For small datasets, the Sort+Merge Join with an Outer Join is a good fit. It behaves most like Join syntax you'd be used to in T-SQL. You can then use a Conditional split to redirect rows that are in your flat file but not in your T-SQL with an expression like ISNULL(table_key_id). You can play with the Sort settings for your columns to change things like case sensitivity. You might also want to use a Derived Column transform before your Sort to clean-up data (i.e. TRIM(text_column)).

One other note, sort your smallest columns first, this can dramatically improve performance for large datasets.

Alternatively, you can do the same thing using a Lookup. This is definitely faster, but can be difficult if you are trying to match on text data.