SLocator: Localizing the Origin of SQL Queries in Database-Backed Web Applications
Why this work is in the frame
A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.
Bibliographic record
Abstract
In database-backed web applications, developers often leverage Object-Relational Mapping (ORM) frameworks for database accesses. ORM frameworks provide an abstraction of the underlying database access details so that developers can focus on implementing the business logic of the application. However, due to the abstraction, developers may not know where and how a problematic SQL query is generated in the application code, causing challenges in debugging database access problems. In this paper, we propose an approach, called SLocator, which locates where a SQL query is generated in the application code. SLocator is a hybrid approach that leverages both static analysis and information retrieval (IR) techniques. SLocator uses static analysis to infer the database access for every possible path in the control flow graph. Then, given a SQL query, SLocator applies IR techniques to find the control flow path (i.e., a sequence of methods called in an interprocedural control flow graph) whose inferred database access has the highest similarity ranking. We implement SLocator for Java’s official ORM API specification (JPA) and evaluate SLocator on seven open source Java applications. We find that SLocator is able to locate the control flow path that generates a SQL query with a Top@1 accuracy ranging from 37.4% to 70% for SQL queries in sessions, and 30.7% to 69.2% for individual SQL queries; and Top@5 ranging from 78.3% to 95.5% for SQL queries in sessions, and 59.1% to 100% for individual SQL queries. We also conduct a study to illustrate how SLocator may be used for locating issues in the database access code.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.002 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.001 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it