Welcome, guest | Sign In | My Account | Store | Cart

This is a rewrite of my text2pdf converter posted as recipe #189858, which can be used to convert pure text files to PDF.

Python, 601 lines
  1
  2
  3
  4
  5
  6
  7
  8
  9
 10
 11
 12
 13
 14
 15
 16
 17
 18
 19
 20
 21
 22
 23
 24
 25
 26
 27
 28
 29
 30
 31
 32
 33
 34
 35
 36
 37
 38
 39
 40
 41
 42
 43
 44
 45
 46
 47
 48
 49
 50
 51
 52
 53
 54
 55
 56
 57
 58
 59
 60
 61
 62
 63
 64
 65
 66
 67
 68
 69
 70
 71
 72
 73
 74
 75
 76
 77
 78
 79
 80
 81
 82
 83
 84
 85
 86
 87
 88
 89
 90
 91
 92
 93
 94
 95
 96
 97
 98
 99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
#! /usr/bin/env python
"""
 pyText2Pdf - Python script to convert plain text files into Adobe
 Acrobat PDF files with support for arbitrary page breaks etc.

 Version 2.0

 Author: Anand B Pillai <abpillai at gmail dot com>
    
"""

# Derived from http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/189858

import sys, os
import string
import time
import optparse
import re

LF_EXTRA=0
LINE_END='\015'
# form feed character (^L)
FF=chr(12)

ENCODING_STR = """\
/Encoding <<
/Differences [ 0 /.notdef /.notdef /.notdef /.notdef
/.notdef /.notdef /.notdef /.notdef /.notdef /.notdef
/.notdef /.notdef /.notdef /.notdef /.notdef /.notdef
/.notdef /.notdef /.notdef /.notdef /.notdef /.notdef
/.notdef /.notdef /.notdef /.notdef /.notdef /.notdef
/.notdef /.notdef /.notdef /.notdef /space /exclam
/quotedbl /numbersign /dollar /percent /ampersand
/quoteright /parenleft /parenright /asterisk /plus /comma
/hyphen /period /slash /zero /one /two /three /four /five
/six /seven /eight /nine /colon /semicolon /less /equal
/greater /question /at /A /B /C /D /E /F /G /H /I /J /K /L
/M /N /O /P /Q /R /S /T /U /V /W /X /Y /Z /bracketleft
/backslash /bracketright /asciicircum /underscore
/quoteleft /a /b /c /d /e /f /g /h /i /j /k /l /m /n /o /p
/q /r /s /t /u /v /w /x /y /z /braceleft /bar /braceright
/asciitilde /.notdef /.notdef /.notdef /.notdef /.notdef
/.notdef /.notdef /.notdef /.notdef /.notdef /.notdef
/.notdef /.notdef /.notdef /.notdef /.notdef /.notdef
/dotlessi /grave /acute /circumflex /tilde /macron /breve
/dotaccent /dieresis /.notdef /ring /cedilla /.notdef
/hungarumlaut /ogonek /caron /space /exclamdown /cent
/sterling /currency /yen /brokenbar /section /dieresis
/copyright /ordfeminine /guillemotleft /logicalnot /hyphen
/registered /macron /degree /plusminus /twosuperior
/threesuperior /acute /mu /paragraph /periodcentered
/cedilla /onesuperior /ordmasculine /guillemotright
/onequarter /onehalf /threequarters /questiondown /Agrave
/Aacute /Acircumflex /Atilde /Adieresis /Aring /AE
/Ccedilla /Egrave /Eacute /Ecircumflex /Edieresis /Igrave
/Iacute /Icircumflex /Idieresis /Eth /Ntilde /Ograve
/Oacute /Ocircumflex /Otilde /Odieresis /multiply /Oslash
/Ugrave /Uacute /Ucircumflex /Udieresis /Yacute /Thorn
/germandbls /agrave /aacute /acircumflex /atilde /adieresis
/aring /ae /ccedilla /egrave /eacute /ecircumflex
/edieresis /igrave /iacute /icircumflex /idieresis /eth
/ntilde /ograve /oacute /ocircumflex /otilde /odieresis
/divide /oslash /ugrave /uacute /ucircumflex /udieresis
/yacute /thorn /ydieresis ]
>>
"""

INTRO="""\
%prog [options] filename

PyText2Pdf  makes a 7-bit clean PDF file from any input file.

