Load bulk bsps

[snake-biscuits]

Most times when I test things, I grab multiple maps at once
This involves typing the same block of code at the start of every session:

>>> import bsp_tool, fnmatch, os
>>> md = "E:/Mod/Titanfall"
>>> maps = {m[:-4]: bsp_tool.load_bsp(os.path.join(md, m)) for m in fnmatch.filter(os.listdir(md), "*.bsp")}

tests.utils.get_test_maps does a similar job, but doesn't overlap well w/ current needs
I would also like to feed the archive to gets w/ get_test_maps, as a pytest flag

tests.test_load_bsp.test_load_bsp

I'd like to break The MegaTest up into more functions
(seeing the name all typed out, it's so ugly; real MuffinMail.MuffinHash.MuffinHash energy)
Being able to target a specific chunk of the MegaTest would be nice
Also means the MegaTest is just for redundancy, new branches can be tested solo

The Apex Archive will require more robust tools

We need to be able to compare iterations of maps (extensions.diff needs more work too)
e.g. Titanfall: Beta -> Titanfall 2: Tech Test (PS4) -> Titanfall 2
And because of how manual the process is rn, Titanfall Engine depots/ are rarely looked at.

[snake-biscuits]

spending all day on tools for The Apex Archive made me realise how bad it is

[snake-biscuits]

Would be nice to have bulk tools for the following branches:

• [Test Maps] vindictus #40
• [Test Maps] vindictus69 #98

Vindictus has over 400 maps, and we have 2 branches to manage w/ 3 StaticProp formats
Also ties into:

• Test maps for Static Props #90
• Unmapped Static Prop lumps #110
• load_bsp branch detection can't tell some formats apart #17

[snake-biscuits]

Loading all Titanfall 2 depots:

>>> import bsp_tool, os, fnmatch
>>> dd = "E:/Mod/Titanfall2/depot"
>>> depots = os.listdir(dd)
>>> sd = "game/r2/maps"
>>> maps = {f"{d}/{m[:-4]}": bsp_tool.load_bsp(os.path.join(dd, d, sd, m))
...         for d in depots
...         for m in fnmatch.filter(os.listdir(os.path.join(dd, d, sd)), "*.bsp")}

Inspecting workshop map listings is equally absurd (1 folder per "mod")
Could probably say the same for .pk3 (Quake III Dreamcast is hell)

[snake-biscuits]

#195 is kinda revisiting this idea
But coming in from the angle of tests

Since we broke out autodetect into it's own script, bsp_tool/__init__.py is much less cluttered
I can see some proper functions for this

def bulk_load_bsp(special_dict: BulkDict, BspClass=None, branch=None):
    if BspClass is None or branch is None:
        use_autodetect = True
    ...

[snake-biscuits]

We need some kind of standard object to return
Not just a {"map_name": bsp} dict
Where a .bsp comes from is important (which Apex season, .pak etc.)

[snake-biscuits]

I've mostly finished a refactor of tests.files that could work for this

>>> from tests import files
>>> search_area = {"Steam": {"Half-Life 2 (Ep1)": ["half-life 2/episodic/maps/"]}}
>>> from bsp_tool.valve import ValveBsp
>>> from bsp_tool.branches.valve import orange_box
>>> files.library_bsps(ValveBsp, {orange_box: search_area})
{"Steam | Half-Life 2 (Ep1) | credits.bsp": <ValveBsp 'credits.bsp' valve.orange_box (VBSP version 19)>,
 "Steam | Half-Life 2 (Ep1) | ep1_c17_00.bsp": <ValveBsp 'ep1_c17_00.bsp' valve.orange_box (VBSP version 20)>,
 ...}

NOTE: no autodetect, since it's for testing branch scripts & BspClasses

Related

• Collect MegaTest files with tests.files #195