?. is a string literal:
?. #=> "."
However, I failed to declare a variable with a name like that:
?some_var = 100 #=> Error
How is?something invalid when ?. is valid?
Asked
Active
Viewed 54 times
0
sawa
- 165,429
- 45
- 277
- 381
Jechol Lee
- 145
- 10
-
@Arup It is not a duplicate. – sawa Mar 08 '15 at 07:25
-
@sawa Ok.. I thought so .. :) – Arup Rakshit Mar 08 '15 at 07:31
2 Answers
2
? cannot describe any string literal; it is valid only for a single character.
Even if ?something were a valid string literal (counter to fact),
?something = ...
will be assignment to a string, which does not make sense. You cannot assign a value to a string.
sawa
- 165,429
- 45
- 277
- 381
1
?a is the same as "a". So it is a value, which belongs on the right hand side of an assignment, not the left hand side. It is not a variable name.
The Syntax exists as a relic from Ruby <=1.9, where it was equivalent to "a".bytes[0] and ?d could be used to shave off one character of code golf. I haven't seen any legitimate use otherwise.
Reactormonk
- 21,472
- 14
- 74
- 123
-
1There was a legitimate use during the transition period from 1.8 to 1.9, because it always returned the same thing that indexing string would return. So, on 1.8 `'Hello'[0] == 72` and `?H == 72` and on 1.9 `'Hello'[0] == 'H'` and `?H == 'H'`, so that it in both cases it is *always* true that `'Hello'[0] == ?H` no matter whether you are on 1.8 or 1.9. – Jörg W Mittag Mar 08 '15 at 13:25