v1.2.0 stable

[tomhea]

Version 1.2.0 changes:

• Flipjump is now a python package. pip install flipjump:
  • Run the new command line tool: fj hello-world.fj.
  • Import from python: from flipjump import assemble_and_run, ....
  • Poetry support via the new pyproject.toml.
• Project reorganization: the flipjump python source files are under flipjump/ (reorganized and split into smaller files), and the stl files are under flipjump/stl/.
• The flipjump_quickstart.py file: easy access to the main flipjump features from Python.
• Improved READMEs and documentation.

It is all tested, and ready to be used on windows or linux (and mac shouldn't be a problem), with python 3.7+.
The only exception is hex.div which still doesn't work (its tests are xfailed), so you can use the slower-but-stable bit.div+casts for division.

---

This discussion was created from the release v1.2.0 stable.

[tomhea]

Update - flipjump is now a python package !

It's much easier to use now - use the new cli tool: fj -h, or just import flipjump :)

@guym0r @coolslutguy420 @ivelieu @lestrozi @vxgmichel @culverit @hmltn2 @jaripekkary @mtve @ebigram @emenel @nst @roma-glushko @Siss3l @tuxcanfly @elhadjx @Peefy @fabiosvm @andrew-johnson-4 @mgmalheiros @frediff @HaraldKorneliussen @adamharmansky @dfischer @CCXXXI @egon-r @Sophia-Cristina @creaktive @lyxal @parched @TonikHorkel @ivannz @FrogBoy2008 @R-Jones @neauoire @idoherman @danytnt @conilas @gavinbm @jakiki6 @Taardisaa @Clinale @zephryn @thisisclosure @kamil-adam @Polarzz @Sherlockcxk @idozo @betedgi

[vxgmichel]

That's great! 🎉

I can now use flipjump.fjm.fjm_reader.Reader to parse .fjm files:

        reader = FjmReader(program)
        width = reader.memory_width
        size = cast(int, max(reader.memory)) + 1
        words = [reader.memory[i] for i in range(size)]

@tomhea Can you confirm that this interface will remain stable?

[vxgmichel]

Use the stable .get_word().

It seems like this API is more designed to be used when running the program since it takes a bit address as an argument, and it includes this "garbage handling" logic that is not useful to me.

What I'm looking for is simply a way to get the full memory as a list of words from an fjm file. Would you mind including a stable reader.get_memory() helper or something similar? That'd be great :)

[tomhea]

Its a bit more complicated than that, as the memory is not always just a list of values, as the memory is not always continuous.

With the segment / reserve ops you'll get different segments in memory, not likely to be adjacent to one another, and some of them might be reserved with 0's. Some of them might be filled with a different fixed value (maybe in a new fjm version).

How would you suggest to handle that? As I'd like to add that get_memory() for you, and I even think it'll be used not just by you.

I thought of a default_dict, from address to value, (for stuff like reserved 0s), but will it be enough for you? What would you suggest?

[vxgmichel]

A defaultdict sounds great! In my specific case I need the following information:

• the memory width
• the memory size (or at least the maximum address that the program is meant to access)
• a mapping to get a word value from a word address

As long as this info is exposed through a public API I'll be happy :)

[tomhea]

Cool, ill think how and implement it, and ill do my best to update here :)

[tomhea]

Hi, I haven't forgot you, though I published v1.2.1 now, I'll solve the #238 issue I opened for that matter probably in v1.2.2.

The API haven't changed in this version.