想要一个正则表达式，它将验证Java变量之间只有一个逗号

Question

I have a regex which I use to validate a string.

String s = SELECT column1,column2 GROUPBY column3 AS AliasName
regex = ^SELECT\s+[[a-zA-Z_$][a-zA-Z_$0-9]*,*]+[a-zA-Z_$][a-zA-Z_$0-9]?\s+GROUPBY\s+[[a-zA-Z_$][a-zA-Z_$0-9]*,*]+[a-zA-Z_$][a-zA-Z_$0-9]?\s+AS\s+[a-zA-Z_$][a-zA-Z_$0-9]*$

This validation works but I need help with changing the regex so it only allows one comma in between each Java variable and doesn't allow it before the first column.

For example, the following String should not match but it does:

String s2 = SELECT ,column1,,column2 GROUPBY column3 AS AliasName

Also, I would like to change it so the SELECT , GROUPBY , and AS could be uppercase or lower case. I know I could use an alternation | but I would prefer a different way.

Answer 1

Try it like this:

(?i)^SELECT ([a-zA-Z_$][a-zA-Z_$0-9]+,)*([a-zA-Z_$][a-zA-Z_$0-9]+) (?:(GROUPBY) ([a-zA-Z_$][a-zA-Z_$0-9]+,)*([a-zA-Z_$][a-zA-Z_$0-9]+))?(?: AS ([a-zA-Z_$][a-zA-Z_$0-9]+))?

Online Demo
Code Sample

The easiest way to allow lowercase too is to switch the regex to case-insensitive using the inline modifier (?i) .

PS: Your query looks incomplete. There is no FROM part? PPS: You may want to look into different solutions since regular expressions cannot fully match and validate SQL syntax. You can try to tokenize the query . Or, even better, use a parser instead. ANTLR has some SQL grammar .