Add XML based vtk datatype

lib/galaxy/config/sample/datatypes_conf.xml.sample

@@ -975,6 +975,7 @@
     <datatype extension="gmsh.msh" type="galaxy.datatypes.constructive_solid_geometry:GmshMsh" display_in_upload="true"/>
     <datatype extension="gmsh.geo" type="galaxy.datatypes.constructive_solid_geometry:GmshGeo" display_in_upload="true"/>
     <datatype extension="zset.geof" type="galaxy.datatypes.constructive_solid_geometry:ZsetGeof" display_in_upload="true"/>
+    <datatype extension="vtkxml" type="galaxy.datatypes.constructive_solid_geometry:VtkXml" display_in_upload="true"/>
     <datatype extension="stl" type="galaxy.datatypes.binary:Binary" mimetype="application/octet-stream" display_in_upload="true" description="STL is a file format native to the stereolithography CAD software created by 3D Systems."/>
     <!-- Povray script -->
     <datatype extension="pov" type="galaxy.datatypes.text:Text" subclass="true" display_in_upload="true"/>
@@ -1033,6 +1034,7 @@
     <sniffer type="galaxy.datatypes.mothur:Axes"/>
     <sniffer type="galaxy.datatypes.constructive_solid_geometry:PlyAscii"/>
     <sniffer type="galaxy.datatypes.constructive_solid_geometry:PlyBinary"/>
+    <sniffer type="galaxy.datatypes.constructive_solid_geometry:VtkXml"/>
     <sniffer type="galaxy.datatypes.constructive_solid_geometry:VtkAscii"/>
     <sniffer type="galaxy.datatypes.constructive_solid_geometry:VtkBinary"/>
     <sniffer type="galaxy.datatypes.constructive_solid_geometry:NeperTess"/>

lib/galaxy/datatypes/constructive_solid_geometry.py

@@ -5,6 +5,7 @@
 """
 
 import abc
+import re
 from typing import (
     List,
     Optional,
@@ -29,6 +30,8 @@
     FilePrefix,
 )
 from galaxy.datatypes.tabular import Tabular
+from galaxy.datatypes.xml import GenericXml
+from xml.etree import ElementTree as ET
 
 if TYPE_CHECKING:
     from io import TextIOBase
@@ -189,9 +192,6 @@ class Vtk:
 
     Binary data must be placed into the file immediately after the newline
     ('\\n') character from the previous ASCII keyword and parameter sequence.
-
-    TODO: only legacy formats are currently supported and support for XML formats
-    should be added.
     """
 
     subtype = ""
@@ -813,3 +813,60 @@ def get_next_line(fh):
         # Discard the rest of the line
         fh.readline()
     return line.strip()
+
+class VtkXml(GenericXml):
+    """Format for defining VTK (XML based) and its sub-datatypes. https://docs.vtk.org/en/latest/design_documents/VTKFileFormats.html"""
+
+    edam_format = "edam:format_2332"
+    file_ext = "vtkxml"
+
+    #The same MetadataElements are also available for legacy VTK datatypes.
+    MetadataElement(name="vtk_version", default=None, desc="Vtk version", readonly=True, optional=True, visible=True)
+    MetadataElement(name="file_format", default=None, desc="File format", readonly=True, optional=True, visible=True)
+    MetadataElement(name="dataset_type", default=None, desc="Dataset type", readonly=True, optional=True, visible=True)
+
+    def extract_version(self, line: str) -> str:      
+        match = re.search(r'version="([^"]+)"', line)
+        if match:
+            return match.group(1)
+        return "?"
+
+    def extract_type(self, line: str) -> str:
+        match = re.search(r'type="([^"]+)"', line)
+        if match:
+            return match.group(1)
+        return "?"
+
+    def set_meta(self, dataset: DatasetProtocol, **kwd) -> None:
+        dataset.metadata.file_format = "XML"; 
+        with open(dataset.get_file_name(), errors="ignore") as file:
+            # first line might be the xml header, so we take two
+            first_line = file.readline()
+            if first_line.startswith("<?xml"):
+                first_line = file.readline()
+            dataset.metadata.vtk_version = self.extract_version(first_line) 
+            dataset.metadata.dataset_type = self.extract_type(first_line)
+
+
+    def set_peek(self, dataset: DatasetProtocol, **kwd) -> None:
+        """Set the peek and blurb text for VTK dataset files."""
+        if not dataset.dataset.purged:
+            dataset.peek = "VTK Dataset file"
+            dataset.blurb = f"type {dataset.metadata.dataset_type} version {dataset.metadata.vtk_version}"
+        else:
+            dataset.peek = "File does not exist"
+            dataset.blurb = "File purged from disk"
+
+    def sniff_prefix(self, file_prefix: FilePrefix) -> bool:
+        """Check for the key string 'VTKFile' to determine if this is a VTK dataset file.
+
+        >>> from galaxy.datatypes.sniff import get_test_fname
+        >>> fname = get_test_fname('data.vtu')
+        >>> VtkXml().sniff(fname)
+        True
+        >>> fname = get_test_fname('solid.xml')
+        >>> VtkXml().sniff(fname)
+        False
+        """
+        return self._has_root_element_in_prefix(file_prefix, "VTKFile")
+

