Skip to content

larsko/pure-api

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

21 Commits
 
 
 
 
 
 
 
 

Repository files navigation

pure-api

This is a light-weight Pure API wrapper written in C#.

REST responses are automatically deserialized to a dynamic JSON object (Note: XML not currently supported). This makes it very easy to use as-is. This code is intended as a quick and dirty demonstration of how to retrieve data from the Pure API. If you need type safety, feel free to implement the model on top and use the overloaded methods.

For an introduction to the Pure API, please refer to the Swagger.io documentation. The documentation can be found at https://<PURE_URL>/ws/api/{version}/api-docs.

Dependencies

  • .NET 4.6+ (+ Microsoft.CSharp for Mono users)
  • Newtonsoft.Json
  • RestSharp

Usage

Create an API key for the Pure API. You will need to pass the API key for every request. It is recommended to store the configuration values below to your App.Config file.

string baseUrl = "<PURE_URL>"; // do not include the relative path, i.e. /ws/api
string apiKey = "<PURE_API_KEY>"; // The API key to use - generate this in Pure
string apiVersion = "512"; // API version to use, e.g. 510, 511, 512, etc.

Instantiate the Pure Client with the URL, API key and version.

var client = new PureClient(baseUrl, apiKey, apiVersion);

Making Requests

Request a particular resource and execute it with the Pure Client.

var request = new PureRequest("persons");

The result returned is a dynamic type.

var result = client.Execute(request);
int count = result.count; // total # persons returned given criteria
Console.WriteLine(result.items[0].name.lastName); // print name of first record

It is possible to deserialize to a concrete type - simply pass the type parameter T:

var result = client.Execute<T>(request);

Tip: http://jsonutils.com/ is a great class generation tool for JSON that can help save you some time.

Harvesting Data

A common use case is to download all data from a particular content family, e.g. research output. The Pure API implements paging for this purpose, making it possible to download content in batches of the desired size.

Using the harvester is easy, simply instantiate it with the API client:

var harvester = new PureHarvester(client);

Then download all content from a certain family, e.g. research output:

harvester.Harvest("research-outputs", data => {
    // Print the UUID of each publication
    foreach(var item in data.items){
        Console.WriteLine(item.uuid);
    }

});

By default harvesting is done in parallel (4 threads), but this can be disabled if serial requests are required. Note: The harvester uses dynamic typing, so please refer to the API reference for the JSON structure.

Endpoints

See the class ContentType for string constants for available endpoints.

var orgs = new PureRequest(ContentType.OrganisationalUnits); // get orgs
var persons = new PureRequest(ContentType.Persons); // get persons

Getting Changes

The Pure API can also return all a list of changes since a given date. This is useful to process incremental updates after having harvested the initial dataset (as opposed to downloading everything from scratch each time).

Each change is a tuple (changeType, uuid, familySystemName, version). Every time content is updated in Pure, the version is incremented and the operation changeType is tracked. When you pass a date to this endpoint, it will return a sequence of changes in chronological order.

Usage is simple:

harvester.GetChanges(new DateTime(2017, 06, 15), data => {
    // Print the number of changes in this batch:
    Console.WriteLine(data.count);
    // Print each tuple.
    foreach(var item in data.items){
        Console.WriteLine($"{item.changeType}: {item.uuid} {item.familySystemName} v.{item.version}");
    }
});

Use this by passing an Action<T> callback to GetChanges and sift through changes as desired. You may want to pass a queue to gather all the changes. Then afterwards it is just a matter of making a serial request (using familySystemName and uuid) for each item in the queue.

The following example shows how to fetch all persons that were created since June 16, 2017. Then each person is downloaded and the UUID is printed.

var results = harvester.FilterChanges(new DateTime(2017, 06, 16), ContentType.Persons, "CREATE");
// harvest all modified records in parallel
harvester.HarvestChangeset(results, item => { 
    Console.Writeline(item.uuid)
}

Releases

No releases published

Packages

No packages published

Languages