Skip to content
7 changes: 6 additions & 1 deletion packages/mysql-parser/README.md
Original file line number Diff line number Diff line change
Expand Up @@ -80,7 +80,7 @@ This requires `bash`, `curl`, `docker`, and `php` and executes the following ste
3. Extract and compact the grammar and tokens, and save them in `src`.

### Building the query corpus
This package includes a testing corpus of about 70,000 MySQL queries extracted
This package includes a testing corpus of about 78,000 MySQL queries extracted
from the MySQL release that corresponds to the generated MySQL grammar version.
To regenerate it, use:

Expand All @@ -94,6 +94,11 @@ This requires `bash`, `git`, and `php`. It performs the following steps:
3. Store the queries under `data/mysql-server-query-corpus/mysql-latest.csv`.

A query that runs in SQL modes that affect the lexer has them in a second column.
A query that the MySQL tests only ever expect to fail with a syntax error has
`syntax_error` in a third column, unless the error comes from outside its own
syntax, such as the text of a `PREPARE` statement. A few queries with syntax
errors that the tests don't expect alone have it too, and the extractor lists
them by name.

## Tests and benchmarks
To run lexer and parser tests and benchmarks, use:
Expand Down
Loading
Loading