Ok, Here's a developer's primer for aget source:

The code is splitted into several files:

main.{c|h}
	* starts a thread for handling signals.
	* gets program options and decides the 
	  program's behviour by setting a few 
 	  global variables.

Aget.{c|h} 
	* get() and resume_get() functions
	  starts the download process. Depending on the
	  protocol, necessary methods are called.
	  
	  resume_get() is for resume jobs. If the there 
	  is a log file of a previous job, the values 
	  are taken from it.

Head.{c|h}
	* Before the actual download, a head request
	  is sent to the server to determine the current
	  status of the file.
          Some values like:
		1. HTTP return code (e.g. 404, 206, 403, 200)
		2. Content-length

	  might be of interest there.

Ftp.{c|h}
       * ftp_get() function which is the thread runner for ftp
 	 downloads.

Download.{c|h}
	* Aget is no more %100 Posix compliant :(
	  pwrite is from X_OPEN standarts. I had to do it, since I needed
	  a way to simultaneously write to the same file without loosing
	  any time while synchronising the threads.
	  
	  For the time, there is only http_get. When aget is ftp-aware
  	  ftp_get will be in this file.


Resume.{c|h}
	* Methods about saving and restoring history files are in this
	  file.

Signal.{c|h}
	* Signal Handling functions.

Misc.{c|h}
	* Other functions which do not fit any of the general categories
 	  are stored here. These are generally helper functions...


For more information read the aget source, and you may contact
murat at enderunix dot org.


Fri Nov 22 07:51:04 EET 2002