Regular expressions are a powerful and (mostly) standardized way of searching, replacing, and parsing text with complex patterns of characters.

In Python, all functionality related to regular expression is in re module.

re.compile(pattern[, flags]): compiles a RegEx into an object. It is recommended to use raw Python strings r'someString' for pattern because it doesn’t require escaping. It can contain an optional prefix ?P<PATTERNNAME> to assign a name to the pattern.

There are some useful flags to customize the behavior of various methods.

  • re.IGNORECASE or re.I enables case-insensitive text operations;
  • re.M or re.MULTILINE enables multi-line environment so that ^ matches the beginning of the string and the beginning of the newline, and $ matches the end of the string and the end of each line.
  • re.A or re.ASCII Make \w, \W, \b, \B, \d, \D, \s and \S perform ASCII-only matching instead of full Unicode matching defaulted in Python 3.
  • re.S or re.DOTALL let . match all characters including newline.
  • re.X or re.VERBOSE enables verbose regular expressions. In verbose regular expression, whitespace (spaces, tabs, carriage returns) and comments (everything follows a #) are ignored. To escape a whitespace, a \ is required. We can write verbose regular expression as a multiline string, and comment every part of it in a separated line. For example, a and b are functionally the same:
    a = re.compile(r"""\d +  # the integral part
                     \.    # the decimal point
                     \d *  # some fractional digits
                      """, re.X)
    b = re.compile(r"\d+\.\d*")
    

Parentheses in pattern string define groups, which can be extracted by group(id) later. For example, ^(\d{3})-(\d{3,8})$ defines two groups. group(0) is the whole match, and group(1) and group(2) are the two matching groups.

By default, regex match is greedy, that is, it will match as many characters as possible. By adding a ? to a pattern`, it becomes non-greedy, i.e. match as few characters as possible.

Functions for Regex

There are two ways to use functions in re module:

  • re.match(pattern, string[, flags])
  • RegexObject.match(string,[ pos[, endpos]]), and RegexObject is precompiled with re.compile.

If a match is found, a match object is returned; otherwise, None is returned. Match object is always evaluated to True, so we can do this:

if re.search(regexp, exr):
	doSomething()
  • match finds pattern from the beginning of the string. RegexObject.match(string[, pos[, endpos]]) will match from pos instead of the beginning.
  • fullmatch performs a match on the whole string.
  • search: finds the first location of pattern in a string.
  • splitsplits string by the occurrences of pattern and return a list. It takes an optional maxsplit parameter, default to 0. It is more flexible than str.split method.
  • findallreturns a list of all the substrings that matched the pattern. Note: it doesn’t return overlapping matches.
  • finditerreturns an iterator of all matching substrings, instead of a list.
  • sub: replaces ALL occurrences of pattern in string with replacement. If no occurrences are found, it acts like a no-op. replacement can be a function that takes a single match object and returns the replacement string.
  • subn returns the number of matches too.

Matching Object

  • group(id) returns one or more subgroups of the match. By default, 0 is the whole match, 1, 2, … are matching groups defined by parentheses. If ?P<name>... syntax (to define group name) is used, name can be passed in to group to get the match. It also takes group(id1, id2, ...) to return a tuple of matching groups.
  • groups returns a tuple containing all the subgroups of the match, from 1 up. it takes a default parameter to define unfound match value (default to None).
  • groupdict returns a dictionary containing all the named (?P<name>...) subgroups of the match, keyed by the subgroup name.
  • start and end returns the indices of the start or end of the match. It takes the optional group ID.
  • span returns the indices of the start and end of the match in a tuple.