ulibos / ndf-parse Goto Github PK

This package allows to parse Eugen Systems ndf files, modify them and write back modified versions as a valid ndf code. It was created to allow easier editing of warno mods than what is currently available with game's own tools.

An example of a script doubling logistics capacity for all vehicles:

import ndf_parse as ndf

# setup mod donor and destination mods
mod = ndf.Mod("path/to/src/mod", "path/to/dst/mod")
# create/update destination mod
mod.check_if_src_is_newer()

with mod.edit(r"GameData\Generated\Gameplay\Gfx\UniteDescriptor.ndf") as source:
   # filter out root descriptors that have supply in them
   for logi_descr in source.match_pattern(
      "TEntityDescriptor(ModulesDescriptors = [TSupplyModuleDescriptor()])"
   ):
      descriptors = logi_descr.v.by_member("ModulesDescriptors").v  # get modules list
      supply_row = descriptors.find_by_cond(  # find supply module
            # safe way to check if row has type and equals the one we search for
            lambda x: getattr(x.v, "type", None) == "TSupplyModuleDescriptor"
      )
      # get capacity row
      supply_capacity_row = supply_row.v.by_member("SupplyCapacity")
      old_capacity = supply_capacity_row.v
      new_capacity = float(old_capacity) * 2  # process value
      supply_capacity_row.v = str(new_capacity)

      print(f"{logi_descr.namespace}: new capacity = {new_capacity}") # log result

• python >= 3.8 (this package was developed on 3.8 and tested on minor versions from 3.8 to 3.12).

pip install ndf-parse

As an alternative, you can download and install this package from github releases.

For usage please refer to the documentation.

This package gets shipped with an ndf.dll containing the tree-sitter language parser. It's linkage is also hardcoded in module's __init__. If you're planning to use this on linux or MacOS (why?..), you will have to build the lib yourself and set an env variable NDF_LIB_PATH=path/to/your/lib.

In order to develop this module you will need to fork this repo, clone the fork and run the following command:

pip install -e "path\to\cloned\repo[dev]"

It will load most dependencies automatically. Only thing you will have to provide manually is an ndf.dll (see below). You can then build a release package using scripts\build_package.bat script. It outputs the result to a build\package folder. By default the build sctipt uses a local library (ndf_parse\bin\ndf.dll). If there is none, it copies one from tree-sitter's default build path to local path. If there is also none then it refuses to build.

black is used for code styling with line length limit == 79. Code is (mostly) type hinted.

.gitignore does not store editor specific excludes, I store those in .git\info\exclude.

• build - temp folder to store build data, untracked
• ndf_parse - package source code
• sphinx - documentation sources
• scripts - mostly scripts for building stuff
• tests - basic testing scripts

To build docs for current commit use scripts\build_docs.bat. Your docs will be in build\docs.

To build docs for all releases follow these steps:

1. Make sure to bump release version and commit all changes related to the latest release (and stash what is left).
2. Tag the release with semver (example: v1.0.5, v is mandatory).
3. Add new tag to publish_tags variable in sphinx\conf.py.
4. Remove build\multiver to ensure clean build.
5. Run python scripts\build_multidocs.py. Result will be in build\multiver.
6. Checkout docs branch.
7. Remove old docs dirs (named by their releases), move new ones, including build\multiver\index.html (but excluding .doctree dirs in each version build, they are not needed for serving), to the root of the repo.
8. Add new stuff to git (be careful not to include junk, there is no .gitignore) and commit.

If you're planning to test scripts from the documentation (the ones in sphinx\code), you will have to setup 2 env variables in your terminal:

set MOD_SRC="path\to\source_mod"
set MOD_DST="path\to\destination_mod"

Currently there are only tests for ndf_parse.model and docs' code snippets and examples. Docstrings code is tested with sphinx via tests\docs_tests.bat, py script with the same name has examples tests and a deprecated version of doctests.

This package looks for an ntf.dll in the following places (descending priority):

1. NDF_LIB_PATH env variable ("C:\custom\path\to\ndf.dll"),
2. default tree-sitter's build path ("%LocalAppData%\tree-sitter\lib\ndf.dll"),
3. a copy bundled with the package ("ndf_parse\bin\ndf.dll").

The repo itself does not hold a prebuilt copy of the library so you'll have to either yank one from a release wheel (it's just a renamed zip) or build one from source.

I have no idea on how frequently I'll be able to respond to those, so expect delays. You might find it easier catching me on WarYes discord or Eugen discord in case you have some blocking issue or a PR.

Created by Ulibos, 2023-2024.