From 80b011ef0a13bb326861f79ba987b4fa04ae4a27 Mon Sep 17 00:00:00 2001 From: Tom Lane Date: Tue, 5 Mar 2013 13:02:30 -0500 Subject: [PATCH] Fix to_char() to use ASCII-only case-folding rules where appropriate. formatting.c used locale-dependent case folding rules in some code paths where the result isn't supposed to be locale-dependent, for example to_char(timestamp, 'DAY'). Since the source data is always just ASCII in these cases, that usually didn't matter ... but it does matter in Turkish locales, which have unusual treatment of "i" and "I". To confuse matters even more, the misbehavior was only visible in UTF8 encoding, because in single-byte encodings we used pg_toupper/pg_tolower which don't have locale-specific behavior for ASCII characters. Fix by providing intentionally ASCII-only case-folding functions and using these where appropriate. Per bug #7913 from Adnan Dursun. Back-patch to all active branches, since it's been like this for a long time. --- src/backend/utils/adt/formatting.c | 170 ++++++++++++++++++++--------- src/include/utils/formatting.h | 4 + 2 files changed, 120 insertions(+), 54 deletions(-) diff --git a/src/backend/utils/adt/formatting.c b/src/backend/utils/adt/formatting.c index 0c3a9f1c42..db5dfca51d 100644 --- a/src/backend/utils/adt/formatting.c +++ b/src/backend/utils/adt/formatting.c @@ -1492,12 +1492,7 @@ str_tolower(const char *buff, size_t nbytes, Oid collid) /* C/POSIX collations use this path regardless of database encoding */ if (lc_ctype_is_c(collid)) { - char *p; - - result = pnstrdup(buff, nbytes); - - for (p = result; *p; p++) - *p = pg_ascii_tolower((unsigned char) *p); + result = asc_tolower(buff, nbytes); } #ifdef USE_WIDE_UPPER_LOWER else if (pg_database_encoding_max_length() > 1) @@ -1617,12 +1612,7 @@ str_toupper(const char *buff, size_t nbytes, Oid collid) /* C/POSIX collations use this path regardless of database encoding */ if (lc_ctype_is_c(collid)) { - char *p; - - result = pnstrdup(buff, nbytes); - - for (p = result; *p; p++) - *p = pg_ascii_toupper((unsigned char) *p); + result = asc_toupper(buff, nbytes); } #ifdef USE_WIDE_UPPER_LOWER else if (pg_database_encoding_max_length() > 1) @@ -1743,23 +1733,7 @@ str_initcap(const char *buff, size_t nbytes, Oid collid) /* C/POSIX collations use this path regardless of database encoding */ if (lc_ctype_is_c(collid)) { - char *p; - - result = pnstrdup(buff, nbytes); - - for (p = result; *p; p++) - { - char c; - - if (wasalnum) - *p = c = pg_ascii_tolower((unsigned char) *p); - else - *p = c = pg_ascii_toupper((unsigned char) *p); - /* we don't trust isalnum() here */ - wasalnum = ((c >= 'A' && c <= 'Z') || - (c >= 'a' && c <= 'z') || - (c >= '0' && c <= '9')); - } + result = asc_initcap(buff, nbytes); } #ifdef USE_WIDE_UPPER_LOWER else if (pg_database_encoding_max_length() > 1) @@ -1886,6 +1860,87 @@ str_initcap(const char *buff, size_t nbytes, Oid collid) return result; } +/* + * ASCII-only lower function + * + * We pass the number of bytes so we can pass varlena and char* + * to this function. The result is a palloc'd, null-terminated string. + */ +char * +asc_tolower(const char *buff, size_t nbytes) +{ + char *result; + char *p; + + if (!buff) + return NULL; + + result = pnstrdup(buff, nbytes); + + for (p = result; *p; p++) + *p = pg_ascii_tolower((unsigned char) *p); + + return result; +} + +/* + * ASCII-only upper function + * + * We pass the number of bytes so we can