On Fri, Apr 14, 2017 at 09:39:48PM +0100, Sami Kerola wrote:
> Recent commit that removed header timestamp from typescript output when
> --quiet option is defined broke scriptreplay. Trouble was that scriptreplay
> always skipped first line of the typescript. But --quiet makes that line to
> be part of what must be printed by scriptreplay.
This explain why I had bad feeing from the patch ;-)
> +static int
> +sfile_has_header(FILE *sfile, FILE *tfile)
> +{
> + double delay;
> + size_t blk, total = 0;
> + char nl;
> + struct stat st;
> +
> + if (fstat(fileno(sfile), &st) < 0)
> + return 0;
> + while (fscanf(tfile, "%lf %zu%c\n", &delay, &blk, &nl) == 3)
> + total += blk;
> + fseek(sfile, 0, SEEK_SET);
> + fseek(tfile, 0, SEEK_SET);
> + if ((size_t)st.st_size <= total)
> + return 0;
> + return 1;
> +}
Not sure about it. It seems like overkill...
I have reverted Rui's change and fixed the man page.
It seems better to keep things simple and stupid... it means keep
typescript file completely independent on the --quit option (for the
both, start and done messages).
Note that Rui also suggested to avoid strftime() for the messages due
to problems with \n (and I have suggested to use ISO times). It's
already committed.
Anyway, I don't like officially supported file formats without any
header. IMHO it would be really nice to add to the typescript:
### script(1) typescript; Version: util-linux v2.30; Date: 2017-04-18 12:10:43+0200
... and maybe more information. The same for timing file.
Karel
--
Karel Zak <kzak@xxxxxxxxxx>
http://karelzak.blogspot.com
--
To unsubscribe from this list: send the line "unsubscribe util-linux" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
