On 2014-07-04 13:27, Florian Lindner wrote:
Hello,I have that piece of code: def _split_block(self, block): cre = [re.compile(r, flags = re.MULTILINE) for r in self.regexps] block = "".join(block) print(block) print("-------------------") for regexp in cre: match = regexp.match(block) for grp in regexp.groupindex: data = match.group(grp) if match else None self.data[grp].append(data) block is a list of strings, terminated by \n. self.regexps: self.regexps = [r"it (?P<coupling_iterations>\d+) .* dt complete yes | write-iteration-checkpoint |", r"it (?P<it_read_ahead>\d+) read ahead" If I run my program it looks like that: it 1 ahadf dt complete yes | write-iteration-checkpoint | Timestep completed ------------------- it 1 read ahead it 2 ahgsaf dt complete yes | write-iteration-checkpoint | Timestep completed ------------------- it 4 read ahead it 3 dfdsag dt complete yes | write-iteration-checkpoint | Timestep completed ------------------- it 9 read ahead it 4 dsfdd dt complete yes | write-iteration-checkpoint | Timestep completed ------------------- it 16 read ahead ------------------- {'it_read_ahead': [None, '1', '4', '9', '16'], 'coupling_iterations': ['1', None, None, None, None]} it_read_ahead is always matched when it should (all blocks but the first). But why is the regexp containing coupling_iterations only matched in the first block? I tried different combinations using re.match vs. re.search and with or without re.MULTILINE.
The character '|' is a metacharacter that separates alternatives. For example, the regex 'a|b' will match 'a' or b'. Your regexes end with '|', which means that they will match an empty string at the start of the target string. -- https://mail.python.org/mailman/listinfo/python-list
