To create a kimono API from any webpage, navigate to the web site you want to get data from and click the kimonify bookmarklet. This will bring up the kimono toolbar:
The toolbar is your central control panel during API creation. It allows you to organize and name your data, view different output formats and finalize/create your API. The buttons/actions available on the toolbar include:
|Data circle||Represents the different datatypes by color and displays the number of items in the current dataset.|
|New datatype||Click this to spawn a new datatype and begin adding to the new dataset|
|Pagination||If you want kimono to follow a 'next'/'more' link to get data from more pages with similar structure, click this then select that link on the page. Find out more by reading about pagination|
|Undo||If you make a mistake, just click undo to step back|
|Extractor view||The default view, showing the web page and your selections|
|Data model view||Shows how the elements you selected will be associated|
|Data preview||Shows a preview of how your data will look in JSON, CSV or RSS|
|Done||Once you've finished selecting the data you want, click done to create your API|
To get data from the page, simply click on it. As you move your mouse around you'll notice that text and images are highlighted in the same color as the data circle in the toolbar. Each time you click on something new, kimono will add that element to your dataset, and also will suggest similar items to you:
Similar elements will be represented in different shades of the current data color. Elements shown in darker shades are more similar than those shown in lighter shades. You can accept or reject the suggested elements by clicking or . You can also select additional elements that are not suggested by clicking on them as you did the first.
It's important that you indicate to kimono when you want to begin selecting a different type of data. You do this by clicking at any time to spawn a new datatype. Each datatype is represented by a new data circle/color in the toolbar.
Only one datatype can be active at any time, and the elements that you click on around the page will be included only in the active dataset, not the others. For example, indicates three different datatypes each with 5 elements selected.
You can name your datatypes using the textfield next to the datatype circles or from the data model view ().
Datatypes can be deleted at any time by clicking the small red delete button that appears at the top right of datatype when using your mouse to hover over the datatype dot:
One of the principles behind the design of kimono is that we don't want you to have to think about the data model. You should be able to click on the information you want from the page, and kimono will figure out the rest.
The built in learning algorithms can associate elements across datatypes and organize them into collections. Click on to view the data model.
This of course doesn't always go as planned. That's why we offer a set of advanced tools to take more control of what data gets selected by kimono. We are adding new functionality to advanced mode all the time, so look out for our feature releases or send us a note at firstname.lastname@example.org if you have suggestions or requests.
After you've made all your selections and named your datatypes and collections, you can preview your data by clicking the button. Data can be viewed or downloaded in JSON, CSV or RSS.
Finally, click to name your API and select a schedule on which you want it to update.
Note that if you select real-time, the API will fetch new data from the target page each time you hit the endpoint which can take a few seconds to respond with data. Your API calls will respond much faster with data from APIs that run on schedule (i.e. Hourly) since kimono responds instantly with the cached version of the data.
That's it. You just created your first API. Click the link displayed to visit the API's detail page. You can see all your APIs by going to 'My APIs' from the menu under your name.