Re: [lm-sensors] [PATCH 1/8] sensord: Remove commandline interface.

[Andre Prendel <andre_prendel@gmx.de>]

On Sun, Apr 19, 2009 at 09:39:54AM +0200, Jean Delvare wrote:
> On Mon, 6 Apr 2009 09:56:18 +0200, Andre Prendel wrote:
> > Remove the hidden commandline interface of sensord.
> > 
> > Sensord can be invoked as an console application. Therefor a link
> > named sensors pointing to sensord is needed. That is very
> > intransparent and IMO useless. Printing sensor values to console is
> > done by the sensors tool.
> 
> I agree. I guess the original idea was that "sensord" would possibly
> ultimately supersede "sensors" but that did not happen and I don't
> think this will ever happen due to the extra dependencies sensord
> requires (librrd).
> 
> > ---
> > 
> >  args.c |   38 +++++++++-----------------------------
> >  1 file changed, 9 insertions(+), 29 deletions(-)
> > 
> > --- quilt-sensors.orig/prog/sensord/args.c	2009-03-24 21:53:35.000000000 +0100
> > +++ quilt-sensors/prog/sensord/args.c	2009-03-24 21:54:17.000000000 +0100
> > @@ -127,19 +127,6 @@
> >  	"the RRD file configuration must EXACTLY match the sensors that are used. If\n"
> >  	"your configuration changes, delete the old RRD file and restart sensord.\n";
> >  
> > -static const char *appSyntax > > -	"  -a, --alarm-scan          -- only scan for alarms\n"
> > -	"  -s, --set                 -- execute set statements (root only)\n"
> > -	"  -r, --rrd-file <file>     -- only update RRD file\n"
> > -	"  -c, --config-file <file>  -- configuration file\n"
> > -	"  -d, --debug               -- display some debug information\n"
> > -	"  -v, --version             -- display version and exit\n"
> > -	"  -h, --help                -- display help and exit\n"
> > -	"\n"
> > -	"Specify the filename `-' to read the config file from stdin.\n"
> > -	"\n"
> > -	"If no chips are specified, all chip info will be printed.\n";
> > -
> >  static const char *daemonShortOptions = "i:l:t:Tf:r:c:p:advhg:";
> >  
> >  static const struct option daemonLongOptions[] = {
> > @@ -159,19 +146,6 @@
> >  	{ NULL, 0, NULL, 0 }
> >  };
> >  
> > -static const char *appShortOptions = "asr:c:dvh";
> > -
> > -static const struct option appLongOptions[] = {
> > -	{ "alarm-scan", no_argument, NULL, 'a' },
> > -	{ "set", no_argument, NULL, 's' },
> > -	{ "rrd-file", required_argument, NULL, 'r' },
> > -	{ "config-file", required_argument, NULL, 'c' },
> > -	{ "debug", no_argument, NULL, 'd' },
> > -	{ "version", no_argument, NULL, 'v' },
> > -	{ "help", no_argument, NULL, 'h' },
> > -	{ NULL, 0, NULL, 0 }
> > -};
> > -
> >  int parseArgs(int argc, char **argv)
> >  {
> >  	int c;
> > @@ -179,8 +153,14 @@
> >  	const struct option *longOptions;
> >  
> >  	isDaemon = (argv[0][strlen (argv[0]) - 1] = 'd');
> > -	shortOptions = isDaemon ? daemonShortOptions : appShortOptions;
> > -	longOptions = isDaemon ? daemonLongOptions : appLongOptions;
> > +	if (!isDaemon) {
> > +		fprintf(stderr, "Sensord no longer runs as an commandline"
> > +			" application.\n");
> > +		return -1;
> > +	}
> > +
> > +	shortOptions = daemonShortOptions;
> > +	longOptions = daemonLongOptions;
> 
> I presume we could go one step further and rename daemonShortOptions
> and daemonLongOptions to shortOptions and longOptions, respectively?
> Or, alternatively, use daemonShortOptions and daemonLongOptions in the
> code below.

That's right. I will send you an updated patch series.

> 
> >  
> >  	while ((c = getopt_long(argc, argv, shortOptions, longOptions, NULL))
> >  	       != EOF) {
> > @@ -235,7 +215,7 @@
> >  			break;
> >  		case 'h':
> >  			printf("Syntax: %s {options} {chips}\n%s", argv[0],
> > -			       isDaemon ? daemonSyntax : appSyntax);
> > +			       daemonSyntax);
> >  			exit(EXIT_SUCCESS);
> >  			break;
> >  		case ':':
> 
> There are many other references to isDaemon left in the code, which
> suggests that there is room for many additional cleanups. Are you going
> to do this?

Will fix this in v2 too. What about the two occurrences at the
beginning of parseArgs? I would prefer leaving that to inform the user
about the change of the behaviour.

Andre

> Thanks,
> -- 
> Jean Delvare

_______________________________________________
lm-sensors mailing list
lm-sensors@lm-sensors.org
http://lists.lm-sensors.org/mailman/listinfo/lm-sensors