Skip to content
API for accessing images
TSQL Scala
Branch: master
Clone or download
This branch is 168 commits ahead, 352 commits behind NDLANO:master.

Latest commit

Fetching latest commit…
Cannot retrieve the latest commit at this time.

Files

Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
project
src
.gitignore
.travis.yml
LICENSE
ReadMe.md
build.properties
build.sbt
build.sh

ReadMe.md

IMAGE API

Build Status

Usage

Creates, updates and returns metadata about an image. It also supports resizing and cropping images on the fly. Implements ElasticSearch for search within the article database. To interact with the API, you need valid security credentials; see Access Tokens usage. To write data to the api, you need write role access. It also has as internal import routines for importing images from the old system to this database.

For a more detailed documentation of the API, please refer to the API documentation (Staging: API documentation).

Developer documentation

Compile: sbt compile

Run tests: sbt test

Create Docker Image: sbt docker

Setting up Minio for local S3-like storage

Minio is used when running image-api locally, so access to AWS S3 isn't necessary. Minio can be run with Docker. Make sure that it is on the same Docker network, and that Minio is aliased minio.gdl-local. image-api expects the existence of a bucket called local.images.gdl. This has to be created manually, using either Minio's web interface or command line tools.

IntegrationTest Tag and sbt run problems

Tests that need a running elasticsearch outside of component, e.g. in your local docker are marked with selfdefined java annotation test tag IntegrationTag in /ndla/article-api/src/test/java/no/ndla/tag/IntegrationTest.java. As of now we have no running elasticserach or tunnel to one on Travis and need to ignore these tests there or the build will fail. Therefore we have the testOptions in Test += Tests.Argument("-l", "no.ndla.tag.IntegrationTest") in build.sbt

sbt "test-only -- -n no.ndla.tag.IntegrationTest"

Working with the fork

This repository is forked from NDLA (called upstream below). We want to keep this repo in sync with upstream to use features and bug fixes from NDLA. We also want to merge some of our own changes with the upstream, so NDLA can benefit from them as well. There are 2 long living branches in this repo: master and ndla. master is the branch with image-api being deployable on the GDL platform, with all the GDL specifics. ndla is synced with the upstream/master, and should never contain any GDL specifics.

Note: Before using any of the git commands below, make sure you've added NDLA's repo as upstream:

git remote add upstream git@github.com:NDLANO/image-api.git

The ndla branch is configured to be protected on GitHub, to avoid accidentally deleting it.

Here's how to work with branches and pull requests to do so:

Case #1: Adding GDL specific features (of no interest to NDLA)

  1. Create feature branch from master, e.g. Issue#123-new_feature.
  2. Open pull request, and merge it into master.

Case #2: Add changes from NDLA to our fork

  1. Run git fetch upstream/master to fetch all new changes from NDLA.
  2. Checkout ndla branch.
  3. Run git merge upstream/master to merge these changes into the ndla branch.
  4. git push.
  5. Open pull request, merging ndla into master.

Case #3: Adding features that should be merged upstream (i.e. they are of interest to NDLA)

  1. Run git fetch upstream/master to fetch all new changes from NDLA.
  2. Checkout ndla branch.
  3. Run git merge upstream/master to merge these changes into the ndla branch.
  4. Checkout a new feature branch from ndla, e.g. Issue#321-new_mutual_beneficial_feature.
  5. Edit and commit changes.
  6. Open pull request, merging Issue#321-new_mutual_beneficial_feature into ndla.
  7. Do internal QA in GDL.
  8. Open pull request, merging ndla into upstream/master.
  9. Let NDLA do QA of this.
  10. When this is done, fetch changes from upstream (see #2) to update our master branch with the same changes.
You can’t perform that action at this time.