[cl-gd-devel] Tiny patch for create-image-from-file

[Left Right]

Hi Hans,

I'm now going through the source to find all places where the pathname
string is used with further I/O, and I found that in
draw-freetype-string the routine for loading font files (or rather
extracting strings from pathnames) differs from other places. I.e.
this is the code in draw-freetype-string

(when (pathnamep font-name)
      (setq font-name (namestring font-name)))

And this is the code in create-image-from-file

(when (pathnamep file-name)
      (setq file-name
              #+:cmu (ext:unix-namestring file-name)
              #-:cmu (namestring file-name)))

I.e. draw-freetype-string seems like it doesn't care about CMUCL - was
this intended? Or, maybe, the exception made for CMUCL was an extra
precaution in the create-image-from-file?

Best.

Oleg

On Fri, Dec 28, 2012 at 11:00 PM, Hans Hübner <hans.huebner at gmail.com> wrote:
> Oleg,
>
> please send a pull request on github.  I like the idea to make cl-gd
> CL pathname aware.
>
> Thanks,
> Hans
>
> On Fri, Dec 28, 2012 at 2:59 PM, Left Right <olegsivokon at gmail.com> wrote:
>> Hello list!
>>
>> I've realized that create-image-from-file will not try to expand
>> logical pathnames. So, I wanted to suggest to change the line:
>>
>> (namestring file-name)
>>
>> to
>>
>> (namestring (translate-logical-pathname file-name))
>>
>> but I also saw that for CMU there is some other function used to read
>> the name, and so long I don't have it installed and never used it - I
>> don't know whether the same will fit there too.
>>
>> This is line 115 in my version of sources (not modified).
>> This shouldn't break older programs because if there are no
>> translation rules, then the original pathname is returned from
>> translate-logical-pathname, but theoretically, it could produce
>> situations when previously some bizarre file name was taken literally
>> instead of being interpreted.
>>
>> Well, that's it.
>>
>> Oh, and this would probably then apply to other functions that call on
>> C functions to do I/O... If this makes sense, I could search for all
>> such functions and export a clean patch for them. Should I?
>>
>> Best.
>>
>> Oleg
>>
>> _______________________________________________
>> cl-gd-devel site list
>> cl-gd-devel at common-lisp.net
>> http://common-lisp.net/mailman/listinfo/cl-gd-devel
>
> _______________________________________________
> cl-gd-devel site list
> cl-gd-devel at common-lisp.net
> http://common-lisp.net/mailman/listinfo/cl-gd-devel