I want programs that I can set loose on a site, and it will provide an access path into the content.
The first, and most obvious such tool, is one that would follow the hierarchical directory structure of host, and build an access path from that.
How it should work:
First Step (and I have made several attempts at this...) -point the program at a single directory. Program reads every file in a directory, creates a list of the content in that directory (maybe with base for a cgi, or put it in that directory). If it is called from cgi we can do relative paths from root, if called outside of a cgi the pathing gets screwed up...wait...chroot camel p. 151!
So: pass the script a directory to read (maybe with filenames/wildcards) and a root
perl docitall c:\www\*.htm c:\www
If it is called from CGI you won't be root, and so the chroot will fail, just let it fail!
Questions: -what file types can we handle? Text/plain Text/html Images cgi scripts: perl, etc.
-what does it mean to handle a file type?
Refinements: Walk a whole directory tree, building a hierachy of index pages.
I am tempted to call it 'Docitall' ?
In order to create an index of the projects in a directory we need some sort of list.
-Could use make... To create the html index we need to know: url, link text, link description.
The url is for the document describing the project. It may contain additional links to the source code, or ?