informatica-python 1.9.9__tar.gz → 1.10.1__tar.gz

{informatica_python-1.9.9 → informatica_python-1.10.1}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: informatica-python
-Version: 1.9.9
+Version: 1.10.1
 Summary: Convert Informatica PowerCenter workflow XML to Python/PySpark code
 Author: Nick
 License: MIT
@@ -414,7 +414,15 @@ The generated `helper_functions.py` provides a complete runtime library:
 
 ## Changelog
 
-### v1.9.8 (Current)
+### v1.10.1 (Current)
+- **Router multi-group output support**: Router transformations now properly handle `<GROUP>` elements with `EXPRESSION` attributes — generates separate filtered DataFrames for each named output group (e.g., `df_rtr_rest_type_per`, `df_rtr_rest_value_per`), not just the DEFAULT group
+- **Connector group routing**: `FROMINSTANCEGROUP` / `TOINSTANCEGROUP` attributes on `CONNECTOR` elements are now parsed and used to wire downstream transforms/targets to the correct Router output group
+- **Joiner group-aware routing**: Joiner transformations now correctly identify master/detail sources through Router group connectors — two different Router output groups feeding a Joiner are properly resolved to their distinct DataFrames
+- **GroupDef expression field**: `GroupDef` model now stores the `EXPRESSION` attribute from `<GROUP>` XML elements
+- **Backward-compatible Router fallback**: Existing `TABLEATTRIBUTE`-based Router group conditions (older XML format) continue to work — the code checks `<GROUP>` elements first, then falls back to `TABLEATTRIBUTE` entries
+- **223 tests** passing
+
+### v1.9.8
 - **NOT(expr) function-call form**: `NOT(ISNULL(x))` now correctly converts to `~(df["x"].isna())` — handles both `NOT ` (with space) and `NOT(` (without space) forms
 - **AND/OR/NOT as field names fix**: Logical operators no longer mangled into `df["AND"]` / `df["OR"]` — conversion moved before field substitution in both `_vec_recursive` fallback and `_vectorize_simple`
 - **Condition tokenizer word-boundary fix**: `_split_condition_tokens` no longer splits on `OR` inside field names like `DeletedIndicator` — verifies preceding character is a real word boundary

{informatica_python-1.9.9 → informatica_python-1.10.1}/README.md

@@ -387,7 +387,15 @@ The generated `helper_functions.py` provides a complete runtime library:
 
 ## Changelog
 
-### v1.9.8 (Current)
+### v1.10.1 (Current)
+- **Router multi-group output support**: Router transformations now properly handle `<GROUP>` elements with `EXPRESSION` attributes — generates separate filtered DataFrames for each named output group (e.g., `df_rtr_rest_type_per`, `df_rtr_rest_value_per`), not just the DEFAULT group
+- **Connector group routing**: `FROMINSTANCEGROUP` / `TOINSTANCEGROUP` attributes on `CONNECTOR` elements are now parsed and used to wire downstream transforms/targets to the correct Router output group
+- **Joiner group-aware routing**: Joiner transformations now correctly identify master/detail sources through Router group connectors — two different Router output groups feeding a Joiner are properly resolved to their distinct DataFrames
+- **GroupDef expression field**: `GroupDef` model now stores the `EXPRESSION` attribute from `<GROUP>` XML elements
+- **Backward-compatible Router fallback**: Existing `TABLEATTRIBUTE`-based Router group conditions (older XML format) continue to work — the code checks `<GROUP>` elements first, then falls back to `TABLEATTRIBUTE` entries
+- **223 tests** passing
+
+### v1.9.8
 - **NOT(expr) function-call form**: `NOT(ISNULL(x))` now correctly converts to `~(df["x"].isna())` — handles both `NOT ` (with space) and `NOT(` (without space) forms
 - **AND/OR/NOT as field names fix**: Logical operators no longer mangled into `df["AND"]` / `df["OR"]` — conversion moved before field substitution in both `_vec_recursive` fallback and `_vectorize_simple`
 - **Condition tokenizer word-boundary fix**: `_split_condition_tokens` no longer splits on `OR` inside field names like `DeletedIndicator` — verifies preceding character is a real word boundary

{informatica_python-1.9.9 → informatica_python-1.10.1}/informatica_python/generators/mapping_gen.py

@@ -54,6 +54,8 @@ def _expand_mapplet_recursive(mapplet, mapplet_map, prefix, depth=0, max_depth=1
             to_instance=new_to,
             to_field=conn.to_field,
             to_instance_type=conn.to_instance_type,
+            from_instance_group=conn.from_instance_group,
+            to_instance_group=conn.to_instance_group,
         ))
 
