Python Tip #95 (of 365):
Don't use re.match(): it's confusing.
I'm not sure I've ever seen re.match() used when it wasn't being used by mistake.
If you think you want re.match(), you probably want either re.āsearch() or re.fullmatch() instead.
When matching a regex against a string, we're usually either trying to:
1. Find a regular expression within a string
2. Match a regular expression against a whole string
re.match() does neither of those!
š§µ (1/3)
To find a match WITHIN a string, you can use re.āsearch().
To match an entire string against a regex, you can use re.fullmatch().
Using re.fullmatch() is the same as using re.āsearch() with ^ and $ used to anchor to the beginning and end of the string.
In other words, these are equivalent:
match = re.āsearch(r"^\w+$", string)
match = re.fullmatch(r"\w+", string)
re.match() is a weird middle ground between re.āsearch() and re.fullmatch().
š§µ (2/3)
But why are they named this way?
Wouldn't it be simpler if re.fullmatch() was called re.match() or re.āsearch() was called re.match() and the current behavior of re.match() didn't exist at all?
I think so. But it's too late for that.
This is all confusing due an accident of history.
re.fullmatch() didn't used to exist but re.match() did. By the time they invented re.fullmatch(), the name "match" was already taken.
Personally, I usually just stick with re.āsearch().
š§µ (3/3)
@treyhunner In 3.15 we've added an re.prefixmatch() as an explicit alias to the confusing re.match().
https://docs.python.org/3.15/library/re.html#prefixmatch-vs-match
#Python #python315 #regex
Trey Hunner š
Hugo van Kemenade
Seth Larson
Stefan Eissing