perf(search): simplify LIKE search (drop LOWER(...) in favour of COLLATE NOCASE)
#695
Conversation
… index) maybe we need to add `COLLATE NOCASE` to the where clause, idk... see: Hochfrequenz/xml-fundamend-python#193
There was a problem hiding this comment.
Pull Request Overview
This PR optimizes database query performance by removing LOWER(...) function calls from SQL LIKE search operations, allowing the database to utilize indexes more effectively. The search values are still converted to lowercase on the application side before being passed to the query.
Key changes:
- Removed
LOWER()SQL function wrapping from column references inLIKEclauses - Maintained
.toLowerCase()on search parameter values to preserve case-insensitive matching
Tip: Customize your code reviews with copilot-instructions.md. Create the file or learn how to get started.
hf-kklein
changed the title
LIKE search (drop LOWER(...) and hence use index)LIKE search (drop LOWER(...) in favour of COLLATE NOCASE)
There was a problem hiding this comment.
Pull Request Overview
Copilot reviewed 1 out of 1 changed files in this pull request and generated 1 comment.
Tip: Customize your code reviews with copilot-instructions.md. Create the file or learn how to get started.
| queryBuilder.andWhere(`al.${columnName} COLLATE NOCASE LIKE :${paramBase}_contains`, { | ||
| [`${paramBase}_contains`]: `%${value.contains.toLowerCase()}%`, |
There was a problem hiding this comment.
The columnName variable is interpolated directly into the SQL string without sanitization, creating a potential SQL injection vulnerability. Consider using a whitelist of allowed column names or TypeORM's column metadata to validate the column name before interpolation.
There was a problem hiding this comment.
was like this before. out of scope for this PR
maybe we need to add
COLLATE NOCASEto the where clause, idk...see: Hochfrequenz/xml-fundamend-python#193
@hf-krechan : man müsste mal aus dem fundamend-pr eine DB bauen und schauen ob das so funktioniert. ich glaube aber, dass aktuell der index auf der spalte nicht genutzt wird und ohne collate nocase nur für ascii-zeichen genutzt würde (siehe auch die description vom fundanemnd-pr)