[jira] [Commented] (OAK-5506) reject item names with unpaired surrogates early
[ https://issues.apache.org/jira/browse/OAK-5506?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16389774#comment-16389774 ] Michael Dürig commented on OAK-5506: [~reschke], I just noticed that we are also using {{java.nio.charset.Charset#decode}} and {{java.nio.charset.Charset#encode}} in various places. Not sure if this is covered with your patches. > reject item names with unpaired surrogates early > > > Key: OAK-5506 > URL: https://issues.apache.org/jira/browse/OAK-5506 > Project: Jackrabbit Oak > Issue Type: Wish > Components: core, jcr, segment-tar >Affects Versions: 1.5.18 >Reporter: Julian Reschke >Priority: Minor > Labels: resilience > Fix For: 1.10 > > Attachments: OAK-5506-01.patch, OAK-5506-02.patch, OAK-5506-4.diff, > OAK-5506-bench.diff, OAK-5506-jcr-level.diff, OAK-5506-name-conversion.diff, > OAK-5506-segment.diff, OAK-5506-segment2.diff, OAK-5506-segment3.diff, > OAK-5506.diff, ValidNamesTest.java > > > Apparently, the following node name is accepted: >{{"foo\ud800"}} > but a subsequent {{getPath()}} call fails: > {noformat} > javax.jcr.InvalidItemStateException: This item [/test_node/foo?] does not > exist anymore > at > org.apache.jackrabbit.oak.jcr.delegate.ItemDelegate.checkAlive(ItemDelegate.java:86) > at > org.apache.jackrabbit.oak.jcr.session.operation.ItemOperation.checkPreconditions(ItemOperation.java:34) > at > org.apache.jackrabbit.oak.jcr.delegate.SessionDelegate.prePerform(SessionDelegate.java:615) > at > org.apache.jackrabbit.oak.jcr.delegate.SessionDelegate.perform(SessionDelegate.java:205) > at > org.apache.jackrabbit.oak.jcr.session.ItemImpl.perform(ItemImpl.java:112) > at > org.apache.jackrabbit.oak.jcr.session.ItemImpl.getPath(ItemImpl.java:140) > at > org.apache.jackrabbit.oak.jcr.session.NodeImpl.getPath(NodeImpl.java:106) > at > org.apache.jackrabbit.oak.jcr.ValidNamesTest.nameTest(ValidNamesTest.java:271) > at > org.apache.jackrabbit.oak.jcr.ValidNamesTest.testUnpairedSurrogate(ValidNamesTest.java:259) > at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method) > at sun.reflect.NativeMethodAccessorImpl.invoke(Unknown Source){noformat} > (test case follows) -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Commented] (OAK-5506) reject item names with unpaired surrogates early
[ https://issues.apache.org/jira/browse/OAK-5506?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16364080#comment-16364080 ] Julian Reschke commented on OAK-5506: - {quote}What about rejecting invalid strings (at least for paths) in the JCR layer, specially when creating / moving nodes? {quote} That's what one of the attached patches tries to do... > reject item names with unpaired surrogates early > > > Key: OAK-5506 > URL: https://issues.apache.org/jira/browse/OAK-5506 > Project: Jackrabbit Oak > Issue Type: Wish > Components: core, jcr, segment-tar >Affects Versions: 1.5.18 >Reporter: Julian Reschke >Priority: Minor > Fix For: 1.10 > > Attachments: OAK-5506-01.patch, OAK-5506-02.patch, OAK-5506-4.diff, > OAK-5506-bench.diff, OAK-5506-jcr-level.diff, OAK-5506-name-conversion.diff, > OAK-5506-segment.diff, OAK-5506-segment2.diff, OAK-5506-segment3.diff, > OAK-5506.diff, ValidNamesTest.java > > > Apparently, the following node name is accepted: >{{"foo\ud800"}} > but a subsequent {{getPath()}} call fails: > {noformat} > javax.jcr.InvalidItemStateException: This item [/test_node/foo?] does not > exist anymore > at > org.apache.jackrabbit.oak.jcr.delegate.ItemDelegate.checkAlive(ItemDelegate.java:86) > at > org.apache.jackrabbit.oak.jcr.session.operation.ItemOperation.checkPreconditions(ItemOperation.java:34) > at > org.apache.jackrabbit.oak.jcr.delegate.SessionDelegate.prePerform(SessionDelegate.java:615) > at > org.apache.jackrabbit.oak.jcr.delegate.SessionDelegate.perform(SessionDelegate.java:205) > at > org.apache.jackrabbit.oak.jcr.session.ItemImpl.perform(ItemImpl.java:112) > at > org.apache.jackrabbit.oak.jcr.session.ItemImpl.getPath(ItemImpl.java:140) > at > org.apache.jackrabbit.oak.jcr.session.NodeImpl.getPath(NodeImpl.java:106) > at > org.apache.jackrabbit.oak.jcr.ValidNamesTest.nameTest(ValidNamesTest.java:271) > at > org.apache.jackrabbit.oak.jcr.ValidNamesTest.testUnpairedSurrogate(ValidNamesTest.java:259) > at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method) > at sun.reflect.NativeMethodAccessorImpl.invoke(Unknown Source){noformat} > (test case follows) -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Commented] (OAK-5506) reject item names with unpaired surrogates early
[ https://issues.apache.org/jira/browse/OAK-5506?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16364070#comment-16364070 ] Thomas Mueller commented on OAK-5506: - What about rejecting invalid strings (at least for paths) in the JCR layer, specially when creating / moving nodes? > reject item names with unpaired surrogates early > > > Key: OAK-5506 > URL: https://issues.apache.org/jira/browse/OAK-5506 > Project: Jackrabbit Oak > Issue Type: Wish > Components: core, jcr, segment-tar >Affects Versions: 1.5.18 >Reporter: Julian Reschke >Priority: Minor > Fix For: 1.10 > > Attachments: OAK-5506-01.patch, OAK-5506-02.patch, OAK-5506-4.diff, > OAK-5506-bench.diff, OAK-5506-jcr-level.diff, OAK-5506-name-conversion.diff, > OAK-5506-segment.diff, OAK-5506-segment2.diff, OAK-5506-segment3.diff, > OAK-5506.diff, ValidNamesTest.java > > > Apparently, the following node name is accepted: >{{"foo\ud800"}} > but a subsequent {{getPath()}} call fails: > {noformat} > javax.jcr.InvalidItemStateException: This item [/test_node/foo?] does not > exist anymore > at > org.apache.jackrabbit.oak.jcr.delegate.ItemDelegate.checkAlive(ItemDelegate.java:86) > at > org.apache.jackrabbit.oak.jcr.session.operation.ItemOperation.checkPreconditions(ItemOperation.java:34) > at > org.apache.jackrabbit.oak.jcr.delegate.SessionDelegate.prePerform(SessionDelegate.java:615) > at > org.apache.jackrabbit.oak.jcr.delegate.SessionDelegate.perform(SessionDelegate.java:205) > at > org.apache.jackrabbit.oak.jcr.session.ItemImpl.perform(ItemImpl.java:112) > at > org.apache.jackrabbit.oak.jcr.session.ItemImpl.getPath(ItemImpl.java:140) > at > org.apache.jackrabbit.oak.jcr.session.NodeImpl.getPath(NodeImpl.java:106) > at > org.apache.jackrabbit.oak.jcr.ValidNamesTest.nameTest(ValidNamesTest.java:271) > at > org.apache.jackrabbit.oak.jcr.ValidNamesTest.testUnpairedSurrogate(ValidNamesTest.java:259) > at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method) > at sun.reflect.NativeMethodAccessorImpl.invoke(Unknown Source){noformat} > (test case follows) -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Commented] (OAK-5506) reject item names with unpaired surrogates early
