Recipe page
The Recipe page is the heart of the HXL Proxy. Here, you define a series of steps (Filters) to transform the data into a different state: for example, you might add or remove columns, change the general shape of the data, produce a report (similar to a pivot table), automatically clean dates and numbers, replace text, and many other steps.
In the typical workflow, you arrive here from the Source page or Tagger page.
When you are finished editing the recipe, the Done filters button will take you to the View page, where you can browse or download the result of your recipe.
Strip text headers: remove the text headers and leaves only the HXL hashtags. This is useful when you need to use HXL data with tools that allow only one header row.
Never cache: show changes in the source data instantly. See the Caching article for details.
Download filename: specify a basename for downloadable data. For example, if you enter "survey-data" as the value, the download filename for CSV data will be "survey-data.csv" and the download filename for JSON data will be "survey-data.json".
Max rows: limit the number of rows that the HXL Proxy will return for a dataset (useful for previews, avoiding DoS attacks, etc.).
For the recipe, you define a series of Filters. Each filter reads the data from the previous filter (or original source), changes it in some way, then passes it on to the next filter (or final output). You can rearrange the filters using the drag handles to the left of each one, or delete them using the [x] box to the right.
See the Filters page for a list of the filter forms available.
As an alternative to defining filters via forms, advanced software developers may choose to use JSON recipes.