def _endswith_tagtable_rest_of_line(text):
return (
# Is the current line the end of record marker?
(None, TT.Word, text, +8, +1),
# Read whatever else is on that line (could be nothing)
(None, TT.AllInSet, TT.invset('\r\n'), +1, +1),
# Get the end of line
("end", TT.Is, '\n', +1, -2), # matches '\n'
(None, TT.Is, '\r', +4, +1),
("end", TT.Is, '\n', +1, -4),
(None, TT.Skip, -1, +1, +1),
("end", TT.Skip, +1, -6, -6),
# Check if EOF (only tests when the end of record line has no \n)
# Only time this should fail is with a bug in TT.
("end", TT.EOF, TT.Here, TT.MatchFail, TT.MatchOk),
# Not the end of record marker, so read to the end of line
(None, TT.AllInSet, TT.invset('\r\n'), +1, +1),
# Check if EOF
(None, TT.EOF, TT.Here, +1, TT.MatchOk),
# Not EOF, so scarf any newlines and try again
(None, TT.AllInSet, TT.set('\r\n'), TT.MatchFail, -10),
)
def match(self, url, datetime=None):
""" Return 1/0 depending on whether the Cookie matches
the given url or not.
datetime is used to check for expiration in case the
Cookie is a temporary one. It defaults to the current
date/time.
"""
url = URL.URL(url)
if self.expires is not None:
if datetime is None:
datetime = DateTime.now()
if self.expires < datetime:
if _debug:
print 'expired'
return 0
if TextTools.prefix(url.path, (self.path, )) is None:
if _debug:
print 'path does not match'
return 0
if TextTools.suffix(url.host, (self.domain, )) is None:
if _debug:
print 'domain does not match'
return 0
return 1
def _startswith_tagtable_rest_of_line(text):
return (
# Ensure the text starts with the given word
("begin", TT.Word, text, TT.MatchFail, +1),
# Read to the end of line
(None, TT.AllInSet, TT.invset('\r\n'), +1, +1),
# Read the end of line
(None, TT.Is, '\n', +1, +4), # matches '\n' or
(None, TT.Is, '\r', +2, +1), # '\r' followed by
(None, TT.Is, '\n', +2, +2), # optional '\n'
# Check if EOF (allow EOF if no EOL found)
(None, TT.EOF, TT.Here, +1, TT.MatchOk),
# Not EOF, so look for the next line starting with text
("begin", TT.Word, text, +1, -5),
# Not what I am looking for, so read to the end of line
(None, TT.AllInSet, TT.invset('\r\n'), +1, +1),
# Read the end of line then test the next line
(None, TT.Is, '\n', +1, -2), # '\n'
(None, TT.Is, '\r', +2, +1), # '\r' followed by
(None, TT.Is, '\n', -4, -4), # optional '\n'
# Allow termination at EOF
(None, TT.EOF, TT.Here, TT.MatchFail, TT.MatchOk),
)
def match(self, url, datetime=None):
""" Return 1/0 depending on whether the Cookie matches
the given url or not.
datetime is used to check for expiration in case the
Cookie is a temporary one. It defaults to the current
date/time.
"""
url = URL.URL(url)
if self.expires is not None:
if datetime is None:
datetime = DateTime.now()
if self.expires < datetime:
if _debug:
print 'expired'
return 0
if TextTools.prefix(url.path, (self.path,)) is None:
if _debug:
print 'path does not match'
return 0
if TextTools.suffix(url.host, (self.domain,)) is None:
if _debug:
print 'domain does not match'
return 0
return 1
def check_assert(text, x, end, tag_words):
result, taglist, pos = TT.tag(text, tag_words, x, end)
if result:
# This succeeded, move forward 1, to be removed later
return x+1
# failed
return x
def check_assert_not(text, x, end, tagtable):
result, taglist, pos = TT.tag(text, tagtable, x, end)
if result:
# This failed
return x
# On success, move forward 1, to be removed later
return x + 1
def check_assert(text, x, end, tag_words):
result, taglist, pos = TT.tag(text, tag_words, x, end)
if result:
# This succeeded, move forward 1, to be removed later
return x + 1
# failed
return x
def _parse_elements(s, tagtable, cont_handler, debug_level, attrlookup):
"""parse the string with the tagtable and send the ContentHandler events
Specifically, it sends the startElement, endElement and characters
events but not startDocument and endDocument.
