[1430] | 1 | #! /usr/bin/env python |
---|
| 2 | # -*- coding: ISO-8859-15 -*- |
---|
| 3 | |
---|
| 4 | # pkpgcounter, a smart software page counter |
---|
| 5 | # |
---|
| 6 | # PyKota - Print Quotas for CUPS and LPRng |
---|
| 7 | # |
---|
| 8 | # (c) 2003-2004 Jerome Alet <alet@librelogiciel.com> |
---|
| 9 | # This program is free software; you can redistribute it and/or modify |
---|
| 10 | # it under the terms of the GNU General Public License as published by |
---|
| 11 | # the Free Software Foundation; either version 2 of the License, or |
---|
| 12 | # (at your option) any later version. |
---|
| 13 | # |
---|
| 14 | # This program is distributed in the hope that it will be useful, |
---|
| 15 | # but WITHOUT ANY WARRANTY; without even the implied warranty of |
---|
| 16 | # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the |
---|
| 17 | # GNU General Public License for more details. |
---|
| 18 | # |
---|
| 19 | # You should have received a copy of the GNU General Public License |
---|
| 20 | # along with this program; if not, write to the Free Software |
---|
| 21 | # Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307, USA. |
---|
| 22 | # |
---|
| 23 | # $Id$ |
---|
| 24 | # |
---|
| 25 | # $Log$ |
---|
[1449] | 26 | # Revision 1.4 2004/05/04 12:21:55 jalet |
---|
| 27 | # Now uses mmap in PCL mode |
---|
| 28 | # |
---|
[1448] | 29 | # Revision 1.3 2004/05/04 04:39:26 jalet |
---|
| 30 | # Better PCL support |
---|
| 31 | # |
---|
[1447] | 32 | # Revision 1.2 2004/05/04 03:14:26 jalet |
---|
| 33 | # fixed copy&paste problem in pkpgcounter |
---|
| 34 | # |
---|
[1430] | 35 | # Revision 1.1 2004/04/08 17:07:42 jalet |
---|
| 36 | # pkpgcounter added |
---|
| 37 | # |
---|
| 38 | # |
---|
| 39 | |
---|
| 40 | import sys |
---|
[1449] | 41 | import os |
---|
| 42 | import mmap |
---|
[1430] | 43 | import tempfile |
---|
| 44 | |
---|
| 45 | def ispostscript(data) : |
---|
| 46 | """Returns 1 if data is PostScript, else 0.""" |
---|
| 47 | if data.startswith("%!") or \ |
---|
| 48 | data.startswith("\004%!") or \ |
---|
| 49 | data.startswith("\033%-12345X%!PS") or \ |
---|
| 50 | ((data[:128].find("\033%-12345X") != -1) and \ |
---|
| 51 | ((data.find("LANGUAGE=POSTSCRIPT") != -1) or \ |
---|
| 52 | (data.find("LANGUAGE = POSTSCRIPT") != -1) or \ |
---|
| 53 | (data.find("LANGUAGE = Postscript") != -1))) : |
---|
| 54 | return 1 |
---|
| 55 | else : |
---|
| 56 | return 0 |
---|
| 57 | |
---|
| 58 | def postscript(infile) : |
---|
| 59 | """Count pages in a PostScript document.""" |
---|
| 60 | pagecount = 0 |
---|
| 61 | pagenum = None |
---|
| 62 | while 1 : |
---|
| 63 | line = infile.readline() |
---|
| 64 | if not line : |
---|
| 65 | break |
---|
| 66 | if line.startswith("%%Page: ") : |
---|
| 67 | pagecount += 1 |
---|
| 68 | return pagecount |
---|
| 69 | |
---|
| 70 | def ispcl(data) : |
---|
| 71 | """Returns 1 if data is PCL, else 0.""" |
---|
| 72 | if data.startswith("\033E\033") or \ |
---|
| 73 | ((data[:128].find("\033%-12345X") != -1) and \ |
---|
| 74 | ((data.find("LANGUAGE=PCL") != -1) or \ |
---|
| 75 | (data.find("LANGUAGE = PCL") != -1) or \ |
---|
| 76 | (data.find("LANGUAGE = Pcl") != -1))) : |
---|
[1447] | 77 | return 1 |
---|
| 78 | else : |
---|
| 79 | return 0 |
---|
[1430] | 80 | |
---|
| 81 | def pcl(infile) : |
---|
| 82 | """Count pages in a PostScript document.""" |
---|
| 83 | # |
---|
| 84 | # Algorithm from pclcount |
---|
| 85 | # (c) 2003, by Eduardo Gielamo Oliveira & Rodolfo Broco Manin |
---|
| 86 | # published under the terms of the GNU General Public Licence v2. |
---|
| 87 | # |
---|
[1449] | 88 | # Backported from C to Python by Jerome Alet, then enhanced |
---|
| 89 | # with more PCL tags detected. I think all the necessary PCL tags |
---|
| 90 | # are recognized to correctly handle PCL5e files wrt their number |
---|
| 91 | # of pages |
---|
[1430] | 92 | # |
---|
[1449] | 93 | infileno = infile.fileno() |
---|
| 94 | infile = mmap.mmap(infileno, os.fstat(infileno).st_size, access=mmap.ACCESS_READ) |
---|
[1448] | 95 | tagsends = { "&n" : "W", "&b": "W", "*i" : "W", "*l" : "W", "*m" : "W", "*v": "W", "*c" : "W", "(f" : "W", "*b" : "VW", "(s" : "W", ")s" : "W", "&p" : "X", "&l" : "X" } |
