RFR: 8057113: (fs) Path should have a method to obtain the filename extension [v3]

[Roger Riggs]

On Thu, 21 Apr 2022 01:21:31 GMT, Brian Burkhalter <bpb at openjdk.org> wrote:

>> Resurrection of the proposal to add a method to obtain the filename extension originated in PR [2319](https://github.com/openjdk/jdk/pull/2319).
>
> Brian Burkhalter has updated the pull request incrementally with one additional commit since the last revision:
> 
>   8057113: Add hasExtension() and replaceExtension()

src/java.base/share/classes/java/nio/file/Path.java line 326:

> 324:             return Optional.empty();
> 325: 
> 326:         Objects.requireNonNull(ext);

Typically, all of the argument checks are done first. 
Though it does create a strange shape for the code when there is no extension. Two passes over the extensions seems like a waste.

src/java.base/share/classes/java/nio/file/Path.java line 355:

> 353:      *         leading and trailing space is ignored; may contain at most
> 354:      *         one dot ('.') which if present must be the first character
> 355:      *         that is not a space

Is there a use case for supplying the new extension with leading or training space.
I think the API is clearer if the supplied extension is treated as a literal.

src/java.base/share/classes/java/nio/file/Path.java line 372:

> 370:         int dotIndex = extension.lastIndexOf('.');
> 371:         if (dotIndex > 0)
> 372:             throw new IllegalArgumentException();

Can this be stricter and not include a conditional "."
It would be more consistent to define the extension as never including the ".".

src/java.base/share/classes/java/nio/file/Path.java line 384:

> 382:         if (thisExtension == null)
> 383:             return extension.isEmpty() ?
> 384:                 this : Path.of(path + "." + extension);

Path.of would revert to the default file system, even if the current Path is from a different FS.
Can the new path be defined from the current FS?

src/java.base/share/classes/java/nio/file/Path.java line 395:

> 393:         StringBuilder sb = new StringBuilder(path.substring(0, dotIndex + 1));
> 394:         sb.append(extension);
> 395:         return Path.of(sb.toString());

Is using StringBuilder more efficient in space than just using String + string concat?

-------------

PR: https://git.openjdk.java.net/jdk/pull/8066