The goal of this project is to implement a golang based RESTapi framework with in-memory datastore. The following is an in-depth walkthrough of this project. This is a demo API, so the "business" intent of it is to support basic CRUD (Create, Read, Update, Delete) operations for an application metadata datastore.
- Build a Golang RESTful API server for application metadata.
- An endpoint to persist application metadata (In memory is fine). The API must support YAML as a valid payload format.
- An endpoint to search application metadata and retrieve a list that matches the query parameters.
# Download project
git clone https://github.com/gandhipr/golang-restapi.git
# Build, test and run
make build
make test
make run ---> API Endpoint : http://localhost:8080
# Format go source code
make fmt
# Genarte mock files using mockgen for mocked interfaces
make generate-mock
├── apiserver
├── apis // Handlers for CRUD operations
├── deleteapi
├── getapi
├── postapi
└── putapi
├── datastore // In-memory datastore and corresponding tests
├── store
├── store_mock
└── store_test
├── messages // Error,success messages categorized based on modules - api, datastore, validators.
├── samples // Sample input yaml files used for testing
├── test
└── apiserver_test // Ent-to-end apiserver test
├── utils // Common methods and structs used in the project
├── validators // Validators defined for input struct
└── apiserver.go // Application server
- gin : implementing http framework and handlers
- gomock and mockgen : in-memory datastore used particularly for testing
- validator : validation framework for input struct
- testing and testify : support for automated testing
- httptest : support for http utilities while testing
- Given below is an example for a yaml that can be used. This project make following criteria to be satisfied:
- All fields are compulsory (non-empty).
- Have implemented restrictions on naming conventions for title and version.
- Should provide valid email-id.
- All fields are compulsory (non-empty).
## metdata.yaml
title: App1
version: 1.0.1
maintainers:
- name: Firstname Lastname
email: [email protected]
company: Upbound Inc.
website: https://upbound.io
source: https://github.com/upbound/repo
license: Apache-2.0
description: |
### blob of markdown
More markdown
- A metadata is identified by title and version. It is possible to have different versions of metadata for same title.
- For feasibility and testing purpose,
yaml file
andbinary data format for yaml
is supported. It is important to specify correct location of the yaml file.
API Endpoint : http://localhost:8080
-
POST apiserver/metadata/
,POST apiserver/metadata/:filepath
:
Store metadata in the store. Returns error if metadata with same title and version is already present. It is important to specify--data-binary
when latter request is used. -
PUT apiserver/metadata/
,PUT apiserver/metadata/:filepath
:
Currently, the update call supports only one functionality of updating existing metadata for given title + version combination. Returns error if no such metadata is present in the datastore. It is important to specify--data-binary
when latter request is used. -
GET apiserver/metadata/
: lists on the metadata in the store. -
GET apiserver/metadata/:title
: lists all the versions of the metadata for a given title. -
GET apiserver/metadata/:title/:version
: returns a specific metadata for a given title, version. -
DELETE apiserver/metadata/:title
: delete all the metadata version for a given title -
DELETE apiserver/metadata/:title/:version
: delete a specific metadata for a given title, version.
This project includes tests for 3 major sections:
- Datastore - unit tests using mock framework
- Validators - unit tests
- End-to-end test for the apiserver - tests for http handlers
- Add support for more fine-grained queries/updates for CRUD operations.
- Support Authentication with user for securing the APIs.
- Write the tests for all APIs.
- Make use of a persistent storage.
- Building a deployment process (automated deployment).