1999-01-13 12:12:04 -04:00
|
|
|
"""text_file
|
|
|
|
|
|
|
|
provides the TextFile class, which gives an interface to text files
|
|
|
|
that (optionally) takes care of stripping comments, ignoring blank
|
|
|
|
lines, and joining lines with backslashes."""
|
|
|
|
|
|
|
|
# created 1999/01/12, Greg Ward
|
|
|
|
|
|
|
|
__revision__ = "$Id$"
|
|
|
|
|
|
|
|
from types import *
|
1999-01-18 13:08:16 -04:00
|
|
|
import sys, os, string, re
|
1999-01-13 12:12:04 -04:00
|
|
|
|
|
|
|
|
|
|
|
class TextFile:
|
|
|
|
|
|
|
|
default_options = { 'strip_comments': 1,
|
|
|
|
'skip_blanks': 1,
|
|
|
|
'join_lines': 0,
|
|
|
|
'lstrip_ws': 0,
|
|
|
|
'rstrip_ws': 1,
|
1999-03-08 17:46:11 -04:00
|
|
|
'collapse_ws': 0,
|
1999-01-13 12:12:04 -04:00
|
|
|
}
|
|
|
|
|
1999-03-23 10:00:06 -04:00
|
|
|
def __init__ (self, filename=None, file=None, **options):
|
|
|
|
|
|
|
|
if filename is None and file is None:
|
|
|
|
raise RuntimeError, \
|
|
|
|
"you must supply either or both of 'filename' and 'file'"
|
1999-01-13 12:12:04 -04:00
|
|
|
|
|
|
|
# set values for all options -- either from client option hash
|
|
|
|
# or fallback to default_options
|
|
|
|
for opt in self.default_options.keys():
|
|
|
|
if options.has_key (opt):
|
1999-08-10 17:09:38 -03:00
|
|
|
setattr (self, opt, options[opt])
|
1999-01-13 12:12:04 -04:00
|
|
|
|
|
|
|
else:
|
|
|
|
setattr (self, opt, self.default_options[opt])
|
|
|
|
|
|
|
|
# sanity check client option hash
|
|
|
|
for opt in options.keys():
|
|
|
|
if not self.default_options.has_key (opt):
|
|
|
|
raise KeyError, "invalid TextFile option '%s'" % opt
|
|
|
|
|
1999-03-23 10:00:06 -04:00
|
|
|
if file is None:
|
|
|
|
self.open (filename)
|
|
|
|
else:
|
1999-01-13 12:12:04 -04:00
|
|
|
self.filename = filename
|
1999-03-23 10:00:06 -04:00
|
|
|
self.file = file
|
|
|
|
self.current_line = 0 # assuming that file is at BOF!
|
1999-03-26 17:48:59 -04:00
|
|
|
|
1999-03-29 14:01:49 -04:00
|
|
|
# 'linebuf' is a stack of lines that will be emptied before we
|
|
|
|
# actually read from the file; it's only populated by an
|
|
|
|
# 'unreadline()' operation
|
|
|
|
self.linebuf = []
|
1999-03-23 10:00:06 -04:00
|
|
|
|
1999-01-13 12:12:04 -04:00
|
|
|
|
1999-03-23 10:00:06 -04:00
|
|
|
def open (self, filename):
|
|
|
|
self.filename = filename
|
1999-01-13 12:12:04 -04:00
|
|
|
self.file = open (self.filename, 'r')
|
|
|
|
self.current_line = 0
|
|
|
|
|
|
|
|
|
|
|
|
def close (self):
|
|
|
|
self.file.close ()
|
|
|
|
self.file = None
|
|
|
|
self.filename = None
|
|
|
|
self.current_line = None
|
|
|
|
|
|
|
|
|
1999-04-15 14:50:19 -03:00
|
|
|
def warn (self, msg, line=None):
|
|
|
|
if line is None:
|
|
|
|
line = self.current_line
|
1999-01-18 13:08:16 -04:00
|
|
|
sys.stderr.write (self.filename + ", ")
|
1999-08-10 17:09:38 -03:00
|
|
|
if type (line) in (ListType, TupleType):
|
1999-04-15 14:50:19 -03:00
|
|
|
sys.stderr.write ("lines %d-%d: " % tuple (line))
|
1999-01-18 13:08:16 -04:00
|
|
|
else:
|
1999-04-15 14:50:19 -03:00
|
|
|
sys.stderr.write ("line %d: " % line)
|
1999-08-10 17:09:38 -03:00
|
|
|
sys.stderr.write (str (msg) + "\n")
|
1999-01-18 13:08:16 -04:00
|
|
|
|
|
|
|
|
1999-01-13 12:12:04 -04:00
|
|
|
def readline (self):