"""
if debug_level:
import Generate
Generate._position = 0
result, taglist, pos = TextTools.tag(s, tagtable, 0, len(s))
# Special case test for the base ContentHandler since I know that
# object does nothing and I want to test the method call overhead.
if isinstance(cont_handler, Dispatch.Dispatcher):
_do_dispatch_callback(s, 0, pos, taglist,
cont_handler._start_table.get, cont_handler,
cont_handler._save_stack,
cont_handler._end_table.get, attrlookup)
elif cont_handler.__class__ != handler.ContentHandler:
# Send any tags to the client (there can be some even if there
_do_callback(s, 0, pos, taglist, cont_handler, attrlookup)
if not result:
if debug_level:
return ParserPositionException(Generate._position)
else:
return ParserPositionException(pos)
elif pos != len(s):
return pos
else:
return None
def _parse_elements(s, tagtable, cont_handler, debug_level, attrlookup):
"""parse the string with the tagtable and send the ContentHandler events
Specifically, it sends the startElement, endElement and characters
events but not startDocument and endDocument.
"""
if debug_level:
import Generate
Generate._position = 0
result, taglist, pos = TextTools.tag(s, tagtable, 0, len(s))
# Special case test for the base ContentHandler since I know that
# object does nothing and I want to test the method call overhead.
if isinstance(cont_handler, Dispatch.Dispatcher):
_do_dispatch_callback(s, 0, pos, taglist,
cont_handler._start_table.get,
cont_handler, cont_handler._save_stack,
cont_handler._end_table.get,
attrlookup)
elif cont_handler.__class__ != handler.ContentHandler:
# Send any tags to the client (there can be some even if there
_do_callback(s, 0, pos, taglist, cont_handler, attrlookup)
if not result:
if debug_level:
return ParserPositionException(Generate._position)
else:
return ParserPositionException(pos)
elif pos != len(s):
return pos
else:
return None
def search_bench(word, text):
iterations = Tools.trange(COUNT)
print ('Searching for all occurences of %r using ...' % word)
t0 = time.time()
so = TextTools.TextSearch(word)
for i in iterations:
l = so.findall(text)
t1 = time.time()
count = len(l)
print (' - mx.TextSearch.TextSearch().findall(): %5.3f ms (%i)' %
((t1 - t0) / COUNT * 1000.0, count))
t0 = time.time()
so = re.compile(word)
for i in iterations:
l = so.findall(text)
t1 = time.time()
count = len(l)
print (' - re.compile().findall(): %5.3f ms (%i)' %
((t1 - t0) / COUNT * 1000.0, count))
t0 = time.time()
for i in iterations:
count = text.count(word)
t1 = time.time()
print (' - text.count(): %5.3f ms (%i)' %
((t1 - t0) / COUNT * 1000.0, count))
class _modinit:
# Reserved URL chars as defined by RFC2396
unsafe_charset = TextTools.set(\
'abcdefghijklmnopqrstuvwxyz'
'ABCDEFGHIJKLMNOPQRSTUVWXYZ'
'0123456789'
'-_.!~*\'()',0)
# Modified version of the above set which includes even fewer
# characters (esp. dots and quotes are not included)
rpc_unsafe_charset = TextTools.set(\
'abcdefghijklmnopqrstuvwxyz'
'ABCDEFGHIJKLMNOPQRSTUVWXYZ'
'0123456789'
'-_()',0)
def _startswith_tagtable_newline(text):
return (
# Ensure the text starts with the given word ...
("begin", TT.Word, text, TT.MatchFail, +1),
# ... followed by the end of line
(None, TT.Is, '\n', +1, +4), # matches '\n' or
(None, TT.Is, '\r', +2, +1), # '\r' followed by
(None, TT.Is, '\n', +2, +2), # optional '\n'
