Enable supports_statement_cache for Db2 SQLAlchemy dialects

ibm_db_sa/base.py

@@ -24,6 +24,8 @@
 from sqlalchemy import types as sa_types
 from sqlalchemy import schema as sa_schema
 from sqlalchemy import util
+from sqlalchemy import exc
+from sqlalchemy.sql.elements import BindParameter
 from sqlalchemy.sql import compiler
 from sqlalchemy.sql import operators
 from sqlalchemy.engine import default
@@ -396,42 +398,87 @@ def visit_mod_binary(self, binary, operator, **kw):
         return "mod(%s, %s)" % (self.process(binary.left),
                                 self.process(binary.right))
 
-    def limit_clause(self, select, **kwargs):
-        limit = select._limit
-        offset = select._offset or 0
-
-        if limit is not None:
-            if offset > 0:
-                return f" LIMIT {limit} OFFSET {offset}"
-            else:
-                return f" LIMIT {limit}"
-        return ""
+    def literalBindsFlagFrom_kw(self, kw=None):
+        """Return True if literal_binds is requested in compile kwargs."""
+        if not kw or not isinstance(kw, dict):
+            return False
+        if kw.get("literal_binds"):
+            return True
+        ck = kw.get("compile_kwargs")
+        if isinstance(ck, dict) and ck.get("literal_binds"):
+            return True
+        return False
+
+    def limit_clause(self, select, **kw):
+        text = ""
+        limit_clause = select._limit_clause
+        offset_clause = select._offset_clause
+        literal_binds = self.literalBindsFlagFrom_kw(kw)
 
-    def visit_select(self, select, **kwargs):
-        limit, offset = select._limit, select._offset
-        sql_ori = compiler.SQLCompiler.visit_select(self, select, **kwargs)
+        def _render_clause(clause):
+            if clause is None:
+                return None
+            if select._simple_int_clause(clause):
+                return self.process(clause.render_literal_execute(), **kw)
+            if literal_binds:
+                if hasattr(clause, "render_literal_execute"):
+                    try:
+                        return self.process(clause.render_literal_execute(), **kw)
+                    except Exception:
+                        pass
+                try:
+                    return self.process(clause, literal_binds=True, **kw)
+                except Exception:
+                    pass
+                try:
+                    if isinstance(clause, BindParameter):
+                        val = getattr(clause, "value", None)
+                        if val is not None:
+                            if isinstance(val, str):
+                                return f"'{val}'"
+                            return str(val)
+                except Exception:
+                    pass
+            try:
+                return self.process(clause, **kw)
+            except Exception as e:
+                raise exc.CompileError(
+                    "dialect 'ibm_db_sa' cannot render LIMIT/OFFSET for this clause; "
+                    "ensure the clause is a simple integer or is processable by the compiler."
+                ) from e
+
+        limit_text = _render_clause(limit_clause)
+        if limit_text is not None:
+            text += " LIMIT %s" % limit_text
+        offset_text = _render_clause(offset_clause)
+        if offset_text is not None:
+            text += " OFFSET %s" % offset_text
+        return text
 
-        if ('LIMIT' in sql_ori.upper()) or ('FETCH FIRST' in sql_ori.upper()):
+    def visit_select(self, select, **kw):
+        sql_ori = compiler.SQLCompiler.visit_select(self, select, **kw)
+        if ("LIMIT" in sql_ori.upper()) or ("FETCH FIRST" in sql_ori.upper()):
             return sql_ori
-
-        if limit is not None:
-            sql = re.sub(r'FETCH FIRST \d+ ROWS ONLY', '', sql_ori, flags=re.IGNORECASE).strip()
-            limit_offset_clause = self.limit_clause(select, **kwargs)
-            sql += limit_offset_clause
-            return sql
-
-        if offset is not None:
+        limit_clause_obj = select._limit_clause
+        offset_clause_obj = select._offset_clause
+        if limit_clause_obj is not None:
+            limit_offset_clause = self.limit_clause(select, **kw)
+            if limit_offset_clause:
+                return sql_ori + limit_offset_clause
+        if offset_clause_obj is not None:
             __rownum = 'Z.__ROWNUM'
-            sql_split = re.split(r"[\s+]FROM ", sql_ori, 1)
+            sql_work = re.sub(r'FETCH FIRST \d+ ROWS ONLY', '', sql_ori, flags=re.IGNORECASE).strip()
+            sql_work = re.sub(r'\s+OFFSET\s+(?:\d+|__\[POSTCOMPILE_[^\]]+\]|:[A-Za-z0-9_]+|\?)\s*$', '', sql_work,
+                              flags=re.IGNORECASE)
+            sql_split = re.split(r"[\s+]FROM ", sql_work, 1)
+            if len(sql_split) < 2:
+                return sql_ori
             sql_sec = " \nFROM %s " % (sql_split[1])
-
             dummyVal = "Z.__db2_"
             sql_pri = ""
-
             sql_sel = "SELECT "
             if select._distinct:
                 sql_sel = "SELECT DISTINCT "
-
             sql_select_token = sql_split[0].split(",")
             i = 0
             while i < len(sql_select_token):
@@ -440,32 +487,41 @@ def visit_select(self, select, **kwargs):
                     sql_pri = f'{sql_pri} {sql_select_token[i]},{sql_select_token[i + 1]},{sql_select_token[i + 2]},{sql_select_token[i + 3]} AS "{dummyVal}{i + 1}",'
                     i += 4
                     continue
-
                 if sql_select_token[i].count(" AS ") == 1:
                     temp_col_alias = sql_select_token[i].split(" AS ")
                     sql_pri = f'{sql_pri} {sql_select_token[i]},'
                     sql_sel = f'{sql_sel} {temp_col_alias[1]},'
                     i += 1
                     continue
-
                 sql_pri = f'{sql_pri} {sql_select_token[i]} AS "{dummyVal}{i + 1}",'
                 sql_sel = f'{sql_sel} "{dummyVal}{i + 1}",'
                 i += 1
-
             sql_pri = sql_pri.rstrip(",")
             sql_pri = f"{sql_pri}{sql_sec}"
             sql_sel = sql_sel.rstrip(",")
             sql = f'{sql_sel}, ( ROW_NUMBER() OVER() ) AS "{__rownum}" FROM ( {sql_pri} ) AS M'
             sql = f'{sql_sel} FROM ( {sql} ) Z WHERE'
 
