RFR [S] 8131034: Cleanup in j.u.regex.Pattern.quote()
ivan.gerasimov at oracle.com
Mon Jul 13 23:51:26 UTC 2015
On 13.07.2015 21:26, Martin Buchholz wrote:
> I agree that normalizedD looks like a typo (but it probably was not),
> but "normalized" is too ambiguous (there are too many different kinds
> of normalization) so a precise term like NFD should be used. Maybe we
> should add a reference to http://unicode.org/reports/tr15/
Okay, what if we have
- * The pattern is converted to normalized form and then a pure group
- * is constructed to match canonical equivalences of the characters.
+ * The pattern is converted to normalized form (NFD, canonical
+ * decomposition) and then a pure group is constructed to match
+ * canonical equivalences of the characters.
+ * @see java.text.Normalizer.Form.NFD
We've already got a link to
http://www.unicode.org/unicode/reports/tr15/tr15-23.html from the doc
I'd rather not duplicate the reference in that comment.
> On Mon, Jul 13, 2015 at 4:29 AM, Ivan Gerasimov
> <ivan.gerasimov at oracle.com <mailto:ivan.gerasimov at oracle.com>> wrote:
> Thank you Paul for your review!
> On 13.07.2015 12:57, Paul Sandoz wrote:
> On Jul 12, 2015, at 6:32 PM, Ivan Gerasimov
> <ivan.gerasimov at oracle.com <mailto:ivan.gerasimov at oracle.com>>
> There's a minor issue in the current implementation of
> Pattern.quote() with possible numeric overflow when
> calculating initial capacity of StringBuilder.
> With the fix, some slight optimizations were done and a
> few typos were fixed.
> Would you please help review the fix?
> BUGURL: https://bugs.openjdk.java.net/browse/JDK-8131034
> Looks ok.
> - * The pattern is converted to normalizedD form and then
> a pure group
> + * The pattern is converted to normalized form and then a
> pure group
> * is constructed to match canonical equivalences of the
> The "D" in "normalizedD form" may well be short hand for
> "canonical decomposition" (Normalizer.Form.NFD) as in
> "normalized canonical decomposition form".
> Hm. Still looks like a typo to me.
> A few lines below normalizedPattern has no extra D.
> Another normalization form, NFKD, also ends with D, so normalizedD
> wouldn't uniquely identify NFD.
> grep didn't find other occurrences of normalizedD abbreviation in
> JDK :)
> Sincerely yours,
> Sincerely yours,
More information about the core-libs-dev