Ambiguous meaning of # in different SQL dialects

[jdorn]

In MySQL, # is the start of a single line comment.
In SQL Server, # is used to prefix temporary table names.
In Microsoft Access SQL, # is used to surround dates.
In Oracle SQL, # is a valid character in an unquoted identifier.

These are just the ones I was able to find with a few minutes of searching.

Currently, it uses the MySQL meaning and highlights/formats it as a comment.

I'm not sure what the correct thing to do is.

[hhellbusch]

Perhaps an optional parameter or config to specify which dialect to use. I'd prefer this as it would allow me to tell the formatter which dialect to use.

[Deltachaos]

For SAP ASE # indicates a temporary table when used as a table name. Would it not be the best to remove the # from comments now?

On the other hand i also prefer @hhellbusch suggestion of giving the formater information about which dialect it should use. Would this be a satisfiable solution to you @jdorn? If so i would implement it that way.

My suggestion is to make the SqlFormater class instanceable and give a dialect as argument in the construct. Then i would create multiple dialect classes which conains the dialect tokens etc.
Then i would deprecate the use of the static format method and would instantiate a new instance there and call the new method on the object.