diff options
| author | Tom Lane <tgl@sss.pgh.pa.us> | 2021-08-09 11:26:34 -0400 |
|---|---|---|
| committer | Tom Lane <tgl@sss.pgh.pa.us> | 2021-08-09 11:26:34 -0400 |
| commit | 0e6aa8747d439bb7f08f95e358f0509c50396785 (patch) | |
| tree | 64977d63a54b4f5c266692966282c6d7aab3aba7 /src/include/regex/regguts.h | |
| parent | 76ad24400d73fa10d527844d50bedf7dacb1e87b (diff) | |
| download | postgresql-0e6aa8747d439bb7f08f95e358f0509c50396785.tar.gz postgresql-0e6aa8747d439bb7f08f95e358f0509c50396785.zip | |
Avoid determining regexp subexpression matches, when possible.
Identifying the precise match locations for parenthesized subexpressions
is a fairly expensive task given the way our regexp engine works, both
at regexp compile time (where we must create an optimized NFA for each
parenthesized subexpression) and at runtime (where determining exact
match locations requires laborious search).
Up to now we've made little attempt to optimize this situation. This
patch identifies cases where we know at compile time that we won't
need to know subexpression match locations, and teaches the regexp
compiler to not bother creating per-subexpression regexps for
parenthesis pairs that are not referenced by backrefs elsewhere in
the regexp. (To preserve semantics, we obviously still have to
pin down the match locations of backref references.) Users could
have obtained the same results before this by being careful to
write "non capturing" parentheses wherever possible, but few people
bother with that.
Discussion: https://postgr.es/m/2219936.1628115334@sss.pgh.pa.us
Diffstat (limited to 'src/include/regex/regguts.h')
| -rw-r--r-- | src/include/regex/regguts.h | 7 |
1 files changed, 4 insertions, 3 deletions
diff --git a/src/include/regex/regguts.h b/src/include/regex/regguts.h index 8db631c83bb..91a52840c47 100644 --- a/src/include/regex/regguts.h +++ b/src/include/regex/regguts.h @@ -477,13 +477,14 @@ struct subre #define MIXED 04 /* mixed preference below */ #define CAP 010 /* capturing parens here or below */ #define BACKR 020 /* back reference here or below */ +#define BRUSE 040 /* is referenced by a back reference */ #define INUSE 0100 /* in use in final tree */ -#define NOPROP 03 /* bits which may not propagate up */ +#define UPPROP (MIXED|CAP|BACKR) /* flags which should propagate up */ #define LMIX(f) ((f)<<2) /* LONGER -> MIXED */ #define SMIX(f) ((f)<<1) /* SHORTER -> MIXED */ -#define UP(f) (((f)&~NOPROP) | (LMIX(f) & SMIX(f) & MIXED)) +#define UP(f) (((f)&UPPROP) | (LMIX(f) & SMIX(f) & MIXED)) #define MESSY(f) ((f)&(MIXED|CAP|BACKR)) -#define PREF(f) ((f)&NOPROP) +#define PREF(f) ((f)&(LONGER|SHORTER)) #define PREF2(f1, f2) ((PREF(f1) != 0) ? PREF(f1) : PREF(f2)) #define COMBINE(f1, f2) (UP((f1)|(f2)) | PREF2(f1, f2)) char latype; /* LATYPE code, if lookaround constraint */ |