Tables

Docs > App Builder > Components > Tables

This page describes advanced features you can use to manipulate table components in your App Builder apps.

Client-side filtering

When you have a full list of items already and you want to filter them, there are multiple methods to do so on the client side.

Filter a table with a text input or search component

One common use case is filtering a table component using the value in a text input component.

For example, if you want to list your dashboards in a table that you can filter using a text input component, you could do the following:

Add a new query using the + button.
Search for “list dashboards” and click the List Dashboards action. Name your query listDashboards0.
Add a text input or search component to your app. Name it searchInput.
Add a table component.
Set the table’s data source property to your data filtered by the text input or search component that you created. In this example, set the data source to the following expression:
```
${listDashboards0?.outputs.dashboards.filter(row => row.title.includes(searchInput.value))}
```

You can type text into the text input component and the rows of table are filtered by that text.

Filter a table with a select component

Another common use case is filtering a table using a select component.

For example, if you want to list your dashboards in a table that you can filter using a select component, you could do the following:

Add a new query using the + button.
Search for “list dashboards” and click the List Dashboards action. Name your query listDashboards0.
Add a select component to your app. Name it selectInput.
Add a table component.
Set the table’s data source property to your data filtered by the select component. In this example, set data source to the following expression:
```
${listDashboards0?.outputs.dashboards.filter(row => row.title.includes(selectInput.value))}
```

You can select a value from the select component and the rows of table are filtered by that value.

Filter query results using a post-query transformation

If you want to filter the results of a query itself, then use those results in your table, perform the following steps:

Add a new query using the + button.
Search for “list dashboards” and click the List Dashboards action. Name your query listDashboards0.
Add a text input or search component to your app. Name it searchInput.
Add a table component and set its data source property to the query that you added.
Expand the Advanced section of the query and find Post-query Tranformation.

Replace return outputs with the following line:

outputs.dashboards.filter(row => row.title.includes(searchInput.value))

You can type text into the text input component and the rows of table are filtered by that text.

If you need the original, untransformed query result, you can reference it as ${listDashboards0.rawOutputs}.

Server-side filtering

In some cases, you might want to filter values server-side and issue new requests when the user enters a value in an input such as a text input component.

In this case, you can enable server-side filtering by editing the query directly.

For example, in the GitHub PR pipeline blueprint, the listOpenedPulls query has an input that gets the following URL:

https://api.github.com/search/issues?q=org:${organizationInput.value}+author:${userNameInput.value}+type:pr+state:open

The GitHub API accepts query parameters for filtering based on organization, author, or pull request type. The preceding query input URL contains template expressions for organizationInput.value, which is the value of the “Organization” text input component, and userNameInput.value, which is the value of the “Username” text input component. If you set the query run settings to auto, the query automatically refreshes when these template expression values change, and the table values update.

Showing a loading indicator

If you want to show a loading indicator on a table while the data is being fetched, you can set the table’s isLoading value equal to the query’s isLoading property. For example:

Follow the steps in filtering with a text input.
In the properties of your table, under Appearance, click the </> next to Is Loading to open the code editor.
Set the table’s isLoading value to the following expression:
```
${listDashboards0.isLoading}
```

The table shows a loading indicator when you type new text into the text input component.

Dynamic table values

You can use the data source property of a table component to dynamically fill table values and constrain which objects are pulled into the table as columns.

For example, the GitHub PR Summarizer blueprint uses a series of GitHub queries to summarize a list of pull requests in a repository. The query uses the data source entry below to constrain the table to 6 columns: title,Summary,updated_at,user,html_url, and state. The highlighted code dynamically populates the user column for each pull request with the author’s avatar and GitHub username.

${(() => {
    const summaryById = Object.fromEntries(
        summarizePulls.outputs.map(({id, summary}) => [id, summary])
    );
    return listPulls.outputs.map(result => {
        const {title, updated_at, user, state, html_url} = result;
        const updatedAt = new Date(result.updated_at);
        let summary;
        if (summarizePulls.isLoading) {
            summary = 'Summarizing';
        } else {
            summary = summaryById[result.id] ?? 'N/A';
        }
        return {
            title: `**${title}**`,
            updated_at: updatedAt.toLocaleString(),
            user: {label: user.login, src: user.avatar_url},
            summary,
            state, html_url};
    })
})()}