This project uses some of the microservice baseline, and has the following dependencies/frameworks:
- Jetty Web Server
- Jersey JAX-RS
- TestNG
- REST-assured: Java DSL for easy testing of REST services
- Maven Shade Plugin
- jose4j
This will start both the server and the database, both running as a docker container.
cd redcross-digital-leksehjelp-backend
mvn clean install
cd Docker
sudo ./test_docker.shWe use MSSQL database. For the development build, it is populated and initiated using the endpoints described below.
ER diagram:
Simple endpoints used for creating a database, tables and filling with some test data, for development purposes. Only the DatabaseInitalizer should be used for prod (together with a migration script), and these endpoints should be removed. The DatabaseInitializer will also initialize the fulltext index and catalogs used for the fulltextsearch. The stoplist is set to system, and the language is set to norwegian.
POST /populate
Initializes the database and popoulates with some test data. To ensure the database is populated with a volunteer,
remember to change the DatabasePopulator.java to insert a valid volunteer.
/populate/stress?total=total
Populates the database with ? random questions. Useful for stress testing.
These endpoints give access to everything relevant to the listing and updating of existing questions, and creating new ones. Most of these endpoints are annotated with @JwtFilter, which means the header needs a valid authorization token from Azure AD. If not the server will simply respond with 401 unauthorized.
Questions have state, to simplify the flow through the status board. The states are as following:
| Index | Name | Description |
|---|---|---|
| 1 | NEW | A question without an answer |
| 2 | WIP | Work in progress, these questions have an answer |
| 3 | ANSWERED | Submitted for approval |
| 4 | APPROVED | Approved by a volunteer, as soon as the question is approved the mail will be sent to the student |
| 5 | PUBLISHED | Published and now visible without JWT token, only available for questions tagged with is_public=1 |
@JwtFilter
GET /questions?state=state&includeAll=includeAll
Lists all the questions with the selected state. If includeAll is true, all questions will return. Default is false, which results in only the questions which have a subject_id, that the logged in volunteer has selected as a subject, will return.
Example response:
[
{
"id":"2",
"questionText":"Hva er meningen med livet?",
"questionDate":"2019-08-08 07:19:55.987",
"title":"Filosofisk spørsmål",
"studentGrade":"10",
"subject":"RLE",
"themes": [
{
"theme": "Etikk",
"id": 1
}]
}
]@JwtFilter
GET /questions/{questionID}
Gets more detailed information of a single question with ID questionID, including answer and files.
Example response:
{
"id":"2",
"state":"2",
"questionText":"Hva er meningen med livet?",
"questionDate":"2019-08-08 07:19:55.987",
"answerText":"42",
"answerDate":"2019-08-09 06:14:53.253",
"title":"Filosofisk spørsmål",
"answer":"42",
"studentGrade":"10",
"subject":"RLE",
"themes": [
{
"theme": "Etikk",
"id": 1
}],
"files": [
{
"share": "azureShare",
"directory": "azureDirectory",
"fileName": "azureFileName",
"fileUrl": "azureFileUrl"
}],
"isSent": "1",
"isPublic": "1"
}GET /questions/public?searchText=searchText&orderByDate=orderByDate&subjectID=subjectID&page=page&grade=grade
Search after searchText in all the public questions, and list the results. Can filter by subjectID and grade. If orderByDate is set to true, the response will be ordered by the date, the default is false, which results in the response being sorted by relevance. The result is pagenated in pages of 10, select the page by using the page parameter. The response will include the number of total rows in the result.
Example response:
[
{
"themes":[],
"questionDate":"2019-08-08 07:19:57.357",
"studentGrade":"10",
"answerText":"4",
"answerDate":"2019-08-08 07:19:57.363",
"subject":"Matte",
"id":"105",
"totalRows":"2",
"title":"Gangespørsmål",
"questionText":"2*2?",
}
]GET /questions/public/{questionID}
Gets more detailed information of a single public question with ID questionID, including answer.
Example response:
{
"id":"2",
"state":"5",
"questionText":"Hva er meningen med livet?",
"questionDate":"2019-08-08 07:19:55.987",
"answerText":"42",
"answerDate":"2019-08-09 06:14:53.253",
"title":"Filosofisk spørsmål",
"answer":"42",
"studentGrade":"10",
"subject":"RLE",
"themes": [
{
"theme": "Etikk",
"id": 1
}]
}@JwtFilter
POST /questions/{questionID}
Edits an exisiting question with the id questionID, this is also where the answer will first be inserted.
