On 4/25/2020 1:39 PM, Jeff Higgins wrote:

On 4/25/20 1:01 PM, Arne Vajhøj wrote:

On 4/25/2020 12:57 PM, Arne Vajhøj wrote:

On 4/24/2020 11:36 PM, Jeff Higgins wrote:

from JLS-3.8 for JavaSE 9
Two identifiers are the same only if they are identical, that is,
have the same
Unicode character for each letter or digit. Identifiers that have
the same external
appearance may yet be different.

from JLS-3.8 for JavaSE 10
Two identifiers are the same only if, after ignoring characters that
are
ignorable, the identifiers have the same Unicode character for each
letter
or digit. An ignorable character is a character for which the method
Character.isIdentifierIgnorable(int) returns true. Identifiers that
have the
same external appearance may yet be different.

But Java 8 javac did ignore them??

So does Java 7 javac.

Java 6 javac gives me an error when trying to create the class file,
which likely means that it is not ignoring these characters.

Thanks. I don't have earlier than 8 easily available.
Was wondering.

I was wondering when Character,isIdentifierIgnorable(int codepoint)
became available. The earliest javadoc that i can find is 5 and is
available there. A Sun press release from 1997 features
"Global language support based on Unicode 2.0 standards" so maybe
available from then?

http://www-evasion.imag.fr/Membres/Francois.Faure/enseignement/ressources/java/jdk1.3/api/index.html

shows it in 1.3 and says "Since: JDK1.1".

Arne

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

On 4/25/20 1:01 PM, Arne Vajhøj wrote:

On 4/25/2020 12:57 PM, Arne Vajhøj wrote:

On 4/24/2020 11:36 PM, Jeff Higgins wrote:

On 4/24/20 10:24 PM, Arne Vajhøj wrote:

On 4/24/2020 8:37 PM, Jeff Higgins wrote:

On 4/24/20 8:29 PM, Arne Vajhøj wrote:

I think we are in the far out corners of the language.

Well, yeah, but still.
And still leaves me questioning motivation for ignorable chars.

We would need to find someone working at SUN back in the mid 90's
to know.

If I were to guess then somebody got the "smart" idea that
the language should not distinguish between names where
the difference could not be seen when type'd/cat'ed and
then someone else later added the \uXXXX feature.

Maybe not that far back.

from JLS-3.8 for JavaSE 9
Two identifiers are the same only if they are identical, that is,
have the same
Unicode character for each letter or digit. Identifiers that have the
same external
appearance may yet be different.

from JLS-3.8 for JavaSE 10
Two identifiers are the same only if, after ignoring characters that are >>> ignorable, the identifiers have the same Unicode character for each
letter
or digit. An ignorable character is a character for which the method
Character.isIdentifierIgnorable(int) returns true. Identifiers that
have the
same external appearance may yet be different.

But Java 8 javac did ignore them??

So does Java 7 javac.

Java 6 javac gives me an error when trying to create the class file,
which likely means that it is not ignoring these characters.

Thanks. I don't have earlier than 8 easily available.
Was wondering.

I was wondering when Character,isIdentifierIgnorable(int codepoint)
became available. The earliest javadoc that i can find is 5 and is
available there. A Sun press release from 1997 features
"Global language support based on Unicode 2.0 standards" so maybe
available from then?

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

On 4/25/20 1:51 PM, Arne Vajhøj wrote:

http://www-evasion.imag.fr/Membres/Francois.Faure/enseignement/ressources/java/jdk1.3/api/index.html

shows it in 1.3 and says "Since: JDK1.1".

Ah, thanks for that. So, since forever.

In one of the bug reports that I mentioned upthread a fellow
compiled his library using maven and javac, then uploaded to
maven central, created a maven project in Eclipse and would
not compile due to a spurious \u00?? in his source!!
So even far flung language features sometime can cause bug bites.

I mention in my post to JDT this from Java VM Spec 14:

Module names may be drawn from the entire Unicode codespace, subject to
the following constraints:

• A module name must not contain any code point in the range
' \u0000 ' to ' \u001F ' inclusive.

But no mention in JLS as far as I can tell?

I wonder if this will be an inconsistency waiting to bite?

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)