1:- module(regex, [ (=~)/2 2 , (\~)/2 3 , op(700,xfx,=~) 4 , op(700,xfx,\~) 5 , regex/4 6 ]). 7:- use_module(library(error), [domain_error/2]). 8:- use_module(library(regex/state), [new_state/3, numbered_captures/2]). 9:- use_module(library(regex/parser), [re//2]). 10:- use_module(library(regex/engine/pp), [engine_match/5]). 11 12 13% operators for matching strings against regular expressions. 14% the syntax is the same used by Perl and Haskell, but Prolog 15% doesn't like '!' in operators so I had to use '\' instead. 16:- op(700,xfx,=~). 17:- op(700,xfx,\~).
Named captures are automatically bound to corresponding named variables in the surrounding scope. For example,
"Hi John" =~ "hi (?<Name>[a-z]+)"/i, Name == "John".
30Text =~ Pattern :- 31 expand_equalstilde(Text =~ Pattern, _, Goal), 32 call(Goal). 33 34expand_equalstilde(Text =~ Pattern, Vars, regex(P,Options,Text,Vars)) :- 35 ( nonvar(Pattern), Pattern = P/Options -> 36 true 37 ; % no explicit options -> 38 P = Pattern, 39 Options = [] 40 ). 41 42% macro expansion giving access to in-scope variables. 43usergoal_expansion(Text =~ Pattern, Goal) :- 44 % is goal expansion wanted? 45 prolog_load_context(module, Module), 46 Module \== regex, % we don't want string interpolation ourselves 47 predicate_property(Module:(_=~_),imported_from(regex)), 48 49 prolog_load_context(variable_names, Vars), 50 expand_equalstilde(Text =~ Pattern, Vars, Goal).
56Text \~ Pattern :-
57 \+ Text =~ Pattern.text
value may either be an atom or a list of codes.
Options can either be an atom or a list of options. If an atom, it's
split into a list of single character atoms which is used as the
Options value. This allows on to use is, for example, instead of
[i,s]. Acceptable options are:
i - case-insensitive (default false)s - let `. match \n` (default false)Captures is unified with a list of captured values, with the leftmost capture first, etc. Each captured value is a list of codes. For example,
?- regex('(a+)(b*)', [], 'aaabbbbb', [A,B]).
A = "aaa",
B = "bbbbb".
Named captures are also supported. In that case, Captures must be
a list of pairs like ['A'=A,'B'=B]. Every named capture in the
pattern must have a corresponding key in Captures. (This is a
temporary restriction and will be removed later).
A brief word on argument order. Prolog convention prefers to place an Options argument as the final argument or as the last one before outputs. However, widely followed regular expression convention places options immediately after the pattern. I chose to follow the latter convention. This argument order benefits higher-order calls like maplist/3 which can do things like:
?- maplist(regex('(a+)(b+)', i), [ab, aab, abb], L).
L = [["a", "b"], ["aa", "b"], ["a", "bb"]].
98regex(Pattern,Options,Text,Captures) :- 99 % normalize text representations 100 text_codes(Text, T), 101 text_codes(Pattern, P0), 102 starts_with_caret(P0,P,StartingCaret), 103 104 % normalize options and captures into a state value 105 new_state(Options, Captures, State0), 106 107 % compile Pattern 108 ( phrase(re(State0,Re),P) -> 109 ( StartingCaret=yes -> 110 once(engine_match(Re, State0, State, T, _)) 111 ; otherwise -> 112 once(regex_no_sugar(Re, State0, State, T, _)) 113 ), 114 ( var(Captures) -> 115 numbered_captures(State, Captures) 116 ; % captures already bound -> 117 true 118 ) 119 ; % invalid pattern -> 120 atom_codes(A, P), 121 domain_error(regex, A) 122 ). 123 124 125starts_with_caret([0'^|P],P,yes) :- % ' syntax highlighter 126 !. 127starts_with_caret(P,P,no). 128 129 130% the heart and soul of regex/4 131regex_no_sugar(Re, State0, State) --> 132 engine_match(Re, State0, State). 133regex_no_sugar(Re, State0, State) --> 134 [_], 135 regex_no_sugar(Re, State0, State).
141text_codes(Atom, Codes) :- 142 atom(Atom), 143 !, 144 atom_codes(Atom, Codes). 145text_codes(String, Codes) :- 146 string(String), 147 !, 148 string_codes(String, Codes). 149text_codes(Codes, Codes)