     for inst in getattr(mapplet, 'instances', []):
@@ -138,6 +140,8 @@ def _inline_mapplets(mapping, folder):
                     to_instance=first_tx,
                     to_field=conn.to_field,
                     to_instance_type=conn.to_instance_type,
+                    from_instance_group=conn.from_instance_group,
+                    to_instance_group=conn.to_instance_group,
                 ))
             else:
                 rewired_connectors.append(conn)
@@ -167,6 +171,8 @@ def _inline_mapplets(mapping, folder):
                     to_instance=conn.to_instance,
                     to_field=conn.to_field,
                     to_instance_type=conn.to_instance_type,
+                    from_instance_group=conn.from_instance_group,
+                    to_instance_group=conn.to_instance_group,
                 ))
             else:
                 rewired_connectors.append(conn)
@@ -730,7 +736,11 @@ def _generate_transformation(lines, tx, connector_graph, source_dfs, transform_m
     input_conns = connector_graph.get("to", {}).get(tx.name, [])
     input_sources = set()
     for c in input_conns:
-        input_sources.add(c.from_instance)
+        if c.from_instance_group:
+            group_key = f"{c.from_instance}:{c.from_instance_group}"
+            input_sources.add(group_key)
+        else:
+            input_sources.add(c.from_instance)
 
     input_df = None
     for src in input_sources:
@@ -739,7 +749,14 @@ def _generate_transformation(lines, tx, connector_graph, source_dfs, transform_m
             break
     if not input_df:
         for src in input_sources:
-            input_df = f"df_{_safe_name(src)}"
+            base = src.split(":")[0] if ":" in src else src
+            if base in source_dfs:
+                input_df = source_dfs[base]
+                break
+    if not input_df:
+        for src in input_sources:
+            base = src.split(":")[0] if ":" in src else src
+            input_df = f"df_{_safe_name(base)}"
             break
     if not input_df:
         input_df = "df_input"
@@ -946,10 +963,11 @@ def _gen_joiner_transform(lines, tx, tx_safe, input_df, input_sources, source_df
         to_field = conn.to_field
         port_to_col[to_field] = conn.from_field
         port_to_col[to_field.lower()] = conn.from_field
+        src_key = f"{conn.from_instance}:{conn.from_instance_group}" if conn.from_instance_group else conn.from_instance
         if to_field in master_fields or to_field.lower() in master_fields_lower:
-            master_src = conn.from_instance
+            master_src = src_key
         elif to_field in detail_fields or to_field.lower() in detail_fields_lower:
-            detail_src = conn.from_instance
+            detail_src = src_key
 
     if left_keys and right_keys and port_to_col:
         left_keys = [port_to_col.get(k, port_to_col.get(k.lower(), k)) for k in left_keys]
@@ -967,9 +985,17 @@ def _gen_joiner_transform(lines, tx, tx_safe, input_df, input_sources, source_df
                 detail_src = s
                 break
 
+    def _resolve_src(key):
+        if key in source_dfs:
+            return source_dfs[key]
+        base = key.split(":")[0] if ":" in key else key
+        if base in source_dfs:
+            return source_dfs[base]
+        return f"df_{_safe_name(base)}"
+
     if master_src and detail_src:
-        df_master = source_dfs.get(master_src, f"df_{_safe_name(master_src)}")
-        df_detail = source_dfs.get(detail_src, f"df_{_safe_name(detail_src)}")
+        df_master = _resolve_src(master_src)
+        df_detail = _resolve_src(detail_src)
 
         lines.append(f"    # Join ({join_type}): {join_condition or 'auto'}")
         if left_keys and right_keys:
@@ -1095,9 +1121,24 @@ def _gen_lookup_transform(lines, tx, tx_safe, input_df, source_dfs, connector_gr
 def _gen_router_transform(lines, tx, tx_safe, input_df, source_dfs):
     lines.append(f"    # Router groups:")
     group_conditions = {}
-    for attr in tx.attributes:
-        if "Group Filter Condition" in attr.name:
-            group_conditions[attr.name] = attr.value
+
+    output_groups = [
+        g for g in tx.groups
+        if g.type.upper() not in ("INPUT", "") and "DEFAULT" not in g.type.upper()
+    ]
+    output_groups.sort(key=lambda g: g.order)
+
+    if output_groups:
+        for g in output_groups:
+            if g.expression and g.expression.strip():
+                group_conditions[g.name] = g.expression
+            else:
+                group_conditions[g.name] = ""
+
+    if not group_conditions:
+        for attr in tx.attributes:
+            if "Group Filter Condition" in attr.name:
+                group_conditions[attr.name] = attr.value
 
