Page iterator Implementation

[muthurathinam]

PageIterator

For a variety of reasons, collections of entities are often split into pages and each page is returned with a URL to the next page. Sometimes, page granularity provided by the API does not match the requirements of the consumer. This task aims to simplify the life of consumers of paged collections.

Iterate over all the messages

async function callingPattern() {
    try {
        // Makes request to fetch mails list. Which is expected to have multiple pages of data.
        let response: PageCollection = await client.api("/me/messages").get();
       // A callback function to be called for every item in the collection. This call back should return boolean indicating whether not to continue the iteration process.
        let callback: PageIteratorCallback = (data) => {
            console.log(data);
            return true;
        };
        // Creating a new page iterator instance with client a graph client instance, page collection response from request and callback
        let pageIterator = new PageIterator(client, response, callback);
        // This iterates the collection until the nextLink is drained out.
        pageIterator.iterate();
    } catch (e) {
        throw e;
    }
}

Stopping and Resuming the iteration

// Populating custom size pages if the api restricts to some maximum size. Lazy loading more data on user prompt or something, stop and resume will do the trick.
async function customSize() {
    try {
        let response: PageCollection = await client.api("/me/messages").get();
        let size = 1000;
        let count = 0;
        let callback: PageIteratorCallback = (data) => {
            console.log(data);
            count++;
            if (count === size) {
                count = 0;
                return false;
            }
            return true;
        };
        let pageIterator = new PageIterator(client, response, callback);
        // This stops iterating over for 1000 entities.
        pageIterator.iterate();

        // Resuming will do start from where it left off and iterate for next 1000 entities.
        // Resume is likely to be called in any user interaction requiring to load more data.
        pageIterator.resume();
    } catch (e) {
        throw e;
    }
}

[MIchaelMainer]

We should support DeltaLink as part of the pageIterator as it uses nextLink to perform the paging until the changes are sync'd

[MIchaelMainer]

I think iterationHelper is supposed to be private.

[muthurathinam]

Added in the spec as private, but I missed in the implementation. will change

[MIchaelMainer]

nit: it's just

[muthurathinam]

will change

[muthurathinam]

@MIchaelMainer Have done the changes that you are requested.

[deepak2016]

Looks good.