test: Add thinking trace parsing tests

This commit is contained in:
2026-03-13 21:53:17 -04:00
parent c5a406eff8
commit ea7b3ae3ae
3 changed files with 166 additions and 3 deletions

53
src/thinking_parser.py Normal file
View File

@@ -0,0 +1,53 @@
import re
from typing import List, Tuple
from src.models import ThinkingSegment
def parse_thinking_trace(text: str) -> Tuple[List[ThinkingSegment], str]:
"""
Parses thinking segments from text and returns (segments, response_content).
Support extraction of thinking traces from <thinking>...</thinking>, <thought>...</thought>,
and blocks prefixed with Thinking:.
"""
segments = []
# 1. Extract <thinking> and <thought> tags
current_text = text
# Combined pattern for tags
tag_pattern = re.compile(r'<(thinking|thought)>(.*?)</\1>', re.DOTALL | re.IGNORECASE)
def extract_tags(txt: str) -> Tuple[List[ThinkingSegment], str]:
found_segments = []
def replace_func(match):
marker = match.group(1).lower()
content = match.group(2).strip()
found_segments.append(ThinkingSegment(content=content, marker=marker))
return ""
remaining = tag_pattern.sub(replace_func, txt)
return found_segments, remaining
tag_segments, remaining = extract_tags(current_text)
segments.extend(tag_segments)
# 2. Extract Thinking: prefix
# This usually appears at the start of a block and ends with a double newline or a response marker.
thinking_colon_pattern = re.compile(r'(?:^|\n)Thinking:\s*(.*?)(?:\n\n|\nResponse:|\nAnswer:|$)', re.DOTALL | re.IGNORECASE)
def extract_colon_blocks(txt: str) -> Tuple[List[ThinkingSegment], str]:
found_segments = []
def replace_func(match):
content = match.group(1).strip()
if content:
found_segments.append(ThinkingSegment(content=content, marker="Thinking:"))
return "\n\n"
res = thinking_colon_pattern.sub(replace_func, txt)
return found_segments, res
colon_segments, final_remaining = extract_colon_blocks(remaining)
segments.extend(colon_segments)
return segments, final_remaining.strip()