WIP: Add schema matcher plugin

CHANGELOG.md

@@ -7,7 +7,9 @@ The format is based on [Keep a Changelog](http://keepachangelog.com/) and this p
 
 ## [Unreleased]
 
-TODO: add at least one Added, Changed, Deprecated, Removed, Fixed or Security section
+### Added
+
+- Added `SchemaMatcherPlugin` interface.
 
 
 ## [4.5.0] 2024-01-10

cmem_plugin_base/dataintegration/ai/schema_matcher.py

@@ -0,0 +1,52 @@
+"""All classes related to schema matcher plugins.
+   WARNING: All classes in this file are preliminary and might be changed."""
+
+from cmem_plugin_base.dataintegration.plugins import PluginBase
+
+
+class MatchingSchema:
+    """The schema that is used by schema matchers."""
+
+    def write_rdf_file(self, path: str, lang: str) -> None:
+        """Write this schema to an RDF file.
+
+        :param path: The target file path.
+        :param lang: The RDF format. Usually, either "N-TRIPLE" or "TURTLE".
+        """
+        # Implementation provided by DataIntegration
+
+
+class Correspondence:
+    """Candidate match between two properties from two schemata."""
+
+    def __init__(self, source: str, target: str, confidence: float):
+        self.source = source
+        self.target = target
+        self.confidence = confidence
+
+    def __str__(self):
+        """Convert to a string representation"""
+        return f"{self.source} - {self.target} ({self.confidence})"
+
+
+class Alignment:
+    """Set of correspondences between two schemata."""
+
+    def __init__(self, matches: list[Correspondence]):
+        self.matches = matches
+
+
+class SchemaMatcherPlugin(PluginBase):
+    """
+    A schema matcher aligns a source dataset with target vocabularies.
+    """
+
+    def match(self, source: MatchingSchema, target: MatchingSchema) -> Alignment:
+        """
+        Aligns a source dataset with target vocabularies.
+
+        :param source: Source schema
+        :param target: Target schema
+
+        :return: The generated alignment.
+        """

cmem_plugin_base/dataintegration/description.py

@@ -14,6 +14,7 @@
     PluginContextParameterType,
 )
 from cmem_plugin_base.dataintegration.utils import generate_id
+from cmem_plugin_base.dataintegration.ai.schema_matcher import SchemaMatcherPlugin
 
 
 class Icon:
@@ -129,6 +130,8 @@ def __init__(  # pylint: disable=too-many-arguments
             self.plugin_type = "WorkflowPlugin"
         elif issubclass(plugin_class, TransformPlugin):
             self.plugin_type = "TransformPlugin"
+        elif issubclass(plugin_class, SchemaMatcherPlugin):
+            self.plugin_type = "SchemaMatcherPlugin"
         else:
             raise ValueError(
                 f"Class {plugin_class.__name__} does not implement a supported "