Re: Extracting name of font from file...

[Erwin Dieterich]

I looked into this once: You don't get any faster from only reading the names.
Reason: The thing most time consuming in looking for the font names is opening
and reading the files, because they are stored on the hard disk. Once they are
in the disk cache everthing is very fast. But this is true only the second
time you read the names. In my opinion you need some status bar.

I append my attempts (little C program etc. etc.)

Regards,
 - Erwin



Am Die, 07 Mär 2000 you wrote: 
>Another
thing to consider... > >When opening a font with FreeType, it loads a bunch of
stuff, initializes >the hinting >engine, etc. If I just want to extract the
font name to put in a menu or >something, that's
>a lot of extra overhead.  If the user has lots of fonts, I had to resort to
>putting up a
>status bar while I generate the list of font names.
>
>I came up with a partial solution which involves doing this once and storing
>the names
>in a reference file.  Then to generate the menu, I only have to pull the
>name from the
>font if it's not already known.  This works fairly well, but I still have to
>give a status
>bar the first time as I generate the ref file.
>
>It would be nice if the open function had a flag to only initialize the name
>table(s).
>

nameread.tar.gz
Description: GNU Zip compressed data