It reads from a named file, and writes the PDF file to a file specified by
the user, otherwise to a file with '.pdf' appended to the input file.

Author: Anand B Pillai."""


class PyText2Pdf(object):
    """ Text2pdf converter in pure Python """
    
    def __init__(self):
        # version number
        self._version="1.3"
        # iso encoding flag
        self._IsoEnc=False
        # formfeeds flag
        self._doFFs=False
        self._progname="PyText2Pdf"
        self._appname = " ".join((self._progname,str(self._version)))
        # default font
        self._font="/Courier"
        # default font size
        self._ptSize=10
        # default vert space
        self._vertSpace=12
        self._lines=0
        # number of characters in a row
        self._cols=80
        self._columns=1
        # page ht
        self._pageHt=792
        # page wd
        self._pageWd=612
        # input file 
        self._ifile=""
        # output file 
        self._ofile=""
        # default tab width
        self._tab=4
        # input file descriptor
        self._ifs=None
        # output file descriptor
        self._ofs=None
        # landscape flag
        self._landscape=False
        # Subject
        self._subject = ''
        # Author
        self._author = ''
        # Keywords
        self._keywords = []
        # Custom regexp  for page breaks
        self._pagebreakre = None
        
        # marker objects
        self._curobj = 5
        self._pageObs = [0]
        self._locations = [0,0,0,0,0,0]
        self._pageNo=0

        # file position marker
        self._fpos=0

    def parse_args(self):
        
        """ Callback function called by argument parser.
        Helps to remove duplicate code """

        if len(sys.argv)<2:
            sys.argv.append('-h')
            
        parser = optparse.OptionParser(usage=INTRO)
        parser.add_option('-o','--output',dest='outfile',help='Direct output to file OUTFILE',metavar='OUTFILE')
        parser.add_option('-f','--font',dest='font',help='Use Postscript font FONT (must be in standard 14, default: Courier)',
                          default='Courier')
        parser.add_option('-I','--isolatin',dest='isolatin',help='Use ISO latin-1 encoding',default=False,action='store_true')
        parser.add_option('-s','--size',dest='fontsize',help='Use font at PTSIZE points (default=>10)',metavar='PTSIZE',default=10)
        parser.add_option('-v','--linespace',dest='linespace',help='Use line spacing LINESPACE (deault 12)',metavar='LINESPACE',default=12)
        parser.add_option('-l','--lines',dest='lines',help='Lines per page (default 60, determined automatically if unspecified)',default=60, metavar=None)
        parser.add_option('-c','--chars',dest='chars',help='Maximum characters per line (default 80)',default=80,metavar=None)
        parser.add_option('-t','--tab',dest='tabspace',help='Spaces per tab character (default 4)',default=4,metavar=None)
        parser.add_option('-F','--ignoreff',dest='formfeed',help='Ignore formfeed character ^L (i.e, accept formfeed characters as pagebreaks)',default=False,action='store_true')
        parser.add_option('-P','--papersize',dest='papersize',help='Set paper size (default is letter, accepted values are "A4" or "A3")')
        parser.add_option('-W','--width',dest='width',help='Independent paper width in points',metavar=None,default=612)
        parser.add_option('-H','--height',dest='height',help='Independent paper height in points',metavar=None,default=792)
        parser.add_option('-2','--twocolumns',dest='twocolumns',help='Format as two columns',metavar=None,default=False,action='store_true')
        parser.add_option('-L','--landscape',dest='landscape',help='Format in landscape mode',metavar=None,default=False,action='store_true')
        parser.add_option('-R','--regexp',dest='pageregexp',help='Regular expression string to determine page breaks (if supplied, this will be used to split text into pages, instead of using line count)',metavar=None)
        parser.add_option('-S','--subject',dest='subject',help='Optional subject for the document',metavar=None)
        parser.add_option('-A','--author',dest='author',help='Optional author for the document',metavar=None)
        parser.add_option('-K','--keywords',dest='keywords',help='Optional list of keywords for the document (separated by commas)',metavar=None)
        

        optlist, args = parser.parse_args()
        # print optlist.__dict__, args

        if len(args)==0:
            sys.exit('Error: input file argument missing')
        elif len(args)>1:
            sys.exit('Error: Too many arguments')            

