An in-process implementation of T-SQL, packaged as a library, suitable for testing data acess code and queries.

• create, alter, drop for
  • Tables
  • Views
  • Procedures
  • Functions
    • Scalar
    • Table-valued
• select, update, insert, delete, truncate, merge with
  • output for DML statements
  • offset ... fetch for queries
  • select ... into
  • insert ... select
  • update ... join
  • distinct
• CTEs (Common Table Expressions)
• if, while statments
• case, between expressions
• try, catch, throw, error_* functions
• All binary operators
• sum, count, rownumber aggregate functions
• Calling user-defined functions
• Input, output and return parameters
• declare, set, access local variables
• Global variables
  • @@rowcount, @@fetch_status, @@identity, @@error
• Cursors
• Uniqueness constraints imposed by indexes
• identity auto-incrementing columns
  • constraints imposed by it
  • ability to set indentity insert on and off
• default values for columns
• implicit type conversions and type checking

• waitfor
• break, continue, goto, labels
• Working Transactions with rollback/commit
• Nested table results
• Customizable collations (everything is case insensitive by default)
• Backup/restore capability
• INFORMATION_SCHEMA and other control tables and procedures
• Optimizations around indexes
• for keyword that specifies output format
• XML types and namespaces
• any special handling of the go statement
• Query plan caching
• Ability to report query plan