Re: Dynamically inspecting Regex OP-Codes at runtime?

[Anonymous Monk]

I asked in Rx-0.53/dump_regexes.pl as a module because I wanted it to simplify YAPE::Regex::Explain -- easier to extend if perl does the tokenizing

I did come accross re::engine::Hooks and Devel::RegExp, but couldn't compile either at the time ... and its all perlreguts to me, but t/re-engine-Hooks-TestDist/TestDist.xs does look promising

[LanX]

thanks for the pointers! =)

I'm still hoping to find something simpler w/o needing to compile XS-modules, preferably within the core.

Maybe my hack in Re^2: Is there really no @LAST_MATCH_START equivalent for named capture groups? is of help for you!

[Anonymous Monk]

I'm still hoping to find something simpler w/o needing to compile XS-modules, preferably within the core.

Pipe dream if you ask me, look at http://perl5.git.perl.org/perl.git/blob#l14597 and http://perl5.git.perl.org/perl.git/blob?f=regcomp.h#l778

I imagine you might be able to peek at a regex structure using B, but you can't get at the debug messages, they're not stored, and you can't hook DEBUG_COMPILE macros

[LanX]

Thx for the links but thats Chinese for me.

I'm not a versatile C programmer and AFAIK even C gurus get depressed when diving into perlguts.

(BTW: the first link is defunc)

> I imagine you might be able to peek at a regex structure using B,

that would be nice...

> but you can't get at the debug messages, they're not stored, and you can't hook DEBUG_COMPILE macros

What exactly do you mean with "debug messages"? The token tree with commands like OPEN1?

Or just message like

anchored "1234" at 0 (checking anchored) minlen 4 
Freeing REx: "(?<C>1)(?<D>2)(3)(?<A>4)"

The latter I don't need! =)

[Anonymous Monk]

forgot the :)