Hi Darren, Glad that's useful. FYI, I was trying to get you an answer using the `&&` Conjunction method as well, but fell short because you want to filter out certain literal words (i.e. negation). That's why I simplified your regex pattern to `<wb> <alpha>+ <wb>`. I trust you know your specifications.
Another approach might be good old `grep`, for which there two different negation forms described in the Docs: https://docs.raku.org/routine/grep ~ % raku -ne 'BEGIN my regex K { null | false | true }; put .words.grep(none / <K> /);' cats_dogs_jays.txt cat dog jay cat dog jay ~ % raku -ne 'BEGIN my regex K { null | false | true }; put .words.grep( {!/ <K> /} );' cats_dogs_jays.txt cat dog jay cat dog jay ~ % raku -ne 'BEGIN my regex K { null | false | true }; put .words.grep( {!/ <K> /} ).grep(/ <[A..Z]> /);' cats_dogs_jays.txt ~ % raku -ne 'BEGIN my regex K { null | false | true }; put .words.grep( {!/ <K> /} ).grep(/ <[a..z]> /);' cats_dogs_jays.txt cat dog jay cat dog jay HTH, Bill. On Sun, Jul 30, 2023 at 9:48 PM Darren Duncan <dar...@darrenduncan.net> wrote: > Thank you William, that boolean condition check option looks like it is in > the > direction of the answer I sought. > > FYI, the reason I spelled out the character class explicitly rather than > using > "<alpha>" was because I wanted it strictly applied to ASCII chars and not > everything Unicode considers a char, and this seemed the best way to be > sure. > > -- Darren Duncan > > On 2023-07-30 9:30 p.m., William Michels via perl6-users wrote: > > Hi Darren (and Marcel), > > > > Two different approaches: > > https://docs.raku.org/language/regexes#Conjunction:_&&; > > <https://docs.raku.org/language/regexes#Conjunction:_&&;> > > > > From the docs: > > /"For example if you have a regex quoted that matches a quoted string, > then `/ > > <quoted> && <-[x]>* /` matches a quoted string that does not contain the > > character `x`."/ > > > > Second approach : > > > https://stackoverflow.com/questions/64909029/is-it-possible-to-do-boolean-assertions-with-raku-regex > > https://docs.raku.org/language/regexes#Regex_Boolean_condition_check > > > > Testing "Regex Boolean condition check" with a one-liner: > > > > ~ % cat cats_dogs_jays.txt > > cat > > dog > > jay > > null cat > > false dog > > true jay > > ~ % raku -ne 'put $/ if $_ ~~ m:g/ ( <wb> <alpha>+ <wb> ) <!{ $0 eq > "null" | > > "false" | "true" }> /;' cats_dogs_jays.txt > > cat > > dog > > jay > > cat > > dog > > jay > > > > > > HTH, Bill. > > > > > > > >> On Jul 30, 2023, at 08:59, Marcel Timmerman wrote: > >> > >> On 30-07-2023 06:21, Darren Duncan wrote: > >>> Hello, I have a Raku regex question. > >>> > >>> See the following: > >>> > >>> token nonquoted_alphanumeric_text > >>> { > >>> <!before [null | false | true] <wb>> > >>> <[ A..Z _ a..z ]> <[ 0..9 A..Z _ a..z ]>* > >>> } > >>> > >>> What I want is for "nonquoted_alphanumeric_text" to match any simple > ASCII > >>> bareword EXCEPT a few special cases indicated in the example. > >>> > >>> I was hoping there might be some better way of specifying this than my > example. > >>> > >>> Is there any more direct way in Raku to say, match this pattern > initially, > >>> but if the result equals these exceptional values then treat it as > having not > >>> matched. > >>> > >>> I am looking for a fully declarative solution in the grammar itself, > not > >>> something involving post-processing. > >>> > >>> Thank you. > >>> > >>> -- Darren Duncan > >>> > >> You can try: <|wb><ALPHA><ALNUM>* > >
