The first working version of tapestry was released a few months ago. Since then, I have been working on implementing features beyond my own requirements and use cases. The version 0.2.0, which includes some of the new capabilities was released last week. Yesterday, a patch version 0.2.1 was released fixing a regression (that originated in 0.1.0).

Bug fixes

• The status command has been fixed to work correctly with one-file-all-queries layout and name-tagging. It was broken ever since both these features were implemented.

Features

• Tapestry now comes with a builtin SQL formatter based on the sqlformat-rs crate. This removes the dependency on pg_format for SQL formatting, although it can still be used optionally (see next point).

• Besides the builtin formatter, three external SQL formatting tools are now supported - pg_format, sql-formatter, sqlfluff.

What's next?

• As mentioned earlier, I don't use some of the newer features myself. So I believe there are still some edge cases that need to be handled. The plan is to write more unit tests and aim for decent coverage, specially for such features.

• I am happy to add support for other external SQL formatters. If there's a particular formatter that you'd like to see supported, feel free to create an issue on github.

• A command for generating boilerplate code for pgTAP tests. I occasionally miss this feature myself, so it's definitely high on my list.

• I recently found out that an equivalent of pgTAP exists for MySQL - mytap. It also seems to be (originally) written by David Wheeler, the author of pgTAP. I think it makes a lot of sense for tapestry to support it.