Add enum values to table_summary and truncate long column values

Author: txsmith

TODO.md

@@ -9,9 +9,3 @@
     - changing DB port
   - search
   - table_summary
-
-
-## schema include/exclude list in config
-
-## Add enum values to table_summary
-## Truncate sampled columns in table_summary

config_example.yaml

@@ -74,5 +74,4 @@ databases:
 settings:
   max_query_timeout: 30
   max_rows_per_query: 500
-  sample_size: 10
   enable_write_operations: false

src/meeseeql/main.py

@@ -103,10 +103,18 @@ async def table_summary(
     database: str,
     table_name: str,
     db_schema: str | None = None,
-    limit: int = 250,
+    limit: int = 100,
     page: int = 1,
 ) -> ToolResult:
-    """Get table structure including columns and foreign keys with pagination"""
+    """Get table structure including columns and foreign keys with pagination
+
+    Args:
+        database: Database name to query
+        table_name: Name of the table to summarize
+        db_schema: Optional schema name (uses default schema if not provided)
+        limit: Maximum items per page (default: 100)
+        page: Page number (default: 1)
+    """
     result = await tools.table_summary(
         get_or_init_db_manager(), database, table_name, db_schema, limit, page
     )

src/meeseeql/tools/sql/mssql/enum_values.sql

@@ -0,0 +1,4 @@
+SELECT
+    NULL as column_name,
+    NULL as enum_values
+LIMIT 0

src/meeseeql/tools/sql/mysql/enum_values.sql

@@ -0,0 +1,13 @@
+SELECT
+    c.column_name,
+    CASE
+        WHEN c.data_type = 'enum' THEN
+            REPLACE(REPLACE(SUBSTRING(c.column_type, 6, CHAR_LENGTH(c.column_type) - 6), '''', ''), ')', '')
+        ELSE NULL
+    END as enum_values
+FROM information_schema.columns c
+WHERE c.table_schema NOT IN ('information_schema', 'mysql', 'performance_schema', 'sys')
+AND LOWER(c.table_name) = LOWER('{{table_name}}')
+AND LOWER(c.table_schema) = LOWER('{{schema_name}}')
+AND c.data_type = 'enum'
+ORDER BY c.ordinal_position

src/meeseeql/tools/sql/postgresql/columns.sql

@@ -1,6 +1,9 @@
 SELECT
     c.column_name,
-    c.data_type,
+    CASE 
+        WHEN c.data_type = 'USER-DEFINED' THEN c.udt_name
+        ELSE c.data_type
+    END as data_type,
     c.is_nullable,
     c.column_default
 FROM information_schema.columns c

src/meeseeql/tools/sql/postgresql/enum_values.sql

@@ -0,0 +1,12 @@
+SELECT
+    c.column_name,
+    string_agg(e.enumlabel, ', ' ORDER BY e.enumsortorder) as enum_values
+FROM information_schema.columns c
+JOIN pg_type t ON t.typname = c.udt_name
+JOIN pg_enum e ON e.enumtypid = t.oid
+WHERE c.table_schema NOT IN ('information_schema', 'pg_catalog')
+AND LOWER(c.table_name) = LOWER('{{table_name}}')
+AND LOWER(c.table_schema) = LOWER('{{schema_name}}')
+AND t.typtype = 'e'
+GROUP BY c.column_name, c.ordinal_position
+ORDER BY c.ordinal_position

src/meeseeql/tools/sql/snowflake/enum_values.sql

@@ -0,0 +1,4 @@
+SELECT
+    NULL as column_name,
+    NULL as enum_values
+LIMIT 0

src/meeseeql/tools/sql/sqlite/enum_values.sql

@@ -0,0 +1,4 @@
+SELECT
+    NULL as column_name,
+    NULL as enum_values
+LIMIT 0

src/meeseeql/tools/table_summary.py

@@ -22,6 +22,7 @@ class ColumnInfo(BaseModel):
     nullable: bool
     default: Any = None
     primary_key: bool = False
+    enum_values: str | None = None
 
 
 class ForeignKey(BaseModel):
@@ -54,6 +55,8 @@ def __str__(self) -> str:
                     parts.insert(-1, "PRIMARY KEY")
                 if col.default is not None:
                     parts.append(f"default: {col.default}")
+                if col.enum_values:
+                    parts.append(f"values: {col.enum_values}")
                 result += f"  {col.name}: {', '.join(parts)}\n"
 
         if self.sample_rows:
@@ -64,7 +67,13 @@ def __str__(self) -> str:
                 result += f"  {'-' * (len(' | '.join(column_names)))}\n"
                 for row in self.sample_rows:
                     row_str = " | ".join(
-                        str(val) if val is not None else "NULL" for val in row
+                        (
+                            # Truncate column values to 50 chars to prevent massive overflow
+                            str(val)[:50] + ("..." if len(str(val)) > 50 else "")
+                            if val is not None
+                            else "NULL"
+                        )
+                        for val in row
                     )
                     result += f"  {row_str}\n"
 
@@ -86,6 +95,29 @@ def __str__(self) -> str:
         return result
 
 
+async def _get_enum_values(
+    db_manager: DatabaseManager,
+    database: str,
+    table_name: str,
+    db_schema: str,
+) -> dict[str, str]:
+    try:
+        dialect = db_manager.get_dialect_name(database)
+        enum_query_template = load_sql_query(dialect, "enum_values")
+        enum_query = enum_query_template.replace("{{table_name}}", table_name).replace(
+            "{{schema_name}}", db_schema
+        )
+
+        transformer = SqlQueryTransformer(enum_query, dialect)
+        transformer.validate_read_only()
+
+        result = await db_manager.execute_query(database, enum_query)
+        rows = result.fetchall()
+        return {row[0]: row[1] for row in rows if row[0] and row[1]}
+    except Exception:
+        return {}
+
+
 async def _get_primary_keys(
     db_manager: DatabaseManager,
     database: str,
@@ -117,6 +149,7 @@ async def _get_columns(
     table_name: str,
     db_schema: str,
     primary_keys: set,
+    enum_values: dict[str, str],
     limit: int,
     offset: int,
 ) -> List[ColumnInfo]:
@@ -152,6 +185,7 @@ async def _get_columns(
                     nullable=nullable,
                     default=column_default,
                     primary_key=is_primary_key,
+                    enum_values=enum_values.get(column_name),
                 )
             )
         return columns
@@ -365,12 +399,13 @@ async def table_summary(
         db_manager, database, table_name, schema_value
     )
 
-    sample_size = db_manager.config.settings.get("sample_size", 5)
+    enum_values = await _get_enum_values(db_manager, database, table_name, schema_value)
+
     sample_response = await execute_query(
         db_manager,
         database,
         f"SELECT * FROM {schema_value}.{table_name}",
-        limit=sample_size,
+        limit=5,
         page=1,
     )
     sample_rows = [list(row.values()) for row in sample_response.rows]
@@ -386,6 +421,7 @@ async def table_summary(
             table_name,
             schema_value,
             primary_keys,
+            enum_values,
             num_columns_to_fetch,
             current_offset,
         )