[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 | # $Log$ |
---|
[1461] | 26 | # Revision 1.7 2004/05/07 23:08:21 jalet |
---|
| 27 | # Skeleton for PCLXL aka PCL6 |
---|
| 28 | # Added the "potential" fix for rastertoprinter's output |
---|
| 29 | # |
---|
[1456] | 30 | # Revision 1.6 2004/05/06 21:19:27 jalet |
---|
| 31 | # Doesn't exit anymore on the first nul byte |
---|
| 32 | # |
---|
[1451] | 33 | # Revision 1.5 2004/05/06 12:37:29 jalet |
---|
| 34 | # pkpgcounter : comments |
---|
| 35 | # pkprinters : when --add is used, existing printers are now skipped. |
---|
| 36 | # |
---|
[1449] | 37 | # Revision 1.4 2004/05/04 12:21:55 jalet |
---|
| 38 | # Now uses mmap in PCL mode |
---|
| 39 | # |
---|
[1448] | 40 | # Revision 1.3 2004/05/04 04:39:26 jalet |
---|
| 41 | # Better PCL support |
---|
| 42 | # |
---|
[1447] | 43 | # Revision 1.2 2004/05/04 03:14:26 jalet |
---|
| 44 | # fixed copy&paste problem in pkpgcounter |
---|
| 45 | # |
---|
[1430] | 46 | # Revision 1.1 2004/04/08 17:07:42 jalet |
---|
| 47 | # pkpgcounter added |
---|
| 48 | # |
---|
| 49 | # |
---|
| 50 | |
---|
| 51 | import sys |
---|
[1449] | 52 | import os |
---|
| 53 | import mmap |
---|
[1430] | 54 | import tempfile |
---|
| 55 | |
---|
| 56 | def ispostscript(data) : |
---|
| 57 | """Returns 1 if data is PostScript, else 0.""" |
---|
| 58 | if data.startswith("%!") or \ |
---|
| 59 | data.startswith("\004%!") or \ |
---|
| 60 | data.startswith("\033%-12345X%!PS") or \ |
---|
| 61 | ((data[:128].find("\033%-12345X") != -1) and \ |
---|
| 62 | ((data.find("LANGUAGE=POSTSCRIPT") != -1) or \ |
---|
| 63 | (data.find("LANGUAGE = POSTSCRIPT") != -1) or \ |
---|
| 64 | (data.find("LANGUAGE = Postscript") != -1))) : |
---|
| 65 | return 1 |
---|
| 66 | else : |
---|
| 67 | return 0 |
---|
| 68 | |
---|
[1461] | 69 | def ispcl(data) : |
---|
| 70 | """Returns 1 if data is PCL, else 0.""" |
---|
| 71 | if data.startswith("\033E\033") or \ |
---|
| 72 | ((data[:128].find("\033%-12345X") != -1) and \ |
---|
| 73 | ((data.find("LANGUAGE=PCL") != -1) or \ |
---|
| 74 | (data.find("LANGUAGE = PCL") != -1) or \ |
---|
| 75 | (data.find("LANGUAGE = Pcl") != -1))) : |
---|
| 76 | return 1 |
---|
| 77 | else : |
---|
| 78 | return 0 |
---|
| 79 | |
---|
| 80 | def ispclxl(data) : |
---|
| 81 | """Returns 1 if data is PCLXL aka PCL6, else 0.""" |
---|
| 82 | if ((data[:128].find("\033%-12345X") != -1) and \ |
---|
| 83 | (data.find(" HP-PCL XL;") != -1) and \ |
---|
| 84 | ((data.find("LANGUAGE=PCLXL") != -1) or \ |
---|
| 85 | (data.find("LANGUAGE = PCLXL") != -1))) : |
---|
| 86 | return 1 |
---|
| 87 | else : |
---|
| 88 | return 0 |
---|
| 89 | |
---|
[1430] | 90 | def postscript(infile) : |
---|
[1461] | 91 | """Count pages in a DSC compliant PostScript document.""" |
---|
[1430] | 92 | pagecount = 0 |
---|
| 93 | pagenum = None |
---|
| 94 | while 1 : |
---|
| 95 | line = infile.readline() |
---|
| 96 | if not line : |
---|
| 97 | break |
---|
| 98 | if line.startswith("%%Page: ") : |
---|
| 99 | pagecount += 1 |
---|
| 100 | return pagecount |
---|
| 101 | |
---|
| 102 | def pcl(infile) : |
---|
[1461] | 103 | """Count pages in a PCL5 document.""" |
---|
[1430] | 104 | # |
---|
| 105 | # Algorithm from pclcount |
---|
| 106 | # (c) 2003, by Eduardo Gielamo Oliveira & Rodolfo Broco Manin |
---|
| 107 | # published under the terms of the GNU General Public Licence v2. |
---|
| 108 | # |
---|
[1449] | 109 | # Backported from C to Python by Jerome Alet, then enhanced |
---|
| 110 | # with more PCL tags detected. I think all the necessary PCL tags |
---|
[1451] | 111 | # are recognized to correctly handle PCL5 files wrt their number |
---|
| 112 | # of pages. The documentation used for this was : |
---|
| 113 | # |
---|
| 114 | # HP PCL/PJL Reference Set |
---|
| 115 | # PCL5 Printer Language Technical Quick Reference Guide |
---|
| 116 | # http://h20000.www2.hp.com/bc/docs/support/SupportManual/bpl13205/bpl13205.pdf |
---|
| 117 | # |
---|
[1449] | 118 | infileno = infile.fileno() |
---|
| 119 | infile = mmap.mmap(infileno, os.fstat(infileno).st_size, access=mmap.ACCESS_READ) |
---|
[1451] | 120 | tagsends = { "&n" : "W", |
---|
| 121 | "&b" : "W", |
---|
| 122 | "*i" : "W", |
---|
| 123 | "*l" : "W", |
---|
| 124 | "*m" : "W", |
---|
| 125 | "*v" : "W", |
---|
| 126 | "*c" : "W", |
---|
| 127 | "(f" : "W", |
---|
| 128 | "*b" : "VW", |
---|
| 129 | "(s" : "W", |
---|
| 130 | ")s" : "W", |
---|
| 131 | "&p" : "X", |
---|
| 132 | "&l" : "X" } |
---|
[1430] | 133 | copies = 1 |
---|
[1461] | 134 | pagecount = resets = 0 |
---|
[1430] | 135 | tag = None |
---|
[1449] | 136 | position = 0 |
---|
[1430] | 137 | while 1 : |
---|
[1449] | 138 | try : |
---|
| 139 | char = infile[position] |
---|
| 140 | except IndexError : # EOF |
---|
| 141 | break |
---|
| 142 | position += 1 |
---|
[1430] | 143 | if char == "\014" : |
---|
| 144 | pagecount += 1 |
---|
| 145 | elif char == "\033" : |
---|
| 146 | # |
---|
[1448] | 147 | # <ESC>*b###W -> Start of a raster data row/block |
---|
| 148 | # <ESC>*b###V -> Start of a raster data plane |
---|
| 149 | # <ESC>*c###W -> Start of a user defined pattern |
---|
| 150 | # <ESC>*i###W -> Start of a viewing illuminant block |
---|
| 151 | # <ESC>*l###W -> Start of a color lookup table |
---|
| 152 | # <ESC>*m###W -> Start of a download dither matrix block |
---|
| 153 | # <ESC>*v###W -> Start of a configure image data block |
---|
[1430] | 154 | # <ESC>(s###W -> Start of a characters description block |
---|
| 155 | # <ESC>)s###W -> Start of a fonts description block |
---|
[1448] | 156 | # <ESC>(f###W -> Start of a symbol set block |
---|
| 157 | # <ESC>&b###W -> Start of configuration data block |
---|
| 158 | # <ESC>&l###X -> Number of copies |
---|
| 159 | # <ESC>&n###W -> Starts an alphanumeric string ID block |
---|
[1430] | 160 | # <ESC>&p###X -> Start of a non printable characters block |
---|
| 161 | # |
---|
[1449] | 162 | tagstart = infile[position] |
---|
| 163 | position += 1 |
---|
[1448] | 164 | if tagstart in "E9=YZ" : # one byte PCL tag |
---|
[1461] | 165 | if tagstart == "E" : |
---|
| 166 | resets += 1 |
---|
[1448] | 167 | continue # skip to next tag |
---|
[1449] | 168 | tag = tagstart + infile[position] |
---|
| 169 | position += 1 |
---|
[1430] | 170 | try : |
---|
| 171 | tagend = tagsends[tag] |
---|
| 172 | except KeyError : |
---|
| 173 | pass # Unsupported PCL tag |
---|
| 174 | else : |
---|
| 175 | # Now read the numeric argument |
---|
| 176 | size = 0 |
---|
| 177 | while 1 : |
---|
[1449] | 178 | char = infile[position] |
