Re: [Maxima-discuss] more about file search paths, share Texinfo, was: duplicate file names, was: New file search patterns

[Raymond T. <toy...@gm...>]

On 7/18/23 11:57 AM, Raymond Toy wrote:

> On 7/18/23 9:25 AM, Robert Dodier wrote:
>> On Tue, Jul 18, 2023 at 7:29 AM Raymond Toy<toy...@gm...>  wrote:
>>
>>> Also, the demo patterns for demo files includes the extensions "dem",
>>> "dm1", "dm2", "dm3", and "dmt".  In the share directory, the only
>>> extension used "dem".  I propose for the share directory, we only look
>>> for "dem".  The user dir will search for all of these (for backward
>>> compatiblity).
>> I dunno. The probability that anyone is actually using files with
>> extensions .dm1, .dm2, .dm3, or .dmt is infinitesimal.

The current search paths are now:

|(%i1) file_search_maxima; (%o1) [/home/toy/.maxima/**/*.mac, 
/home/toy/.maxima/**/*.wxm, 
/home/toy/src/sourceforge/maxima/share/**/*.mac, 
/home/toy/src/sourceforge/maxima/src/*.mac, 
/home/toy/src/sourceforge/maxima/*.mac] (%i2) file_search_lisp; (%o2) 
[/home/toy/.maxima/**/*.sse2f, /home/toy/.maxima/**/*.lisp, 
/home/toy/src/sourceforge/maxima/share/**/*.lisp, 
/home/toy/src/sourceforge/maxima/src/*.sse2f, 
/home/toy/src/sourceforge/maxima/src/*.lisp, 
/home/toy/src/sourceforge/maxima/*.sse2f, 
/home/toy/src/sourceforge/maxima/*.lisp] (%i3) file_search_demo; (%o3) 
[/home/toy/src/sourceforge/maxima/share/**/*.dem, 
/home/toy/src/sourceforge/maxima/demo/*.dem] (%i4) file_search_usage; 
(%o4) [/home/toy/src/sourceforge/maxima/share/**/*.usg, 
/home/toy/src/sourceforge/maxima/doc/*.usg] |

I think these are ok. I’m still concerned that ~/.maxima/binary exists 
and will currently be searched because it never was before. I guess this 
is only a problem if you want to load a file that happens to name a fasl 
in ~/.maxima/binary. AFAICT, fasls are stored here because some package 
used mk:defsystem to save the fasls there. The comments say that done 
because it’s the only fairly reasonable location that is likely to be 
writable by the user.

printfile is a weird function to me because it looks in 
file_search_usage for files to print. But that’s a different issue.

In any case, If there are no objections, I’ll probably be merging this soon.

​