Skip to content

Releases: katholiek-onderwijs-vlaanderen/sri4node

v2.3.35

25 Apr 13:32
Compare
Choose a tag to compare

version 2.3.35 (25-04-2024)

Fixed

  • The next link of a list resource would not be url-encoded correctly if the key contained a comma.
    This is fixed now, and a test case has been added to prevent regressions.

v2.3.34

28 Mar 14:13
Compare
Choose a tag to compare

Added

  • close() method to the Plugin interface, which is called when close() is called on the sri4node server instance

Changed

  • RELEASENOTES.md tries to follow the format as defined in keepachangelog.com now
    (but if we change the format of the title, we'd probably have to fix the npm release script as well!)

Fixed

  • Startup would fail if 'resources' in the configuration was an empty array

v2.3.33

26 Mar 13:45
Compare
Choose a tag to compare

version 2.3.33 (26-03-2024)

  • package.json fix exports['.'].import so that the ESM module can be imported correctly
    by using import * as sri4node from 'sri4node';
    (was pointing to ./dist/sri4node.esm.js instead of ./dist/sri4node.esm.mjs)

v2.3.32

01 Mar 10:45
Compare
Choose a tag to compare
  • Version compatible with postgres > 12
    A small change was needed to vsko_resource_version_inc_function() that gets created a startup if
    it does not exist yet. That means that existing api's will work on postgres 15, but if this
    trigger function would be removed, the api would not be able to start with older sri4node versions.

  • Work on testing: it is now easy to test against various combinations of nodejs and postgres.
    Also added a github workflow to matrix test multiple combinations on each commit.

    Example: npm run test:on_docker -- [--continue] 16,11-alpine 16,12-alpine 18,12 20,15

v2.3.31

01 Mar 10:43
Compare
Choose a tag to compare
Version 2.3.31

Changes:

v2.3.30

19 Feb 11:09
Compare
Choose a tag to compare

version 2.3.30 (19-02-2024)

  • All default filters use EXISTS instead of IN queries now. This is a performance improvement for large datasets.
  • Fixed information_schema query at start-up (would always retunr true since v2.3.29).
  • Updated dependencies, solving an issue with json-stream-stringify where the stream would not be closed properly in some cases.
  • Added auto-formatting with prettier and lint-staged.

v2.3.29

15 Feb 08:13
Compare
Choose a tag to compare

Improved information_schema query at start-up (using EXISTS instead of IN)

v2.3.28: Version 2.3.28

14 Nov 06:35
Compare
Choose a tag to compare

Bugfix: avoid sri4node crash in case of a database connection error during a db transaction.

v2.3.27

13 Oct 09:40
Compare
Choose a tag to compare
  • Improved support for more complex json-schema's (containing oneOff/anyOf/allOf at the root for example)
  • It used to be that storing a string in a JSONB column would not work,
    but that does work now, so it's easier to not worry too much anymore about the datatype
    and use jsonb column for almost everything
  • Internal improvements:
    • Improved 'npm run test' to only run a subset of all the tests. You can now run things like:
      npm run test ./testBatch.ts ./testPutAndPatch.ts
    • A few typing improvements.
    • Updated signature of 'query' handlers, so they always have the same signature
      (defaultFilter, queryUtils, relationsFilter + the user's own filters all have
      the same signature now)

v2.3.26

09 Oct 12:41
Compare
Choose a tag to compare
  • startUp hook changed: it gets db and pgp as paramters now instead of a 'half-baked' sriServerInstance
    and is called even earlier as a consequence.
    • Also added tests to check whether the startUp hook gets called, and whether we are able to do
      db changes here.
  • Change to version update triggers: they don't containe the schema name anymore.
    • Changed the created trigger's name in order to only contain the table name
    • Improved the code that was checking whether that trigger already existed + added test cases.
    • Added some code to cleanup OLD version update triggers that contained the schema name.
      This could lead to (harmless because identical) double triggers when dumping an api's
      schema to another schema, which is confusing.
  • Added sriRequest.outStream.write('') in streaming custom JSON routes to force sending the headers
    early even if nothing has been written to the stream yet.