summaryrefslogtreecommitdiffstats
path: root/Lib/pipes.py
blob: 26750f6ab88d8f43465710bee33936997bbc1553 (plain)
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
"""Conversion pipeline templates.

The problem:
------------

Suppose you have some data that you want to convert to another format,
such as from GIF image format to PPM image format.  Maybe the
conversion involves several steps (e.g. piping it through compress or
uuencode).  Some of the conversion steps may require that their input
is a disk file, others may be able to read standard input; similar for
their output.  The input to the entire conversion may also be read
from a disk file or from an open file, and similar for its output.

The module lets you construct a pipeline template by sticking one or
more conversion steps together.  It will take care of creating and
removing temporary files if they are necessary to hold intermediate
data.  You can then use the template to do conversions from many
different sources to many different destinations.  The temporary
file names used are different each time the template is used.

The templates are objects so you can create templates for many
different conversion steps and store them in a dictionary, for
instance.


Directions:
-----------

To create a template:
    t = Template()

To add a conversion step to a template:
   t.append(command, kind)
where kind is a string of two characters: the first is '-' if the
command reads its standard input or 'f' if it requires a file; the
second likewise for the output. The command must be valid /bin/sh
syntax.  If input or output files are required, they are passed as
$IN and $OUT; otherwise, it must be  possible to use the command in
a pipeline.

To add a conversion step at the beginning:
   t.prepend(command, kind)

To convert a file to another file using a template:
  sts = t.copy(infile, outfile)
If infile or outfile are the empty string, standard input is read or
standard output is written, respectively.  The return value is the
exit status of the conversion pipeline.

To open a file for reading or writing through a conversion pipeline:
   fp = t.open(file, mode)
where mode is 'r' to read the file, or 'w' to write it -- just like
for the built-in function open() or for os.popen().

To create a new template object initialized to a given one:
   t2 = t.clone()
"""                                     # '


import re
import os
import tempfile
import string

__all__ = ["Template"]

# Conversion step kinds

FILEIN_FILEOUT = 'ff'                   # Must read & write real files
STDIN_FILEOUT  = '-f'                   # Must write a real file
FILEIN_STDOUT  = 'f-'                   # Must read a real file
STDIN_STDOUT   = '--'                   # Normal pipeline element
SOURCE         = '.-'                   # Must be first, writes stdout
SINK           = '-.'                   # Must be last, reads stdin

stepkinds = [FILEIN_FILEOUT, STDIN_FILEOUT, FILEIN_STDOUT, STDIN_STDOUT, \
             SOURCE, SINK]


class Template:
    """Class representing a pipeline template."""

    def __init__(self):
        """Template() returns a fresh pipeline template."""
        self.debugging = 0
        self.reset()

    def __repr__(self):
        """t.__repr__() implements repr(t)."""
        return '<Template instance, steps=%r>' % (self.steps,)

    def reset(self):
        """t.reset() restores a pipeline template to its initial state."""
        self.steps = []

    def clone(self):
        """t.clone() returns a new pipeline template with identical
        initial state as the current one."""
        t = Template()
        t.steps = self.steps[:]
        t.debugging = self.debugging
        return t

    def debug(self, flag):
        """t.debug(flag) turns debugging on or off."""
        self.debugging = flag

    def append(self, cmd, kind):
        """t.append(cmd, kind) adds a new step at the end."""
        if type(cmd) is not type(''):
            raise TypeError, \
                  'Template.append: cmd must be a string'
        if kind not in stepkinds:
            raise ValueError, \
                  'Template.append: bad kind %r' % (kind,)
        if kind == SOURCE:
            raise ValueError, \
                  'Template.append: SOURCE can only be prepended'
        if self.steps and self.steps[-1][1] == SINK:
            raise ValueError, \
                  'Template.append: already ends with SINK'
        if kind[0] == 'f' and not re.search(r'\$IN\b', cmd):
            raise ValueError, \
                  'Template.append: missing $IN in cmd'
        if kind[1] == 'f' and not re.search(r'\$OUT\b', cmd):
            raise ValueError, \
                  'Template.append: missing $OUT in cmd'
        self.steps.append((cmd, kind))

    def prepend(self, cmd, kind):
        """t.prepend(cmd, kind) adds a new step at the front."""
        if type(cmd) is not type(''):
            raise TypeError, \
                  'Template.prepend: cmd must be a string'
        if kind not in stepkinds:
            raise ValueError, \
                  'Template.prepend: bad kind %r' % (kind,)
        if kind == SINK:
            raise ValueError, \
                  'Template.prepend: SINK can only be appended'
        if self.steps and self.steps[0][1] == SOURCE:
            raise ValueError, \
                  'Template.prepend: already begins with SOURCE'
        if kind[0] == 'f' and not re.search(r'\$IN\b', cmd):
            raise ValueError, \
                  'Template.prepend: missing $IN in cmd'
        if kind[1] == 'f' and not re.search(r'\$OUT\b', cmd):
            raise ValueError, \
                  'Template.prepend: missing $OUT in cmd'
        self.steps.insert(0, (cmd, kind))