     remaining_mask_parts = []
     if group_conditions:
@@ -1108,15 +1149,21 @@ def _gen_router_transform(lines, tx, tx_safe, input_df, source_dfs):
                 expr_py = f"pd.Series(True, index={input_df}.index)"
             mask_var = f"_router_mask_{tx_safe}_{i}"
             lines.append(f"    {mask_var} = {expr_py}  # {gname}")
-            lines.append(f"    df_{tx_safe}_group{i} = {input_df}[{mask_var}].copy()")
-            source_dfs[f"{tx.name}_group{i}"] = f"df_{tx_safe}_group{i}"
+            group_df_name = f"df_{tx_safe}_{_safe_name(gname)}"
+            lines.append(f"    {group_df_name} = {input_df}[{mask_var}].copy()")
+            source_dfs[f"{tx.name}:{gname}"] = group_df_name
             remaining_mask_parts.append(f"~{mask_var}")
+
+    default_groups = [g for g in tx.groups if "DEFAULT" in g.type.upper()]
+    default_name = default_groups[0].name if default_groups else "DEFAULT"
+
     if remaining_mask_parts:
         lines.append(f"    _router_default_mask = {' & '.join(remaining_mask_parts)}")
-        lines.append(f"    df_{tx_safe} = {input_df}[_router_default_mask].copy()  # Default group")
+        lines.append(f"    df_{tx_safe} = {input_df}[_router_default_mask].copy()  # Default group ({default_name})")
     else:
-        lines.append(f"    df_{tx_safe} = {input_df}.copy()  # Default group")
+        lines.append(f"    df_{tx_safe} = {input_df}.copy()  # Default group ({default_name})")
     source_dfs[tx.name] = f"df_{tx_safe}"
+    source_dfs[f"{tx.name}:{default_name}"] = f"df_{tx_safe}"
 
 
 def _gen_union_transform(lines, tx, tx_safe, input_sources, source_dfs, data_lib="pandas"):
@@ -1448,6 +1495,11 @@ def _generate_target_write(lines, tgt_name, tgt_def, connector_graph, source_dfs
     to_conns = connector_graph.get("to", {}).get(tgt_name, [])
     input_df = None
     for c in to_conns:
+        if c.from_instance_group:
+            group_key = f"{c.from_instance}:{c.from_instance_group}"
+            if group_key in source_dfs:
+                input_df = source_dfs[group_key]
+                break
         if c.from_instance in source_dfs:
             input_df = source_dfs[c.from_instance]
             break

{informatica_python-1.9.9 → informatica_python-1.10.1}/informatica_python/models.py

@@ -80,6 +80,7 @@ class GroupDef:
     type: str = ""
     description: str = ""
     order: int = 0
+    expression: str = ""
     fields: List[FieldDef] = field(default_factory=list)
 
 
@@ -274,6 +275,8 @@ class ConnectorDef:
     to_field: str
     to_instance: str
     to_instance_type: str
+    from_instance_group: str = ""
+    to_instance_group: str = ""
 
 
 @dataclass

{informatica_python-1.9.9 → informatica_python-1.10.1}/informatica_python/parser.py

@@ -240,6 +240,7 @@ class InformaticaParser:
             type=self._attr(elem, "TYPE"),
             description=self._attr(elem, "DESCRIPTION"),
             order=self._int_attr(elem, "ORDER"),
+            expression=self._attr(elem, "EXPRESSION"),
         )
         for fld in elem.findall("SOURCEFIELD"):
             grp.fields.append(self._parse_source_field(fld))
@@ -580,6 +581,8 @@ class InformaticaParser:
             to_field=self._attr(elem, "TOFIELD"),
             to_instance=self._attr(elem, "TOINSTANCE"),
             to_instance_type=self._attr(elem, "TOINSTANCETYPE"),
+            from_instance_group=self._attr(elem, "FROMINSTANCEGROUP"),
+            to_instance_group=self._attr(elem, "TOINSTANCEGROUP"),
         )
 
     def _parse_instance(self, elem) -> InstanceDef:

{informatica_python-1.9.9 → informatica_python-1.10.1}/informatica_python.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: informatica-python
-Version: 1.9.9
+Version: 1.10.1
 Summary: Convert Informatica PowerCenter workflow XML to Python/PySpark code
 Author: Nick
 License: MIT
@@ -414,7 +414,15 @@ The generated `helper_functions.py` provides a complete runtime library:
 
 ## Changelog
 
-### v1.9.8 (Current)
+### v1.10.1 (Current)
+- **Router multi-group output support**: Router transformations now properly handle `<GROUP>` elements with `EXPRESSION` attributes — generates separate filtered DataFrames for each named output group (e.g., `df_rtr_rest_type_per`, `df_rtr_rest_value_per`), not just the DEFAULT group
+- **Connector group routing**: `FROMINSTANCEGROUP` / `TOINSTANCEGROUP` attributes on `CONNECTOR` elements are now parsed and used to wire downstream transforms/targets to the correct Router output group
+- **Joiner group-aware routing**: Joiner transformations now correctly identify master/detail sources through Router group connectors — two different Router output groups feeding a Joiner are properly resolved to their distinct DataFrames
+- **GroupDef expression field**: `GroupDef` model now stores the `EXPRESSION` attribute from `<GROUP>` XML elements
+- **Backward-compatible Router fallback**: Existing `TABLEATTRIBUTE`-based Router group conditions (older XML format) continue to work — the code checks `<GROUP>` elements first, then falls back to `TABLEATTRIBUTE` entries
+- **223 tests** passing
+
+### v1.9.8
 - **NOT(expr) function-call form**: `NOT(ISNULL(x))` now correctly converts to `~(df["x"].isna())` — handles both `NOT ` (with space) and `NOT(` (without space) forms
 - **AND/OR/NOT as field names fix**: Logical operators no longer mangled into `df["AND"]` / `df["OR"]` — conversion moved before field substitution in both `_vec_recursive` fallback and `_vectorize_simple`
 - **Condition tokenizer word-boundary fix**: `_split_condition_tokens` no longer splits on `OR` inside field names like `DeletedIndicator` — verifies preceding character is a real word boundary

