Does AroCC support parsing a []u8 in some way?

[heysokam]

Title, essentially.

Lets say I have a source code slice in memory.
How would I go about setting up what the parser requires to do its job?

I saw Parser.parse(PP), but it requires a preprocessor object that in turn requires a Compilation which takes a folder... 🤔
Am I missing something? How would one go about doing this?

[Vexu]

Compilation takes a folder that should be used as cwd, it should probably use a Driver.Filesystem to provide even better encapsulation with Filesystem.fake. Easiest way to get a Compilation is to just pass it std.fs.cwd(). Then you can use Compilation.addSourceFromBuffer() to turn your []u8 into a Souce that can be passed to Preprocessor.preprocessSources().

The minimal steps to turn a source buffer into an AST would look something like this:

var comp = aro.Compilation.initDefault(gpa, std.fs.cwd());
defer comp.deinit();

const your_source = try comp.addSourceFromBuffer("<your buffer>", your_buffer);
const builtin = try comp.generateBuiltinMacros(.include_system_defines, null);

var pp = try aro.Preprocessor.initDefault(&comp);
defer pp.deinit();

try pp.preprocessSources(&.{ source, builtin });

var tree = try pp.parse();
defer tree.deinit();

[heysokam]

Thats so useful! Ty!