Merge pull request #42 from IGNF/move-writer-params

Move trick to pass params to a writer to las_info

CHANGELOG.md

@@ -1,3 +1,6 @@
+# dev
+- refactor tool to propagate header infos from one pipeline to another to use it by itself
+
 # 1.5.1
 - fix add_buffer: propagate header infos from input to the output
 - update pdal.Writer params to make sure input format is forwarded except for the specified parameters

pdaltools/las_add_buffer.py

@@ -1,11 +1,14 @@
 import argparse
 import logging
 import os
-from typing import Dict, List
+from typing import List
 
 import pdal
 
-from pdaltools.las_info import get_buffered_bounds_from_filename
+from pdaltools.las_info import (
+    get_buffered_bounds_from_filename,
+    get_writer_parameters_from_reader_metadata,
+)
 from pdaltools.las_merge import create_list
 
 
@@ -46,40 +49,6 @@ def create_las_with_buffer(
     )
 
 
-def get_writer_parameters_from_reader_metadata(metadata: Dict, a_srs=None) -> Dict:
-    """As pdal las writers does not permit to pass easily metadata from one file as
-    parameters for a writer, use a trick to generate writer parameters from the
-    reader metadata of a previous pipeline:
-    This function uses the metadata from the reader of a pipeline to provide parameters
-    to pass to the writer of another pipeline
-
-    To be removed once https://github.com/PDAL/python/issues/147 is solved
-
-    Args:
-        metadata (Dict): metadata of an executed pipeline (that can be accessed using pipeline.metadata)
-    Returns:
-        Dict: parameters to pass to a pdal writer
-    """
-
-    reader_metadata = metadata["metadata"]["readers.las"]
-
-    params = {
-        "major_version": reader_metadata["major_version"],
-        "minor_version": reader_metadata["minor_version"],
-        "global_encoding": reader_metadata["global_encoding"],
-        "extra_dims": "all",
-        "scale_x": reader_metadata["scale_x"],
-        "scale_y": reader_metadata["scale_y"],
-        "scale_z": reader_metadata["scale_z"],
-        "offset_x": reader_metadata["offset_x"],
-        "offset_y": reader_metadata["offset_y"],
-        "offset_z": reader_metadata["offset_z"],
-        "dataformat_id": reader_metadata["dataformat_id"],
-        "a_srs": a_srs if a_srs else reader_metadata["comp_spatialreference"],
-    }
-    return params
-
-
 def las_merge_and_crop(
     input_dir: str,
     tile_filename: str,

pdaltools/las_info.py

@@ -1,7 +1,7 @@
 import json
 import logging
 import os
-from typing import Tuple
+from typing import Dict, Tuple
 
 import osgeo.osr as osr
 import pdal
@@ -151,3 +151,37 @@ def get_buffered_bounds_from_filename(
     ys = [minY - buffer_width, maxY + buffer_width]
 
     return (xs, ys)
+
+
+def get_writer_parameters_from_reader_metadata(metadata: Dict, a_srs=None) -> Dict:
+    """As pdal las writers does not permit to pass easily metadata from one file as
+    parameters for a writer, use a trick to generate writer parameters from the
+    reader metadata of a previous pipeline:
+    This function uses the metadata from the reader of a pipeline to provide parameters
+    to pass to the writer of another pipeline
+
+    To be removed once https://github.com/PDAL/python/issues/147 is solved
+
+    Args:
+        metadata (Dict): metadata of an executed pipeline (that can be accessed using pipeline.metadata)
+    Returns:
+        Dict: parameters to pass to a pdal writer
+    """
+
+    reader_metadata = metadata["metadata"]["readers.las"]
+
+    params = {
+        "major_version": reader_metadata["major_version"],
+        "minor_version": reader_metadata["minor_version"],
+        "global_encoding": reader_metadata["global_encoding"],
+        "extra_dims": "all",
+        "scale_x": reader_metadata["scale_x"],
+        "scale_y": reader_metadata["scale_y"],
+        "scale_z": reader_metadata["scale_z"],
+        "offset_x": reader_metadata["offset_x"],
+        "offset_y": reader_metadata["offset_y"],
+        "offset_z": reader_metadata["offset_z"],
+        "dataformat_id": reader_metadata["dataformat_id"],
+        "a_srs": a_srs if a_srs else reader_metadata["comp_spatialreference"],
+    }
+    return params

test/test_las_info.py

@@ -1,6 +1,7 @@
 import os
 import test.utils as tu
 
+import pdal
 import pytest
 
 from pdaltools import las_info
@@ -90,3 +91,43 @@ def test_get_buffered_bounds_from_filename_with_buffer():
     )
     assert xs == [770550 - buffer_width, 770600 + buffer_width]
     assert ys == [6277550 - buffer_width, 6277600 + buffer_width]
+
+
+def test_get_writer_parameters_from_reader_metadata():
+    output_file = os.path.join(TMP_PATH, "writer_with_parameters.las")
+    output_expected = os.path.join(TMP_PATH, "writer_with_forward.las")
+
+    # First generate the expected output from a pipeline in a single part:
+    pipeline = pdal.Pipeline()
+    pipeline |= pdal.Reader.las(filename=INPUT_FILE)
+    pipeline |= pdal.Writer.las(
+        filename=output_expected,
+        forward="all",
+    )
+    pipeline.execute()
+
+    # Use pdal info summary to get metadata about the las file format and not only those of the data
+    out_expected_metadata = tu.get_pdal_infos_summary(output_expected)["summary"]["metadata"]
+
+    # Generate output from a pipeline separated in 2 parts
+    pipeline1 = pdal.Pipeline()
+    pipeline1 |= pdal.Reader.las(filename=INPUT_FILE)
+    # At this point a useful pipeline wou
+    pipeline1.execute()
+    out_data = pipeline.arrays[0]
+    metadata = pipeline.metadata
+
+    # Here in the expected usecase, you would do stuff to modify your data
+    params = las_info.get_writer_parameters_from_reader_metadata(metadata)
+    pipeline2 = pdal.Pipeline(arrays=[out_data])
+    pipeline2 |= pdal.Writer(filename=output_file, forward="all", **params)
+    pipeline2.execute()
+
+    out_metadata = tu.get_pdal_infos_summary(output_file)["summary"]["metadata"]
+
+    # Pop metadata that we don't expect to be the same
+    for key in ["creation_year", "creation_doy", "software_id"]:
+        out_metadata.pop(key)
+        out_expected_metadata.pop(key)
+
+    assert out_metadata == out_expected_metadata