Editing your details extractor

With the details extractor we created, we might want to go back in and add hidden values or clean up other values within Import.io before getting our data output.

You can click Edit to get back to the Edit view of an extractor. After adding more columns, you can add find values that might be originally by turning off the styling of a page. To this, go to the Advanced options, choose Page, and then disable CSS. The styling can be turned back on by re-enabling CSS.

When training an extractor, it will save how the page looks when you first create it. If the page changes, you can click the Refresh button (1) to add an updated version of the page and edit the Train on the additional URLs button (2).

Train with additional URLs will reveal a screen that allows you to add one or more pages to edit such as https://www.yelp.com/biz/los-gatos-coffee-roasting-company-los-gatos, to check the training of your extractor and improve it if needed. Once you add a second page, you can switch between the two pages to re-select data points if needed. The extractor will now use both pages to understand which data point you are selecting.

Besides adding data columns, you can edit existing data columns as well. Clicking the drop-down arrow next to a column name will reveal the column setting menu

From the column settings, you can use Set regular expression to match and filter out a string. For example, try selecting all of the text under "More business info". Then go to Set regular expression and set Match to Google Pay\s+(.+) and Replace to $1. This will match the string that contains Google Pay and then output the Yes or No value in the data column. We can then use Duplicate column and change the Match to Apple Pay\s+(.+) to create a separate column for Apple Pay.

Before saving the extractor, you can drag and drop the data columns to rearrange the order and go back to the Data view to check the output. Once you are done, click Save and run or Save only.

Now with the details extractor finished, you will want to move on to creating a listings extractor.