        self._ifile = args[0]
        
        d = optlist.__dict__
        if d.get('isolatin'): self._IsoEnc=True
        if d.get('formfeed'): self._doFFs = True
        if d.get('twocolumns'): self._columns = 2
        if d.get('landscape'): self._landscape = True

        self._font = '/' + d.get('font')
        psize = d.get('papersize')
        if psize=='A4':
            self._pageWd=595
            self._pageHt=842
        elif psize=='A3':
            self._pageWd=842
            self._pageHt=1190

        fsize = int(d.get('fontsize'))
        if fsize < 1: fsize = 1
        self._ptSize = fsize

        lspace = int(d.get('linespace'))
        if lspace<1: lspace = 1
        self._vertSpace = lspace

        lines = int(d.get('lines'))
        if lines<1: lines = 1
        self._lines = int(lines)

        chars = int(d.get('chars'))
        if chars<4: chars = 4
        self._cols = chars

        tab = int(d.get('tabspace'))
        if tab<1: tab = 1
        self._tab = tab

        w = int(d.get('width'))
        if w<72: w=72
        self._pageWd = w

        h = int(d.get('height'))
        if h<72: h=72
        self._pageHt = h

        # Very optional args
        author = d.get('author')
        if author: self._author = author

        subject = d.get('subject')
        if subject: self._subject = subject

        keywords = d.get('keywords')
        if keywords:
            self._keywords = keywords.split(',')

        pagebreak = d.get('pageregexp')
        if pagebreak:
            self._pagebreakre = re.compile(pagebreak, re.UNICODE|re.IGNORECASE)
        
        outfile = d.get('outfile')
        if outfile: self._ofile = outfile
        
        if self._landscape:
            print 'Landscape option on...'
        if self._columns==2:
            print 'Printing in two columns...'
        if self._doFFs:
            print 'Ignoring form feed character...'
        if self._IsoEnc:
            print 'Using ISO Latin Encoding...'

        print 'Using font',self._font[1:],'size =', self._ptSize

    def writestr(self, str):
        """ Write string to output file descriptor.
        All output operations go through this function.
        We keep the current file position also here"""

        # update current file position
        self._fpos += len(str)
        for x in range(0, len(str)):
            if str[x] == '\n':
                self._fpos += LF_EXTRA
        try:
            self._ofs.write(str)
        except IOError, e:
            print e
            return -1

        return 0
            
    def convert(self):
        """ Perform the actual conversion """
    
        if self._landscape:
            # swap page width & height
            tmp = self._pageHt
            self._pageHt = self._pageWd
            self._pageWd = tmp

        if self._lines==0:
            self._lines = (self._pageHt - 72)/self._vertSpace
        if self._lines < 1:
            self._lines=1
        
        try:
            self._ifs=open(self._ifile)
        except IOError, (strerror, errno):
            print 'Error: Could not open file to read --->', self._ifile
            sys.exit(3)

        if self._ofile=="":
            self._ofile = os.path.splitext(self._ifile)[0] + '.pdf'

        try:
            self._ofs = open(self._ofile, 'wb')
        except IOError, (strerror, errno):
            print 'Error: Could not open file to write --->', self._ofile
            sys.exit(3)

        print 'Input file=>',self._ifile
        print 'Writing pdf file',self._ofile, '...'
        self.writeheader()
        self.writepages()
        self.writerest()

        print 'Wrote file', self._ofile
        self._ifs.close()
        self._ofs.close()
        return 0

    def writeheader(self):
        """Write the PDF header"""

        ws = self.writestr

        title = self._ifile
        
        t=time.localtime()
        timestr=str(time.strftime("D:%Y%m%d%H%M%S", t))
        ws("%PDF-1.4\n")
        self._locations[1] = self._fpos
        ws("1 0 obj\n")
        ws("<<\n")

        buf = "".join(("/Creator (", self._appname, " By Anand B Pillai )\n"))
        ws(buf)
        buf = "".join(("/CreationDate (", timestr, ")\n"))
        ws(buf)
        buf = "".join(("/Producer (", self._appname, "(\\251 Anand B Pillai))\n"))
        ws(buf)
        if self._subject:
            title = self._subject
            buf = "".join(("/Subject (",self._subject,")\n"))
            ws(buf)
        if self._author:
            buf = "".join(("/Author (",self._author,")\n"))
            ws(buf)
        if self._keywords:
            buf = "".join(("/Keywords (",' '.join(self._keywords),")\n"))
            ws(buf)

