merge-duplicates slice and did a first polish

src/repo_registry/storage/sqlite.py

@@ -643,6 +643,132 @@ class RegistryStore:
             target_capability_id=target_capability_id,
         )
 
+    def merge_candidate_ability(
+        self,
+        repository_id: int,
+        analysis_run_id: int,
+        source_ability_id: int,
+        target_ability_id: int,
+    ) -> None:
+        if source_ability_id == target_ability_id:
+            raise ValueError("source and target candidate ability must be different")
+        self._ensure_candidate_row(
+            table="candidate_abilities",
+            label="target candidate ability",
+            repository_id=repository_id,
+            analysis_run_id=analysis_run_id,
+            candidate_id=target_ability_id,
+        )
+        with self.connect() as connection:
+            cursor = connection.execute(
+                """
+                UPDATE candidate_abilities
+                SET status = 'merged'
+                WHERE id = ?
+                  AND repository_id = ?
+                  AND analysis_run_id = ?
+                  AND status = 'candidate'
+                """,
+                (source_ability_id, repository_id, analysis_run_id),
+            )
+            if cursor.rowcount == 0:
+                raise NotFoundError(
+                    "source candidate ability "
+                    f"{source_ability_id} was not found for repository "
+                    f"{repository_id} analysis run {analysis_run_id}"
+                )
+            connection.execute(
+                """
+                UPDATE candidate_capabilities
+                SET ability_id = ?
+                WHERE ability_id = ? AND repository_id = ? AND analysis_run_id = ?
+                """,
+                (target_ability_id, source_ability_id, repository_id, analysis_run_id),
+            )
+
+    def merge_candidate_capability(
+        self,
+        repository_id: int,
+        analysis_run_id: int,
+        source_capability_id: int,
+        target_capability_id: int,
+    ) -> None:
+        if source_capability_id == target_capability_id:
+            raise ValueError("source and target candidate capability must be different")
+        self._ensure_candidate_row(
+            table="candidate_capabilities",
+            label="target candidate capability",
+            repository_id=repository_id,
+            analysis_run_id=analysis_run_id,
+            candidate_id=target_capability_id,
+        )
+        with self.connect() as connection:
+            cursor = connection.execute(
+                """
+                UPDATE candidate_capabilities
+                SET status = 'merged'
+                WHERE id = ?
+                  AND repository_id = ?
+                  AND analysis_run_id = ?
+                  AND status = 'candidate'
+                """,
+                (source_capability_id, repository_id, analysis_run_id),
+            )
+            if cursor.rowcount == 0:
+                raise NotFoundError(
+                    "source candidate capability "
+                    f"{source_capability_id} was not found for repository "
+                    f"{repository_id} analysis run {analysis_run_id}"
+                )
+            for table in ("candidate_features", "candidate_evidence"):
+                connection.execute(
+                    f"""
+                    UPDATE {table}
+                    SET capability_id = ?
+                    WHERE capability_id = ?
+                      AND repository_id = ?
+                      AND analysis_run_id = ?
+                    """,
+                    (
+                        target_capability_id,
+                        source_capability_id,
+                        repository_id,
+                        analysis_run_id,
+                    ),
+                )
+
+    def merge_candidate_feature(
+        self,
+        repository_id: int,
+        analysis_run_id: int,
+        source_feature_id: int,
+        target_feature_id: int,
+    ) -> None:
+        self._merge_candidate_leaf(
+            table="candidate_features",
+            label="candidate feature",
+            repository_id=repository_id,
+            analysis_run_id=analysis_run_id,
+            source_id=source_feature_id,
+            target_id=target_feature_id,
+        )
+
+    def merge_candidate_evidence(
+        self,
+        repository_id: int,
+        analysis_run_id: int,
+        source_evidence_id: int,
+        target_evidence_id: int,
+    ) -> None:
+        self._merge_candidate_leaf(
+            table="candidate_evidence",
+            label="candidate evidence",
+            repository_id=repository_id,
+            analysis_run_id=analysis_run_id,
+            source_id=source_evidence_id,
+            target_id=target_evidence_id,
+        )
+
     def _ensure_candidate_row(
         self,
         *,
@@ -703,6 +829,43 @@ class RegistryStore:
                 f"{repository_id} analysis run {analysis_run_id}"
             )
 
+    def _merge_candidate_leaf(
+        self,
+        *,
+        table: str,
+        label: str,
+        repository_id: int,
+        analysis_run_id: int,
+        source_id: int,
+        target_id: int,
+    ) -> None:
+        if source_id == target_id:
+            raise ValueError(f"source and target {label} must be different")
+        self._ensure_candidate_row(
+            table=table,
+            label=f"target {label}",
+            repository_id=repository_id,
+            analysis_run_id=analysis_run_id,
+            candidate_id=target_id,
+        )
+        with self.connect() as connection:
+            cursor = connection.execute(
+                f"""
+                UPDATE {table}
+                SET status = 'merged'
+                WHERE id = ?
+                  AND repository_id = ?
+                  AND analysis_run_id = ?
+                  AND status = 'candidate'
+                """,
+                (source_id, repository_id, analysis_run_id),
+            )
+        if cursor.rowcount == 0:
+            raise NotFoundError(
+                f"source {label} {source_id} was not found for repository "
+                f"{repository_id} analysis run {analysis_run_id}"
+            )
+
     def _reject_candidate_leaf(
         self,
         *,