pass varlena and char* + * to this function. The result is a palloc'd, null-terminated string. + */ +char * +asc_toupper(const char *buff, size_t nbytes) +{ + char *result; + char *p; + + if (!buff) + return NULL; + + result = pnstrdup(buff, nbytes); + + for (p = result; *p; p++) + *p = pg_ascii_toupper((unsigned char) *p); + + return result; +} + +/* + * ASCII-only initcap function + * + * We pass the number of bytes so we can pass varlena and char* + * to this function. The result is a palloc'd, null-terminated string. + */ +char * +asc_initcap(const char *buff, size_t nbytes) +{ + char *result; + char *p; + int wasalnum = false; + + if (!buff) + return NULL; + + result = pnstrdup(buff, nbytes); + + for (p = result; *p; p++) + { + char c; + + if (wasalnum) + *p = c = pg_ascii_tolower((unsigned char) *p); + else + *p = c = pg_ascii_toupper((unsigned char) *p); + /* we don't trust isalnum() here */ + wasalnum = ((c >= 'A' && c <= 'Z') || + (c >= 'a' && c <= 'z') || + (c >= '0' && c <= '9')); + } + + return result; +} + /* convenience routines for when the input is null-terminated */ static char * @@ -1906,6 +1961,20 @@ str_initcap_z(const char *buff, Oid collid) return str_initcap(buff, strlen(buff), collid); } +static char * +asc_tolower_z(const char *buff) +{ + return asc_tolower(buff, strlen(buff)); +} + +static char * +asc_toupper_z(const char *buff) +{ + return asc_toupper(buff, strlen(buff)); +} + +/* asc_initcap_z is not currently needed */ + /* ---------- * Skip TM / th in FROM_CHAR @@ -2418,7 +2487,8 @@ DCH_to_char(FormatNode *node, bool is_interval, TmToChar *in, char *out, Oid col INVALID_FOR_INTERVAL; if (tmtcTzn(in)) { - char *p = str_tolower_z(tmtcTzn(in), collid); + /* We assume here that timezone names aren't localized */ + char *p = asc_tolower_z(tmtcTzn(in)); strcpy(s, p); pfree(p); @@ -2465,7 +2535,7 @@ DCH_to_char(FormatNode *node, bool is_interval, TmToChar *in, char *out, Oid col strcpy(s, str_toupper_z(localized_full_months[tm->tm_mon - 1], collid)); else sprintf(s, "%*s", S_FM(n->suffix) ? 0 : -9, - str_toupper_z(months_full[tm->tm_mon - 1], collid)); + asc_toupper_z(months_full[tm->tm_mon - 1])); s += strlen(s); break; case DCH_Month: @@ -2475,7 +2545,8 @@ DCH_to_char(FormatNode *node, bool is_interval, TmToChar *in, char *out, Oid col if (S_TM(n->suffix)) strcpy(s, str_initcap_z(localized_full_months[tm->tm_mon - 1], collid)); else - sprintf(s, "%*s", S_FM(n->suffix) ? 0 : -9, months_full[tm->tm_mon - 1]); + sprintf(s, "%*s", S_FM(n->suffix) ? 0 : -9, + months_full[tm->tm_mon - 1]); s += strlen(s); break; case DCH_month: @@ -2485,10 +2556,8 @@ DCH_to_char(FormatNode *node, bool is_interval, TmToChar *in, char *out, Oid col if (S_TM(n->suffix)) strcpy(s, str_tolower_z(localized_full_months[tm->tm_mon - 1], collid)); else - { - sprintf(s, "%*s", S_FM(n->suffix) ? 0 : -9, months_full[tm->tm_mon - 1]); - *s = pg_tolower((unsigned char) *s); - } + sprintf(s, "%*s", S_FM(n->suffix) ? 