umadump

https://github.com/werseter/umadump
Admin
umadump is a runtime memory reader and data exporter for Uma Musume Pretty Derby that extracts game...
Tokens:121,767
Snippets:745
Trust Score:6.4
Update:1 week ago
Show doc for...
Tokens:
Raw
### CLI usage examples

Source: https://github.com/werseter/umadump/blob/master/README.md

Various command-line flags for running umadump in live mode, daemon mode, or processing minidumps.

```powershell
# Live mode (attaches to running game process)
python main.py

# Live mode, run once and exit
python main.py --rerun-mode once

# Live daemon mode, polling every 2 seconds and writing changed outputs
python main.py --rerun-mode daemon --poll-interval 2

# Dev mode from full-memory minidump
python main.py --minidump "D:\path\to\dump.dmp" --metadata-path "D:\path\to\global-metadata.dat"

# Run schema validation only, then exit
python main.py --minidump "D:\path\to\dump.dmp" --validate-only

# Skip the startup GitHub release lookup
python main.py --no-update-check
```

--------------------------------

### Metadata path configuration

Source: https://github.com/werseter/umadump/blob/master/README.md

The default location for the global metadata file, derived from the game executable directory.

```text
<exe_dir>/<ExeName>_Data/il2cpp_data/Metadata/global-metadata.dat
```

--------------------------------

### Registering Schema and Runtime Validatable Classes

Source: https://github.com/werseter/umadump/blob/master/README.md

Use @register_schema_validatable for startup metadata cross-checks and @register_runtime_validatable for additional per-access runtime guards.

```python
@register_schema_validatable("Gallop::WorkSkillData.AcquiredSkill")
class AcquiredSkillFields(CStructureDataclass):
    ...


@register_runtime_validatable("Gallop::WorkDataManager")
class WorkDataManagerWrapper(CStructureDataclass):
    _il2cpp_obj: RuntimeIl2CppObject
    ...
```

--------------------------------

### jq commands using output_validation.jq for API and export comparison

Source: https://github.com/werseter/umadump/blob/master/README.md

Run jq with the output_validation module to normalize an API response (career_load_api) or a dump file (pass_through). Use files representing the same observation; the contract does not select matching captures or perform the comparison.

```sh
jq -L . 'include "output_validation"; career_load_api' response.json
jq -L . 'include "output_validation"; pass_through' turn_024_010.json
```

--------------------------------

### Registering Enums and Defining Enum Fields

Source: https://github.com/werseter/umadump/blob/master/README.md

Enums are registered with a storage type, and fields can be defined using C_Enum or C_EnumIn to preserve semantics and enable validation.

```python
@register_enum("Gallop::RaceDefine.RunningStyle", storage_type=c_uint8)
class RunningStyle(SafeIntEnum):
    None_ = 0
    Nige = 1
    # ...


class RaceFields(CStructureDataclass):
    runningStyle: C_Enum[RunningStyle]
    encryptedState: C_EnumIn[RunningStyle, ObscuredInt]
```
Tokens:
Raw