GoodData Apache Ossie Converter

Bidirectional converter between GoodData's declarative Logical Data Model (LDM) and the Apache Ossie semantic model specification.

Features

• GoodData → Ossie: Convert a GoodData declarative LDM JSON to Ossie semantic model YAML
• Ossie → GoodData: Convert an Ossie semantic model YAML to GoodData declarative LDM JSON
• Preserves GoodData-specific metadata (labels, date granularities, geo types) via Ossie custom_extensions
• Generates dual-dialect expressions (ANSI_SQL + MAQL) for fields

Usage

import json
import yaml
from ossie_gooddata import gooddata_to_osi, osi_to_gooddata
from ossie_gooddata.models import gd_model_from_dict, gd_model_to_dict

# GoodData → Apache Ossie
with open("gooddata_ldm.json") as f:
    gd_model = gd_model_from_dict(json.load(f))
osi_model = gooddata_to_osi(gd_model, model_name="my_model")
with open("osi_model.yaml", "w") as f:
    yaml.dump(osi_model, f, default_flow_style=False)

# Apache Ossie → GoodData
with open("osi_model.yaml") as f:
    osi_data = yaml.safe_load(f)
gd_model = osi_to_gooddata(osi_data, data_source_id="my_datasource")
with open("gooddata_ldm.json", "w") as f:
    json.dump(gd_model_to_dict(gd_model), f, indent=2)

Development

uv sync --group dev
uv run pytest

Concept Mapping

Limitations

• Metrics are not converted. GoodData metrics use MAQL, a context-aware metric language where dimensionality and filters are applied at report time. The current Ossie metric model is SQL-expression-based and cannot represent this paradigm.
• AggregatedFacts are not yet supported.
• Workspace data filters are not mapped.