{informatica_python-1.9.9 → informatica_python-1.10.1}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "informatica-python"
-version = "1.9.9"
+version = "1.10.1"
 description = "Convert Informatica PowerCenter workflow XML to Python/PySpark code"
 readme = "README.md"
 license = {text = "MIT"}

{informatica_python-1.9.9 → informatica_python-1.10.1}/tests/test_integration.py

@@ -2911,3 +2911,302 @@ class TestConcatWithLtrimRtrim(unittest.TestCase):
         result = convert_expression_vectorized(expr)
         assert "+" in result
         assert "||" not in result
+
+
+class TestRouterGroupElements(unittest.TestCase):
+
+    ROUTER_GROUP_XML = '''<?xml version="1.0" encoding="UTF-8"?>
+<!DOCTYPE POWERMART SYSTEM "powrmart.dtd">
+<POWERMART CREATION_DATE="01/01/2025" REPOSITORY_VERSION="1">
+<REPOSITORY NAME="repo" VERSION="1" CODEPAGE="UTF-8" DATABASETYPE="Oracle">
+<FOLDER NAME="TEST" OWNER="admin">
+  <SOURCE NAME="SRC" DATABASETYPE="Flat File" DBDNAME="SRC">
+    <FLATFILE DELIMITEDBY="COMMA" HEADERROWPRESENT="YES" PADBYTES="NO" ROWDELIMITER="\\n"/>
+    <SOURCEFIELD NAME="ID" DATATYPE="integer" PRECISION="10" SCALE="0" NULLABLE="NOTNULL" KEYTYPE="PRIMARY KEY" FIELDNUMBER="1"/>
+    <SOURCEFIELD NAME="Party_Type" DATATYPE="string" PRECISION="10" SCALE="0" NULLABLE="NULL" KEYTYPE="NOT A KEY" FIELDNUMBER="2"/>
+    <SOURCEFIELD NAME="Attrib_Name" DATATYPE="string" PRECISION="50" SCALE="0" NULLABLE="NULL" KEYTYPE="NOT A KEY" FIELDNUMBER="3"/>
+  </SOURCE>
+  <TARGET NAME="TGT_PER_TYPE" DATABASETYPE="Flat File">
+    <TARGETFIELD NAME="ID" DATATYPE="integer" PRECISION="10" SCALE="0" NULLABLE="NULL" KEYTYPE="NOT A KEY" FIELDNUMBER="1"/>
+  </TARGET>
+  <TARGET NAME="TGT_PER_VALUE" DATABASETYPE="Flat File">
+    <TARGETFIELD NAME="ID" DATATYPE="integer" PRECISION="10" SCALE="0" NULLABLE="NULL" KEYTYPE="NOT A KEY" FIELDNUMBER="1"/>
+  </TARGET>
+  <TARGET NAME="TGT_ORG_TYPE" DATABASETYPE="Flat File">
+    <TARGETFIELD NAME="ID" DATATYPE="integer" PRECISION="10" SCALE="0" NULLABLE="NULL" KEYTYPE="NOT A KEY" FIELDNUMBER="1"/>
+  </TARGET>
+  <TARGET NAME="TGT_DEFAULT" DATABASETYPE="Flat File">
+    <TARGETFIELD NAME="ID" DATATYPE="integer" PRECISION="10" SCALE="0" NULLABLE="NULL" KEYTYPE="NOT A KEY" FIELDNUMBER="1"/>
+  </TARGET>
+  <MAPPING NAME="m_router_groups_test" ISVALID="YES">
+    <TRANSFORMATION NAME="SQ_SRC" TYPE="Source Qualifier" REUSABLE="NO">
+      <TRANSFORMFIELD NAME="ID" DATATYPE="integer" PRECISION="10" SCALE="0" PORTTYPE="OUTPUT"/>
+      <TRANSFORMFIELD NAME="Party_Type" DATATYPE="string" PRECISION="10" SCALE="0" PORTTYPE="OUTPUT"/>
+      <TRANSFORMFIELD NAME="Attrib_Name" DATATYPE="string" PRECISION="50" SCALE="0" PORTTYPE="OUTPUT"/>
+    </TRANSFORMATION>
+    <TRANSFORMATION NAME="RTR_SPLIT" TYPE="Router" REUSABLE="NO">
+      <TRANSFORMFIELD NAME="ID" DATATYPE="integer" PRECISION="10" SCALE="0" PORTTYPE="INPUT/OUTPUT"/>
+      <TRANSFORMFIELD NAME="Party_Type" DATATYPE="string" PRECISION="10" SCALE="0" PORTTYPE="INPUT/OUTPUT"/>
+      <TRANSFORMFIELD NAME="Attrib_Name" DATATYPE="string" PRECISION="50" SCALE="0" PORTTYPE="INPUT/OUTPUT"/>
+      <GROUP DESCRIPTION="" NAME="INPUT" ORDER="1" TYPE="INPUT"/>
+      <GROUP DESCRIPTION="" EXPRESSION="Party_Type = &apos;PER&apos; AND Attrib_Name = &apos;RESTRICTION_TYPE&apos;" NAME="Rest_Type_PER" ORDER="2" TYPE="OUTPUT"/>
+      <GROUP DESCRIPTION="" EXPRESSION="Party_Type = &apos;PER&apos; AND Attrib_Name = &apos;RESTRICTION_VALUE&apos;" NAME="Rest_Value_PER" ORDER="3" TYPE="OUTPUT"/>