lib/galaxy/datatypes/test/data.vtu

@@ -0,0 +1,41 @@
+<?xml version="1.0"?>
+<VTKFile type="UnstructuredGrid" version="0.1" byte_order="LittleEndian" header_type="UInt32" compressor="vtkZLibDataCompressor">
+  <UnstructuredGrid>
+    <Piece NumberOfPoints="21" NumberOfCells="21">
+      <PointData>
+        <DataArray type="Float64" Name="P_BC_FLOW" format="binary" RangeMin="-2" RangeMax="0">
+          AQAAAACAAACoAAAAFgAAAA==eJxjYCAKHCBAE1JPqvno4gwABYYGwQ==
+        </DataArray>
+        <DataArray type="UInt64" Name="bulk_node_ids" format="binary" RangeMin="0" RangeMax="70">
+          AQAAAACAAACoAAAAOwAAAA==eJwtxRECgDAAAMBJkDQomQySBk2SYNKg/v+kQXdyIfySJ2fP3r34cHTx6tObqy/fbn7c/frzAL7YAwQ=
+        </DataArray>
+      </PointData>
+      <CellData>
+      </CellData>
+      <Points>
+        <DataArray type="Float64" Name="Points" NumberOfComponents="3" format="binary" RangeMin="42.42640687119285" RangeMax="94.33981132056604">
+          AQAAAACAAAD4AQAAYwAAAA==eJxjYAABOwcw1RAIoeEAXVzlAIo4gx929XBxqHq4fjT1GOLo6j1xqPdEVQ+31wWHe2Di6OajqccQRzffDlU9hjiaerh5Jmjm+KGKY4QDurgnqjiGv9DVu+BQbwcXBwAoLCke
+          <InformationKey name="L2_NORM_RANGE" location="vtkDataArray" length="2">
+            <Value index="0">
+              42.426406871
+            </Value>
+            <Value index="1">
+              94.339811321
+            </Value>
+          </InformationKey>
+        </DataArray>
+      </Points>
+      <Cells>
+        <DataArray type="Int64" Name="connectivity" format="binary" RangeMin="0" RangeMax="20">
+          AQAAAACAAACoAAAAMQAAAA==eJwtxUECQCAAALAQRSLx/7d2sF0Wwm/y7MXRqzcnZ+8+XHy6+vLt5sfdrz8PMMgA0w==
+        </DataArray>
+        <DataArray type="Int64" Name="offsets" format="binary" RangeMin="1" RangeMax="21">
+          AQAAAACAAACoAAAAMgAAAA==eJwtxbcBgCAAADArFsAC8v+rDiRLxqGbPHvx6uDNuw+fjk7Ovnz78evi6s/NPzgAAOg=
+        </DataArray>
+        <DataArray type="UInt8" Name="types" format="binary" RangeMin="1" RangeMax="1">
+          AQAAAACAAAAVAAAACwAAAA==eJxjZMQCAAD8ABY=
+        </DataArray>
+      </Cells>
+    </Piece>
+  </UnstructuredGrid>
+</VTKFile>

lib/galaxy/datatypes/test/solid.xml

@@ -0,0 +1,25 @@
+<phase>
+    <type>Solid</type>
+    <properties>
+        <property>
+            <name>density</name>
+            <type>Constant</type>
+            <value>2300</value>
+        </property>
+        <property>
+            <name>thermal_conductivity</name>
+            <type>Constant</type>
+            <value>1.9</value>
+        </property>
+        <property>
+            <name>specific_heat_capacity</name>
+            <type>Constant</type>
+            <value>800</value>
+        </property>
+        <property>
+            <name>thermal_expansivity</name>
+            <type>Constant</type>
+            <value>1.7e-5</value>
+        </property>
+    </properties>
+</phase>

test/unit/data/datatypes/test_constructive_solid_geometry.py

@@ -0,0 +1,18 @@
+
+from galaxy.datatypes.constructive_solid_geometry import VtkXml 
+from .util import (
+    get_input_files,
+    MockDataset,
+)
+
+def test_vtkXml_set_meta():
+        vtkXml = VtkXml()
+        with get_input_files("data.vtu") as input_files:
+            dataset = MockDataset(1)
+            dataset.set_file_name(input_files[0])
+
+            vtkXml.set_meta(dataset)
+
+        assert dataset.metadata.vtk_version == "0.1"  
+        assert dataset.metadata.file_format == "XML"  
+        assert dataset.metadata.dataset_type == "UnstructuredGrid"