well ok i guess no one cares in the proggin' forum but i've solved my own problem and if anyone else is interested its quite simple:
<original executable><file><toc><int len><int headercnt>
The <toc> section is a bunch of headers, each describing a file, and the length of the file.
<char 10><int len><filename><char 13>
The <len> element is the integer length of the <toc> section.
The <headercnt> section is an integer count of how many headers there are in the <toc> section.
With this, all one must do to read the additional data at the end of the executable is seek to the end minus 8 bytes, read 2 integers, seek back to end minus len+8 read the headers, seek back to file offset, output data to file.
And to top it off, all I've done is create an archive format with the headers on the *end*. So they should prolly be called footers i guess but whatever.
i coded 3 apps to handle the file type. I think I'm gonna call it 'Tick'.
This one embeds a file into an existing executable and adds Tick data. Sends result to stdout
embedball <exec> <file>
This one reads information about the Tick file and prints it out, in the fashion of
File "/x/x/xxx", format correct.
TOC length, 31 bytes.
2 headers.
Header # 0:
type: File (10)
length: 123789
data: somefile
...
readball <tickball>
This one extracts all files except for the first one (the executable). So if one were to write an interface that supported multiple files in Tickballs, it would be supported. I probably will later, when I need it.
extractball <tickball>
Yay! I learned something completely new in one day!
MICROSUCK OWNS! we are becoming the anti-microsoft monopoly!