        if title:
            buf = "".join(("/Title (", title, ")\n"))
            ws(buf)

        ws(">>\n")
        ws("endobj\n")
    
        self._locations[2] = self._fpos

        ws("2 0 obj\n")
        ws("<<\n")
        ws("/Type /Catalog\n")
        ws("/Pages 3 0 R\n")
        ws(">>\n")
        ws("endobj\n")
        
        self._locations[4] = self._fpos
        ws("4 0 obj\n")
        ws("<<\n")
        buf = "".join(("/BaseFont ", str(self._font), " /Encoding /WinAnsiEncoding /Name /F1 /Subtype /Type1 /Type /Font >>\n"))
        ws(buf)
    
        if self._IsoEnc:
            ws(ENCODING_STR)
            
        ws(">>\n")
        ws("endobj\n")
        
        self._locations[5] = self._fpos
        
        ws("5 0 obj\n")
        ws("<<\n")
        ws("  /Font << /F1 4 0 R >>\n")
        ws("  /ProcSet [ /PDF /Text ]\n")
        ws(">>\n")
        ws("endobj\n")
    
    def startpage(self):
        """ Start a page of data """

        ws = self.writestr
        
        self._pageNo += 1
        self._curobj += 1

        self._locations.append(self._fpos)
        self._locations[self._curobj]=self._fpos
    
        self._pageObs.append(self._curobj)
        self._pageObs[self._pageNo] = self._curobj
        
        buf = "".join((str(self._curobj), " 0 obj\n"))

        ws(buf)
        ws("<<\n")
        ws("/Type /Page\n")
        ws("/Parent 3 0 R\n")
        ws("/Resources 5 0 R\n")

        self._curobj += 1
        buf = "".join(("/Contents ", str(self._curobj), " 0 R\n"))
        ws(buf)
        ws(">>\n")
        ws("endobj\n")
        
        self._locations.append(self._fpos)
        self._locations[self._curobj] = self._fpos

        buf = "".join((str(self._curobj), " 0 obj\n"))
        ws(buf)
        ws("<<\n")
        
        buf = "".join(("/Length ", str(self._curobj + 1), " 0 R\n"))
        ws(buf)
        ws(">>\n")
        ws("stream\n")
        strmPos = self._fpos
    
        ws("BT\n");
        buf = "".join(("/F1 ", str(self._ptSize), " Tf\n"))
        ws(buf)
        buf = "".join(("1 0 0 1 50 ", str(self._pageHt - 40), " Tm\n"))
        ws(buf)
        buf = "".join((str(self._vertSpace), " TL\n"))
        ws(buf)
    
        return strmPos

    def endpage(self, streamStart):
        """End a page of data """
        
        ws = self.writestr

        ws("ET\n")
        streamEnd = self._fpos
        ws("endstream\n")
        ws("endobj\n")
    
        self._curobj += 1
        self._locations.append(self._fpos)
        self._locations[self._curobj] = self._fpos
    
        buf = "".join((str(self._curobj), " 0 obj\n"))
        ws(buf)
        buf = "".join((str(streamEnd - streamStart), '\n'))
        ws(buf)
        ws('endobj\n')
    
    def writepages(self):
        """Write pages as PDF"""
        
        ws = self.writestr

        beginstream=0
        lineNo, charNo=0,0
        ch, column=0,0
        padding,i=0,0
        atEOF=0
        linebuf = ''
        
        while not atEOF:
            beginstream = self.startpage()
            column=1
            
            while column <= self._columns:
                column += 1
                atFF=0
                atBOP=0
                lineNo=0
                # Special flag for regexp page break
                pagebreak = False
                
                while lineNo < self._lines and not atFF and not atEOF and not pagebreak:
                    linebuf = ''
                    lineNo += 1
                    ws("(")
                    charNo=0
                    
                    while charNo < self._cols:
                        charNo += 1
                        ch = self._ifs.read(1)
                        cond = ((ch != '\n') and not(ch==FF and self._doFFs) and (ch != ''))
                        if not cond:
                            # See if this dude matches the pagebreak regexp
                            if self._pagebreakre and self._pagebreakre.search(linebuf.strip()):
                                pagebreak = True
                                
