Fix more multibyte issues in ltree.

Commit 84d5efa7e3 missed some multibyte issues caused by short-circuit
logic in the callers. The callers assumed that if the predicate string
is longer than the label string, then it couldn't possibly be a match,
but it can be when using case-insensitive matching (LVAR_INCASE) if
casefolding changes the byte length.

Fix by refactoring to get rid of the short-circuit logic as well as
the function pointer, and consolidate the logic in a replacement
function ltree_label_match().

Discussion: https://postgr.es/m/02c6ef6cf56a5013ede61ad03c7a26affd27d449.camel@j-davis.com
Backpatch-through: 14

diff --git a/contrib/ltree/lquery_op.c b/contrib/ltree/lquery_op.c
index 0adcdd8ff2ac63248df4cacfa87b69515804bf0f..e6a1969c3d39a0f418e372f0ad0835f9bdc1f189 100644 (file)
--- a/contrib/ltree/lquery_op.c
+++ b/contrib/ltree/lquery_op.c
@@ -41,8 +41,7 @@ getlexeme(char *start, char *end, int *len)
 }
 
 bool
-compare_subnode(ltree_level *t, char *qn, int len,
-                               ltree_prefix_eq_func prefix_eq, bool anyend)
+compare_subnode(ltree_level *t, char *qn, int len, bool prefix, bool ci)
 {
        char       *endt = t->name + t->len;
        char       *endq = qn + len;
@@ -57,10 +56,8 @@ compare_subnode(ltree_level *t, char *qn, int len,
                isok = false;
                while ((tn = getlexeme(tn, endt, &lent)) != NULL)
                {
-                       if ((lent == lenq || (lent > lenq && anyend)) &&
-                               (*prefix_eq) (qn, lenq, tn, lent))
+                       if (ltree_label_match(qn, lenq, tn, lent, prefix, ci))
                        {
-
                                isok = true;
                                break;
                        }
@@ -76,64 +73,69 @@ compare_subnode(ltree_level *t, char *qn, int len,
 }
 
 /*
- * Check if 'a' is a prefix of 'b'.
- */
-bool
-ltree_prefix_eq(const char *a, size_t a_sz, const char *b, size_t b_sz)
-{
-       if (a_sz > b_sz)
-               return false;
-       else
-               return (strncmp(a, b, a_sz) == 0);
-}
-
-/*
- * Case-insensitive check if 'a' is a prefix of 'b'.
+ * Check if the label matches the predicate string. If 'prefix' is true, then
+ * the predicate string is treated as a prefix. If 'ci' is true, then the
+ * predicate string is case-insensitive (and locale-aware).
  */
 bool
-ltree_prefix_eq_ci(const char *a, size_t a_sz, const char *b, size_t b_sz)
+ltree_label_match(const char *pred, size_t pred_len, const char *label,
+                                 size_t label_len, bool prefix, bool ci)
 {
        static pg_locale_t locale = NULL;
-       size_t          al_sz = a_sz + 1;
-       size_t          al_len;
-       char       *al = palloc(al_sz);
-       size_t          bl_sz = b_sz + 1;
-       size_t          bl_len;
-       char       *bl = palloc(bl_sz);
+       char       *fpred;                      /* casefolded predicate */
+       size_t          fpred_len = pred_len;
+       char       *flabel;                     /* casefolded label */
+       size_t          flabel_len = label_len;
+       size_t          len;
        bool            res;
 
+       /* fast path for binary match or binary prefix match */
+       if ((pred_len == label_len || (prefix && pred_len < label_len)) &&
+               strncmp(pred, label, pred_len) == 0)
+               return true;
+       else if (!ci)
+               return false;
+
+       /*
+        * Slow path for case-insensitive comparison: case fold and then compare.
+        * This path is necessary even if pred_len > label_len, because the byte
+        * lengths may change after casefolding.
+        */
        if (!locale)
                locale = pg_database_locale();
 
-       /* casefold both a and b */
-
-       al_len = pg_strfold(al, al_sz, a, a_sz, locale);
-       if (al_len + 1 > al_sz)
+       fpred = palloc(fpred_len + 1);
+       len = pg_strfold(fpred, fpred_len + 1, pred, pred_len, locale);
+       if (len > fpred_len)
        {
                /* grow buffer if needed and retry */
-               al_sz = al_len + 1;
-               al = repalloc(al, al_sz);
-               al_len = pg_strfold(al, al_sz, a, a_sz, locale);
-               Assert(al_len + 1 <= al_sz);
+               fpred_len = len;
+               fpred = repalloc(fpred, fpred_len + 1);
+               len = pg_strfold(fpred, fpred_len + 1, pred, pred_len, locale);
        }
+       Assert(len <= fpred_len);
+       fpred_len = len;
 
-       bl_len = pg_strfold(bl, bl_sz, b, b_sz, locale);
-       if (bl_len + 1 > bl_sz)
+       flabel = palloc(flabel_len + 1);
+       len = pg_strfold(flabel, flabel_len + 1, label, label_len, locale);
+       if (len > flabel_len)
        {
                /* grow buffer if needed and retry */
-               bl_sz = bl_len + 1;
-               bl = repalloc(bl, bl_sz);
-               bl_len = pg_strfold(bl, bl_sz, b, b_sz, locale);
-               Assert(bl_len + 1 <= bl_sz);
+               flabel_len = len;
+               flabel = repalloc(flabel, flabel_len + 1);
+               len = pg_strfold(flabel, flabel_len + 1, label, label_len, locale);
        }
+       Assert(len <= flabel_len);
+       flabel_len = len;
 
