discours.io/core
5
0
Fork 0
Code Pull Requests Actions Projects Activity

fix(search.py): store all results in cash not only first offset
All checks were successful
Deploy on push / deploy (push) Successful in 52s
Details

Browse Source
This commit is contained in:
Stepan Vladovskiy 2025-04-03 13:10:53 -03:00
parent a0db5707c4
commit bc4ec79240
1 changed files with 18 additions and 16 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

34
services/search.py
View File

@ -23,6 +23,8 @@ SEARCH_MIN_SCORE = float(os.environ.get("SEARCH_MIN_SCORE", "0.1"))
SEARCH_PREFETCH_SIZE = int(os.environ.get("SEARCH_PREFETCH_SIZE", "200")) SEARCH_PREFETCH_SIZE = int(os.environ.get("SEARCH_PREFETCH_SIZE", "200"))
SEARCH_USE_REDIS = bool(os.environ.get("SEARCH_USE_REDIS", "true").lower() in ["true", "1", "yes"]) SEARCH_USE_REDIS = bool(os.environ.get("SEARCH_USE_REDIS", "true").lower() in ["true", "1", "yes"])
search_offset = 0
# Import Redis client if Redis caching is enabled # Import Redis client if Redis caching is enabled
if SEARCH_USE_REDIS: if SEARCH_USE_REDIS:
try: try:
@ -573,10 +575,13 @@ class SearchService:
search_limit = limit search_limit = limit
search_offset = offset search_offset = offset
# If cache is enabled, prefetch more results to store in cache # Always prefetch full results when caching is enabled
if SEARCH_CACHE_ENABLED and offset == 0: if SEARCH_CACHE_ENABLED:
search_limit = SEARCH_PREFETCH_SIZE # Fetch more results to cache search_limit = SEARCH_PREFETCH_SIZE # Always fetch a large set
search_offset = 0 # Always start from beginning for cache search_offset = 0 # Always start from beginning
else:
search_limit = limit
search_offset = offset
logger.info(f"Sending search request: text='{text}', limit={search_limit}, offset={search_offset}") logger.info(f"Sending search request: text='{text}', limit={search_limit}, offset={search_offset}")
response = await self.client.post( response = await self.client.post(
@ -585,7 +590,7 @@ class SearchService:
) )
response.raise_for_status() response.raise_for_status()
logger.info(f"Raw search response: {response.text}") # logger.info(f"Raw search response: {response.text}")
result = response.json() result = response.json()
logger.info(f"Parsed search response: {result}") logger.info(f"Parsed search response: {result}")
@ -617,18 +622,15 @@ class SearchService:
logger.info(f"Sample result: {formatted_results[0]}") logger.info(f"Sample result: {formatted_results[0]}")
else: else:
logger.warning(f"No results found for '{text}'") logger.warning(f"No results found for '{text}'")
# Store full results in cache if caching is enabled if SEARCH_CACHE_ENABLED:
if SEARCH_CACHE_ENABLED and offset == 0: logger.info(f"Storing {len(formatted_results)} results in cache for query '{text}'")
# Store normal sorted results await self.cache.store(text, formatted_results) # Return the proper page slice from the full results stored in cache end_idx = offset + limit
await self.cache.store(text, formatted_results) end_idx = offset + limit
page_results = formatted_results[offset:end_idx]
# Return only the requested page logger.info(f"Returning results from {offset} to {end_idx} (of {len(formatted_results)} total)")
if limit < len(formatted_results): return page_results
page_results = formatted_results[:limit]
logger.info(f"Returning first page of {len(page_results)} results " +
f"(out of {len(formatted_results)} total)")
return page_results
return formatted_results return formatted_results
except Exception as e: except Exception as e: