objavi/book_utils.py

   1 # Part of Objavi2, which turns html manuals into books
   2 #
   3 # Copyright (C) 2009 Douglas Bagnall
   4 #
   5 # This program is free software; you can redistribute it and/or modify
   6 # it under the terms of the GNU General Public License as published by
   7 # the Free Software Foundation; either version 2 of the License, or
   8 # (at your option) any later version.
   9 #
  10 # This program is distributed in the hope that it will be useful,
  11 # but WITHOUT ANY WARRANTY; without even the implied warranty of
  12 # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  13 # GNU General Public License for more details.
  14 #
  15 # You should have received a copy of the GNU General Public License along
  16 # with this program; if not, write to the Free Software Foundation, Inc.,
  17 # 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
  18 """
  19 General utility functions.
  20 """
  21
  22 import os, sys
  23 import shutil
  24 import time, re
  25 from subprocess import Popen, PIPE
  26
  27 #from objavi.fmbook import log
  28 from objavi import config
  29
  30 #general, non-cgi functions
  31 def init_log():
  32     """Try to redirect stderr to the log file.  If it doesn't work,
  33     leave stderr as it is."""
  34     if config.REDIRECT_LOG:
  35         logfile = os.path.join(config.LOGDIR, 'objavi.log')
  36         try:
  37             size = os.stat(logfile).st_size
  38             if size > config.LOG_ROTATE_SIZE:
  39                 oldlog = os.path.join(config.LOGDIR, time.strftime('objavi-%Y-%m-%d+%H-%M-%S.log'))
  40                 f = open(logfile, 'a')
  41                 print >> f, "CLOSING LOG at size %s, renaming to %s" % (size, oldlog)
  42                 f.close()
  43                 os.rename(logfile, oldlog)
  44         except OSError, e:
  45             log(e) # goes to original stderr
  46         try:
  47             f = open(logfile, 'a')
  48             sys.stderr.flush()
  49             os.dup2(f.fileno(), sys.stderr.fileno())
  50             # reassign the object as well as dup()ing, in case it closes down out of scope.
  51             sys.stderr = f
  52         except (IOError, OSError), e:
  53             log(e)
  54             return
  55
  56 def log(*messages, **kwargs):
  57     """Send the messages to the appropriate place (stderr, or syslog).
  58     If a <debug> keyword is specified, the message is only printed if
  59     its value is in the global DEBUG_MODES."""
  60     if 'debug' not in kwargs or config.DEBUG_ALL or kwargs['debug'] in config.DEBUG_MODES:
  61         for m in messages:
  62             try:
  63                 print >> sys.stderr, m
  64             except Exception:
  65                 print >> sys.stderr, repr(m)
  66         sys.stderr.flush()
  67
  68 def log_types(*args, **kwargs):
  69     """Log the type of the messages as well as their value (size constrained)."""
  70     size = kwargs.get('size', 50)
  71     for a in args:
  72         try:
  73             s = ("%s" % (a,))
  74         except Exception:
  75             try:
  76                 s = ("%r" % (a,))
  77             except Exception:
  78                 s = '<UNPRINTABLE!>'
  79         log("%s: %s" % (type(a), s[:size]))
  80
  81 def make_book_name(book, server, suffix='.pdf', timestamp=None):
  82     lang = guess_lang(server, book)
  83     book = ''.join(x for x in book if x.isalnum())
  84     if timestamp is None:
  85         timestamp = time.strftime('%Y.%m.%d-%H.%M.%S')
  86     return '%s-%s-%s%s' % (book, lang,
  87                            timestamp,
  88                            suffix)
  89
  90 def guess_lang(server, book):
  91     if server is None:
  92         server = config.DEFAULT_SERVER
  93     lang = config.SERVER_DEFAULTS[server].get('lang')
  94     if lang is None and '_' in book:
  95         lang = book[book.rindex('_') + 1:]
  96     return lang
  97
  98 def guess_text_dir(server, book):
  99     try:
 100         dir = config.SERVER_DEFAULTS[server]['dir']
 101     except KeyError:
 102         dir = None
 103     if dir not in ('LTR', 'RTL'):
 104         log("server %s, book %s: no specified dir (%s)" %(server, book, dir))
 105         if '_' in book:
 106             lang = book[book.rindex('_') + 1:]
 107             dir = config.LANGUAGE_DIR.get(lang, 'LTR')
 108         elif '.' in server:
 109             lang = server[:server.index('.')]
 110             dir = config.LANGUAGE_DIR.get(lang, 'LTR')
 111         else:
 112             dir = 'LTR'
 113     log("server %s, book %s: found dir %s" %(server, book, dir))
 114     return str(dir)
 115
 116 def run(cmd):
 117     try:
 118         p = Popen(cmd, stdout=PIPE, stderr=PIPE)
 119         out, err = p.communicate()
 120     except Exception:
 121         log("Failed on command: %r" % cmd)
 122         raise
 123     log("%s\n%s returned %s and produced\nstdout:%s\nstderr:%s" %
 124         (' '.join(cmd), cmd[0], p.poll(), out, err))
 125     return p.poll()
 126
 127 def shift_file(fn, dir, backup='~'):
 128     """Shift a file and save backup (only works on same filesystem)"""
 129     log("shifting file %r to %s" % (fn, dir))
 130     base = os.path.basename(fn)
 131     dest = os.path.join(dir, base)
 132     if backup and os.path.exists(dest):
 133         os.rename(dest, dest + backup)
 134     shutil.move(fn, dest)
 135     return dest
 136
 137 class ObjaviError(Exception):
 138     pass
 139
 140
 141 def decode_html_entities(text):
 142     for encoding in ('ascii', 'utf-8', 'iso-8859-1'):
 143         try:
 144             text = text.decode(encoding)
 145             break
 146         except UnicodeDecodeError:
 147             continue
 148     def fixup(m):
 149         entity = m.group(0)
 150         try:
 151             if entity[:3] == "&#x":
 152                 return unichr(int(entity[3:-1], 16))
 153             elif entity[:2] == "&#":
 154                 return unichr(int(entity[2:-1]))
 155         except ValueError:
 156             log("ignoring bad entity %s" % entity)
 157         return entity
 158     return re.sub("&#?[0-9a-fA-F]+;", fixup, text).encode('utf-8')