Re: [classlib][regex] Is this a issue of java.util.regex.Matcher.region(int, int);
Hi Nikolay, I will look to commit 713 later on today. Been a bit busy on other things. Best regards, George Nikolay Kuznetsov wrote: Hello Spark Shen, Thank you very much, do I need to raise a new JIRA for this issue? Using the test case I give, No, for this problem we already have couple of issues in JIRA: http://issues.apache.org/jira/browse/HARMONY-713 http://issues.apache.org/jira/browse/HARMONY-759 I've updated first one with the fix, and asked to close second one as duplicate. You may want to try suggested fix with your example and/or suggest you testcase as a patch to unit tests(You'll find Richard's test over there). Also, thank you very much for your help with debugging this issue, but the guess was not completely correct. SecuenceSet is a part of the pattern compiled(Pattern.compile("c")), and it's not affected by the regions, at least not it's content, which should always remain "c". region() method affects underlying match result, which SequenceSet should communicate doing find, but it didn't, that was a problem. Thank you. Nik. - Terms of use : http://incubator.apache.org/harmony/mailing.html To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED] - Terms of use : http://incubator.apache.org/harmony/mailing.html To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [classlib][regex] Is this a issue of java.util.regex.Matcher.region(int, int);
Hello Spark Shen, Thank you very much, do I need to raise a new JIRA for this issue? Using the test case I give, No, for this problem we already have couple of issues in JIRA: http://issues.apache.org/jira/browse/HARMONY-713 http://issues.apache.org/jira/browse/HARMONY-759 I've updated first one with the fix, and asked to close second one as duplicate. You may want to try suggested fix with your example and/or suggest you testcase as a patch to unit tests(You'll find Richard's test over there). Also, thank you very much for your help with debugging this issue, but the guess was not completely correct. SecuenceSet is a part of the pattern compiled(Pattern.compile("c")), and it's not affected by the regions, at least not it's content, which should always remain "c". region() method affects underlying match result, which SequenceSet should communicate doing find, but it didn't, that was a problem. Thank you. Nik. - Terms of use : http://incubator.apache.org/harmony/mailing.html To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [classlib][regex] Is this a issue of java.util.regex.Matcher.region(int, int);
Nikolay Kuznetsov 写道: Hello Spark Shen, It seems that Matcher.region(int, int) will not have any effects when called. This can be demonstrated by the following test case there is an issue quite similar to one you have raised filled by Richard: http://issues.apache.org/jira/browse/HARMONY-759 I'll try to provide a fix today, thanks for reminding. BTW, there are other issues about java.util.regex. If necessary, can we disscuss all these issues case by case. Surely, we can discuss issues if needed, but if smth just does not work, please fill an issue in JIRA, I'm looking after regex issues. If you need some problems to fix urgently, let me know separately. Thank you very much, do I need to raise a new JIRA for this issue? Using the test case I give, I debug into the code. The calling stack is as follow: SequenceSet.indexOf(CharSequence, int) SequenceSet.find(int, CharSequence, MatchResultImpl) line: 69 SingleSet.find(int, CharSequence, MatchResultImpl) line: 52 Matcher.findAt(int) line: 317 Matcher.find(int) line: 304 Matcher.find() line: 342 in indexOf line 115 : int last = string.charAt(charCount - 1); the value of string is "c". I figure out this string variable is a member of SequenceSet. Since the region in my test case setting is (0, 2), I am not sure whether the string value here should be smth like "a". Since SequenceSet is construct when Pattern.complile() and this members value of the constructed SequenceSet seems never change ever since, does it make sense that when region is called, value of SequeceSet should also be updated? Hope this guess would be helpful. :-) Correct me if I am wrong. Best regards Yours sincerely, Nik. - Terms of use : http://incubator.apache.org/harmony/mailing.html To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED] -- Spark Shen China Software Development Lab, IBM - Terms of use : http://incubator.apache.org/harmony/mailing.html To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [classlib][regex] Is this a issue of java.util.regex.Matcher.region(int, int);
Hello Spark Shen, It seems that Matcher.region(int, int) will not have any effects when called. This can be demonstrated by the following test case there is an issue quite similar to one you have raised filled by Richard: http://issues.apache.org/jira/browse/HARMONY-759 I'll try to provide a fix today, thanks for reminding. BTW, there are other issues about java.util.regex. If necessary, can we disscuss all these issues case by case. Surely, we can discuss issues if needed, but if smth just does not work, please fill an issue in JIRA, I'm looking after regex issues. If you need some problems to fix urgently, let me know separately. Yours sincerely, Nik. - Terms of use : http://incubator.apache.org/harmony/mailing.html To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [classlib][regex] Is this a issue of java.util.regex.Matcher.region(int, int);
Spark Shen 写道: Hi All: It seems that Matcher.region(int, int) will not have any effects when called. This can be demonstrated by the following test case: public void test_RegionII() { CharBuffer buffer = CharBuffer.wrap(new char[] { 'a', 'a', 'b', 'b', '.', 'c' }); Matcher matcher = Pattern.compile("c").matcher(buffer); //$NON-NLS-1$ matcher.region(0, 2); assertFalse(matcher.find()); } This test case pass on RI, but fail on Harmony. Can I infer that, harmony ignores region settting when trying to find the specified pattern? Correct me if I am wrong. Best regards BTW, there are other issues about java.util.regex. If necessary, can we disscuss all these issues case by case. -- Spark Shen China Software Development Lab, IBM - Terms of use : http://incubator.apache.org/harmony/mailing.html To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]