Add Delete / Update modifiers for MySQL #1254

[OlivierCavadenti]

• Fixes Update ignore and delete ignore not parseable #1254

• Add IGNORE parser/deparse for DELETE and UPDATE.

• I find also two modifiers not available for MySQL so I also add LOW_PRIORITY modifier for DELETE and UPDATE, and QUICK modifier for DELETE.

• Documentation : https://dev.mysql.com/doc/refman/8.0/en/update.html https://dev.mysql.com/doc/refman/8.0/en/delete.html

• Example queries:

UPDATE IGNORE table1 A SET A.columna = 'XXX'
DELETE IGNORE FROM tablename

UPDATE LOW_PRIORITY table1 A SET A.columna = 'XXX'
DELETE LOW_PRIORITY FROM tablename
DELETE QUICK FROM tablename

[manticore-projects]

Pardon me, I do not understand that line: would the token not always be "LOW_PRIORITY" and so modifierPriority = UpdateModifierPriority.LOW_PRIORITY?

Why would we not write this as

[ <K_LOW_PRIORITY> { modifierPriority = UpdateModifierPriority.LOW_PRIORITY; } ]

Am I missing something please?

[OlivierCavadenti]

I fix that ok. I copy the logic from INSERT priority but they have more values...

[manticore-projects]

Apologies, I was not verbose enough. You can replace:

[(tk = <K_LOW_PRIORITY>)
    {if (tk!=null)
        modifierPriority = DeleteModifierPriority.LOW_PRIORITY;
    }]

with a simple

[ <K_LOW_PRIORITY> { modifierPriority = UpdateModifierPriority.LOW_PRIORITY; } ]

Just showing the possibilities, you can leave as it is and let the compiler do the work.

[OlivierCavadenti]

My fault sry ^^. Do I need something in this PR regarding your comment here ? #1382 (comment)

[manticore-projects]

Do I need something in this PR regarding your comment here ? #1382 (comment)

No, you did good.
I just used your contribution as an example, why we need to work on the Reserved Keyword Management in general and it is not certain, if @wumpz will like my idea. If he does, then your new token QUICK will be handled automatically. If he does not, we can add it any time later to the Allowed Keywords.

Managing Reserved Keywords in a JavaCC Grammar is a bit weird and not straight forward.

[OlivierCavadenti]

Ok no problem, thanks for the review anyway.

[manticore-projects]

Thank you so much for the contribution, good stuff!
Please consider 2 amendments:

1. I left a comment on a possible code simplification, if my assumption was wrong please disregard

2. more importantly: the PR fails the CI Test as shown in https://app.codacy.com/gh/JSQLParser/JSqlParser/pullRequest?prid=8341437
   It is mostly on Code Style and annoying stuff, but important for house keeping. Please clean out this list and push the changes. Thank you!