From 189c34e5f58f6fb0df14075533bc5551e686eb68 Mon Sep 17 00:00:00 2001
From: Kevin Longe <273747162+kvlonge@users.noreply.github.com>
Date: Sun, 8 Nov 2026 12:43:08 +0000
Subject: [PATCH] docs: say which dialects check, fix and format read

A T-SQL, BigQuery or other unread file is reported as unavailable; the
configuration page now names the dialects each command reads.
---
 .../skill/references/docs/configuration.md            | 11 +++++++----
 website/src/content/docs/docs/configuration.mdx       | 11 +++++++----
 2 files changed, 14 insertions(+), 8 deletions(-)

diff --git a/crates/sqlclean-cli/skill/references/docs/configuration.md b/crates/sqlclean-cli/skill/references/docs/configuration.md
index ac2a08f..9beb205 100644
--- a/crates/sqlclean-cli/skill/references/docs/configuration.md
+++ b/crates/sqlclean-cli/skill/references/docs/configuration.md
@@ -123,16 +123,19 @@ SELECT [order id] FROM orders
 error[SQLCLEAN999]: SQL could not be parsed as duckdb: Parse error at line 1, column 17: Expected RBracket, got Var ('id') near [SELECT [order id] FROM orders]
 ```
 
-SQLClean reads SQL with SQLQuack's parsers. A file in a dialect SQLQuack does not parse yet, such
-as `mysql`, is not checked, fixed or formatted: `check` reports it once as a warning (in JSON, under
-`unavailable`), and it never fails a command or counts as a problem.
+SQLClean reads SQL with SQLQuack's parsers. `check` and `fix` read `duckdb`, `generic`, `postgres`,
+`snowflake` and `sqlite` files, and `format` lays out `duckdb` and `snowflake` files; a `generic`,
+`postgres` or `sqlite` file is checked and fixed but left as written by `format`. A file in any
+other dialect, such as `mysql` or `tsql`, is not checked, fixed or formatted: `check` reports it once
+as a warning (in JSON, under `unavailable`), `format` leaves it unchanged with a warning, and it
+never fails a command or counts as a problem.
 
 ```text
 warning: SQL lint is unavailable for the mysql dialect: SQLQuack does not parse it yet
   --> reports/q.sql
 ```
 
-Supported dialects, case-insensitive:
+Dialect names SQLClean accepts, case-insensitive:
 
 | Dialect | Also accepted |
 |---------|---------------|
diff --git a/website/src/content/docs/docs/configuration.mdx b/website/src/content/docs/docs/configuration.mdx
index 1d098fb..f9ac8f7 100644
--- a/website/src/content/docs/docs/configuration.mdx
+++ b/website/src/content/docs/docs/configuration.mdx
@@ -112,16 +112,19 @@ SELECT [order id] FROM orders
 error[SQLCLEAN999]: SQL could not be parsed as duckdb: Parse error at line 1, column 17: Expected RBracket, got Var ('id') near [SELECT [order id] FROM orders]
 ```
 
-SQLClean reads SQL with SQLQuack's parsers. A file in a dialect SQLQuack does not parse yet, such
-as `mysql`, is not checked, fixed or formatted: `check` reports it once as a warning (in JSON, under
-`unavailable`), and it never fails a command or counts as a problem.
+SQLClean reads SQL with SQLQuack's parsers. `check` and `fix` read `duckdb`, `generic`, `postgres`,
+`snowflake` and `sqlite` files, and `format` lays out `duckdb` and `snowflake` files; a `generic`,
+`postgres` or `sqlite` file is checked and fixed but left as written by `format`. A file in any
+other dialect, such as `mysql` or `tsql`, is not checked, fixed or formatted: `check` reports it once
+as a warning (in JSON, under `unavailable`), `format` leaves it unchanged with a warning, and it
+never fails a command or counts as a problem.
 
 ```text
 warning: SQL lint is unavailable for the mysql dialect: SQLQuack does not parse it yet
   --> reports/q.sql
 ```
 
-Supported dialects, case-insensitive:
+Dialect names SQLClean accepts, case-insensitive:
 
 | Dialect | Also accepted |
 |---------|---------------|