---|
| 179 | position += 1 |
---|
[1430] | 180 | if not char.isdigit() : |
---|
| 181 | break |
---|
| 182 | size = (size * 10) + int(char) |
---|
[1448] | 183 | if char in tagend : |
---|
[1430] | 184 | if tag == "&l" : |
---|
| 185 | copies = size |
---|
| 186 | else : |
---|
| 187 | # doing a read will prevent the seek |
---|
| 188 | # for unseekable streams. |
---|
| 189 | # we just ignore the block anyway. |
---|
[1448] | 190 | if tag == "&n" : |
---|
| 191 | # we have to take care of the operation id byte |
---|
| 192 | # which is before the string itself |
---|
| 193 | size += 1 |
---|
[1449] | 194 | position += size |
---|
[1461] | 195 | |
---|
| 196 | # if pagecount is still 0, we will return the number |
---|
| 197 | # of resets instead of the number of form feed characters. |
---|
| 198 | # but the number of resets is always at least 2 with a valid |
---|
| 199 | # pcl file : one at the very start and one at the very end |
---|
| 200 | # of the job's data. So we substract 2 from the number of |
---|
| 201 | # resets. And since on our test data we needed to substract |
---|
| 202 | # 1 more, we finally substract 3, and will test several |
---|
| 203 | # PCL files with this. If resets < 2, then the file is |
---|
| 204 | # probably not a valid PCL file, so we return 0 |
---|
| 205 | if not pagecount : |
---|
| 206 | return copies * (resets - 3) * (resets > 2) |
---|
| 207 | else : |
---|
| 208 | return copies * pagecount |
---|
[1430] | 209 | |
---|
[1461] | 210 | def pclxl(infile) : |
---|
| 211 | """Count pages in a PCL6 aka PCLXL document.""" |
---|
| 212 | raise TypeError, "Sorry but PCL6, aka PCLXL, is not supported yet.\n" |
---|
| 213 | |
---|
[1430] | 214 | def smartpagecounter(filename) : |
---|
| 215 | """Autodetects file format and returns number of pages.""" |
---|
| 216 | if filename == "-" : |
---|
| 217 | # we must read from stdin |
---|
| 218 | # but since stdin is not seekable, we have to use a temporary |
---|
| 219 | # file instead. |
---|
| 220 | infile = tempfile.TemporaryFile() |
---|
| 221 | while 1 : |
---|
| 222 | data = sys.stdin.read(256 * 1024) |
---|
| 223 | if not data : |
---|
| 224 | break |
---|
| 225 | infile.write(data) |
---|
| 226 | infile.flush() |
---|
| 227 | infile.seek(0) |
---|
| 228 | else : |
---|
| 229 | # normal file |
---|
| 230 | infile = open(filename, "rb") |
---|
| 231 | |
---|
| 232 | # Try to detect file type by reading first block of datas |
---|
| 233 | firstblock = infile.read(1024) |
---|
| 234 | infile.seek(0) |
---|
| 235 | if ispostscript(firstblock) : |
---|
| 236 | size = postscript(infile) |
---|
[1461] | 237 | elif ispclxl(firstblock) : |
---|
| 238 | size = pclxl(infile) |
---|
[1430] | 239 | elif ispcl(firstblock) : |
---|
| 240 | size = pcl(infile) |
---|
| 241 | else : |
---|
| 242 | sys.stderr.write("ERROR : Unknown file format for %s\n" % filename) |
---|
| 243 | size = 0 |
---|
| 244 | infile.close() |
---|
| 245 | return size |
---|
| 246 | |
---|
| 247 | if __name__ == "__main__" : |
---|
| 248 | if len(sys.argv) < 2 : |
---|
| 249 | sys.argv.append("-") |
---|
| 250 | |
---|
| 251 | totalsize = 0 |
---|
| 252 | for arg in sys.argv[1:] : |
---|
| 253 | totalsize += smartpagecounter(arg) |
---|
| 254 | print "%s" % totalsize |
---|