    def open(self, file, rw):
        """t.open(file, rw) returns a pipe or file object open for
        reading or writing; the file is the other end of the pipeline."""
        if rw == 'r':
            return self.open_r(file)
        if rw == 'w':
            return self.open_w(file)
        raise ValueError, \
              'Template.open: rw must be \'r\' or \'w\', not %r' % (rw,)

    def open_r(self, file):
        """t.open_r(file) and t.open_w(file) implement
        t.open(file, 'r') and t.open(file, 'w') respectively."""
        if not self.steps:
            return open(file, 'r')
        if self.steps[-1][1] == SINK:
            raise ValueError, \
                  'Template.open_r: pipeline ends width SINK'
        cmd = self.makepipeline(file, '')
        return os.popen(cmd, 'r')

    def open_w(self, file):
        if not self.steps:
            return open(file, 'w')
        if self.steps[0][1] == SOURCE:
            raise ValueError, \
                  'Template.open_w: pipeline begins with SOURCE'
        cmd = self.makepipeline('', file)
        return os.popen(cmd, 'w')

    def copy(self, infile, outfile):
        return os.system(self.makepipeline(infile, outfile))

    def makepipeline(self, infile, outfile):
        cmd = makepipeline(infile, self.steps, outfile)
        if self.debugging:
            print cmd
            cmd = 'set -x; ' + cmd
        return cmd


def makepipeline(infile, steps, outfile):
    # Build a list with for each command:
    # [input filename or '', command string, kind, output filename or '']

    list = []
    for cmd, kind in steps:
        list.append(['', cmd, kind, ''])
    #
    # Make sure there is at least one step
    #
    if not list:
        list.append(['', 'cat', '--', ''])
    #
    # Take care of the input and output ends
    #
    [cmd, kind] = list[0][1:3]
    if kind[0] == 'f' and not infile:
        list.insert(0, ['', 'cat', '--', ''])
    list[0][0] = infile
    #
    [cmd, kind] = list[-1][1:3]
    if kind[1] == 'f' and not outfile:
        list.append(['', 'cat', '--', ''])
    list[-1][-1] = outfile
    #
    # Invent temporary files to connect stages that need files
    #
    garbage = []
    for i in range(1, len(list)):
        lkind = list[i-1][2]
        rkind = list[i][2]
        if lkind[1] == 'f' or rkind[0] == 'f':
            (fd, temp) = tempfile.mkstemp()
            os.close(fd)
            garbage.append(temp)
            list[i-1][-1] = list[i][0] = temp
    #
    for item in list:
        [inf, cmd, kind, outf] = item
        if kind[1] == 'f':
            cmd = 'OUT=' + quote(outf) + '; ' + cmd
        if kind[0] == 'f':
            cmd = 'IN=' + quote(inf) + '; ' + cmd
        if kind[0] == '-' and inf:
            cmd = cmd + ' <' + quote(inf)
        if kind[1] == '-' and outf:
            cmd = cmd + ' >' + quote(outf)
        item[1] = cmd
    #
    cmdlist = list[0][1]
    for item in list[1:]:
        [cmd, kind] = item[1:3]
        if item[0] == '':
            if 'f' in kind:
                cmd = '{ ' + cmd + '; }'
            cmdlist = cmdlist + ' |\n' + cmd
        else:
            cmdlist = cmdlist + '\n' + cmd
    #
    if garbage:
        rmcmd = 'rm -f'
        for file in garbage:
            rmcmd = rmcmd + ' ' + quote(file)
        trapcmd = 'trap ' + quote(rmcmd + '; exit') + ' 1 2 3 13 14 15'
        cmdlist = trapcmd + '\n' + cmdlist + '\n' + rmcmd
    #
    return cmdlist


# Reliably quote a string as a single argument for /bin/sh

# Safe unquoted
_safechars = frozenset(string.ascii_letters + string.digits + '@%_-+=:,./')

def quote(file):
    """Return a shell-escaped version of the file string."""
    for c in file:
        if c not in _safechars:
            break
    else:
        if not file:
            return "''"
        return file
    # use single quotes, and put single quotes into double quotes
    # the string $'b is then quoted as '$'"'"'b'
    return "'" + file.replace("'", "'\"'\"'") + "'"