Including Text In Your Executable

[Axpen]

Ok i've officially searched the boards and now know theres nothing relevant to this on cprogramming.com, and im not sure enough how to word what it is im looking for to get relevant results from google.

My problem is I wish to make my own sort of installer, not nessecerily Win32 specific, since the implementation should be relatively OS neutral in theory. My problem is just this, how do installers put the files including other executables into 1 file, it would be EASY to have a parser executable then have an archive that holds all the data, but this isnt exactly feasable when you want a single installer.exe.

What I want to know is how you insert binary data irrelevant to the executables operation into your exe file. I am pretty sure you cant just tack it on the end with some signifying code to identify it, or can you? Now I know DLL's would do it, but thats defeating the purpose once again since all installers dont have an exe and a dll.

Any help on how to include the files to be installed in the single file would be great, thanks,
Alex

[dwks]

Maybe something like

free installer cross-platform windows linux

keywords for google.

Link: here.

[Axpen]

Yeah I know about already made ones, my question is how to include text or binary in your executable and still have it run, ie I know you cant just add it to then end and expect it to still run.

Well actually I take that back I added ABC to the end of the executable, IE literally 0x41 0x42 0x43, using a hex editor. It still ran, making my computer freeze and crash the first time, but thats been happening without even running the program, so just running the exe may have triggered it without any sort of conflict cause by the ``invalid insertion''.

Now lets say this really works without any sort of trouble, my 2 questions are, is the behavior considered undefined my computer proffessors, is it considered bad practice to just ``tack on'' bytes to the end of an executable? My other question is just this, does this actually work in practice, could I expect this to work on a linux machine, mac and windows without crashing?

I guess really this topic has little to do with help on installation programs, as the title states, I just want to know if its considered bad practice with undefined behavior to do this adding of bytes after an executables last byte?

If this works ultimately I want to make a program that adds files after an executable and at the last byte/s (depending on how big the files are) have a size counter, and have my program read up the file that many bytes. The executable that the files are injected into will preform these operations on its self and add the files to your computer and do registry stuff according to the commands and files tacked on to the exe. This will effectively make it an installation file by the fact that all its stuff is in 1 file.

Anywho any help will be appriciated,
Alex

[anonytmouse]

Search: "site:codeproject.com append data exe"

[Lateralus]

Great question. I was just thinking about this myself the other day.

[Jaken Veina]

Well, I suppose you could simply create a variable that stores the contents of that file as text, or as an array, or whatever, and that is initialized when it is created, then have your exe create a new file and write that variable to the newly created file. Although, this can't be how professionals do it cause that would consume an abhorrent amount of memory, both in the exe file, and the running of it. But if you've only got a few small files to install, I suppose it would work.