Google Cloud Search connectors are software programs allowing Google Cloud Search to operate on third-party data. The Content Connector SDK is a wrapper around the REST API allowing one to quickly create connectors.
There are two types of connectors: content connectors and identity connectors. Content connectors are used to traverse a repository and index the data so that Google Cloud Search can effectively search that data. For more information on building content connectors using Google Cloud Search connector SDK refer to https://developers.google.com/cloud-search/docs/guides/content-connector.
Identity connectors are used to map your enterprise's identities and group rosters to the Google accounts and groups used by Google Cloud Search. For more information on building identity connectors using Google Cloud Search connector SDK, refer to https://developers.google.com/cloud-search/docs/guides/identity-connector.
Building the Google Cloud Search Connector SDK requires the following development tools:
- Java SE Development Kit (JDK) version 1.8.0_20 or greater
- Apache Maven version 3.3.0 or greater.
-
Clone the SDK repository from GitHub:
git clone https://github.com/google-cloudsearch/connector-sdk.git cd connector-sdk
-
Checkout the desired version of the SDK:
git checkout tags/v1-0.0.5
-
Install the SDK components:
mvn install
Maven commands:
mvn install
- download required dependencies, build and install sdk jar as local maven repositorymvn clean
- deletes generated "target" directories.mvn compile
- compiles source.mvn test
- runs tests.mvn package
- compiles, tests, and creates the jar files and a distribution zip file.mvn -Pjavadoc
package -mvn package
withtarget/javadoc
and-javadoc.jar
file(s).
The integration tests check the correct interoperability between the connector SDK and the Cloud Search APIs. They are run using the Failsafe plug-in and follow its naming conventions (i.e., they are suffixed with "IT").
To run them, build and install the SDK, then cd
into the it/
subdirectory, and issue the
command
mvn verify \
-DskipITs=false \
-Dapi.test.serviceAccountPrivateKeyFile=<PATH_TO_SERVICE_ACCOUNT_KEY> \
-Dapi.test.sourceId=<DATA_SOURCE_ID>
-Dapi.test.identitySourceId=<IDENTITY_SOURCE_ID> \
-Dapi.test.customerId=<CUSTOMER_ID> \
-Dapi.test.domain=<COMPANY_DOMAIN>
where
-
api.test.serviceAccountKey
: path to JSON file containing the credentials of a service account that can access the APIs. -
api.test.sourceId
is the ID of the data source to which the test will sync the data. -
api.test.identity.identitySourceId
: ID of a Cloud Identity source. Preferably, this should be a source only used for testing. Temporary groups will be created on it during the execution of the test. -
api.test.identity.customerId
: the customer ID. -
api.test.identity.domain
: the domain associated with a Cloud Identity customer. Temporary users will be created on this domain during the execution of the test.
It's necessary to set a specific schema for the tests to run, otherwise a check during the set-up of the tests will fail. The schema can be found in it/schema.json (or it/schema.protoascii).
The following is an example of a complete command:
mvn verify \
-DskipITs=false \
-Dapi.test.serviceAccountPrivateKeyFile=key.json \
-Dapi.test.sourceId=01cb7dfca117fbb591360a2b6e46912e \
-Dapi.test.identitySourceId=01cb7dfca117fbb591360a2a5b64177c \
-Dapi.test.customerId=C14e3bmn2 \
-Dapi.test.domain=company.com
Adding -Dit.test=<IT class>#<method>
to the command line allows to run only a specific
test.
Notice also that will it's possible to run only unit tests (using
-DskipITs=true
) it's not obvious how to run only integration tests
(-DskipTests=true
skips both unit and integration tests). A work-around
for this is specifying -Dtest=NONE -DfailIfNoTests=false
(assuming that
there's no test class called NONE
) plus the appropriate flags to run
integration tests. For example:
mvn verify \
-Dtest=NONE -DfailIfNoTests=false \
-DskipITs=false \
-Dit.test=SomeTest \
...
A code coverage report for unit tests is generated for each module as part of the "test" goal. The
output directory for the report is <module>/target/site/jacoco/
.