From 830f891a7475a53d4a55bd84613697994c6b890d Mon Sep 17 00:00:00 2001
From: Amy <1530140574@qq.com>
Date: Sat, 18 Oct 2025 11:58:40 +0800
Subject: [PATCH] Fix json in md when use quesion classifier node (#26992)

Co-authored-by: autofix-ci[bot] <114827586+autofix-ci[bot]@users.noreply.github.com>
Co-authored-by: crazywoola <100913391+crazywoola@users.noreply.github.com>
---
 .../question_classifier_node.py               |  3 +++
 api/libs/json_in_md_parser.py                 | 14 +++++++++----
 .../unit_tests/libs/test_json_in_md_parser.py | 21 +++++++++++++++++++
 3 files changed, 34 insertions(+), 4 deletions(-)

diff --git a/api/core/workflow/nodes/question_classifier/question_classifier_node.py b/api/core/workflow/nodes/question_classifier/question_classifier_node.py
index 592a6566fd..31b1cd4966 100644
--- a/api/core/workflow/nodes/question_classifier/question_classifier_node.py
+++ b/api/core/workflow/nodes/question_classifier/question_classifier_node.py
@@ -1,4 +1,5 @@
 import json
+import re
 from collections.abc import Mapping, Sequence
 from typing import TYPE_CHECKING, Any
 
@@ -194,6 +195,8 @@ class QuestionClassifierNode(Node):
 
             category_name = node_data.classes[0].name
             category_id = node_data.classes[0].id
+            if "<think>" in result_text:
+                result_text = re.sub(r"<think[^>]*>[\s\S]*?</think>", "", result_text, flags=re.IGNORECASE)
             result_text_json = parse_and_check_json_markdown(result_text, [])
             # result_text_json = json.loads(result_text.strip('```JSON\n'))
             if "category_name" in result_text_json and "category_id" in result_text_json:
diff --git a/api/libs/json_in_md_parser.py b/api/libs/json_in_md_parser.py
index 0c642041bf..310e677747 100644
--- a/api/libs/json_in_md_parser.py
+++ b/api/libs/json_in_md_parser.py
@@ -6,22 +6,22 @@ from core.llm_generator.output_parser.errors import OutputParserError
 def parse_json_markdown(json_string: str):
     # Get json from the backticks/braces
     json_string = json_string.strip()
-    starts = ["```json", "```", "``", "`", "{"]
-    ends = ["```", "``", "`", "}"]
+    starts = ["```json", "```", "``", "`", "{", "["]
+    ends = ["```", "``", "`", "}", "]"]
     end_index = -1
     start_index = 0
     parsed: dict = {}
     for s in starts:
         start_index = json_string.find(s)
         if start_index != -1:
-            if json_string[start_index] != "{":
+            if json_string[start_index] not in ("{", "["):
                 start_index += len(s)
             break
     if start_index != -1:
         for e in ends:
             end_index = json_string.rfind(e, start_index)
             if end_index != -1:
-                if json_string[end_index] == "}":
+                if json_string[end_index] in ("}", "]"):
                     end_index += 1
                 break
     if start_index != -1 and end_index != -1 and start_index < end_index:
@@ -38,6 +38,12 @@ def parse_and_check_json_markdown(text: str, expected_keys: list[str]):
         json_obj = parse_json_markdown(text)
     except json.JSONDecodeError as e:
         raise OutputParserError(f"got invalid json object. error: {e}")
+
+    if isinstance(json_obj, list):
+        if len(json_obj) == 1 and isinstance(json_obj[0], dict):
+            json_obj = json_obj[0]
+        else:
+            raise OutputParserError(f"got invalid return object. obj:{json_obj}")
     for key in expected_keys:
         if key not in json_obj:
             raise OutputParserError(
diff --git a/api/tests/unit_tests/libs/test_json_in_md_parser.py b/api/tests/unit_tests/libs/test_json_in_md_parser.py
index 53fd0bea16..953f203e89 100644
--- a/api/tests/unit_tests/libs/test_json_in_md_parser.py
+++ b/api/tests/unit_tests/libs/test_json_in_md_parser.py
@@ -86,3 +86,24 @@ def test_parse_and_check_json_markdown_multiple_blocks_fails():
     # opening fence to the last closing fence, causing JSON decode failure.
     with pytest.raises(OutputParserError):
         parse_and_check_json_markdown(src, [])
+
+
+def test_parse_and_check_json_markdown_handles_think_fenced_and_raw_variants():
+    expected = {"keywords": ["2"], "category_id": "2", "category_name": "2"}
+    cases = [
+        """
+        ```json
+        [{"keywords": ["2"], "category_id": "2", "category_name": "2"}]
+        ```, error: Expecting value: line 1 column 1 (char 0)
+        """,
+        """
+        ```json
+        {"keywords": ["2"], "category_id": "2", "category_name": "2"}
+        ```, error: Extra data: line 2 column 5 (char 66)
+        """,
+        '{"keywords": ["2"], "category_id": "2", "category_name": "2"}',
+        '[{"keywords": ["2"], "category_id": "2", "category_name": "2"}]',
+    ]
+    for src in cases:
+        obj = parse_and_check_json_markdown(src, ["keywords", "category_id", "category_name"])
+        assert obj == expected