---|
[1430] | 96 | copies = 1 |
---|
| 97 | pagecount = 0 |
---|
| 98 | tag = None |
---|
[1449] | 99 | position = 0 |
---|
[1430] | 100 | while 1 : |
---|
[1449] | 101 | try : |
---|
| 102 | char = infile[position] |
---|
| 103 | except IndexError : # EOF |
---|
| 104 | break |
---|
| 105 | position += 1 |
---|
[1430] | 106 | if not char : |
---|
| 107 | break |
---|
| 108 | if char == "\014" : |
---|
| 109 | pagecount += 1 |
---|
| 110 | elif char == "\033" : |
---|
| 111 | # |
---|
[1448] | 112 | # <ESC>*b###W -> Start of a raster data row/block |
---|
| 113 | # <ESC>*b###V -> Start of a raster data plane |
---|
| 114 | # <ESC>*c###W -> Start of a user defined pattern |
---|
| 115 | # <ESC>*i###W -> Start of a viewing illuminant block |
---|
| 116 | # <ESC>*l###W -> Start of a color lookup table |
---|
| 117 | # <ESC>*m###W -> Start of a download dither matrix block |
---|
| 118 | # <ESC>*v###W -> Start of a configure image data block |
---|
[1430] | 119 | # <ESC>(s###W -> Start of a characters description block |
---|
| 120 | # <ESC>)s###W -> Start of a fonts description block |
---|
[1448] | 121 | # <ESC>(f###W -> Start of a symbol set block |
---|
| 122 | # <ESC>&b###W -> Start of configuration data block |
---|
| 123 | # <ESC>&l###X -> Number of copies |
---|
| 124 | # <ESC>&n###W -> Starts an alphanumeric string ID block |
---|
[1430] | 125 | # <ESC>&p###X -> Start of a non printable characters block |
---|
| 126 | # |
---|
[1449] | 127 | tagstart = infile[position] |
---|
| 128 | position += 1 |
---|
[1448] | 129 | if tagstart in "E9=YZ" : # one byte PCL tag |
---|
| 130 | continue # skip to next tag |
---|
[1449] | 131 | tag = tagstart + infile[position] |
---|
| 132 | position += 1 |
---|
[1430] | 133 | try : |
---|
| 134 | tagend = tagsends[tag] |
---|
| 135 | except KeyError : |
---|
| 136 | pass # Unsupported PCL tag |
---|
| 137 | else : |
---|
| 138 | # Now read the numeric argument |
---|
| 139 | size = 0 |
---|
| 140 | while 1 : |
---|
[1449] | 141 | char = infile[position] |
---|
| 142 | position += 1 |
---|
[1430] | 143 | if not char.isdigit() : |
---|
| 144 | break |
---|
| 145 | size = (size * 10) + int(char) |
---|
[1448] | 146 | if char in tagend : |
---|
[1430] | 147 | if tag == "&l" : |
---|
| 148 | copies = size |
---|
| 149 | else : |
---|
| 150 | # doing a read will prevent the seek |
---|
| 151 | # for unseekable streams. |
---|
| 152 | # we just ignore the block anyway. |
---|
[1448] | 153 | if tag == "&n" : |
---|
| 154 | # we have to take care of the operation id byte |
---|
| 155 | # which is before the string itself |
---|
| 156 | size += 1 |
---|
[1449] | 157 | position += size |
---|
[1430] | 158 | return copies * pagecount |
---|
| 159 | |
---|
| 160 | def smartpagecounter(filename) : |
---|
| 161 | """Autodetects file format and returns number of pages.""" |
---|
| 162 | if filename == "-" : |
---|
| 163 | # we must read from stdin |
---|
| 164 | # but since stdin is not seekable, we have to use a temporary |
---|
| 165 | # file instead. |
---|
| 166 | infile = tempfile.TemporaryFile() |
---|
| 167 | while 1 : |
---|
| 168 | data = sys.stdin.read(256 * 1024) |
---|
| 169 | if not data : |
---|
| 170 | break |
---|
| 171 | infile.write(data) |
---|
| 172 | infile.flush() |
---|
| 173 | infile.seek(0) |
---|
| 174 | else : |
---|
| 175 | # normal file |
---|
| 176 | infile = open(filename, "rb") |
---|
| 177 | |
---|
| 178 | # Try to detect file type by reading first block of datas |
---|
| 179 | firstblock = infile.read(1024) |
---|
| 180 | infile.seek(0) |
---|
| 181 | if ispostscript(firstblock) : |
---|
| 182 | size = postscript(infile) |
---|
| 183 | elif ispcl(firstblock) : |
---|
| 184 | size = pcl(infile) |
---|
| 185 | else : |
---|
| 186 | sys.stderr.write("ERROR : Unknown file format for %s\n" % filename) |
---|
| 187 | size = 0 |
---|
| 188 | infile.close() |
---|
| 189 | return size |
---|
| 190 | |
---|
| 191 | if __name__ == "__main__" : |
---|
| 192 | if len(sys.argv) < 2 : |
---|
| 193 | sys.argv.append("-") |
---|
| 194 | |
---|
| 195 | totalsize = 0 |
---|
| 196 | for arg in sys.argv[1:] : |
---|
| 197 | totalsize += smartpagecounter(arg) |
---|
| 198 | print "%s" % totalsize |
---|