/Demo/comparisons/regextest.py

 1#! /usr/bin/env python
 2
 3# 1)  Regular Expressions Test
 4#
 5#     Read a file of (extended per egrep) regular expressions (one per line),
 6#     and apply those to all files whose names are listed on the command line.
 7#     Basically, an 'egrep -f' simulator.  Test it with 20 "vt100" patterns
 8#     against a five /etc/termcap files.  Tests using more elaborate patters
 9#     would also be interesting.  Your code should not break if given hundreds
10#     of regular expressions or binary files to scan.
11
12# This implementation:
13# - combines all patterns into a single one using ( ... | ... | ... )
14# - reads patterns from stdin, scans files given as command line arguments
15# - produces output in the format <file>:<lineno>:<line>
16# - is only about 2.5 times as slow as egrep (though I couldn't run
17#   Tom's test -- this system, a vanilla SGI, only has /etc/terminfo)
18
19import string
20import sys
21import re
22
23def main():
24    pats = map(chomp, sys.stdin.readlines())
25    bigpat = '(' + '|'.join(pats) + ')'
26    prog = re.compile(bigpat)
27
28    for file in sys.argv[1:]:
29        try:
30            fp = open(file, 'r')
31        except IOError, msg:
32            print "%s: %s" % (file, msg)
33            continue
34        lineno = 0
35        while 1:
36            line = fp.readline()
37            if not line:
38                break
39            lineno = lineno + 1
40            if prog.search(line):
41                print "%s:%s:%s" % (file, lineno, line),
42
43def chomp(s):
44    return s.rstrip('\n')
45
46if __name__ == '__main__':
47    main()