lf.Template to support rendered text post processing.

langfun/core/template.py

@@ -280,6 +280,10 @@ def _preprocess_template(self, template_str: str) -> str:
     """
     return template_str
 
+  def _postprocess_rendered(self, rendered_text: str) -> str:
+    """Postprocesses the rendered text."""
+    return rendered_text
+
   def vars(
       self,
       specified: bool | None = None,
@@ -441,7 +445,9 @@ def render(
                 # natural language when they are directly returned as rendering
                 # elements in the template.
                 with modality.format_modality_as_ref():
-                  rendered_text = self._template.render(**inputs)
+                  rendered_text = self._postprocess_rendered(
+                      self._template.render(**inputs)
+                  )
 
               # Carry the modality references passed from the constructor.
               # This is to support modality objects that is already rendered

langfun/core/template_test.py

@@ -191,6 +191,22 @@ def _preprocess_template(self, template_str: str) -> str:
         'Google is good'
     )
 
+  def test_postprocess_rendered(self):
+
+    class MyTemplate(Template):
+      """My template with postprocess.
+
+      $COMPANY {{x}}
+      """
+
+      def _postprocess_rendered(self, rendered_text: str) -> str:
+        return rendered_text.replace('$COMPANY', 'Google')
+
+    self.assertEqual(
+        MyTemplate(x='is $COMPANY').render(),
+        'Google is Google'
+    )
+
 
 class FromValueTest(unittest.TestCase):