+      <GROUP DESCRIPTION="" EXPRESSION="Party_Type = &apos;ORG&apos; AND Attrib_Name = &apos;RESTRICTION_TYPE&apos;" NAME="Rest_Type_ORG" ORDER="4" TYPE="OUTPUT"/>
+      <GROUP DESCRIPTION="" NAME="DEFAULT1" ORDER="5" TYPE="OUTPUT/DEFAULT"/>
+    </TRANSFORMATION>
+    <INSTANCE NAME="SRC" TYPE="Source Definition" TRANSFORMATION_NAME="SRC"/>
+    <INSTANCE NAME="SQ_SRC" TYPE="Source Qualifier" TRANSFORMATION_NAME="SQ_SRC"/>
+    <INSTANCE NAME="RTR_SPLIT" TYPE="Router" TRANSFORMATION_NAME="RTR_SPLIT"/>
+    <INSTANCE NAME="TGT_PER_TYPE" TYPE="Target Definition" TRANSFORMATION_NAME="TGT_PER_TYPE"/>
+    <INSTANCE NAME="TGT_PER_VALUE" TYPE="Target Definition" TRANSFORMATION_NAME="TGT_PER_VALUE"/>
+    <INSTANCE NAME="TGT_ORG_TYPE" TYPE="Target Definition" TRANSFORMATION_NAME="TGT_ORG_TYPE"/>
+    <INSTANCE NAME="TGT_DEFAULT" TYPE="Target Definition" TRANSFORMATION_NAME="TGT_DEFAULT"/>
+    <CONNECTOR FROMINSTANCE="SRC" FROMFIELD="ID" TOINSTANCE="SQ_SRC" TOFIELD="ID"/>
+    <CONNECTOR FROMINSTANCE="SRC" FROMFIELD="Party_Type" TOINSTANCE="SQ_SRC" TOFIELD="Party_Type"/>
+    <CONNECTOR FROMINSTANCE="SRC" FROMFIELD="Attrib_Name" TOINSTANCE="SQ_SRC" TOFIELD="Attrib_Name"/>
+    <CONNECTOR FROMINSTANCE="SQ_SRC" FROMFIELD="ID" TOINSTANCE="RTR_SPLIT" TOFIELD="ID"/>
+    <CONNECTOR FROMINSTANCE="SQ_SRC" FROMFIELD="Party_Type" TOINSTANCE="RTR_SPLIT" TOFIELD="Party_Type"/>
+    <CONNECTOR FROMINSTANCE="SQ_SRC" FROMFIELD="Attrib_Name" TOINSTANCE="RTR_SPLIT" TOFIELD="Attrib_Name"/>
+    <CONNECTOR FROMINSTANCE="RTR_SPLIT" FROMINSTANCEGROUP="Rest_Type_PER" FROMFIELD="ID" TOINSTANCE="TGT_PER_TYPE" TOFIELD="ID"/>
+    <CONNECTOR FROMINSTANCE="RTR_SPLIT" FROMINSTANCEGROUP="Rest_Value_PER" FROMFIELD="ID" TOINSTANCE="TGT_PER_VALUE" TOFIELD="ID"/>
+    <CONNECTOR FROMINSTANCE="RTR_SPLIT" FROMINSTANCEGROUP="Rest_Type_ORG" FROMFIELD="ID" TOINSTANCE="TGT_ORG_TYPE" TOFIELD="ID"/>
+    <CONNECTOR FROMINSTANCE="RTR_SPLIT" FROMINSTANCEGROUP="DEFAULT1" FROMFIELD="ID" TOINSTANCE="TGT_DEFAULT" TOFIELD="ID"/>
+  </MAPPING>
+  <CONFIG NAME="default_session_config"/>
+  <WORKFLOW NAME="wf_router_groups_test" ISVALID="YES">
+    <TASK NAME="Start" REUSABLE="NO" TYPE="Start"/>
+    <SESSION NAME="s_m_router_groups_test" ISVALID="YES" REUSABLE="NO" MAPPINGNAME="m_router_groups_test">
+      <CONFIGREFERENCE REFOBJECTNAME="default_session_config" TYPE="Session config"/>
+    </SESSION>
+    <TASKINSTANCE NAME="Start" TASKNAME="Start" TASKTYPE="Start"/>
+    <TASKINSTANCE NAME="s_m_router_groups_test" TASKNAME="s_m_router_groups_test" TASKTYPE="Session"/>
+    <WORKFLOWLINK FROMTASK="Start" TOTASK="s_m_router_groups_test"/>
+  </WORKFLOW>
+</FOLDER>
+</REPOSITORY>
+</POWERMART>'''
+
+    def test_router_generates_all_named_groups(self):
+        converter = InformaticaConverter()
+        tmpdir = tempfile.mkdtemp()
+        try:
+            converter.convert_string(self.ROUTER_GROUP_XML, output_dir=tmpdir)
+            for fn in os.listdir(tmpdir):
+                if fn.startswith("mapping_") and fn.endswith(".py"):
+                    with open(os.path.join(tmpdir, fn)) as f:
+                        code = f.read()
+                    assert "Rest_Type_PER" in code, "Should have Rest_Type_PER group"
+                    assert "Rest_Value_PER" in code, "Should have Rest_Value_PER group"
+                    assert "Rest_Type_ORG" in code, "Should have Rest_Type_ORG group"
+                    assert "_router_mask_" in code, "Should have router masks"
+                    assert "Default group" in code, "Should have default group"
