Search and Replace: Ii/II doesn't work for IIi

[Jerry]

[Using 21.80.0200]

Came across this unexpected behavior which I don't think is right, unless I am misunderstanding something.

1. Run xyplorer /fresh
2. Create file named "Aaa Iii Xxx"
3. Use Search and Replace to rename Ii/II. It works and you now have "AAa IIi Xxx".
4. Repeat step 3, and nothing changes. Why doesn't it change the remaining Ii to II ?

Same thing happens if you try search and replace Aa/AA or Xx/XX.

[highend]

At least I can confirm it. Don will (probably) fix it when he has time...

[admin]

Use switch /c to make the comparison case sensitive: Ii/II /c

[Jerry]

Use switch /c to make the comparison case sensitive: Ii/II /c

Ok, that works for dealing with the second case. I assume you regard the inconsistency as undesirable.

[admin]

Certainly an interesting case. This will handle it in the next beta:

Code: Select all

    * File | Rename Special | Search and Replace...: If the replacement just 
      changes the CASE of the replaced string (eg: re/Re) then the matching is 
      now case-sensitive by default. You don't have to pass the /c switch for 
      case-sensitive matching.

[admin]

Well, I shot too fast. That "fix" would break old code and would also limit the functionality of Search and Replace.

What we have here might be undesirable but it's not an inconsistency but a logical outcome of case-insensitive matching and the standard replacement rules (which are not recursive = do not replace again what has been replaced). ATM I see no way to improve the situation. The /c switch is the way to go for you.

[Jerry]

What we have here might be undesirable but it's not an inconsistency but a logical outcome of case-insensitive matching and the standard replacement rules (which are not recursive = do not replace again what has been replaced). ATM I see no way to improve the situation. The /c switch is the way to go for you.

I understand the issue of recursive replacement; I would not expect the second replacement to take place while doing the first one. But in this example, there are 2 separately initiated, independent search/replace actions, each for a single replacement. The fact that the second replacement fails is clearly inconsistent. It makes one think that there is something special about doing the replacement when the pattern is followed by a space, but that surely shouldn't be the case.

[admin]

Nope, it has nothing to do with the space. Since your replace operation is case-insensitive in both cases the red pair of "ii" is replaced by "II". In the second run you don't note it because it is already "II":

"AAa IIi Xxx"

[Jerry]

Nope, it has nothing to do with the space. Since your replace operation is case-insensitive in both cases the red pair of "ii" is replaced by "II". In the second run you don't note it because it is already "II":
"AAa IIi Xxx"

So you are saying that since case sensitivity is off, the second independent search/replace action is matching and replacing "II" and not the remaining "Ii" so that in effect, there is no apparent change. Is that right?

[admin]

Yep