Hi David

On 18 Aug 2000, at 19:57, David Norris wrote:

> I think ishtml() might qualify as a bug.  It doesn't seem to help
> anything.  Do you see any problems with assuming everything to be HTML? 
> No one seems to mention whether they think it is good or bad.  As
> SWISH-E becomes more powerful I think assuming plain text is very
> limiting.
> 
I totally agree. I am thinking on adding more directives to the config 
file. On of them could be:

DefaultFileType Value

Possible values are: txt, html, xml, wap ...
If Value is html, ishtml() can always return 1.
To maintain backwards compatibility, the default value should be txt

> For example, all filters are assumed to be text.  I am using many
> filters which return HTML.
> 

In the same way. We can extend FileFilter to:
FileFilter  <file-ext> <filter-program> <file-type>
If no file-type is given, then DefaultFileType should be used

> 
> I plan to spend some time on the stemmer.c and soundex.c this weekend. 
> I have been busy during the week.

Good luck.

cu
Jose