feat: Add full-text index type support (#51)

* Add full-text index type support

* Add full-text index type support

---------

Co-authored-by: xxsc0529 <xxsc0529@users.noreply.github.com>

Author: xxsc0529

pyobvector/client/fts_index_param.py

@@ -1,28 +1,58 @@
 """A module to specify fts index parameters"""
 from enum import Enum
-from typing import List, Optional
+from typing import List, Optional, Union
 
 class FtsParser(Enum):
+    """Built-in full-text search parser types supported by OceanBase"""
     IK = 0
     NGRAM = 1
+    NGRAM2 = 2  # NGRAM2 parser (supported from V4.3.5 BP2+)
+    BASIC_ENGLISH = 3  # Basic English parser
+    JIEBA = 4  # jieba parser
 
 
 class FtsIndexParam:
+    """Full-text search index parameter.
+    
+    Args:
+        index_name: Index name
+        field_names: List of field names to create full-text index on
+        parser_type: Parser type, can be FtsParser enum or string (for custom parsers)
+                    If None, uses default Space parser
+    """
     def __init__(
         self,
         index_name: str,
         field_names: List[str],
-        parser_type: Optional[FtsParser],
+        parser_type: Optional[Union[FtsParser, str]] = None,
     ):
         self.index_name = index_name
         self.field_names = field_names
         self.parser_type = parser_type
 
-    def param_str(self) -> str | None:
-        if self.parser_type == FtsParser.IK:
-            return "ik"
-        if self.parser_type == FtsParser.NGRAM:
-            return "ngram"
+    def param_str(self) -> Optional[str]:
+        """Convert parser type to string format for SQL."""
+        if self.parser_type is None:
+            return None  # Default Space parser, no need to specify
+        
+        if isinstance(self.parser_type, str):
+            # Custom parser name (e.g., "thai_ftparser")
+            return self.parser_type.lower()
+        
+        if isinstance(self.parser_type, FtsParser):
+            if self.parser_type == FtsParser.IK:
+                return "ik"
+            if self.parser_type == FtsParser.NGRAM:
+                return "ngram"
+            if self.parser_type == FtsParser.NGRAM2:
+                return "ngram2"
+            if self.parser_type == FtsParser.BASIC_ENGLISH:
+                return "beng"
+            if self.parser_type == FtsParser.JIEBA:
+                return "jieba"
+            # Raise exception for unrecognized FtsParser enum values
+            raise ValueError(f"Unrecognized FtsParser enum value: {self.parser_type}")
+        
         return None
 
     def __iter__(self):
@@ -34,7 +64,7 @@ def __iter__(self):
     def __str__(self):
         return str(dict(self))
 
-    def __eq__(self, other: None):
+    def __eq__(self, other: object) -> bool:
         if isinstance(other, self.__class__):
             return dict(self) == dict(other)
 

pyobvector/client/ob_vec_json_table_client.py

@@ -884,12 +884,19 @@ def _handle_jtable_dml_select(
                 identifier.args['quoted'] = False
                 col.args['table'] = identifier
 
-        join_clause = parse_one(f"from t1, {json_table_str}")
-        join_node = join_clause.args['joins'][0]
-        if 'joins' in ast.args.keys():
-            ast.args['joins'].append(join_node)
-        else:
-            ast.args['joins'] = [join_node]
+        # Manually create the JOIN node for json_table
+        # In some versions of sqlglot, comma-separated tables may not be parsed as
+        # explicit JOINS, so we directly parse the json_table expression and create a JOIN node
+        # explicitly
+        json_table_expr = parse_one(json_table_str, dialect="oceanbase")
+        
+        join_node = exp.Join()
+        join_node.args['this'] = json_table_expr
+        join_node.args['kind'] = None  # CROSS JOIN (implicit join with comma)
+        
+        if 'joins' not in ast.args:
+            ast.args['joins'] = []
+        ast.args['joins'].append(join_node)
 
         if real_user_id:
             extra_filter_str = f"{JSON_TABLE_DATA_TABLE_NAME}.user_id = '{real_user_id}' AND {JSON_TABLE_DATA_TABLE_NAME}.jtable_name = '{table_name}'"