Building the project

This project is built with Maven. Source code is available on GitHub. Thanks to JetBrains for the IntelliJ IDEA License!

Jet Brains

Clone the project

Use git to clone the project locally:

git clone
cd fscrawler

Build the artifact

To build the project, run:

mvn clean package

The final artifacts are available in distribution/esX/target directory where X is the elasticsearch major version target.


To build it faster (without tests), run:

mvn clean package -DskipTests

Integration tests

When running from the command line with mvn integration tests are ran against all supported versions. This is done by running a Docker instance of elasticsearch using the expected version.

A HTTP server is also started on port 8080 during the integration tests, alternatively the assigned port can be set with argument.

Run tests from your IDE

To run integration tests from your IDE, you need to start tests in fscrawler-it-common module. But you need first to specify the Maven profile to use and rebuild the project.

  • es-7x for Elasticsearch 7.x
  • es-6x for Elasticsearch 6.x

Run tests with an external cluster

To run the test suite against an elasticsearch instance running locally, just run:

mvn verify -pl fr.pilato.elasticsearch.crawler:fscrawler-it-v7 -Dtests.cluster.url=http://localhost:9200


If you want to run against a version 6, run:

mvn verify -pl fr.pilato.elasticsearch.crawler:fscrawler-it-v6 -Dtests.cluster.url=http://localhost:9200


If you are using a secured instance, use tests.cluster.user, tests.cluster.pass and tests.cluster.url:

mvn verify -pl fr.pilato.elasticsearch.crawler:fscrawler-it-v7 \
    -Dtests.cluster.user=elastic \
    -Dtests.cluster.pass=changeme \
    -Dtests.cluster.url= \


To run tests against another instance (ie. running on Elasticsearch service by Elastic, you can also use tests.cluster.url to set where elasticsearch is running:

mvn verify -pl fr.pilato.elasticsearch.crawler:fscrawler-it-v7 \
    -Dtests.cluster.user=elastic \
    -Dtests.cluster.pass=changeme \

Or even easier, you can use the Cloud ID available on you Cloud Console:

mvn verify -pl fr.pilato.elasticsearch.crawler:fscrawler-it-v7 \
    -Dtests.cluster.user=elastic \
    -Dtests.cluster.pass=changeme \

Using security feature

Integration tests are run by default against a standard Elasticsearch cluster, which means with no security feature activated.

New in version 2.7.

You can run all the integration tests against a secured cluster by using the security profile:

mvn verify -Psecurity

Note that secured tests are using by default changeme as the password. You can change this by using tests.cluster.pass option:

mvn verify -Psecurity -Dtests.cluster.pass=mystrongpassword

Testing Workplace Search connector

To test the Workplace Search connector, some manual steps needs to be performed as you need to start Enterprise Search and create manually the custom source as there is no API yet to do that.

Running the integration tests will remove everything you have indexed so far in the workplace local instance.

New in version 2.7.

  • Run the following steps:

    cd contrib/docker-compose-workplacesearch
    docker-compose up
  • Wait for it to start and open http://localhost:3002/ws.

  • Enter enterprise_search as the login and changeme as the password.

  • Click on “Add sources” button and choose Custom API.

  • Name it fscrawler and click on “Create Custom API Source” button.

  • Copy the “Access Token” value. We will mention it as ACCESS_TOKEN for the rest of this documentation.

  • Copy the “Key” value. We will mention it as KEY for the rest of this documentation.

  • You can now run in another terminal:

    mvn verify -pl fr.pilato.elasticsearch.crawler:fscrawler-it-v7 \
        -Dtests.workplace.access_token=ACCESS_TOKEN \
  • Then you should be able to see the documents in http://localhost:3002/ws/search


If you want to modify the look, go to the source and choose “Display Settings”. Adapt the settings accordingly.


In “Result Detail” tab, add the missing fields. And click on “Save”.


To run Workplace Search tests against another instance (ie. running on Enterprise Search service by Elastic, you can also use tests.workplace.url to set where Enterprise Search is running:

mvn verify -pl fr.pilato.elasticsearch.crawler:fscrawler-it-v7 \
    -Dtests.cluster.user=elastic \
    -Dtests.cluster.pass=changeme \
    -Dtests.workplace.url= \
    -Dtests.workplace.access_token=ACCESS_TOKEN \

Tests options

Some options are available from the command line when running the tests:

  • tests.leaveTemporary leaves temporary files after tests. false by default.
  • tests.parallelism how many JVM to launch in parallel for tests. Set to auto by default
    which means that it depends on the number of processors you have.
  • tests.output what should be displayed to the console while running tests. By default it is set to
    onError but can be set to always
  • tests.verbose false by default
  • tests.seed if you need to reproduce a specific failure using the exact same random seed
  • tests.timeoutSuite how long a single can run. It’s set by default to 600000 which means 5 minutes.
  • tests.locale by default it’s set to random but you can force the locale to use.
  • tests.timezone by default it’s set to random but you can force the timezone to use.

For example:

mvn install -rf :fscrawler-it -Dtests.output=always

Check for vulnerabilities (CVE)

The project is using OSS Sonatype service to check for known vulnerabilities. This is ran during the verify phase.

Sonatype provides this service but with a anonymous account, you might be limited by the number of tests you can run during a given period.

If you have an existing account, you can use it to bypass this limit for anonymous users by setting sonatype.username and sonatype.password:

mvn verify -DskipTests \ \

If you want to skip the check, you can run with

mvn clean install