[BUGFIX] Fix regex pattern for Mistral Tool Call (#29918)

Signed-off-by: juliendenize <[email protected]>
(cherry picked from commit 1b1e35a)

Author: juliendenize

tests/models/language/generation/test_mistral.py

@@ -315,3 +315,38 @@ def get_vocab():
     assert json.loads(parsed.tool_calls[0].function.arguments) == args_dict
     # No additional content outside the tool call should be returned.
     assert parsed.content is None
+
+    # multiple calls
+    multiple_args_dict = [
+        {
+            "city": "Dallas",
+            "state": "TX",
+            "unit": "fahrenheit",
+            "sub_dict": {"foo": "bar", "inner": {"x": 1, "y": 2}},
+        },
+        {},
+        {"a": 0},
+        {"a": 1, "b": "c"},
+    ]
+    names = ["get_current_weather", "get_current_weather_2", "random", "random_2"]
+
+    model_output = "".join(
+        [
+            f"{parser.bot_token}{name}{json.dumps(args)}"
+            for name, args in zip(names, multiple_args_dict)
+        ]
+    )
+
+    parsed = parser.extract_tool_calls(model_output, None)
+
+    # Assertions: the tool call is detected and the full nested JSON is parsed
+    # without truncation.
+    assert parsed.tools_called
+    assert len(parsed.tool_calls) == len(multiple_args_dict)
+
+    for i, tool_call in enumerate(parsed.tool_calls):
+        assert MistralToolCall.is_valid_id(tool_call.id)
+        assert tool_call.function.name == names[i]
+        assert json.loads(tool_call.function.arguments) == multiple_args_dict[i]
+        # No additional content outside the tool call should be returned.
+        assert parsed.content is None

vllm/entrypoints/openai/tool_parsers/mistral_tool_parser.py

@@ -80,7 +80,7 @@ def __init__(self, tokenizer: TokenizerLike):
         self.tool_call_regex = re.compile(r"\[{.*}\]", re.DOTALL)
         if _is_fn_name_regex_support(self.model_tokenizer):
             self.fn_name_regex = re.compile(
-                r"([a-zA-Z0-9_-]+)(\{[\s\S]*?\})(?=\s*$|,|\s)?", re.DOTALL
+                r"([a-zA-Z0-9_-]+)(\{[\s\S]*?\}+)", re.DOTALL
             )
         else:
             self.fn_name_regex = None