|
|
|
|
|
1999-03-29 14:01:49 -04:00
|
|
|
# If any "unread" lines waiting in 'linebuf', return the top
|
|
|
|
# one. (We don't actually buffer read-ahead data -- lines only
|
|
|
|
# get put in 'linebuf' if the client explicitly does an
|
|
|
|
# 'unreadline()'.
|
|
|
|
if self.linebuf:
|
|
|
|
line = self.linebuf[-1]
|
|
|
|
del self.linebuf[-1]
|
|
|
|
return line
|
|
|
|
|
1999-01-13 12:12:04 -04:00
|
|
|
buildup_line = ''
|
|
|
|
|
|
|
|
while 1:
|
1999-08-10 17:09:38 -03:00
|
|
|
# read the line, make it None if EOF
|
1999-01-13 12:12:04 -04:00
|
|
|
line = self.file.readline()
|
1999-08-10 17:09:38 -03:00
|
|
|
if line == '': line = None
|
|
|
|
|
1999-01-13 12:12:04 -04:00
|
|
|
if self.strip_comments and line:
|
1999-08-10 17:09:38 -03:00
|
|
|
|
|
|
|
# Look for the first "#" in the line. If none, never
|
|
|
|
# mind. If we find one and it's the first character, or
|
|
|
|
# is not preceded by "\", then it starts a comment --
|
|
|
|
# strip the comment, strip whitespace before it, and
|
|
|
|
# carry on. Otherwise, it's just an escaped "#", so
|
|
|
|
# unescape it (and any other escaped "#"'s that might be
|
|
|
|
# lurking in there) and otherwise leave the line alone.
|
|
|
|
|
|
|
|
pos = string.find (line, "#")
|
|
|
|
if pos == -1: # no "#" -- no comments
|
|
|
|
pass
|
|
|
|
elif pos == 0 or line[pos-1] != "\\": # it's a comment
|
|
|
|
# Have to preserve the trailing newline; if
|
|
|
|
# stripping comments resulted in an empty line, we'd
|
|
|
|
# have no way to distinguish end-of-file! (NB. this
|
|
|
|
# means that if the final line is all comment and
|
|
|
|
# has to trailing newline, we will think that it's
|
|
|
|
# EOF; I think that's OK.)
|
|
|
|
has_newline = (line[-1] == '\n')
|
|
|
|
line = line[0:pos]
|
|
|
|
if has_newline: line = line + '\n'
|
|
|
|
|
|
|
|
else: # it's an escaped "#"
|
|
|
|
line = string.replace (line, "\\#", "#")
|
|
|
|
|
1999-01-13 12:12:04 -04:00
|
|
|
|
|
|
|
# did previous line end with a backslash? then accumulate
|
|
|
|
if self.join_lines and buildup_line:
|
|
|
|
# oops: end of file
|
1999-08-10 17:09:38 -03:00
|
|
|
if line is None:
|
1999-01-13 12:12:04 -04:00
|
|
|
self.warn ("continuation line immediately precedes "
|
|
|
|
"end-of-file")
|
|
|
|
return buildup_line
|
|
|
|
|
|
|
|
line = buildup_line + line
|
|
|
|
|
|
|
|
# careful: pay attention to line number when incrementing it
|
|
|
|
if type (self.current_line) is ListType:
|
|
|
|
self.current_line[1] = self.current_line[1] + 1
|
|
|
|
else:
|
|
|
|
self.current_line = [self.current_line, self.current_line+1]
|
|
|
|
# just an ordinary line, read it as usual
|
|
|
|
else:
|
1999-08-10 17:09:38 -03:00
|
|
|
if line is None: # eof
|
1999-01-13 12:12:04 -04:00
|
|
|
return None
|
|
|
|
|
|
|
|
# still have to be careful about incrementing the line number!
|
|
|
|
if type (self.current_line) is ListType:
|
|
|
|
self.current_line = self.current_line[1] + 1
|
|
|
|
else:
|
|
|
|
self.current_line = self.current_line + 1
|
1999-03-29 14:01:49 -04:00
|
|
|
|
1999-01-13 12:12:04 -04:00
|
|
|
|
|
|
|
# strip whitespace however the client wants (leading and
|
|
|
|
# trailing, or one or the other, or neither)
|
|
|
|
if self.lstrip_ws and self.rstrip_ws:
|
|
|
|
line = string.strip (line)
|
|
|
|
else:
|
|
|
|
if self.lstrip_ws:
|
|
|
|
line = string.lstrip (line)
|
|
|
|
if self.rstrip_ws:
|
|
|
|
line = string.rstrip (line)
|
|
|
|
|
|
|
|
# blank line (whether we rstrip'ed or not)? skip to next line
|
|
|
|
# if appropriate
|
|
|
|
if line == '' or line == '\n' and self.skip_blanks:
|
|
|
|
continue
|
|
|
|
|
|
|
|
if self.join_lines:
|
|
|
|
if line[-1] == '\\':
|
|
|
|
buildup_line = line[:-1]
|
|
|
|
continue
|
|
|
|
|
|
|
|
if line[-2:] == '\\\n':
|
|
|
|
buildup_line = line[0:-2] + '\n'
