cardosofelipe/syndarix
1
0
Fork 0
forked from cardosofelipe/fast-next-template
Code Issues 18 Pull Requests Wiki Activity

chore(context): refactor for consistency, optimize formatting, and simplify logic

Browse Source 
- Cleaned up unnecessary comments in `__all__` definitions for better readability.
- Adjusted indentation and formatting across modules for improved clarity (e.g., long lines, logical grouping).
- Simplified conditional expressions and inline comments for context scoring and ranking.
- Replaced some hard-coded values with type-safe annotations (e.g., `ClassVar`).
- Removed unused imports and ensured consistent usage across test files.
- Updated `test_score_not_cached_on_context` to clarify caching behavior.
- Improved truncation strategy logic and marker handling.
This commit is contained in:
Felipe Cardoso
2026-01-04 15:23:14 +01:00
parent 9e54f16e56
commit 2bea057fb1
26 changed files with 226 additions and 273 deletions
Download Patch File Download Diff File Expand all files Collapse all files

22
backend/app/services/context/scoring/composite.py
View File

@@ -92,7 +92,9 @@ class CompositeScorer:
# Per-context locks to prevent race conditions during parallel scoring
# Uses WeakValueDictionary so locks are garbage collected when not in use
self._context_locks: WeakValueDictionary[str, asyncio.Lock] = WeakValueDictionary()
self._context_locks: WeakValueDictionary[str, asyncio.Lock] = (
WeakValueDictionary()
)
self._locks_lock = asyncio.Lock() # Lock to protect _context_locks access
def set_mcp_manager(self, mcp_manager: "MCPClientManager") -> None:
@@ -207,17 +209,14 @@ class CompositeScorer:
ScoredContext with all scores
"""
# Get lock for this specific context to prevent race conditions
# within concurrent scoring operations for the same query
context_lock = await self._get_context_lock(context.id)
async with context_lock:
# Check if context already has a score (inside lock to prevent races)
if context._score is not None:
return ScoredContext(
context=context,
composite_score=context._score,
)
# Compute individual scores in parallel
# Note: We do NOT cache scores on the context because scores are
# query-dependent. Caching without considering the query would
# return incorrect scores for different queries.
relevance_task = self._relevance_scorer.score(context, query, **kwargs)
recency_task = self._recency_scorer.score(context, query, **kwargs)
priority_task = self._priority_scorer.score(context, query, **kwargs)
@@ -240,9 +239,6 @@ class CompositeScorer:
else:
composite = 0.0
# Cache the score on the context (now safe - inside lock)
context._score = composite
return ScoredContext(
context=context,
composite_score=composite,
@@ -271,9 +267,7 @@ class CompositeScorer:
List of ScoredContext (same order as input)
"""
if parallel:
tasks = [
self.score_with_details(ctx, query, **kwargs) for ctx in contexts
]
tasks = [self.score_with_details(ctx, query, **kwargs) for ctx in contexts]
return await asyncio.gather(*tasks)
else:
results = []

12
backend/app/services/context/scoring/priority.py
View File

@@ -4,7 +4,7 @@ Priority Scorer for Context Management.
Scores context based on assigned priority levels.
"""
from typing import Any
from typing import Any, ClassVar
from ..types import BaseContext, ContextType
from .base import BaseScorer
@@ -19,11 +19,11 @@ class PriorityScorer(BaseScorer):
"""
# Default priority bonuses by context type
DEFAULT_TYPE_BONUSES: dict[ContextType, float] = {
ContextType.SYSTEM: 0.2, # System prompts get a boost
ContextType.TASK: 0.15, # Current task is important
ContextType.TOOL: 0.1, # Recent tool results matter
ContextType.KNOWLEDGE: 0.0, # Knowledge scored by relevance
DEFAULT_TYPE_BONUSES: ClassVar[dict[ContextType, float]] = {
ContextType.SYSTEM: 0.2, # System prompts get a boost
ContextType.TASK: 0.15, # Current task is important
ContextType.TOOL: 0.1, # Recent tool results matter
ContextType.KNOWLEDGE: 0.0, # Knowledge scored by relevance
ContextType.CONVERSATION: 0.0, # Conversation scored by recency
}

21
backend/app/services/context/scoring/relevance.py
View File

@@ -85,7 +85,10 @@ class RelevanceScorer(BaseScorer):
Relevance score between 0.0 and 1.0
"""
# 1. Check for pre-computed relevance score
if isinstance(context, KnowledgeContext) and context.relevance_score is not None:
if (
isinstance(context, KnowledgeContext)
and context.relevance_score is not None
):
return self.normalize_score(context.relevance_score)
# 2. Check metadata for score
@@ -95,14 +98,19 @@ class RelevanceScorer(BaseScorer):
if "score" in context.metadata:
return self.normalize_score(context.metadata["score"])
# 3. Try MCP-based semantic similarity
# 3. Try MCP-based semantic similarity (if compute_similarity tool is available)
# Note: This requires the knowledge-base MCP server to implement compute_similarity
if self._mcp is not None:
try:
score = await self._compute_semantic_similarity(context, query)
if score is not None:
return score
except Exception as e:
logger.debug(f"Semantic scoring failed, using fallback: {e}")
# Log at debug level since this is expected if compute_similarity
# tool is not implemented in the Knowledge Base server
logger.debug(
f"Semantic scoring unavailable, using keyword fallback: {e}"
)
# 4. Fall back to keyword matching
return self._compute_keyword_score(context, query)
@@ -122,6 +130,9 @@ class RelevanceScorer(BaseScorer):
Returns:
Similarity score or None if unavailable
"""
if self._mcp is None:
return None
try:
# Use Knowledge Base's search capability to compute similarity
result = await self._mcp.call_tool(
@@ -129,7 +140,9 @@ class RelevanceScorer(BaseScorer):
tool="compute_similarity",
args={
"text1": query,
"text2": context.content[: self._semantic_max_chars], # Limit content length
"text2": context.content[
: self._semantic_max_chars
], # Limit content length
},
)