diff options
| author | Tom Lane <tgl@sss.pgh.pa.us> | 2010-05-08 16:39:53 +0000 |
|---|---|---|
| committer | Tom Lane <tgl@sss.pgh.pa.us> | 2010-05-08 16:39:53 +0000 |
| commit | 54cd4f04576833abc394e131288bf3dd7dcf4806 (patch) | |
| tree | 0772e1bedbca8466701b6a116e065e7a00959ebd /src/backend/utils/adt/datetime.c | |
| parent | 71a185a24d573dc1449777ff9fa8f3020af6f13c (diff) | |
| download | postgresql-54cd4f04576833abc394e131288bf3dd7dcf4806.tar.gz postgresql-54cd4f04576833abc394e131288bf3dd7dcf4806.zip | |
Work around a subtle portability problem in use of printf %s format.
Depending on which spec you read, field widths and precisions in %s may be
counted either in bytes or characters. Our code was assuming bytes, which
is wrong at least for glibc's implementation, and in any case libc might
have a different idea of the prevailing encoding than we do. Hence, for
portable results we must avoid using anything more complex than just "%s"
unless the string to be printed is known to be all-ASCII.
This patch fixes the cases I could find, including the psql formatting
failure reported by Hernan Gonzalez. In HEAD only, I also added comments
to some places where it appears safe to continue using "%.*s".
Diffstat (limited to 'src/backend/utils/adt/datetime.c')
| -rw-r--r-- | src/backend/utils/adt/datetime.c | 11 |
1 files changed, 10 insertions, 1 deletions
diff --git a/src/backend/utils/adt/datetime.c b/src/backend/utils/adt/datetime.c index f9e40f115fc..743ca8345d6 100644 --- a/src/backend/utils/adt/datetime.c +++ b/src/backend/utils/adt/datetime.c @@ -8,7 +8,7 @@ * * * IDENTIFICATION - * $PostgreSQL: pgsql/src/backend/utils/adt/datetime.c,v 1.210 2010/01/02 16:57:53 momjian Exp $ + * $PostgreSQL: pgsql/src/backend/utils/adt/datetime.c,v 1.211 2010/05/08 16:39:51 tgl Exp $ * *------------------------------------------------------------------------- */ @@ -3740,6 +3740,14 @@ EncodeDateTime(struct pg_tm * tm, fsec_t fsec, int *tzp, char **tzn, int style, AppendTimestampSeconds(str + strlen(str), tm, fsec); + /* + * Note: the uses of %.*s in this function would be unportable + * if the timezone names ever contain non-ASCII characters, + * since some platforms think the string length is measured + * in characters not bytes. However, all TZ abbreviations in + * the Olson database are plain ASCII. + */ + if (tzp != NULL && tm->tm_isdst >= 0) { if (*tzn != NULL) @@ -4091,6 +4099,7 @@ CheckDateTokenTable(const char *tablename, const datetkn *base, int nel) { if (strncmp(base[i - 1].token, base[i].token, TOKMAXLEN) >= 0) { + /* %.*s is safe since all our tokens are ASCII */ elog(LOG, "ordering error in %s table: \"%.*s\" >= \"%.*s\"", tablename, TOKMAXLEN, base[i - 1].token, |