CI / test (postgres, odbc, Driver=PostgreSQL Unicode;Server=127.0.0.1;Port=5432;Database=sqlpage;UID=root;PWD=Password123!, true) (push) Has been cancelled
Add a discovery request counter to FakeOidcProvider and assert it
increments after the background refresh completes. Remove unused
set_discovery_delay method.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Replace std::time::Instant with tokio::time::Instant in OidcSnapshot so
that tokio::time::pause()/advance() controls elapsed time in tests.
Remove force_expire() — tests advance time past MAX_REFRESH_INTERVAL
instead. Simplify slow discovery test from ~70 to ~30 lines.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Replace tokio::sync::RwLock<ClientWithTime> with std::sync::RwLock<Arc<OidcSnapshot>>.
The std lock makes it structurally impossible to hold across await points.
Readers clone an Arc (nanoseconds) and use it freely — no lock contention.
Many previously-async functions become synchronous (get_token_claims,
build_auth_url, handle_unauthenticated_request, handle_oidc_logout, etc).
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- OIDC provider metadata refreshes now run in a background task via
spawn_local, never blocking incoming HTTP requests.
- Multiple concurrent refresh triggers are deduplicated via an AtomicBool.
- The write lock on the OIDC client is only held briefly to swap data,
not during the upstream HTTP call.
- Add a body-read timeout to OIDC HTTP requests to prevent hangs when
the provider stalls after sending headers.
- Add tests for both scenarios: slow discovery and slow token endpoint.
Fixes https://github.com/sqlpage/SQLPage/issues/1231
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* add timeout to OIDC HTTP response body read
A stalled OIDC provider that sends HTTP headers but never completes the
body would cause response.body().await to hang forever, freezing the
entire SQLPage process. Add a 5-second timeout on the body stream read
using awc's ClientResponse::timeout().
partially Fixes#1231
* simplify test comment
* simplify test: use token_endpoint_delay instead of Notify gate
* simplify test: use tokio time pause + auto-advance instead of select
* use spawn_local + advance instead of select to detect hang
* use Notify sync point instead of yield loop for deterministic test
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* simplify test: replace Notify+yields with sleep+time-advance
Remove the Notify synchronization and yield_now() calls. Instead,
use a small real-time sleep for TCP to complete, then pause+advance
tokio time. Assert the actual response status instead of is_finished().
* fix clippy: remove unnecessary mut on response
* improve sqlpage function argument warnings with source context
Made-with: Cursor
* Param extraction: Result-based errors, single caller message, no CompoundIdentifier special case
- expr_to_stmt_param returns Result<StmtParam, ExprToParamError>; error carries only line + kind (UnsupportedExpr, UnemulatedFunction, NamedArgs)
- function_args_to_stmt_params logs one formatted message (ctx.format_param_error) then returns Err
- Single unsupported-expr arm; expr_summary() used for description
- Rename ParamWarnContext to ParamExtractContext
Made-with: Cursor
* Surface param extraction error in parse result; add parse_sql error-message tests; remove are_params_extractable
- When func_call_to_param returns StmtParam::Error, store it and have extract_parameters return Err so parse yields ParsedStatement::Error with specialized message
- Add test_parse_sql_unsupported_expr_in_sqlpage_arg and test_parse_sql_unemulated_function_in_sqlpage_arg
- Remove dead are_params_extractable and its unused import
Made-with: Cursor
* Refactor sqlpage function argument error messages to match user expectations
- Overhauled ExprToParamError formatting to construct exact user-friendly descriptions.
- Removed superfluous anyhow::Context prefixes in func_call_to_param.
- Passed source_path properly through validate_function_calls to ensure file line numbers populate the new error template accurately.
- Renamed error test file to match its dynamic error output signature.
- Removed redundant mut mutability warnings on parsing logic loops.
* improve error messages
* Refactor SqlPageFunctionError representation to clean up 'syntax error' wrappers
- Replaced stringly-typed anyhow errors with a strongly typed SqlPageFunctionError.
- Removed source_path threading completely from the parameter extraction phases, conforming to better separation of concerns.
- Appended file path prefix dynamically at the evaluation stage in clone_anyhow_err strictly when downcasting to SqlPageFunctionError.
