pfbfont(self.filename)
 441
 442     def getflags(self):
 443         if self.font is None:
 444             self.mkfontfile()
 445         return self.font.getflags()
 446
 447     def outputPDF(self, file, writer, registry):
 448         if self.font is None:
 449             self.mkfontfile()
 450         if self.strip:
 451             # XXX: access to the encoding file
 452             if self.encodingfilename:
 453                 encodingfile = type1font.encodingfile(self.encodingfilename, self.encodingfilename)
 454                 usedglyphs = [encodingfile.decode(char)[1:] for char in self.usedchars.keys()]
 455             else:
 456                 self.font._encoding()
 457                 usedglyphs = [self.font.encoding.decode(char) for char in self.usedchars.keys()]
 458             strippedfont = self.font.getstrippedfont(usedglyphs)
 459         else:
 460             strippedfont = self.font
 461         strippedfont.outputPDF(file, writer)
 462
 463
 464 class PDFencoding(PDFobject):
 465
 466     def __init__(self, encoding):
 467         PDFobject.__init__(self, "encoding", encoding.name)
 468         self.encoding = encoding
 469
 470     def outputPDF(self, file, writer, registry):
 471         encodingfile = type1font.encodingfile(self.encoding.name, self.encoding.filename)
 472         encodingfile.outputPDF(file, writer, registry)
 473
 474
 475 class PDFwriter:
 476
 477     def __init__(self, document, filename,
 478                        title=None, author=None, subject=None, keywords=None,
 479                        fullscreen=0, writebbox=0, compress=1, compresslevel=6):
 480         if not filename.endswith(".pdf"):
 481             filename = filename + ".pdf"
 482         try:
 483             file = open(filename, "wb")
 484         except IOError:
 485             raise IOError("cannot open output file")
 486
 487         self.title = title
 488         self.author = author
 489         self.subject = subject
 490         self.keywords = keywords
 491         self.fullscreen = fullscreen
 492         self.writebbox = writebbox
 493         if compress and not haszlib:
 494             compress = 0
 495             warnings.warn("compression disabled due to missing zlib module")
 496         self.compress = compress
 497         self.compresslevel = compresslevel
 498
 499         file.write("%%PDF-1.4\n%%%s%s%s%s\n" % (chr(195), chr(182), chr(195), chr(169)))
 500
 501         # the PDFcatalog class automatically builds up the pdfobjects from a document
 502         registry = PDFregistry()
 503         catalog = PDFcatalog(document, registry)
 504         registry.add(catalog)
 505         registry.write(file, self, catalog)
 506         file.close()
 507
 508
 509 class compressedstream:
 510
 511     def __init__(self, file, compresslevel):
 512         self.file = file
 513         self.compressobj = zlib.compressobj(compresslevel)
 514
 515     def write(self, string):
 516         self.file.write(self.compressobj.compress(string))
 517
 518     def flush(self):
 519         self.file.write(self.compressobj.flush())
 520
 521
 522 class context:
 523
 524     def __init__(self):
 525         self.linewidth_pt = None
 526         # XXX there are both stroke and fill color spaces
 527         self.colorspace = None
 528         self.strokeattr = 1
 529         self.fillattr = 1
 530         self.font = None
 531         self.textregion = 0
 532
 533     def __call__(self, **kwargs):
 534         newcontext = copy.copy(self)
 535         for key, value in kwargs.items():
 536             setattr(newcontext, key, value)
 537         return newcontext