|
|
|
|
continue
|
|
|
|
|
1999-03-08 17:46:11 -04:00
|
|
|
# collapse internal whitespace (*after* joining lines!)
|
|
|
|
if self.collapse_ws:
|
|
|
|
line = re.sub (r'(\S)\s+(\S)', r'\1 \2', line)
|
|
|
|
|
1999-01-13 12:12:04 -04:00
|
|
|
# well, I guess there's some actual content there: return it
|
|
|
|
return line
|
|
|
|
|
|
|
|
# end readline
|
|
|
|
|
|
|
|
|
|
|
|
def readlines (self):
|
|
|
|
lines = []
|
|
|
|
while 1:
|
|
|
|
line = self.readline()
|
|
|
|
if line is None:
|
|
|
|
return lines
|
|
|
|
lines.append (line)
|
|
|
|
|
|
|
|
|
1999-03-29 14:01:49 -04:00
|
|
|
def unreadline (self, line):
|
|
|
|
self.linebuf.append (line)
|
|
|
|
|
|
|
|
|
1999-01-13 12:12:04 -04:00
|
|
|
if __name__ == "__main__":
|
|
|
|
test_data = """# test file
|
|
|
|
|
|
|
|
line 3 \\
|
|
|
|
continues on next line
|
|
|
|
"""
|
|
|
|
|
|
|
|
# result 1: no fancy options
|
|
|
|
result1 = map (lambda x: x + "\n", string.split (test_data, "\n")[0:-1])
|
|
|
|
|
|
|
|
# result 2: just strip comments
|
|
|
|
result2 = ["\n", "\n", "line 3 \\\n", "continues on next line\n"]
|
|
|
|
|
|
|
|
# result 3: just strip blank lines
|
|
|
|
result3 = ["# test file\n", "line 3 \\\n", "continues on next line\n"]
|
|
|
|
|
|
|
|
# result 4: default, strip comments, blank lines, and trailing whitespace
|
|
|
|
result4 = ["line 3 \\", "continues on next line"]
|
|
|
|
|
|
|
|
# result 5: full processing, strip comments and blanks, plus join lines
|
|
|
|
result5 = ["line 3 continues on next line"]
|
|
|
|
|
|
|
|
def test_input (count, description, file, expected_result):
|
|
|
|
result = file.readlines ()
|
|
|
|
# result = string.join (result, '')
|
|
|
|
if result == expected_result:
|
|
|
|
print "ok %d (%s)" % (count, description)
|
|
|
|
else:
|
|
|
|
print "not ok %d (%s):" % (count, description)
|
|
|
|
print "** expected:"
|
|
|
|
print expected_result
|
|
|
|
print "** received:"
|
|
|
|
print result
|
|
|
|
|
|
|
|
|
|
|
|
filename = "test.txt"
|
|
|
|
out_file = open (filename, "w")
|
|
|
|
out_file.write (test_data)
|
|
|
|
out_file.close ()
|
|
|
|
|
|
|
|
in_file = TextFile (filename, strip_comments=0, skip_blanks=0,
|
1999-08-10 17:09:38 -03:00
|
|
|
lstrip_ws=0, rstrip_ws=0)
|
1999-01-13 12:12:04 -04:00
|
|
|
test_input (1, "no processing", in_file, result1)
|
|
|
|
|
|
|
|
in_file = TextFile (filename, strip_comments=1, skip_blanks=0,
|
1999-08-10 17:09:38 -03:00
|
|
|
lstrip_ws=0, rstrip_ws=0)
|
1999-01-13 12:12:04 -04:00
|
|
|
test_input (2, "strip comments", in_file, result2)
|
|
|
|
|
|
|
|
in_file = TextFile (filename, strip_comments=0, skip_blanks=1,
|
1999-08-10 17:09:38 -03:00
|
|
|
lstrip_ws=0, rstrip_ws=0)
|
1999-01-13 12:12:04 -04:00
|
|
|
test_input (3, "strip blanks", in_file, result3)
|
|
|
|
|
|
|
|
in_file = TextFile (filename)
|
|
|
|
test_input (4, "default processing", in_file, result4)
|
|
|
|
|
|
|
|
in_file = TextFile (filename, strip_comments=1, skip_blanks=1,
|
|
|
|
join_lines=1, rstrip_ws=1)
|
|
|
|
test_input (5, "full processing", in_file, result5)
|
|
|
|
|
|
|
|
os.remove (filename)
|
|
|
|
|