- Removed the confusing generic 'Caused by: x.sql contains a syntax error...' wrapper from actual function logic errors.
* Remove redundant 'reorganize' hint from error message
* readd deleted test
* split parameter extraction logic into a separate file
* Add to changelog
* Ignore invalid SQLPAGE_PORT values from Kubernetes service environment variables
Co-authored-by: contact <contact@ophir.dev>
* Use custom visitor for port deserialization instead of serde_json::Value
Co-authored-by: contact <contact@ophir.dev>
* Fix clippy warnings about uninlined format args
Co-authored-by: contact <contact@ophir.dev>
* Add robust port deserialization for Kubernetes
---------
Co-authored-by: Cursor Agent <cursoragent@cursor.com>
* Format protected and public paths with site prefix
Update protected and public paths to include site prefix.
* Fix formatting of paths in OIDC module
* Validate that paths start with '/' in config
* Refactor path handling for OIDC configuration
* Update OIDC paths configuration details
* code fix+clippy hints+formatting
CI / test (postgres, odbc, Driver=PostgreSQL Unicode;Server=127.0.0.1;Port=5432;Database=sqlpage;UID=root;PWD=Password123!, true) (push) Has been cancelled
Introduces a new configuration option `cache_stale_duration_ms` which
allows users to control how long files are cached before their freshness
is checked. This provides more fine-grained control over caching
behavior, especially in production environments.
* add a duckdb docker image
* Add DuckDB ODBC driver installation and configuration
Installs DuckDB ODBC driver and configures system-wide ODBC settings to
enable DuckDB database connectivity for SQLPage.
* add duckdb odbc info to readme
* Fix CI digest artifact matching
* Temporarily publish docker images on branch
* Allow docker publish on branch (fix)
* Temporarily run CI on branch pushes
* Disable branch publish; fix duckdb inspect tag
* Temporarily enable CI push on branch
* Fix minimal tag suffix in docker_push
* Disable temporary branch CI push
* Add Oracle DB (free) and ODBC CI support
This change adds support for testing with Oracle DB (using the free version `gvenzl/oracle-free:slim`) in the CI pipeline. It:
- Updates `.github/workflows/ci.yml` to include a new matrix entry for Oracle DB.
- Adds steps to install the Oracle Instant Client and ODBC driver in the CI runner.
- Configures `odbcinst.ini` to register the Oracle ODBC driver.
- Updates `docker-compose.yml` to include the Oracle DB service definition.
* Fix CI: Remove libaio1 dependency
`libaio1` is not available in the ubuntu-latest environment used by GitHub Actions (which likely uses a newer Ubuntu version where `libaio1` is replaced by `libaio1t64` or similar, or it is transitively installed). Removing explicit installation to fix the CI failure.
* Fix CI: Update Oracle Instant Client to 21.14
The previous version 21.10.0.0.0-1 seems to be no longer available at the specified URL (404 Not Found). Updated to 21.14.0.0.0-1 which was verified to exist.
* Fix CI: Install libaio1t64 for Oracle Instant Client
Oracle Instant Client requires `libaio.so.1`, which is provided by the `libaio1t64` package in newer Ubuntu versions (like 24.04). Installing this package should resolve the "cannot open shared object file: No such file or directory" error.
* Fix CI: Symlink libaio.so.1 for Oracle Instant Client
On Ubuntu 24.04 (Noble), `libaio1t64` installs the library as `libaio.so.1t64`, but Oracle Instant Client explicitly looks for `libaio.so.1`. Creating a symlink fixes this loading issue.
* Fix CI: Install libodbcinst2 and unixodbc for Oracle driver
The Oracle ODBC driver requires `libodbcinst.so.2` which is provided by `libodbcinst2`. Installing `unixodbc` ensures the full ODBC stack is available.
* Fix CI: Remove redundant wget installation
`wget` is pre-installed on GitHub Actions runners, so the explicit installation via `apt-get` is unnecessary.
* Fix CI: Create 'root' user in Oracle DB
The Oracle test run failed with ORA-01017 (invalid credentials) because the 'root' user does not exist by default in the Oracle database.
This commit:
1. Installs Oracle SQL*Plus (instantclient-sqlplus) in the CI environment.
2. Adds a step to connect as SYSDBA and create the 'root' user with the required permissions and password, matching the test configuration.
