Question about card names

Question about card names

[Kuninori Morimoto]

Hi ALSA ML

snd_soc_instantiate_card() setups card->snd_card's
"shortname", "longname", "driver" names. (= 1,2,3).

But, it cares only card->snd_card->driver (= 4).
I think we need to care about all naming, but why only driver ??

	static int snd_soc_instantiate_card(struct snd_soc_card *card)
	{
		...
1=>		snprintf(card->snd_card->shortname, sizeof(card->snd_card->shortname),
			 "%s", card->name);
2=>		snprintf(card->snd_card->longname, sizeof(card->snd_card->longname),
			 "%s", card->long_name ? card->long_name : card->name);
3=>		snprintf(card->snd_card->driver, sizeof(card->snd_card->driver),
			 "%s", card->driver_name ? card->driver_name : card->name);
		for (i = 0; i < ARRAY_SIZE(card->snd_card->driver); i++) {
			switch (card->snd_card->driver[i]) {
x=>			case '_':
x=>			case '-':
			case '\0':
				break;
			default:
				if (!isalnum(card->snd_card->driver[i]))
4=>					card->snd_card->driver[i] = '_';
				break;
			}
		}
		...
	}

And, it seems in this loop, it breaks check if it finds "_" or "-" (= x).
If my understanding was correct,
the naming "abcd??efg" will be fixed to "abcd__efg",
but, it do nothing if name was for example "-abcd??efg".
Is this bug ??

Thank you for your help !!
Best regards
---
Kuninori Morimoto

Re: Question about card names

[Kuninori Morimoto]

Hi

> 		for (i = 0; i < ARRAY_SIZE(card->snd_card->driver); i++) {
> 			switch (card->snd_card->driver[i]) {
> x=>			case '_':
> x=>			case '-':
> 			case '\0':
> 				break;
> 			default:
> 				if (!isalnum(card->snd_card->driver[i]))
> 4=>					card->snd_card->driver[i] = '_';
> 				break;
> 			}
(snip)
> And, it seems in this loop, it breaks check if it finds "_" or "-" (= x).
> If my understanding was correct,
> the naming "abcd??efg" will be fixed to "abcd__efg",
> but, it do nothing if name was for example "-abcd??efg".
> Is this bug ??

Grr
This break is for "switch", not for "for"...


Thank you for your help !!
Best regards
---
Kuninori Morimoto

Re: Question about card names

[Takashi Iwai]

On Wed, 29 May 2019 06:41:35 +0200,
Kuninori Morimoto wrote:
> 
> 
> Hi ALSA ML
> 
> snd_soc_instantiate_card() setups card->snd_card's
> "shortname", "longname", "driver" names. (= 1,2,3).
> 
> But, it cares only card->snd_card->driver (= 4).
> I think we need to care about all naming, but why only driver ??

Do you mean the normalization of strings, or do you suggest to make
mandatory?  The normalization isn't needed for other name strings.
The driver name is somewhat special, as it's used as a key for
searches in the user-space.

> 	static int snd_soc_instantiate_card(struct snd_soc_card *card)
> 	{
> 		...
> 1=>		snprintf(card->snd_card->shortname, sizeof(card->snd_card->shortname),
> 			 "%s", card->name);
> 2=>		snprintf(card->snd_card->longname, sizeof(card->snd_card->longname),
> 			 "%s", card->long_name ? card->long_name : card->name);
> 3=>		snprintf(card->snd_card->driver, sizeof(card->snd_card->driver),
> 			 "%s", card->driver_name ? card->driver_name : card->name);
> 		for (i = 0; i < ARRAY_SIZE(card->snd_card->driver); i++) {
> 			switch (card->snd_card->driver[i]) {
> x=>			case '_':
> x=>			case '-':
> 			case '\0':
> 				break;
> 			default:
> 				if (!isalnum(card->snd_card->driver[i]))
> 4=>					card->snd_card->driver[i] = '_';
> 				break;
> 			}
> 		}
> 		...
> 	}
> 
> And, it seems in this loop, it breaks check if it finds "_" or "-" (= x).
> If my understanding was correct,
> the naming "abcd??efg" will be fixed to "abcd__efg",
> but, it do nothing if name was for example "-abcd??efg".

We do replace the two '?' in the latter case, too.  It's a break
statement in a switch block, i.e. just skip the case.


Takashi

Re: Question about card names

[Kuninori Morimoto]

Hi

> > snd_soc_instantiate_card() setups card->snd_card's
> > "shortname", "longname", "driver" names. (= 1,2,3).
> > 
> > But, it cares only card->snd_card->driver (= 4).
> > I think we need to care about all naming, but why only driver ??
> 
> Do you mean the normalization of strings, or do you suggest to make
> mandatory?  The normalization isn't needed for other name strings.
> The driver name is somewhat special, as it's used as a key for
> searches in the user-space.

Ahh, OK, I see.
Thank you for clearing it.
I want to have this explain at there (= I will post it, soon)

> > And, it seems in this loop, it breaks check if it finds "_" or "-" (= x).
> > If my understanding was correct,
> > the naming "abcd??efg" will be fixed to "abcd__efg",
> > but, it do nothing if name was for example "-abcd??efg".
> 
> We do replace the two '?' in the latter case, too.  It's a break
> statement in a switch block, i.e. just skip the case.

Thanks.
My head is very spaghetting now...

Thank you for your help !!
Best regards
---
Kuninori Morimoto