On Fri, 15 Mar 2002, Martin McCormick wrote: > I must add, here, that I am very impressed with the > commenting that has been done by the previous authors. When I > figure out what I am going to do, I must continue the tradition.
Yes, the commenting of many details is good; but it's more difficult to get the whole picture. In order to understand a lot of the stuff under WWW/..., I believe it's still useful to refer to the Libwww documentation. Not for the details, but for concepts and overall structure. Libwww has become a very different package from what is included (greatly modified) in the lynx code - most of the descriptions don't apply to lynx, caveat lector - but I believe the descriptions of esp. HTAnchor and HTStream objects are still useful. Also perhaps HTStreamStack mechanism. HTProtocol OTOH is quite different. You'll find various overview documents about Libwww starting from <http://www.w3.org/Library/> (then follow "Documentation" link). > I am running lynx under gdb and setting breakpoints at > various places to observe how the whole process works. > > Usually, when I have written C programs, I must do this > process many times even on my own code. I would be lost without > gdb. That was also my experience, it took lots of gdb'ing to get an idea of what-calls-what in Lynx. It's not so easy to see in the source how (e.g.) HTML_* functions get called and where they get their input from - because those calls are made through struct members that are function pointers. That's part of how the libwww code implements object-oriented concepts in C (and the docs referred above should help understand it). Klaus ; To UNSUBSCRIBE: Send "unsubscribe lynx-dev" to [EMAIL PROTECTED]
