Java deleteCharAt with ++ or -- [duplicate]

Question

I have the following code:

public static void main(String[] args) { StringBuilder phoneNumber = new StringBuilder("828-707-5634"); for(int i = 0; i < phoneNumber.length(); i++) { if (phoneNumber.charAt(i) == '-') phoneNumber.deleteCharAt(i--); } System.out.print(phoneNumber); }

Indeed this removes the dash from the phone number, but intuitively I would have expected it to remove the character just before the dash (because of the --). How/why does this work? Additionally, I find that if I use ++ or omit the increment/decrement entirely the dashes are still removed.

you are changing the object during inspecting its content in the for loop.... — ΦXocę 웃 Пepeúpa ツ
– ΦXocę 웃 Пepeúpa ツ, Commented Oct 31, 2017 at 12:52
i-- feeds the unchanged i to the method and afterwards decrements the value. It is the same as if you would call deleteCharAt(i); i--;. The --i operator does decrement it before feeding the method. — Zabuzard
– Zabuzard, Commented Oct 31, 2017 at 12:53
If you want it to delete de character before the dash, use --i — Artur Trapp
– Artur Trapp, Commented Oct 31, 2017 at 12:53
To be fair, this is a "correct" solution to decrement the index since you update the String directly. But I prefer to simply read from the end, that way I don't have to bother with the new length. — AxelH
– AxelH, Commented Oct 31, 2017 at 13:08

Bathsheba · Accepted Answer · 2017-10-31 13:14:52Z

deleteCharAt receives the unincremented value of i, since the value of the expression i-- is the unincremented value. That's why your code works, even though you are relying on the increment expression of the for loop to compensate for a rather contrived and meaningless decrease in i. That all seems rather crude to me, and having i as -1 which will happen if the first character is a dash will raise eyebrows during debugging.

It's probably best to use the replace method in java.lang.String. See replace String with another in java

Else I'd prefer

for (int i = 0; i < phoneNumber.length(); /*intentionally empty*/){ if (phoneNumber.charAt(i) == '-'){ phoneNumber.deleteCharAt(i); } else { ++i; } }

although this is still an O(N²) solution to an O(N) problem due to the repeated evaluation of length()! That can be solved by running the loop backwards, which will also reduce the number of element shuffles:

for (int i = -1 + phoneNumber.length(); i >= 0; /*intentionally empty*/){ if (phoneNumber.charAt(i) == '-'){ phoneNumber.deleteCharAt(i); } else { --i; } }

"Indeed this removes the dash from the phone number, but intuitively I would have expected it to remove the character just before the dash (because of the --) Lumbee " Seems your first statement answer OP question perfectly.
Since you've added your prefered solution, for this kind of code I prefer with for(int i = phoneNumber.length() -1; i >= 0; --i) ...
You shoud always decrement here, if not might end up with an OutOfBoundsException and we already checked the character that will come in that position PS :I like the -1 before the length, simple but much more readable

Collectives™ on Stack Overflow

Java deleteCharAt with ++ or -- [duplicate]

1 Answer 1

4 Comments

Linked

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

4 Comments

Linked

Related