* Fix CI: Use container env vars for Oracle user
Instead of manually creating the 'root' user with SQL*Plus, rely on the `gvenzl/oracle-free` container's initialization features (`APP_USER` and `APP_USER_PASSWORD`) which were already configured in `docker-compose.yml`.
Removed the SQL*Plus installation and user creation steps from the CI workflow.
* Fix CI: Use FREEPDB1 service for Oracle user
The `APP_USER` ('root') created by the `gvenzl/oracle-free` container is created in the `FREEPDB1` pluggable database, not the `FREE` container database. Updated the connection string to use `FREEPDB1`.
* Fix Oracle DB test failures
- `sql_to_json`: Handle Oracle's behavior where empty strings are treated as NULL.
- `filesystem`: Explicitly CAST the timestamp parameter to TIMESTAMP in `was_modified_query` for ODBC (Oracle) to avoid ORA-01843 (invalid month) errors caused by implicit string-to-date conversion mismatches.
* Fix Oracle tests by accounting for known behavior
- Revert generic ODBC changes in `filesystem.rs`.
- Skip `test_sql_file_read_utf8` for ODBC connections (Oracle) because fixing the `ORA-01843` error would require Oracle-specific SQL (e.g. `CAST` or `TO_TIMESTAMP`) in the main codebase which is undesirable for a generic ODBC implementation.
- Keep the test adjustment in `sql_to_json.rs` but clarify the comment that we are assuming ODBC implies Oracle in this test context for the empty string behavior.
* Fix Oracle tests: Identify Oracle by connection string
Instead of relying on `sqlx::any::AnyKind::Odbc` (which applies to any ODBC database), detect Oracle specifically by checking if the connection string contains "Oracle". This allows applying Oracle-specific test logic (like skipping tests with implicit timestamp conversions or handling empty strings as NULL) without incorrectly affecting other ODBC databases.
* use oracle dialect when talking to oracle
* remove stupid ai comment
* update oracle odbc installation steps
* fix odbc installation path
* cast variables to varchar(4000) in oracle
* clippy
* remove long backtraces from ci
* fixed csv upload test for oracle
* update tests for oracle
* properly quote sqlpage-generated col names
* fix test syntax for oracle
* clippy
* remove as but keep alias
---------
Co-authored-by: Cursor Agent <cursoragent@cursor.com>
* fix: handle temp file removal race condition in concurrent initialization
When multiple SQLPage instances start simultaneously in the same directory,
they can encounter a race condition during initialization. The create_default_database()
function creates a temporary file to test directory writability, then removes it.
If multiple instances try to remove the same file concurrently, some panic with
'No such file or directory'.
This commit replaces the .expect() panic with graceful error handling using
if let Err(). The writability test has already succeeded by the time we try
to remove the file, so whether another instance removed it is irrelevant.
Includes a test that spawns 10 concurrent threads initializing AppConfig
to verify no panics occur.
ref #1183
* cargo fmt
* the error may have another cause
* Remove concurrent initialization test
Removed the test for concurrent initialization. The test did not work
---------
Co-authored-by: lovasoa <contact@ophir.dev>
* fix(oidc): respect site_prefix in OIDC redirect and logout URLs
This change ensures that when `site_prefix` is configured, the OIDC redirect URI and logout URI include this prefix.
Previously, `site_prefix` was ignored, causing OIDC callbacks to fail when the application was served under a sub-path.
- Added `site_prefix` to `OidcConfig`.
- Updated `make_oidc_client` to prepend `site_prefix` to the redirect URI.
- Updated `handle_request` to match paths with `site_prefix` included.
- Updated `validate_redirect_url` to respect the prefix when verifying redirect targets.
- Added a regression test `test_oidc_with_site_prefix`.
* Refactor: Update dependencies and remove unused crates
This commit updates several dependencies to their latest versions and removes unused crates to streamline the project.
Co-authored-by: contact <contact@ophir.dev>
* removed unused config
---------
Co-authored-by: Cursor Agent <cursoragent@cursor.com>
This commit refactors the OIDC tests to use a more robust fake OIDC provider and improves the logout URL generation to correctly handle the scheme.
Fixes https://github.com/sqlpage/SQLPage/issues/1174
Ophir LOJKINE