Re: Magic is useless!

[James Henstridge <james daa com au>]

Raphael Bosshard wrote:

> Another problem with magic filecontent/filetype-identification is 
> indexing. It´s allmost impossible to search for a specific filetype.
>
> ie. "Search for Abiword-files", "Search for HTML-Documents"
>
> It could be done, yes. But is is not very resource-friendly...
>
> As I said before: Magic is no solution, its barely a hack... 

No one is saying that it is a solution that should be used to the 
exclusion of all others.  If you have other reliable file typing 
information (such as file extensions, metadata, etc), you should 
probably use that.  But in some cases it is all you can use (eg. 
anonymous data streams).

If you want to search for all html documents, you would probably be best 
off searching by file extension.

James.

--
Email: james daa com au
WWW:   http://www.daa.com.au/~james/