-       if (al_len > bl_len)
-               res = false;
+       if ((fpred_len == flabel_len || (prefix && fpred_len < flabel_len)) &&
+               strncmp(fpred, flabel, fpred_len) == 0)
+               res = true;
        else
-               res = (strncmp(al, bl, al_len) == 0);
+               res = false;
 
-       pfree(al);
-       pfree(bl);
+       pfree(fpred);
+       pfree(flabel);
 
        return res;
 }
@@ -158,19 +160,16 @@ checkLevel(lquery_level *curq, ltree_level *curt)
 
        for (int i = 0; i < curq->numvar; i++)
        {
-               ltree_prefix_eq_func prefix_eq;
-
-               prefix_eq = (curvar->flag & LVAR_INCASE) ? ltree_prefix_eq_ci : ltree_prefix_eq;
+               bool            prefix = (curvar->flag & LVAR_ANYEND);
+               bool            ci = (curvar->flag & LVAR_INCASE);
 
                if (curvar->flag & LVAR_SUBLEXEME)
                {
-                       if (compare_subnode(curt, curvar->name, curvar->len, prefix_eq,
-                                                               (curvar->flag & LVAR_ANYEND)))
+                       if (compare_subnode(curt, curvar->name, curvar->len, prefix, ci))
                                return success;
                }
-               else if ((curvar->len == curt->len ||
-                                 (curt->len > curvar->len && (curvar->flag & LVAR_ANYEND))) &&
-                                (*prefix_eq) (curvar->name, curvar->len, curt->name, curt->len))
+               else if (ltree_label_match(curvar->name, curvar->len, curt->name,
+                                                                  curt->len, prefix, ci))
                        return success;
 
                curvar = LVAR_NEXT(curvar);

diff --git a/contrib/ltree/ltree.h b/contrib/ltree/ltree.h
index b0ded40eba9b6732ff05a6e7c3d4a6b9c27e3345..226c1cb211501c96ea3fecb3db5a4f631f77901d 100644 (file)
--- a/contrib/ltree/ltree.h
+++ b/contrib/ltree/ltree.h
@@ -157,8 +157,6 @@ typedef struct
        char            data[FLEXIBLE_ARRAY_MEMBER];
 } ltxtquery;
 
-typedef bool (*ltree_prefix_eq_func) (const char *, size_t, const char *, size_t);
-
 #define HDRSIZEQT              MAXALIGN(VARHDRSZ + sizeof(int32))
 #define COMPUTESIZE(size,lenofoperand) ( HDRSIZEQT + (size) * sizeof(ITEM) + (lenofoperand) )
 #define LTXTQUERY_TOO_BIG(size,lenofoperand) \
@@ -209,11 +207,11 @@ bool              ltree_execute(ITEM *curitem, void *checkval,
 
 int                    ltree_compare(const ltree *a, const ltree *b);
 bool           inner_isparent(const ltree *c, const ltree *p);
-bool           compare_subnode(ltree_level *t, char *qn, int len,
-                                                       ltree_prefix_eq_func prefix_eq, bool anyend);
+bool           compare_subnode(ltree_level *t, char *qn, int len, bool prefix, bool ci);
 ltree     *lca_inner(ltree **a, int len);
-bool           ltree_prefix_eq(const char *a, size_t a_sz, const char *b, size_t b_sz);
-bool           ltree_prefix_eq_ci(const char *a, size_t a_sz, const char *b, size_t b_sz);
+bool           ltree_label_match(const char *pred, size_t pred_len,
+                                                         const char *label, size_t label_len,
+                                                         bool prefix, bool ci);
 
 /* fmgr macros for ltree objects */
 #define DatumGetLtreeP(X)                      ((ltree *) PG_DETOAST_DATUM(X))

diff --git a/contrib/ltree/ltxtquery_op.c b/contrib/ltree/ltxtquery_op.c
index 3dcbab2c4846017bf3fa30ed8c28d73f40ef38ad..0e6612ff77a428113fed2cf745a93d8e4229b56e 100644 (file)
--- a/contrib/ltree/ltxtquery_op.c
+++ b/contrib/ltree/ltxtquery_op.c
@@ -58,19 +58,18 @@ checkcondition_str(void *checkval, ITEM *val)
        ltree_level *level = LTREE_FIRST(((CHKVAL *) checkval)->node);
        int                     tlen = ((CHKVAL *) checkval)->node->numlevel;
        char       *op = ((CHKVAL *) checkval)->operand + val->distance;
-       ltree_prefix_eq_func prefix_eq;
+       bool            prefix = (val->flag & LVAR_ANYEND);
+       bool            ci = (val->flag & LVAR_INCASE);
 
-       prefix_eq = (val->flag & LVAR_INCASE) ? ltree_prefix_eq_ci : ltree_prefix_eq;
        while (tlen > 0)
        {
                if (val->flag & LVAR_SUBLEXEME)
                {
-                       if (compare_subnode(level, op, val->length, prefix_eq, (val->flag & LVAR_ANYEND)))
+                       if (compare_subnode(level, op, val->length, prefix, ci))
                                return true;
                }
-               else if ((val->length == level->len ||
-                                 (level->len > val->length && (val->flag & LVAR_ANYEND))) &&
-                                (*prefix_eq) (op, val->length, level->name, level->len))
+               else if (ltree_label_match(op, val->length, level->name, level->len,
+                                                                  prefix, ci))
                        return true;
 
                tlen--;