0 : -9, + asc_tolower_z(months_full[tm->tm_mon - 1])); s += strlen(s); break; case DCH_MON: @@ -2498,7 +2567,7 @@ DCH_to_char(FormatNode *node, bool is_interval, TmToChar *in, char *out, Oid col if (S_TM(n->suffix)) strcpy(s, str_toupper_z(localized_abbrev_months[tm->tm_mon - 1], collid)); else - strcpy(s, str_toupper_z(months[tm->tm_mon - 1], collid)); + strcpy(s, asc_toupper_z(months[tm->tm_mon - 1])); s += strlen(s); break; case DCH_Mon: @@ -2518,10 +2587,7 @@ DCH_to_char(FormatNode *node, bool is_interval, TmToChar *in, char *out, Oid col if (S_TM(n->suffix)) strcpy(s, str_tolower_z(localized_abbrev_months[tm->tm_mon - 1], collid)); else - { - strcpy(s, months[tm->tm_mon - 1]); - *s = pg_tolower((unsigned char) *s); - } + strcpy(s, asc_tolower_z(months[tm->tm_mon - 1])); s += strlen(s); break; case DCH_MM: @@ -2536,7 +2602,7 @@ DCH_to_char(FormatNode *node, bool is_interval, TmToChar *in, char *out, Oid col strcpy(s, str_toupper_z(localized_full_days[tm->tm_wday], collid)); else sprintf(s, "%*s", S_FM(n->suffix) ? 0 : -9, - str_toupper_z(days[tm->tm_wday], collid)); + asc_toupper_z(days[tm->tm_wday])); s += strlen(s); break; case DCH_Day: @@ -2544,7 +2610,8 @@ DCH_to_char(FormatNode *node, bool is_interval, TmToChar *in, char *out, Oid col if (S_TM(n->suffix)) strcpy(s, str_initcap_z(localized_full_days[tm->tm_wday], collid)); else - sprintf(s, "%*s", S_FM(n->suffix) ? 0 : -9, days[tm->tm_wday]); + sprintf(s, "%*s", S_FM(n->suffix) ? 0 : -9, + days[tm->tm_wday]); s += strlen(s); break; case DCH_day: @@ -2552,10 +2619,8 @@ DCH_to_char(FormatNode *node, bool is_interval, TmToChar *in, char *out, Oid col if (S_TM(n->suffix)) strcpy(s, str_tolower_z(localized_full_days[tm->tm_wday], collid)); else - { - sprintf(s, "%*s", S_FM(n->suffix) ? 0 : -9, days[tm->tm_wday]); - *s = pg_tolower((unsigned char) *s); - } + sprintf(s, "%*s", S_FM(n->suffix) ? 0 : -9, + asc_tolower_z(days[tm->tm_wday])); s += strlen(s); break; case DCH_DY: @@ -2563,7 +2628,7 @@ DCH_to_char(FormatNode *node, bool is_interval, TmToChar *in, char *out, Oid col if (S_TM(n->suffix)) strcpy(s, str_toupper_z(localized_abbrev_days[tm->tm_wday], collid)); else - strcpy(s, str_toupper_z(days_short[tm->tm_wday], collid)); + strcpy(s, asc_toupper_z(days_short[tm->tm_wday])); s += strlen(s); break; case DCH_Dy: @@ -2579,10 +2644,7 @@ DCH_to_char(FormatNode *node, bool is_interval, TmToChar *in, char *out, Oid col if (S_TM(n->suffix)) strcpy(s, str_tolower_z(localized_abbrev_days[tm->tm_wday], collid)); else - { - strcpy(s, days_short[tm->tm_wday]); - *s = pg_tolower((unsigned char) *s); - } + strcpy(s, asc_tolower_z(days_short[tm->tm_wday])); s += strlen(s); break; case DCH_DDD: @@ -4690,12 +4752,12 @@ NUM_processor(FormatNode *node, NUMDesc *Num, char *inout, char *number, case NUM_rn: if (IS_FILLMODE(Np->Num)) { - strcpy(Np->inout_p, str_tolower_z(Np->number_p, collid)); + strcpy(Np->inout_p, asc_tolower_z(Np->number_p)); Np->inout_p += strlen(Np->inout_p) - 1; } else { - sprintf(Np->inout_p, "%15s", str_tolower_z(Np->number_p, collid)); + sprintf(Np->inout_p, "%15s", asc_tolower_z(Np->number_p)); Np->inout_p += strlen(Np->inout_p) - 1; } break; diff --git a/src/include/utils/formatting.h b/src/include/utils/formatting.h index cc47a7e255..107a95c134 100644 --- a/src/include/utils/formatting.h +++ b/src/include/utils/formatting.h @@ -24,6 +24,10 @@ extern char *str_tolower(const char *buff, size_t nbytes, Oid collid); extern char *str_toupper(const char *buff, size_t nbytes, Oid collid); extern char *str_initcap(const char *buff, size_t nbytes, Oid collid); +extern char *asc_tolower(const char *buff, size_t nbytes); +extern char *asc_toupper(const char *buff, size_t nbytes); +extern char *asc_initcap(const char *buff, size_t nbytes); + extern Datum timestamp_to_char(PG_FUNCTION_ARGS); extern Datum timestamptz_to_char(PG_FUNCTION_ARGS); extern Datum interval_to_char(PG_FUNCTION_ARGS);