Re: BUG #18711: Attempting a connection with a database name longer than 63 characters now fails

On Fri, Nov 22, 2024 at 02:23:47PM -0500, Tom Lane wrote:
> +    if (strlen(dbname) < NAMEDATALEN)
> +    {
> +        /* Typical, easy case: no truncation needed */
> +        tuple = GetDatabaseTupleInternal(relation, dbname);
> +    }
> +    else if (!IS_HIGHBIT_SET(dbname[NAMEDATALEN - 1]) ||
> +             !IS_HIGHBIT_SET(dbname[NAMEDATALEN - 2]))
> +    {
> +        /* Also easy: truncation at NAMEDATALEN - 1 cannot break an MB char */
> +        memcpy(tname, dbname, sizeof(tname));
> +        tname[NAMEDATALEN - 1] = '\0';
> +        tuple = GetDatabaseTupleInternal(relation, tname);

I had some time to think about this issue and I now realize the test
above is correct, but I couldn't figure out why it was correct before.
If we want to use this test, which I now think is fine, I suggest the
following comment:

    If we put a NULL byte at byte offset NAMEDATALEN - 1, we don't want
    to break a multi-byte character when doing this.  If byte offset
    NAMEDATALEN - 1 does not have its high bit set, we can be sure
    we will not break a multi-byte character during the overwrite.
    Also, if NAMEDATALEN - 2 does not have its high bit set, then
    NAMEDATALEN - 1 is either a single-byte non-ASCII character,
    or is the _start_ of a multi-byte character, so it is also safe
    to overwrite it without breaking a multi-byte character.

Sorry I did not figure this out earlier.

-- 
  Bruce Momjian  <bruce@momjian.us>        https://momjian.us
  EDB                                      https://enterprisedb.com

  When a patient asks the doctor, "Am I going to die?", he means 
  "Am I going to die soon?"