refactor(core): route integrations through adapter factory

harness/tht/cli/lsh_cmd.py

@@ -32,20 +32,30 @@ def build_cmd(config: Path = CONFIG_OPT) -> None:
     annotations = Annotations.from_yaml(annotations_path(cfg))
 
     typer.echo("Estrazione valori (i più frequenti) dalle colonne testuali eligible...")
-    if cfg.database.transport == "rest":
-        from tht.db.sampling import unique_values_for_lsh_rest
-        from tht.rest.client import RestClient
+    from tht.adapters.factory import build_dwh
+    from tht.db.sampling import SkippedColumn, TruncatedColumn, is_text_type
+    from tht.mschema.eligibility import effective_eligibility
 
-        values, skipped, truncated = unique_values_for_lsh_rest(
-            RestClient(cfg.rest), physical, cfg.lsh, annotations
-        )
-    else:
-        from tht.db.connection import make_engine
-        from tht.db.sampling import unique_values_for_lsh
-
-        values, skipped, truncated = unique_values_for_lsh(
-            make_engine(cfg.database), physical, cfg.lsh, annotations
-        )
+    dwh = build_dwh(cfg)
+    values: dict[str, dict[str, list[str]]] = {}
+    skipped: list[SkippedColumn] = []
+    truncated: list[TruncatedColumn] = []
+    for table_name, table in physical.tables.items():
+        table_ann = annotations.tables.get(table_name)
+        for column_name, column in table.columns.items():
+            ann_col = table_ann.columns.get(column_name) if table_ann else None
+            if not is_text_type(column.type) or not effective_eligibility(column, ann_col)[0]:
+                continue
+            try:
+                distinct = dwh.distinct_values(table_name, column_name)
+            except Exception as e:
+                skipped.append(SkippedColumn(table_name, column_name, f"errore: {e}"))
+                continue
+            vals = [str(value) for value in distinct.values if value not in (None, "")]
+            if vals:
+                values.setdefault(table_name, {})[column_name] = vals
+            if distinct.truncated or len(vals) >= cfg.lsh.max_values_per_column:
+                truncated.append(TruncatedColumn(table_name, column_name, len(vals)))
     n_values = sum(len(v) for t in values.values() for v in t.values())
     typer.echo(f"  {n_values} valori da {sum(len(t) for t in values.values())} colonne")
     for s in skipped: