Getting values of named subpatterns in erlang

Question

Hi I have a following regexp and value

2> re:run("first second", "^(?<foo>\\w+) (?<bar>\\w+)$", [{capture, [foo, bar], list}]).
{match,["first","second"]}
3> 

Here I matched foo with "first" and bar with "second". The problem is in my app (url mapper), I do not know how many named sub-patterns there will be and what their names will be. So I want them to be matched something like

2> re:magic_run("first second", "^(?<foo>\\w+) (?<bar>\\w+)$" ).
{match,[{foo, "first"},{bar, "second"}]}

My concern is not the output format. I want to able to match values with subpattern names. Is there a way to pair the values with subpatterns?

Answer 1

For people who are interested in this in 2022. You can get the values for the named capture groups like this:

1> re:run("cat",".+(?P<A>at)", [{capture,['A'],list}]).
{match,["at"]}
2> re:run("catlololo",".+(?P<A>at)(?P<C>lol.+)", [{capture,['A','C'],list}]).
{match,["at","lololo"]}

The names of the capture groups after the capture specify which values to read. ref: https://www.erlang.org/doc/man/re.html#run-3

Answer 2

If your string is not regular, do not use regular expressions. Especially in your case I would highly recommend using a String.split()-method of some kind (never programmed Erlang, but I think someone might have implemented this).