> ## Documentation Index
> Fetch the complete documentation index at: https://axiom.co/docs/llms.txt
> Use this file to discover all available pages before exploring further.

<AgentInstructions>

## Submitting Feedback

If you encounter incorrect, outdated, or confusing documentation on this page, submit feedback:

POST https://axiom.co/docs/feedback

```json
{
  "path": "/apl/scalar-functions/sql-functions/format-sql",
  "feedback": "Description of the issue"
}
```

Only submit feedback when you have something specific and actionable to report.

</AgentInstructions>

# format_sql

> This page explains how to use the format_sql function in APL.

## Introduction

The `format_sql` function converts the structured dictionary produced by [`parse_sql`](/apl/scalar-functions/sql-functions/parse-sql) back into a SQL string. Use it to normalize SQL formatting, validate that a parsed query round-trips correctly, or reconstruct a SQL statement after modifying its parsed representation.

`format_sql` is most useful as the second step in a parse-then-reconstruct pipeline: first parse a SQL string into a structured dictionary with `parse_sql`, optionally inspect or transform the result, and then call `format_sql` to produce a clean, normalized SQL string.

## For users of other query languages

If you come from other query languages, this section explains how to adjust your existing queries to achieve the same results in APL.

<AccordionGroup>
  <Accordion title="Splunk SPL users">
    Splunk has no equivalent to `format_sql`. There is no native way in SPL to reconstruct a SQL string from a parsed representation. In APL, `format_sql` takes the output of `parse_sql` and reconstructs the original SQL statement.

    <CodeGroup>
      ```sql Splunk example theme={null}
      -- No direct Splunk equivalent
      ```

      ```kusto APL equivalent theme={null}
      ['sample-http-logs']
      | take 1
      | extend parsed = parse_sql('SELECT id, status FROM logs WHERE status = 500')
      | project formatted = format_sql(parsed)
      ```
    </CodeGroup>
  </Accordion>

  <Accordion title="ANSI SQL users">
    ANSI SQL has no built-in function to reconstruct a SQL string from a parsed representation. `format_sql` is unique to APL and works as the inverse of `parse_sql`.

    <CodeGroup>
      ```sql SQL example theme={null}
      -- No direct SQL equivalent
      ```

      ```kusto APL equivalent theme={null}
      ['sample-http-logs']
      | take 1
      | extend parsed = parse_sql('SELECT id, status FROM logs WHERE status = 500')
      | project formatted = format_sql(parsed)
      ```
    </CodeGroup>
  </Accordion>
</AccordionGroup>

## Usage

### Syntax

```kusto theme={null}
format_sql(parsed_sql_model)
```

### Parameters

| Name               | Type       | Required | Description                                        |
| ------------------ | ---------- | -------- | -------------------------------------------------- |
| parsed\_sql\_model | dictionary | Yes      | The structured data model returned by `parse_sql`. |

### Returns

A string containing the SQL statement reconstructed from the provided data model.

## Example

Parse a SQL query representing a slow-query log entry and reconstruct it to verify the round-trip.

**Query**

```kusto wrap theme={null}
print formatted = format_sql(parse_sql('SELECT id, status, uri FROM requests WHERE req_duration_ms > 1000 ORDER BY req_duration_ms DESC'))
```

[Run in Playground](https://play.axiom.co/axiom-play-qf1k/query?initForm=%7B%22apl%22%3A%22print%20formatted%20%3D%20format_sql%28parse_sql%28%27SELECT%20id%2C%20status%2C%20uri%20FROM%20requests%20WHERE%20req_duration_ms%20%3E%201000%20ORDER%20BY%20req_duration_ms%20DESC%27%29%29%22%7D)

**Output**

```json wrap theme={null}
{
    "formatted": "select id, status, uri from requests where req_duration_ms > 1000 order by req_duration_ms desc"
}
```

The query parses the SQL string and then reconstructs it with `format_sql`, confirming that the parse-and-reconstruct pipeline preserves the original statement's structure.

## List of related functions

* [parse\_sql](/apl/scalar-functions/sql-functions/parse-sql): Parses a SQL statement string into a structured dictionary. `format_sql` is the inverse of `parse_sql`.
* [parse\_json](/apl/scalar-functions/string-functions/parse-json): Parses a JSON string into a dynamic dictionary. Use `parse_json` when your data contains JSON rather than SQL.