On Wednesday 12 March 2008 23:38:46 [EMAIL PROTECTED] wrote:
> Author: duff
> Date: Wed Mar 12 23:38:44 2008
> New Revision: 26345
>
> Modified:
> trunk/languages/perl6/src/pmc/perl6str.pmc
>
> Log:
> [rakudo] rough cut implementation of auto-{increment,decrement} for
> Perl6Str
>
>
> Modified: trunk/languages/perl6/src/pmc/perl6str.pmc
> ===========================================================================
>=== --- trunk/languages/perl6/src/pmc/perl6str.pmc (original)
> +++ trunk/languages/perl6/src/pmc/perl6str.pmc Wed Mar 12 23:38:44 2008
> @@ -165,18 +165,136 @@
> =item C<void decrement()>
>
> Increment/decrement the string magically according to S03 rules.
> -For the moment these are stub implementations to avoid throwing
> -exceptions when we attempt to increment/decrement a string.
> +
> +These implementations are very ASCII oriented. They assume that the
> +alphabet is contiguous and that there aren't any other characters on
> +either side of the letters or digits that return true for C<isalpha()> or
> +C<isdigit()> respectively.
>
> */
>
> - void increment() {
> - /* do nothing */
> + VTABLE void increment() {
> + STRING* str = PMC_str_val(SELF);
> + const char *start = str->strstart;
> + const char* const end = start + string_length(INTERP, str);
> +
> + char* substr; /* pointer to the substring we're going to
> increment */ + INTVAL len; /* length of the substring */
> + INTVAL i, carry;
> +
> + const char* s = start; /* start of the current incrementable
> substring */
If you find yourself tempted to add a comment to explain what the variable
name stands for, fix the variable name instead -- it's a lot easier to read
the subsequent code without scrolling back up to read comment flags every
time you can't tell what 'ep' means.
I'm tidying these PMCs now, so I might get to revising this code in the next
few days... but if I don't, feel free to use something like "prev_end"
and "prev_start", for example.
-- c