-            if offset != 0:
-                sql = f'{sql} "{__rownum}" > {offset}'
-            if offset != 0 and limit is not None:
+            def _process_clause_text(clause):
+                if clause is None:
+                    return None
+                if select._simple_int_clause(clause):
+                    return self.process(clause.render_literal_execute(), **kw)
+                else:
+                    return self.process(clause, **kw)
+
+            offset_text = _process_clause_text(offset_clause_obj)
+            limit_text = _process_clause_text(limit_clause_obj)
+            if offset_text is not None:
+                sql = f'{sql} "{__rownum}" > {offset_text}'
+            if offset_text is not None and limit_text is not None:
                 sql = f'{sql} AND '
-            if limit is not None:
-                sql = f'{sql} "{__rownum}" <= {offset + limit}'
+            if limit_text is not None:
+                if offset_text is not None:
+                    sql = f'{sql} "{__rownum}" <= ({offset_text} + {limit_text})'
+                else:
+                    sql = f'{sql} "{__rownum}" <= {limit_text}'
             return f"( {sql} )"
-
         return sql_ori
 
     def visit_sequence(self, sequence, **kw):
@@ -753,7 +809,7 @@ class DB2Dialect(default.DefaultDialect):
     supports_sane_multi_rowcount = True
     supports_native_decimal = False
     supports_native_boolean = False
-    supports_statement_cache = False
+    supports_statement_cache = True
     preexecute_sequences = False
     supports_alter = True
     supports_sequences = True

ibm_db_sa/ibm_db.py

@@ -99,7 +99,7 @@ def get_result_proxy(self):
 class DB2Dialect_ibm_db(DB2Dialect):
     driver = 'ibm_db_sa'
     supports_unicode_statements = True
-    supports_statement_cache = False
+    supports_statement_cache = True
     supports_sane_rowcount = True
     supports_sane_multi_rowcount = False
     supports_native_decimal = False

ibm_db_sa/pyodbc.py

@@ -30,7 +30,7 @@ class DB2Dialect_pyodbc(PyODBCConnector, DB2Dialect):
     supports_unicode_statements = True
     supports_char_length = True
     supports_native_decimal = False
-    supports_statement_cache = False
+    supports_statement_cache = True
 
     execution_ctx_cls = DB2ExecutionContext_pyodbc
 

ibm_db_sa/reflection.py

@@ -29,6 +29,7 @@
 
 class CoerceUnicode(sa_types.TypeDecorator):
     impl = sa_types.Unicode
+    cache_ok = True
 
     def process_bind_param(self, value, dialect):
         if isinstance(value, str):

pyproject.toml

@@ -1,3 +1,3 @@
 [build-system]
-requires = ["setuptools>=42", "wheel"]
+requires = ["setuptools>=42", "wheel", "packaging>=20.0"]
 build-backend = "setuptools.build_meta"

setup.py

@@ -11,9 +11,9 @@
 
 readme = os.path.join(os.path.dirname(__file__), 'README.md')
 if 'USE_PYODBC' in os.environ and os.environ['USE_PYODBC'] == '1':
-    require = ['sqlalchemy>=0.7.3']
+    require = ['sqlalchemy>=0.7.3', 'packaging>=20.0']
 else:
-    require = ['sqlalchemy>=0.7.3','ibm_db>=2.0.0']
+    require = ['sqlalchemy>=0.7.3','ibm_db>=2.0.0', 'packaging>=20.0']
 
 
 setup(