Fix/issue 216 bot utterance (Enable Parser to Handle Multiline Strings and Newlines)

nemoguardrails/actions/llm/generation.py

@@ -14,6 +14,7 @@
 # limitations under the License.
 
 """A set of actions for generating various types of completions using an LLMs."""
+
 import asyncio
 import logging
 import random
@@ -910,6 +911,7 @@ async def generate_bot_message(
             log.info(f"Generated bot message: {bot_utterance}")
 
         if bot_utterance:
+            bot_utterance = clean_utterance_content(bot_utterance)
             # In streaming mode, we also push this.
             if streaming_handler:
                 await streaming_handler.push_chunk(bot_utterance)
@@ -1279,3 +1281,21 @@ async def generate_intent_steps_message(
             return ActionResult(
                 events=[new_event_dict("BotMessage", text=text)],
             )
+
+
+def clean_utterance_content(utterance: str) -> str:
+    """
+    Clean an utterance by performing the following operations:
+     - replacing "\\n" with "\n".
+
+    Args:
+        utterance (str): The utterance to clean.
+
+    Returns:
+        str: The cleaned utterance.
+    """
+    if utterance:
+        # If "\n" is used inside a predefined message, it will be returned as is as part of the message.
+        # It should be translated to an actual \n character.
+        utterance = utterance.replace("\\n", "\n")
+    return utterance

nemoguardrails/colang/v1_0/lang/utils.py

@@ -86,9 +86,38 @@ def get_numbered_lines(content: str):
     i = 0
     multiline_comment = False
     current_comment = None
+    multiline_string = False
+    current_string = None
     while i < len(raw_lines):
         raw_line = raw_lines[i].strip()
 
+        # handle multiline string
+        if multiline_string:
+            current_string += "\n" + raw_line
+            if raw_line.endswith('"'):
+                multiline_string = False
+                lines.append(
+                    {
+                        "text": current_string,
+                        "number": i + 1,
+                        "indentation": multiline_indentation,
+                        "comment": current_comment,
+                    }
+                )
+                current_string = None
+            i += 1
+            continue
+        if (
+            raw_line.startswith('"')
+            and not raw_line.startswith('"""')
+            and not raw_line.endswith('"')
+        ):
+            multiline_string = True
+            current_string = raw_line
+            multiline_indentation = len(raw_lines[i]) - len(raw_line.lstrip())
+            i += 1
+            continue
+
         # If we have a line comment, we record it
         if raw_line.startswith("#"):
             if current_comment is None:

tests/test_issue_216.py

@@ -0,0 +1,79 @@
+# SPDX-FileCopyrightText: Copyright (c) 2023 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
+# SPDX-License-Identifier: Apache-2.0
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+# http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+import pytest
+
+from nemoguardrails import LLMRails, RailsConfig
+
+
+@pytest.mark.asyncio
+async def test_new_line_in_bot_message():
+    config = RailsConfig.from_content(
+        colang_content="""
+        define user express greeting
+            "hello"
+
+        define flow
+            user express greeting
+            bot express greeting
+
+        define bot express greeting
+            "Hello World!\\n NewLine World!"
+        """,
+        config={
+            "models": [],
+            "rails": {"dialog": {"user_messages": {"embeddings_only": True}}},
+        },
+    )
+    rails = LLMRails(config)
+
+    # This should not raise an exception
+    res = await rails.generate_async(
+        messages=[
+            {"role": "user", "content": "hi!"},
+            {"role": "assistant", "content": "hi!"},
+            {"role": "user", "content": "hi!"},
+        ]
+    )
+
+    assert res["content"] == "Hello World!\n NewLine World!"
+
+
+@pytest.mark.asyncio
+async def test_new_line_in_user_message():
+    config = RailsConfig.from_content(
+        colang_content="""
+        define user express greeting
+            "hello\nworld"
+        define flow
+            user express greeting
+            bot express greeting
+        define bot express greeting
+            "Hello World!\\n NewLine World!"
+        """,
+        config={
+            "models": [],
+            "rails": {"dialog": {"user_messages": {"embeddings_only": True}}},
+        },
+    )
+    rails = LLMRails(config)
+    # This should not raise an exception
+    await rails.generate_async(
+        messages=[
+            {"role": "user", "content": "hi!"},
+            {"role": "assistant", "content": "hi!"},
+            {"role": "user", "content": "hi!"},
+        ]
+    )