+                    assert "_router_default_mask" in code, "Should have default mask"
+                    break
+        finally:
+            shutil.rmtree(tmpdir)
+
+    def test_router_group_creates_separate_dataframes(self):
+        converter = InformaticaConverter()
+        tmpdir = tempfile.mkdtemp()
+        try:
+            converter.convert_string(self.ROUTER_GROUP_XML, output_dir=tmpdir)
+            for fn in os.listdir(tmpdir):
+                if fn.startswith("mapping_") and fn.endswith(".py"):
+                    with open(os.path.join(tmpdir, fn)) as f:
+                        code = f.read()
+                    assert "df_rtr_split_rest_type_per" in code, "Should create df for Rest_Type_PER"
+                    assert "df_rtr_split_rest_value_per" in code, "Should create df for Rest_Value_PER"
+                    assert "df_rtr_split_rest_type_org" in code, "Should create df for Rest_Type_ORG"
+                    assert ".copy()" in code, "Should copy DataFrames"
+                    break
+        finally:
+            shutil.rmtree(tmpdir)
+
+    def test_router_group_connector_resolution(self):
+        converter = InformaticaConverter()
+        tmpdir = tempfile.mkdtemp()
+        try:
+            converter.convert_string(self.ROUTER_GROUP_XML, output_dir=tmpdir)
+            for fn in os.listdir(tmpdir):
+                if fn.startswith("mapping_") and fn.endswith(".py"):
+                    with open(os.path.join(tmpdir, fn)) as f:
+                        code = f.read()
+                    assert "df_rtr_split_rest_type_per" in code
+                    assert "df_rtr_split_rest_value_per" in code
+                    break
+        finally:
+            shutil.rmtree(tmpdir)
+
+    def test_router_default_excludes_all_groups(self):
+        converter = InformaticaConverter()
+        tmpdir = tempfile.mkdtemp()
+        try:
+            converter.convert_string(self.ROUTER_GROUP_XML, output_dir=tmpdir)
+            for fn in os.listdir(tmpdir):
+                if fn.startswith("mapping_") and fn.endswith(".py"):
+                    with open(os.path.join(tmpdir, fn)) as f:
+                        code = f.read()
+                    assert "~_router_mask_rtr_split_0" in code, "Default should negate first group mask"
+                    assert "~_router_mask_rtr_split_1" in code, "Default should negate second group mask"
+                    assert "~_router_mask_rtr_split_2" in code, "Default should negate third group mask"
+                    break
+        finally:
+            shutil.rmtree(tmpdir)
+
+
+class TestRouterGroupParsing(unittest.TestCase):
+
+    def test_group_expression_parsed(self):
+        xml = '''<?xml version="1.0" encoding="UTF-8"?>
+<!DOCTYPE POWERMART SYSTEM "powrmart.dtd">
+<POWERMART CREATION_DATE="01/01/2025" REPOSITORY_VERSION="1">
+<REPOSITORY NAME="repo" VERSION="1" CODEPAGE="UTF-8" DATABASETYPE="Oracle">
+<FOLDER NAME="TEST" OWNER="admin">
+  <MAPPING NAME="m_test" ISVALID="YES">
+    <TRANSFORMATION NAME="RTR" TYPE="Router" REUSABLE="NO">
+      <TRANSFORMFIELD NAME="X" DATATYPE="string" PRECISION="10" SCALE="0" PORTTYPE="INPUT/OUTPUT"/>
+      <GROUP NAME="INPUT" ORDER="1" TYPE="INPUT"/>
+      <GROUP NAME="GRP_A" ORDER="2" TYPE="OUTPUT" EXPRESSION="X = &apos;A&apos;"/>
+      <GROUP NAME="DEFAULT1" ORDER="3" TYPE="OUTPUT/DEFAULT"/>
+    </TRANSFORMATION>
+  </MAPPING>
+</FOLDER>
+</REPOSITORY>
+</POWERMART>'''
+        from informatica_python.parser import InformaticaParser
