MySQL Bugs: #82540: Names for keyword identifier parser rules can be clearer

Bug #82540	Names for keyword identifier parser rules can be clearer
Submitted:	11 Aug 2016 8:24	Modified:	11 Aug 2016 14:08
Reporter:	Martin Hansson	Email Updates:
Status:	Closed	Impact on me:	None
Category:	MySQL Server: Parser	Severity:	S3 (Non-critical)
Version:		OS:	Any
Assigned to:		CPU Architecture:	Any

Description:
This is a pure refactoring which only renames things.

We have a couple of parser rules to handle when non-sql keywords are used as identifiers. The sets of allowed keywords differ between roles, labels and variables. But the rules have abstract names that have nothing to do with their purpose, and they read awkwardly in English.

How to repeat:
It is currently hard to figure out the meaning of e.g. keyword_role_aux1.

Suggested fix:
Rename the rules as follows:

keyword -> ident_keyword (more consistent with <ident>, the only rule producing it, also more descriptive)

keyword_role -> role_keyword (reads like English)

keyword_sp -> label_keyword (more consistent with <label_ident> which producing it)

keyword_role_aux1 -> role_or_label_keyword (describes what it is used for)

keyword_role_aux2 -> role_or_ident_keyword (describes what it is used for)

The names currently have a lot of comments that will hardly be needed if we use these names instead.

Posted by developer:
 
Fixed in 8.0.0.

Code refactoring. No changelog entry needed.