Open Rails Shape Packer

[pwillard]

After seeing the tools coming from Scott lately, I decided to take some inspiration from it and share some of my own. So here we go.


Screenshot 2026-08-20 120008.jpg
Background: I have always disliked FFEDITC_UNICODE because of its clunkiness and its inability to work on more than one file at a time. While I have nothing but praise for everyone who solved this problem before me (Making a better FFEDITC_UNICODE Utility), I figured I'd see what I could come up with.

So here is what I have: 2 programs, ORZIP and ORSP.

* ORZIP -- is a command-line replacement for FFEDITC_UNICODE.exe
* ORSP (Open Rails Shape Packer) -- basically provides the same function as FFEDITC_SUB by Slipperman; it's just my take on it. It's a Graphical front end to ORZIP

Ok, I lied, 3 programs total.

There are actually 3 Tools involved, since I also found getting the old Paul Gausden "Shape File Manager 2.5" HTA file (HTML and JavaScript) a bit problematic to run on Windows 11. Sooo... we have; SFM3

SFM3 is now available as a front end to ORZIP, providing the functionality of SFM2.5 tool.

All 3 programs are supplied as Windows executable programs with documentation.

In addition, all programs are made available on GitHub along with their source code. So you won't lose the source code and end up with only an executable file you can't change.

1. Orzip -- https://github.com/pwillard/orzip/releases

2. ORSP -- https://github.com/pwillard/ORSP/releases -- Includes Orzip.exe

3. SFM3 -- https://github.com/pwillard/SFM3/releases -- Includes Orzip.exe

Pete Willard

Note: These have only been lightly tested by me, so there are no guarantees there won't be issues. If you have difficulties, report it.

[scottb613]

Hi Pete,

VERY COOL - hah - those colors and widgets look familiar...

NICE...


Regards,
Scott

[pwillard]

I can make them any way I like, of course... but I had this weird idea to try and look like your tools. Like there was a grand plan and all.

[pwillard]

More about SFM3

What it looks like.


Screenshot 2026-08-20 153348.jpg

[yellowdog]

Can you name it something other than Packer? Don't want to feed Eric's football delusions any further. Otherwise, brilliant work, Pete, thanks!

[eric]

Can you name it something other than Packer? Don't want to feed Eric's football delusions any further.

Lol... that was going to be my icon suggestion.

[pwillard]

Look, I struggle with naming stuff... It's an affliction.

[pwillard]

Also, keep an eye out for updates in the RELEASES section. As I find issues and fix them, updates will happen.

* Like... I found some compressed S files in my trainset folder with characters AFTER the end of the file... which is sort of wrong. Orzip complains a bit about it. I found 11 files like this; oddly, they were all from the same content creator.
* I also found a problem with how I handled Animated Wiper files... so a fix for that is incoming as well.


*** Updates Have been completed

[pwillard]

Where to find the latest. https://github.com/pwillard/ORSP/releases

[pwillard]

I'm constantly reminded why I never became a programmer.

With ORZIP sort of done... I'm realizing, from lack of experience, that I may be over-sanitizing the whitespace.

I'm trying to test all the possible scenarios I can think of, and I remembered that Shape Converter works on S files. Sure enough, that tool errors out on uncompressed files from ORZIP in some conditions. While ORZIP follows what I thought was a standard for SIMIS text files, Shape Converter's parser doesn't see it that way, or rather, is too strict?. ORZIP isn't broken, per se; all the data is there, but Shape Converter has a brittle whitespace expectation that ORZIP doesn't adhere to, making the converted text easier to read but an error to Shape Converter. ORTS is not complaining about the files.

This "strictness" probably isn't limited to Shape Converter; it's likely an issue with Polymaster as well (Wild Guess, since I don't have or use Polymaster), so I likely need to adopt a less free-form approach to output formatting. We are always victims of our past, I guess.

So I guess I have 3 choices. (Well, really only one, but you get the idea.)

1) Do nothing, since Shape Converter is not on the beloved apps list of many creators.
2) Fix Shape Converter (seriously, it's possible; we have the code, but really?)
3) Adjust ORZIP to match the strict parser constraints of Shape Converter so it won't error out.

[EldoradoRailroad]

Dear Mr. Willard,

So I guess I have 3 choices. (Well, really only one, but you get the idea.)

1) Do nothing, since Shape Converter is not on the beloved apps list of many creators.
2) Fix Shape Converter (seriously, it's possible; we have the code, but really?)
3) Adjust ORZIP to match the strict parser constraints of Shape Converter so it won't error out.

Yes, the tedium of catching all the exceptions, that have gone before us!

Not that my replies seem to have much staying power but:

1) It seems that at least here there is more and more easiness about things like Shape Converter. In the past I have massaged .s source files to get something of use from this program. I am sure I am not the only one. So why fight the trend. I would massage the output to "help" Shape Converter. Otherwise we will just keep on using FFEDITC_UNICODE.exe!

2) Likewise from above, yeah, why not. Some of the constraints with Mingw could be massaged with a python re-write, even if the output would be a bit slower. Any problems with the interpreted python code could be further fixed by others.

3) See 1) and 2)

Not on the list, but nice GUI interface, so full marks PLUS for that Pete!

The times they are a changing,
Steve

[pwillard]

Blame/Praise Scott for the interface concept, as I followed the look and feel of his designs in an attempt to have some cosistency across tools in our toolkit.
I did eventually go with option 3 of course and uploaded the results.

[GokuMK]

What did you use as an object source for conversions binary -> unicode? As without a schema it is not that obvious.

[pwillard]

I'm still working out the final details, as I'm still finding some files that cause issues. I started with the details from the UTILS folder from FFEDITC_UNICODE and ZLIB python library. The shapes I used as samples were DASH8.s and US2Freightcar4.s