# Check if EOF instead of a newline (allow EOF if found)
# Otherwise, this means the line starts with the text but
# doesn't have a successive newline.
# XXX BUG! When looking for "A\n" should not fail on "AA\n"!
(None, TT.EOF, TT.Here, TT.MatchFail, TT.MatchOk),
# Look for the next line starting with text
("begin", TT.Word, text, +1, -4),
# Not what I am looking for, so read to the end of line
(None, TT.AllInSet, TT.invset('\r\n'), +1, +1),
# Read the end of line then test the next line
(None, TT.Is, '\n', +1, -2), # '\n'
(None, TT.Is, '\r', +2, +1), # '\r' followed by
(None, TT.Is, '\n', -4, -4), # optional '\n'
# Allow termination at EOF
(None, TT.EOF, TT.Here, TT.MatchFail, TT.MatchOk),
)
def check_assert_not(text, x, end, tagtable):
result, taglist, pos = TT.tag(text, tagtable, x, end)
if result:
# This failed
return x
# On success, move forward 1, to be removed later
return x + 1
def _find_begin_positions(text, tagtable):
success, tags, pos = TT.tag(text, tagtable)
# print "XXX", success, tags, pos, len(text)
if not success:
raise ReaderError("invalid format starting with %s" % repr(text[:50]))
if pos != len(text):
raise ReaderError, \
"could not parse to end of text (ended at %d of %d)" % \
(pos, len(text))
return [tag[1] for tag in tags]
def _endswith_tagtable_newline(text):
return (
# Is the current line the end of record marker?
(None, TT.Word, text, +6, +1),
# Make sure it ends the line
("end", TT.Is, '\n', +1, -1), # matches '\n'
(None, TT.Is, '\r', +4, +1),
("end", TT.Is, '\n', +1, -3),
(None, TT.Skip, -1, +1, +1),
("end", TT.Skip, +1, -5, -5),
# Not the end of record marker, so read to the end of line
(None, TT.AllInSet, TT.invset('\r\n'), +1, +1),
# Check if EOF
(None, TT.EOF, TT.Here, +1, TT.MatchOk),
# Not EOF, so scarf any newlines
(None, TT.AllInSet, TT.set('\r\n'), TT.MatchFail, -8),
)
def call(self, text, x, end):
# Called by 'TextTools.Call' to detect a match.
# I do the full match here and store the results for later use.
# If successful, I return x+1, else return x+0 (the +1/-1 trick)
min_count, max_count = self._get_ranges()
assert min_count == max_count, \
"cannot have different sizes: %s %s" % (min_count, max_count)
tagtable = self.tagtable * min_count
result, taglist, pos = TT.tag(text, tagtable, x, end)
if result == 1:
# Store the taglist for later use
self.taglist = taglist
return pos + 1 # +1 because {0} is allowed; Skip -1 later
else:
self.taglist = None
return x
def call(self, text, x, end):
# Called by 'TextTools.Call' to detect a match.
# I do the full match here and store the results for later use.
# If successful, I return x+1, else return x+0 (the +1/-1 trick)
min_count, max_count = self._get_ranges()
assert min_count == max_count, \
"cannot have different sizes: %s %s" % (min_count, max_count)
tagtable = self.tagtable * min_count
result, taglist, pos = TT.tag(text, tagtable, x, end)
if result == 1:
# Store the taglist for later use
self.taglist = taglist
return pos + 1 # +1 because {0} is allowed; Skip -1 later
else:
self.taglist = None
return x
def generate_dot(expression, genstate):
return [(None, TT.IsInSet, TT.invset('\n')), ]
def normalize(s):
return string.join(TT.splitlines(s), "\n")
def generate_dot(expression, genstate):
return [
(None, TT.IsInSet, TT.invset('\n')),
]
fake = self.text + "\n"
reader = StartsWith(self.infile, self.text, self.sizehint,
fake + self.lookahead)
rec = reader.next()
rec = rec[len(fake):] # remove the fake data
self.infile, self.lookahead = reader.remainder()
self.found = 1
return rec
def remainder(self):
return self.infile, self.lookahead
# Tag the last byte of every newline
_tag_lines_tagtable = (
# Skip non-newline characters
(None, TT.AllInSet, TT.invset('\r\n'), +1, +1),
# Check if newline
("newline", TT.Is, '\n', +1, -1), # can be '\n'
(None, TT.Is, '\r', +3, +1), # or start a '\r' followed by ..
("newline", TT.Is, '\n', +1, -3), # .. an optional '\n'
("newline", TT.Skip, 0, -4, -4), # get here with just an '\r'
(None, TT.EOF, TT.Here, -5, TT.MatchOk), # stop at end of text
)
class CountLines(RecordReader):
"""Read a specified (fixed) number of lines"""
def __init__(self, infile, count, sizehint = SIZEHINT, lookahead = ""):
assert count > 0, "CountLines reader must read at least one line"
assert lookahead > 0, "Must read at least a character at a time"
def normalize(s):
return string.join(TT.splitlines(s), "\n")