+        parser = InformaticaParser()
+        pm = parser.parse_string(xml)
+        mapping = pm.repositories[0].folders[0].mappings[0]
+        rtr = mapping.transformations[0]
+        assert len(rtr.groups) == 3
+        grp_a = [g for g in rtr.groups if g.name == "GRP_A"][0]
+        assert "X = " in grp_a.expression
+        assert grp_a.type == "OUTPUT"
+        default_g = [g for g in rtr.groups if g.name == "DEFAULT1"][0]
+        assert "DEFAULT" in default_g.type.upper()
+
+    def test_connector_group_attributes_parsed(self):
+        xml = '''<?xml version="1.0" encoding="UTF-8"?>
+<!DOCTYPE POWERMART SYSTEM "powrmart.dtd">
+<POWERMART CREATION_DATE="01/01/2025" REPOSITORY_VERSION="1">
+<REPOSITORY NAME="repo" VERSION="1" CODEPAGE="UTF-8" DATABASETYPE="Oracle">
+<FOLDER NAME="TEST" OWNER="admin">
+  <MAPPING NAME="m_test" ISVALID="YES">
+    <TRANSFORMATION NAME="RTR" TYPE="Router" REUSABLE="NO">
+      <TRANSFORMFIELD NAME="X" DATATYPE="string" PRECISION="10" SCALE="0" PORTTYPE="INPUT/OUTPUT"/>
+    </TRANSFORMATION>
+    <CONNECTOR FROMINSTANCE="RTR" FROMINSTANCEGROUP="GRP_A" FROMFIELD="X" TOINSTANCE="TGT" TOFIELD="X"/>
+  </MAPPING>
+</FOLDER>
+</REPOSITORY>
+</POWERMART>'''
+        from informatica_python.parser import InformaticaParser
+        parser = InformaticaParser()
+        pm = parser.parse_string(xml)
+        mapping = pm.repositories[0].folders[0].mappings[0]
+        conn = mapping.connectors[0]
+        assert conn.from_instance_group == "GRP_A"
+
+
+class TestNotInsideIIF(unittest.TestCase):
+
+    def test_not_isnull_in_iif(self):
+        expr = "IIF(NOT(ISNULL(STATUS)), 'HAS_STATUS', 'NO_STATUS')"
+        result = convert_expression_vectorized(expr)
+        assert "isna" in result or "isnull" in result
+        assert "~" in result
+        assert "np.where" in result
+
+    def test_not_isnull_standalone(self):
+        from informatica_python.utils.expression_converter import convert_filter_vectorized
+        result = convert_filter_vectorized("NOT(ISNULL(FIELD1))", "df")
+        assert "~" in result
+        assert "isna" in result
+
+    def test_not_equals_in_iif(self):
+        expr = "IIF(NOT(STATUS = 'ACTIVE'), 'INACTIVE', 'ACTIVE')"
+        result = convert_expression_vectorized(expr)
+        assert "np.where" in result
+        assert "~" in result
+
+
+class TestLookupSqlOverride(unittest.TestCase):
+
+    LOOKUP_SQL_XML = '''<?xml version="1.0" encoding="UTF-8"?>
+<!DOCTYPE POWERMART SYSTEM "powrmart.dtd">
+<POWERMART CREATION_DATE="01/01/2025" REPOSITORY_VERSION="1">
+<REPOSITORY NAME="repo" VERSION="1" CODEPAGE="UTF-8" DATABASETYPE="Oracle">
+<FOLDER NAME="TEST" OWNER="admin">
+  <SOURCE NAME="SRC" DATABASETYPE="Flat File" DBDNAME="SRC">
+    <FLATFILE DELIMITEDBY="COMMA" HEADERROWPRESENT="YES" PADBYTES="NO" ROWDELIMITER="\\n"/>
+    <SOURCEFIELD NAME="ID" DATATYPE="integer" PRECISION="10" SCALE="0" NULLABLE="NOTNULL" KEYTYPE="PRIMARY KEY" FIELDNUMBER="1"/>
+    <SOURCEFIELD NAME="CODE" DATATYPE="string" PRECISION="10" SCALE="0" NULLABLE="NULL" KEYTYPE="NOT A KEY" FIELDNUMBER="2"/>
+  </SOURCE>
+  <TARGET NAME="TGT" DATABASETYPE="Flat File">
+    <TARGETFIELD NAME="ID" DATATYPE="integer" PRECISION="10" SCALE="0" NULLABLE="NULL" KEYTYPE="NOT A KEY" FIELDNUMBER="1"/>
+    <TARGETFIELD NAME="DESC" DATATYPE="string" PRECISION="50" SCALE="0" NULLABLE="NULL" KEYTYPE="NOT A KEY" FIELDNUMBER="2"/>
+  </TARGET>
+  <MAPPING NAME="m_lookup_sql_test" ISVALID="YES">
+    <TRANSFORMATION NAME="SQ_SRC" TYPE="Source Qualifier" REUSABLE="NO">
