Randy, all,
I'm running into a problem where something that I want to be part of a
description of a struct member or function parameter gets parsed as a
new section. You can easily see the problem by running below test file
through kernel-doc, clearly the intent is not to have a new section in
the middle of function argument descriptions, but I have no idea how I
could fix this.
I'd appreciate if you could take a look.
Thanks,
Johannes
--- here's the file ---
/**
* fn - test
*
* @a: This is just a non-sense and totally useless dummy function
* argument: it has absolutely no effect.
*/
void fn(int a)
{
}
On Thu, 19 Aug 2010 15:20:19 +0200 Johannes Berg wrote:
> Randy, all,
>
> I'm running into a problem where something that I want to be part of a
> description of a struct member or function parameter gets parsed as a
> new section. You can easily see the problem by running below test file
> through kernel-doc, clearly the intent is not to have a new section in
> the middle of function argument descriptions, but I have no idea how I
> could fix this.
>
> I'd appreciate if you could take a look.
>
> Thanks,
> Johannes
>
> --- here's the file ---
> /**
> * fn - test
> *
> * @a: This is just a non-sense and totally useless dummy function
> * argument: it has absolutely no effect.
> */
> void fn(int a)
> {
> }
Colons (':') are bad. They cause a section instance.
I don't know how to fix it (inside kernel-doc or docproc or xmlto or wherever).
Sometimes I change colon to "--" or "." or something that makes some sense,
but colon is still the better choice if it would work.
The word "example" has a similar problem. It causes an <informalexample>
markup to begin, but it's not always terminated properly.
See recent commit 3d2be54bab7efd97b642838b9c883869ca5b98be.
---
~Randy
*** Remember to use Documentation/SubmitChecklist when testing your code ***
On Thu, 2010-08-19 at 08:30 -0700, Randy Dunlap wrote:
> > --- here's the file ---
> > /**
> > * fn - test
> > *
> > * @a: This is just a non-sense and totally useless dummy function
> > * argument: it has absolutely no effect.
> > */
> > void fn(int a)
> > {
> > }
>
> Colons (':') are bad. They cause a section instance.
Yeah, that's what I found.
> I don't know how to fix it (inside kernel-doc or docproc or xmlto or wherever).
> Sometimes I change colon to "--" or "." or something that makes some sense,
> but colon is still the better choice if it would work.
Ok.
Actually, I can think of something. When we have a line that starts with
this kind of pattern, assume that the next line is a continuation of it
and do not pattern match it, unless there was a blank line. I guess that
needs a special case for @...: though.
johannes