Add AutoCAD DXF data source

README.md

@@ -6,6 +6,7 @@
 Introduced in Spark 4.x, Python Data Source API allows you to create PySpark Data Sources leveraging long standing python libraries for handling unique file types or specialized interfaces with spark read, readStream, write and writeStream APIs.
 | Data Source Name | Purpose |
 | --- | --- |
+| [dxf](dxf/README.md) | Read AutoCAD DXF drawing exchange files |
 | [zipdcm](zipdcm/README.md) | Read DICOM files from Zip file archives |
 
 ## Documentation
@@ -27,5 +28,6 @@ Please see our [installation guide](./INSTALL.md)
 
 | Datasource | Package    | Purpose                           | License     | Source                               |
 | ---------- | ---------- | --------------------------------- | ----------- | ------------------------------------ |
+| dxf        | ezdxf      | Python library for DXF files      | MIT         | https://github.com/mozman/ezdxf      |
 | zipdcm     | pydicom	  | Python api for DICOM files        |	MIT	        | https://github.com/pydicom/pydicom   |
 | zipdcm     | pylibjpeg  | Decoding / Encoding pixel formats | GPLv3 & MIT | https://github.com/pydicom/pylibjpeg |

dxf/Makefile

@@ -0,0 +1,28 @@
+.PHONY: dev test unit style check
+
+all: clean style test 
+
+clean: ## Remove build artifacts and cache files
+	rm -rf build/
+	rm -rf dist/
+	rm -rf *.egg-info/
+	rm -rf htmlcov/
+	rm -rf .coverage
+	rm -rf coverage.xml
+	rm -rf .pytest_cache/
+	rm -rf .mypy_cache/
+	rm -rf .ruff_cache/
+	find . -type d -name __pycache__ -delete
+	find . -type f -name "*.pyc" -delete
+
+test:
+	pip install -r requirements.txt
+	pytest .
+
+dev:
+	pip install -r requirements.txt
+
+style:
+	pre-commit run --all-files
+
+check: style test

dxf/README.md

@@ -0,0 +1,75 @@
+# DXF Data Source
+
+A PySpark Python Data Source for reading AutoCAD DXF (Drawing Exchange Format) files using the [ezdxf](https://ezdxf.readthedocs.io/en/stable/) library.
+
+DXF is an open exchange format for CAD systems. This data source extracts geometric entities from DXF files into a tabular format suitable for analysis in Spark.
+
+## Schema
+
+| Column | Type | Description |
+|--------|------|-------------|
+| file_path | STRING | Path to the source DXF file |
+| entity_type | STRING | DXF entity type (LINE, CIRCLE, ARC, TEXT, etc.) |
+| layer | STRING | The layer the entity belongs to |
+| handle | STRING | Unique entity handle within the DXF file |
+| attributes | STRING | JSON string containing entity-specific attributes |
+
+## Supported Entity Types
+
+| Entity Type | Attributes |
+|-------------|-----------|
+| LINE | start_x/y/z, end_x/y/z |
+| CIRCLE | center_x/y/z, radius |
+| ARC | center_x/y/z, radius, start_angle, end_angle |
+| POINT | location_x/y/z |
+| TEXT | text, insert_x/y/z, height, rotation |
+| MTEXT | text, insert_x/y/z, height |
+| LWPOLYLINE | points (array), is_closed |
+| POLYLINE | vertices (array), is_closed |
+| SPLINE | control_points (array), degree, is_closed |
+| ELLIPSE | center_x/y/z, major_axis_x/y/z, ratio, start_param, end_param |
+| INSERT | block_name, insert_x/y/z, x/y/z_scale, rotation |
+| DIMENSION | dimtype, defpoint_x/y/z |
+| HATCH | pattern_name, solid_fill |
+
+## Usage
+
+```python
+# Register the data source
+from dxf_datasource import DXFDataSource
+spark.dataSource.register(DXFDataSource)
+
+# Read all entities from DXF files
+df = spark.read.format("dxf").option("path", "/path/to/files").load()
+
+# Filter by layer at read time
+df = spark.read.format("dxf") \
+    .option("path", "/path/to/files") \
+    .option("layerFilter", "walls") \
+    .load()
+
+# Extract attributes from JSON
+from pyspark.sql.functions import get_json_object, col
+
+circles = df.filter(col("entity_type") == "CIRCLE") \
+    .select(
+        "layer",
+        get_json_object(col("attributes"), "$.center_x").alias("cx"),
+        get_json_object(col("attributes"), "$.center_y").alias("cy"),
+        get_json_object(col("attributes"), "$.radius").alias("radius"),
+    )
+```
+
+## Options
+
+| Option | Default | Description |
+|--------|---------|-------------|
+| path | (required) | Path to DXF file(s) or directory |
+| pathGlobFilter | `*.dxf` | Glob pattern for file matching |
+| numPartitions | 4 | Number of partitions to split files across |
+| recursiveFileLookup | false | Recursively search subdirectories |
+| layerFilter | (all layers) | Filter entities by layer name |
+
+## Dependencies
+
+- [ezdxf](https://ezdxf.readthedocs.io/en/stable/) - Python library for reading/writing DXF files

dxf/requirements.txt

@@ -0,0 +1 @@
+ezdxf==1.4.2