Example payload:
{
"themes":[],
"studentGrade":"10",
"subject":"Matte",
"isPublic":"true",
"files":[],
"id":"1",
"state":2,
"title":"Addisjon",
"questionText":"Hva er 2+2?",
"answerText":"4",
"questionId":"1"
}
Illegal state transistions (implemented in the EndpointUtils class):
State {1, 2, 3, 4, 5} -> 1
State {1, 2} -> 4
State {1, 2, 3} -> 5
State {1, 2, 3, 4} -> 5 hvis is_public = 0
POST /questions
Posts a new question. Note this is a public endpoint.
Example payload:
{
"email":"[email protected]",
"studentGrade":"12",
"subjectID":5,
"questionText":"Hvorfor faller epler?",
"isPublic":true,
"files":[],
"themes": [
{
"theme":"Mekanikk",
"id":17
},
{
"theme":"Tyngdekraft",
"id":15}]
}
These endpoints give access to everything relevant to the volunteers, including registering new volunteers in the database. All of these endpoints are annotated with @JwtFilter, which means the header needs a valid authorization token from Azure AD. If not the server will simply respond with 401 unauthorized.
@JwtFilter
GET /volunteers
Lists all volunteers.
Example response:
[
{
"id":"2",
"name":"Morten",
"bio_text":"Min profil.",
"img_url":"URL til mitt profilbilde",
"email":"[email protected]"
}
]@JwtFilter
GET /volunteers/self
Gets information about the logged in volunteer.
Example response:
{
"id":"2",
"name":"Morten",
"bio_text":"Min profil.",
"img_url":"URL til mitt profilbilde",
"email":"[email protected]"
}@JwtFilter
POST /volunteers
Updates the information of the logged in volunteer.
Example payload:
{
"name":"Morten",
"bio_text":"Min profil.",
"img_url":"URL til mitt profilbilde",
"email":"[email protected]"
}@JwtFilter
GET /volunteers/subjects
Lists the subjects of the logged in volunteer.
Example response:
[
{
"id":"2",
"subject":"Norsk",
"is_mestring":"1"
}
]@JwtFilter
POST /volunteers/subjects
Changes the subjects of the logged in volunteer.
Example response:
{
"subjects": [1, 7, 4, 18]
}GET /subjects?isMestring=isMestring
Lists all subjects with themes. Can filter by isMestring.
@JwtFilter
POST /subjects
Creates a new subject.
GET /subjects?isMestring=isMestring
Lists all subjects with themes. Can filter by isMestring.
@JwtFilter
POST /subjects
Creates a new subject.
@JwtFilter
GET /feedback?includeAll=includeAll
Lists all feedback. If includeAll is true, all questions will return. Default is false, which results in only the feedback which have a subject_id that the logged in volunteer has selected as a subject.
@JwtFilter
POST /feedback/{feedbackID}/delete
Deletes the feedback with the ID feedbackID.
POST /question/{questionID}
Adds new feedback to the question with ID questionID.
GET /question/{questionID}
Gets all feedback to the question with ID questionID.
This baseline is set up without Spring, Constretto or similar tools.
Instead the properties are read through the class PropertiesHelper in
the Main class.
Properties are read from the following locations, in the following order:
application.propertiesfromclasspath(resources/application.properties).config_override/application.propertiesfile.
The default logback.xml in this repository only appends statements to STDOUT.
This is intended, and covers the two main use cases:
- Running locally, and in IDE
- Running in ECS with the awslogs Log Driver
If you wish to run the application with logging to file, you should use a properly set up file appender:
<?xml version="1.0" encoding="UTF-8"?>
<configuration scan="true" scanPeriod="60 seconds">
<property name="LOG_DIR" value="logs/"/>
<property name="appName" value="app"/>
<appender name="logfile" class="ch.qos.logback.core.rolling.RollingFileAppender">
<file>${LOG_DIR}/${appName}.log</file>
<rollingPolicy class="ch.qos.logback.core.rolling.TimeBasedRollingPolicy">
<fileNamePattern>${LOG_DIR}/${appName}-%d{yyyy-MM-dd}.%i.log</fileNamePattern>
<maxHistory>7</maxHistory>
<totalSizeCap>250MB</totalSizeCap>
<timeBasedFileNamingAndTriggeringPolicy
class="ch.qos.logback.core.rolling.SizeAndTimeBasedFNATP">
<maxFileSize>50MB</maxFileSize>
</timeBasedFileNamingAndTriggeringPolicy>
</rollingPolicy>
<encoder>
<pattern>%d{yyyy-MM-dd'T'HH:mm:ss.SSSZ} [%thread] %-5level %logger{35} - %msg%n</pattern>
</encoder>
</appender>
<logger name="org.eclipse.jetty" level="WARN"/>
<logger name="no.capraconsulting" level="INFO"/>
<root level="info">
<appender-ref ref="logfile"/>
</root>
</configuration>