+      <TRANSFORMFIELD NAME="ID" DATATYPE="integer" PRECISION="10" SCALE="0" PORTTYPE="OUTPUT"/>
+      <TRANSFORMFIELD NAME="CODE" DATATYPE="string" PRECISION="10" SCALE="0" PORTTYPE="OUTPUT"/>
+    </TRANSFORMATION>
+    <TRANSFORMATION NAME="LKP_CODES" TYPE="Lookup Procedure" REUSABLE="NO">
+      <TRANSFORMFIELD NAME="CODE_IN" DATATYPE="string" PRECISION="10" SCALE="0" PORTTYPE="INPUT"/>
+      <TRANSFORMFIELD NAME="DESC_OUT" DATATYPE="string" PRECISION="50" SCALE="0" PORTTYPE="OUTPUT"/>
+      <TABLEATTRIBUTE NAME="Lookup Sql Override" VALUE="SELECT CODE, DESCRIPTION FROM REF_CODES WHERE ACTIVE_FLAG = &apos;Y&apos;"/>
+      <TABLEATTRIBUTE NAME="Lookup table name" VALUE="REF_CODES"/>
+    </TRANSFORMATION>
+    <INSTANCE NAME="SRC" TYPE="Source Definition" TRANSFORMATION_NAME="SRC"/>
+    <INSTANCE NAME="SQ_SRC" TYPE="Source Qualifier" TRANSFORMATION_NAME="SQ_SRC"/>
+    <INSTANCE NAME="LKP_CODES" TYPE="Lookup Procedure" TRANSFORMATION_NAME="LKP_CODES"/>
+    <INSTANCE NAME="TGT" TYPE="Target Definition" TRANSFORMATION_NAME="TGT"/>
+    <CONNECTOR FROMINSTANCE="SRC" FROMFIELD="ID" TOINSTANCE="SQ_SRC" TOFIELD="ID"/>
+    <CONNECTOR FROMINSTANCE="SRC" FROMFIELD="CODE" TOINSTANCE="SQ_SRC" TOFIELD="CODE"/>
+    <CONNECTOR FROMINSTANCE="SQ_SRC" FROMFIELD="ID" TOINSTANCE="LKP_CODES" TOFIELD="CODE_IN"/>
+    <CONNECTOR FROMINSTANCE="LKP_CODES" FROMFIELD="DESC_OUT" TOINSTANCE="TGT" TOFIELD="DESC"/>
+    <CONNECTOR FROMINSTANCE="SQ_SRC" FROMFIELD="ID" TOINSTANCE="TGT" TOFIELD="ID"/>
+  </MAPPING>
+  <CONFIG NAME="default_session_config"/>
+  <WORKFLOW NAME="wf_lookup_sql_test" ISVALID="YES">
+    <TASK NAME="Start" REUSABLE="NO" TYPE="Start"/>
+    <SESSION NAME="s_m_lookup_sql_test" ISVALID="YES" REUSABLE="NO" MAPPINGNAME="m_lookup_sql_test">
+      <CONFIGREFERENCE REFOBJECTNAME="default_session_config" TYPE="Session config"/>
+    </SESSION>
+    <TASKINSTANCE NAME="Start" TASKNAME="Start" TASKTYPE="Start"/>
+    <TASKINSTANCE NAME="s_m_lookup_sql_test" TASKNAME="s_m_lookup_sql_test" TASKTYPE="Session"/>
+    <WORKFLOWLINK FROMTASK="Start" TOTASK="s_m_lookup_sql_test"/>
+  </WORKFLOW>
+</FOLDER>
+</REPOSITORY>
+</POWERMART>'''
+
+    def test_lookup_sql_override_applied(self):
+        converter = InformaticaConverter()
+        tmpdir = tempfile.mkdtemp()
+        try:
+            converter.convert_string(self.LOOKUP_SQL_XML, output_dir=tmpdir)
+            for fn in os.listdir(tmpdir):
+                if fn.startswith("mapping_") and fn.endswith(".py"):
+                    with open(os.path.join(tmpdir, fn)) as f:
+                        code = f.read()
+                    assert "SELECT CODE, DESCRIPTION FROM REF_CODES" in code, \
+                        "Lookup SQL Override should appear in generated code"
+                    assert "read_from_db" in code, \
+                        "Should use read_from_db with the override SQL"
+                    break
+        finally:
+            shutil.rmtree(tmpdir)
+
+    def test_lookup_sql_in_all_sql_queries(self):
+        converter = InformaticaConverter()
+        tmpdir = tempfile.mkdtemp()
+        try:
+            converter.convert_string(self.LOOKUP_SQL_XML, output_dir=tmpdir)
+            sql_file = os.path.join(tmpdir, "all_sql_queries.sql")
+            if os.path.exists(sql_file):
+                with open(sql_file) as f:
+                    sql = f.read()
+                assert "REF_CODES" in sql, "SQL file should contain lookup SQL"
+        finally:
+            shutil.rmtree(tmpdir)