Semi-Structured Data

Semi-Structured Data in Datameer

[Video Transcript]

Hi. Let’s take a look at Datameer’s ability to explore, parse and extract semi-structured data such as JSON and variant columns using a NoCode Wizardlike approach versus the typical tedious custom coding approaches.

Here we have a table that contains some weather data in a JSON or a variant column, as you can see from the bottom of your screen.

Let’s extract some information from this semi-structured data. To do so, click on the plus button right next to the table and choose new SQL view to open the SQL editor.

You can then edit the actual SQL query to add an additional column by adding a comma right next to the select star and then click on the required JSON at the bottom.

To open the JSON exploring dialog or wizard, you can choose the required JSON attributes you’d like to extract, for example, in this case, the city name, and it displays the actual JSON expression path.

And then, you can choose or click on the Insert Path expression button to automatically insert the path into the query.

You can then format the query if needed and then run the query to get your results. As you can see, I have successfully extracted the JSON City Name attribute at the bottom of your screen.

You can do something similar if you want for a JSON object versus an individual attribute. To do so, let’s add another column to the query and then click on the actual exploring dialog and then choose all of the object lines required to extract that column.

And then Similar button at the bottom exists to insert that expression into the query and then run the query again successfully to both extract the attribute as well as the object from the semi-structured data.

Thank you.