On Fri, 17 Dec 1999, Alexis Mikhailov wrote:
> Thank you. And here is a simple HtFile class (attach 1) I mentioned in
> one of my letters. I'm not insisting on integrating it or something,
> just in case it can be useful. This class tries to read file in in
> one operation, so it needs a little additional function to String
> class (attach 2). Probably it is better to read by one line, than it
> will be possible to open FIFO's as well as regular files. Attach 3
> contains a little change to URL class (there is no need to strip
> index.html from file:// URLs). And at last attach 4 contains changes
> for Document class to create HtFile object on file:// URLs.
I only glanced at this, but the last two look fine. As for the HtFile
class, there are two things I thought of while looking at the code--first,
some people might want directories to return links to the files (we can do
this later) and second, we need the MIME-magic code even more now.
I'm not so sure I like the setLength() method for strings. It seems to
defeat the point of abstraction. Granted, there are times when you might
not want extra allocation because you won't modify the string. Maybe this
should be a subclass of String? (Buffer?)
> And another one thing. I think that CanBeParsed, isParsable, etc. stuff
> does not belong to HtHTTP class. AFAIR it reports about parsability of
> document using only content_type. So it probably should be standalone
> utility class.
Yeah, or maybe part of the Parsable class...
-Geoff Hutchison
Williams Students Online
http://wso.williams.edu/
------------------------------------
To unsubscribe from the htdig3-dev mailing list, send a message to
[EMAIL PROTECTED]
You will receive a message to confirm this.
