Re: How to document a type of int / float?
| From: | greg at chiaraquartet dot net | Date: | Fri, 12 Mar 2004 17:28:30 +0000 |
| Subject: | Re: How to document a type of int / float? | ||
| References: | 1 2 3 | Groups: | php.pear.doc |
| Request: | Send a blank email to pear-doc+get-2980@lists.php.net to get a copy of this message | ||
Hi,
I would prefer that authors update their source to include the <?php ?>, as
this is better documentation practice. I am of the opinion from a few years of
experience now that the less magic there is in interpretation of docs, the
better. I think that either QA or the doc group could hassle authors to do
this in the same manner they were hassled to handle get_class() problems :)
I have good news for parse errors - phpDocumentor 2.0 will be using PHP_Parser,
which generates the exact same parse errors that PHP's parser generates,
because it uses a generated parser based on zend_language_parser.y. So, within
a year's time, expect that this will catch all the parse errors prior to commit.
Greg
Quoting Mehdi Achour <didou@keliglia.com>:
> greg@chiaraquartet.net wrote:
> > Hi,
> >
> > I use | as the separator, not / int|float. phpDocumentor parses | to
> > mean "either or" as in "either int or float", and also will check to
> > see if
> any
> > of the words are classes.
> >
> > I'll get these issues and the & fixed in 1.3.0 final, which will be out
> > sometime in the next month, I think
>
> great !
> One more thing, is the examples generated by phpDocumentor ? if so, can
> you add <?php and ?> around the PHP examples ?
> This will help cleaning the manual from parse errors in the example
> (turning short_open_tag to Off and calling php -l on each xml file)
>
> thank you in advance !
>
> didou
>
> >
> > Greg
> >
> > Quoting Stefan Neufeind <stefan@neufeind.net>:
> >
> >
> >>Hi,
> >>
> >>in reference to the recent discussion raised by didou about a
> >>&type.int/float; in the image-graph-docs I wonder how this should be
> >>documented in phpdoc. We could surely define all these occurences as
> >>type mixed, however I don't think this is a benefit to the
> >>documentation. But mentioning only int or float also makes no sense,
> >>since the function can accept both without problems - despite other
> >>functions that maybe demand an int.
> >>
> >>A similar question appeared to me at places where either an int or a
> >>string may be supplied. What's the best way to document this? Using
> >>mixed everywhere and just giving "mixed here means int or string" in
> >>the comment doesn't sound like a good solution for me.
> >>
> >>
> >>Regards,
> >> Stefan
> >>
>
>