                            linebuf = ''
                            break
                        else:
                            linebuf = linebuf + ch

                        if ord(ch) >= 32 and ord(ch) <= 127:
                            if ch == '(' or ch == ')' or ch == '\\':
                                ws("\\")
                            ws(ch)
                        else:
                            if ord(ch) == 9:
                                padding =self._tab - ((charNo - 1) % self._tab)
                                for i in range(padding):
                                    ws(" ")
                                charNo += (padding -1)
                            else:
                                if ch != FF:
                                    # write \xxx form for dodgy character
                                    buf = "".join(('\\', ch))
                                    ws(buf)
                                else:
                                    # dont print anything for a FF
                                    charNo -= 1

                    ws(")'\n")
                    if ch == FF:
                        atFF=1
                    if lineNo == self._lines:
                        atBOP=1
                        
                    if atBOP:
                        pos=0
                        ch = self._ifs.read(1)
                        pos= self._ifs.tell()
                        if ch == FF:
                            ch = self._ifs.read(1)
                            pos=self._ifs.tell()
                        # python's EOF signature
                        if ch == '':
                            atEOF=1
                        else:
                            # push position back by one char
                            self._ifs.seek(pos-1)

                    elif atFF:
                        ch = self._ifs.read(1)
                        pos=self._ifs.tell()
                        if ch == '':
                            atEOF=1
                        else:
                            self._ifs.seek(pos-1)

                if column < self._columns:
                    buf = "".join(("1 0 0 1 ",
                                   str((self._pageWd/2 + 25)),
                                   " ",
                                   str(self._pageHt - 40),
                                   " Tm\n"))
                    ws(buf)

            self.endpage(beginstream)

    def writerest(self):
        """Finish the file"""

        ws = self.writestr
        self._locations[3] = self._fpos
    
        ws("3 0 obj\n")
        ws("<<\n")
        ws("/Type /Pages\n")
        buf = "".join(("/Count ", str(self._pageNo), "\n"))
        ws(buf)
        buf = "".join(("/MediaBox [ 0 0 ", str(self._pageWd), " ", str(self._pageHt), " ]\n"))
        ws(buf)
        ws("/Kids [ ")
    
        for i in range(1, self._pageNo+1):
            buf = "".join((str(self._pageObs[i]), " 0 R "))
            ws(buf)

        ws("]\n")
        ws(">>\n")
        ws("endobj\n")
        
        xref = self._fpos
        ws("xref\n")
        buf = "".join(("0 ", str((self._curobj) + 1), "\n"))
        ws(buf)
        buf = "".join(("0000000000 65535 f ", str(LINE_END)))
        ws(buf)

        for i in range(1, self._curobj + 1):
            val = self._locations[i]
            buf = "".join((string.zfill(str(val), 10), " 00000 n ", str(LINE_END)))
            ws(buf)

        ws("trailer\n")
        ws("<<\n")
        buf = "".join(("/Size ", str(self._curobj + 1), "\n"))
        ws(buf)
        ws("/Root 2 0 R\n")
        ws("/Info 1 0 R\n")
        ws(">>\n")
        
        ws("startxref\n")
        buf = "".join((str(xref), "\n"))
        ws(buf)
        ws("%%EOF\n")
        

def main():
    
    pdfclass=PyText2Pdf()
    pdfclass.parse_args()
    pdfclass.convert()

if __name__ == "__main__":
    main()

This one is a modification of a previous recipe which provided a pure Python text to PDF converter.

This version modifies argument parsing using optparse and provides a few additional options to add subject, author, keywords etc. There is also an option to provide arbitrary page breaks using regular expressions. This is useful to convert documents such as RFCs to PDF.

Example...

$text2pdf.py rfc_3917.txt -S "Netflow RFC" -A "Quittek, et al." -K netflow,rfc,networking,ipfix -R "[Page\s+\d+]"

2 comments

Charles Kellum 16 years, 6 months ago  # | flag

Worked turn-key for me. two minutes later, I have a solution for creating PDF files for my output from network ip scans. Thanks

shiva 8 years, 11 months ago  # | flag

This is awesome, working like a charm .....................thanks Anand. I wish to use your code for a project i'm up to.