[ https://issues.apache.org/jira/browse/OAK-5506?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16360873#comment-16360873 ] Julian Reschke commented on OAK-5506: - FWIW, I just noticed that the various (R)DBs vary in their behavior: - H2DB and Derby roundtrip any string - PostgreSQL rejects the invalid string early - DB2 and Oracle fail the same way as segment store (they persist the replacement character) - MySQL and SQLServer fail the same as DB2 and Oracle, but here it's the RDBDocumentStore's fault, because the ID column is binary, and we transform to byte sequences ourselves ...will file individual tickets > reject item names with unpaired surrogates early > > > Key: OAK-5506 > URL: https://issues.apache.org/jira/browse/OAK-5506 > Project: Jackrabbit Oak > Issue Type: Wish > Components: core, jcr, segment-tar >Affects Versions: 1.5.18 >Reporter: Julian Reschke >Priority: Minor > Fix For: 1.10 > > Attachments: OAK-5506-01.patch, OAK-5506-02.patch, OAK-5506-4.diff, > OAK-5506-bench.diff, OAK-5506-jcr-level.diff, OAK-5506-name-conversion.diff, > OAK-5506-segment.diff, OAK-5506-segment2.diff, OAK-5506-segment3.diff, > OAK-5506.diff, ValidNamesTest.java > > > Apparently, the following node name is accepted: >{{"foo\ud800"}} > but a subsequent {{getPath()}} call fails: > {noformat} > javax.jcr.InvalidItemStateException: This item [/test_node/foo?] does not > exist anymore > at > org.apache.jackrabbit.oak.jcr.delegate.ItemDelegate.checkAlive(ItemDelegate.java:86) > at > org.apache.jackrabbit.oak.jcr.session.operation.ItemOperation.checkPreconditions(ItemOperation.java:34) > at > org.apache.jackrabbit.oak.jcr.delegate.SessionDelegate.prePerform(SessionDelegate.java:615) > at > org.apache.jackrabbit.oak.jcr.delegate.SessionDelegate.perform(SessionDelegate.java:205) > at > org.apache.jackrabbit.oak.jcr.session.ItemImpl.perform(ItemImpl.java:112) > at > org.apache.jackrabbit.oak.jcr.session.ItemImpl.getPath(ItemImpl.java:140) > at > org.apache.jackrabbit.oak.jcr.session.NodeImpl.getPath(NodeImpl.java:106) > at > org.apache.jackrabbit.oak.jcr.ValidNamesTest.nameTest(ValidNamesTest.java:271) > at > org.apache.jackrabbit.oak.jcr.ValidNamesTest.testUnpairedSurrogate(ValidNamesTest.java:259) > at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method) > at sun.reflect.NativeMethodAccessorImpl.invoke(Unknown Source){noformat} > (test case follows) -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Commented] (OAK-5506) reject item names with unpaired surrogates early
[ https://issues.apache.org/jira/browse/OAK-5506?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16358016#comment-16358016 ] Julian Reschke commented on OAK-5506: - https://issues.apache.org/jira/secure/attachment/12909901/OAK-5506-jcr-level.diff has JCR-level changes for rejecting the characters. > reject item names with unpaired surrogates early > > > Key: OAK-5506 > URL: https://issues.apache.org/jira/browse/OAK-5506 > Project: Jackrabbit Oak > Issue Type: Wish > Components: core, jcr, segment-tar >Affects Versions: 1.5.18 >Reporter: Julian Reschke >Priority: Minor > Fix For: 1.10 > > Attachments: OAK-5506-01.patch, OAK-5506-02.patch, OAK-5506-4.diff, > OAK-5506-bench.diff, OAK-5506-jcr-level.diff, OAK-5506-name-conversion.diff, > OAK-5506-segment.diff, OAK-5506-segment2.diff, OAK-5506-segment3.diff, > OAK-5506.diff, ValidNamesTest.java > > > Apparently, the following node name is accepted: >{{"foo\ud800"}} > but a subsequent {{getPath()}} call fails: > {noformat} > javax.jcr.InvalidItemStateException: This item [/test_node/foo?] does not > exist anymore > at > org.apache.jackrabbit.oak.jcr.delegate.ItemDelegate.checkAlive(ItemDelegate.java:86) > at > org.apache.jackrabbit.oak.jcr.session.operation.ItemOperation.checkPreconditions(ItemOperation.java:34) > at > org.apache.jackrabbit.oak.jcr.delegate.SessionDelegate.prePerform(SessionDelegate.java:615) > at > org.apache.jackrabbit.oak.jcr.delegate.SessionDelegate.perform(SessionDelegate.java:205) > at > org.apache.jackrabbit.oak.jcr.session.ItemImpl.perform(ItemImpl.java:112) > at > org.apache.jackrabbit.oak.jcr.session.ItemImpl.getPath(ItemImpl.java:140) > at > org.apache.jackrabbit.oak.jcr.session.NodeImpl.getPath(NodeImpl.java:106) > at > org.apache.jackrabbit.oak.jcr.ValidNamesTest.nameTest(ValidNamesTest.java:271) > at > org.apache.jackrabbit.oak.jcr.ValidNamesTest.testUnpairedSurrogate(ValidNamesTest.java:259) > at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method) > at sun.reflect.NativeMethodAccessorImpl.invoke(Unknown Source){noformat} > (test case follows) -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Commented] (OAK-5506) reject item names with unpaired surrogates early
[ https://issues.apache.org/jira/browse/OAK-5506?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16354109#comment-16354109 ] Julian Reschke commented on OAK-5506: - Created https://issues.apache.org/jira/browse/OAK-7249 to track addition of utility class and related tests. > reject item names with unpaired surrogates early > > > Key: OAK-5506 > URL: https://issues.apache.org/jira/browse/OAK-5506 > Project: Jackrabbit Oak > Issue Type: Wish > Components: core, jcr, segment-tar >Affects Versions: 1.5.18 >Reporter: Julian Reschke >Priority: Minor > Fix For: 1.10 > > Attachments: OAK-5506-01.patch, OAK-5506-02.patch, OAK-5506-4.diff, > OAK-5506-bench.diff, OAK-5506-name-conversion.diff, OAK-5506-segment.diff, > OAK-5506-segment2.diff, OAK-5506-segment3.diff, OAK-5506.diff, > ValidNamesTest.java > > > Apparently, the following node name is accepted: >{{"foo\ud800"}} > but a subsequent {{getPath()}} call fails: > {noformat} > javax.jcr.InvalidItemStateException: This item [/test_node/foo?] does not > exist anymore > at > org.apache.jackrabbit.oak.jcr.delegate.ItemDelegate.checkAlive(ItemDelegate.java:86) > at > org.apache.jackrabbit.oak.jcr.session.operation.ItemOperation.checkPreconditions(ItemOperation.java:34) > at > org.apache.jackrabbit.oak.jcr.delegate.SessionDelegate.prePerform(SessionDelegate.java:615) > at > org.apache.jackrabbit.oak.jcr.delegate.SessionDelegate.perform(SessionDelegate.java:205) > at > org.apache.jackrabbit.oak.jcr.session.ItemImpl.perform(ItemImpl.java:112) > at > org.apache.jackrabbit.oak.jcr.session.ItemImpl.getPath(ItemImpl.java:140) > at > org.apache.jackrabbit.oak.jcr.session.NodeImpl.getPath(NodeImpl.java:106) > at > org.apache.jackrabbit.oak.jcr.ValidNamesTest.nameTest(ValidNamesTest.java:271) > at > org.apache.jackrabbit.oak.jcr.ValidNamesTest.testUnpairedSurrogate(ValidNamesTest.java:259) > at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method) > at sun.reflect.NativeMethodAccessorImpl.invoke(Unknown Source){noformat} > (test case follows) -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Commented] (OAK-5506) reject item names with unpaired surrogates early
[ https://issues.apache.org/jira/browse/OAK-5506?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16353806#comment-16353806 ] Michael Dürig commented on OAK-5506: re. b), let's leave it like it is for now and figure it out while writing the migration related test. > reject item names with unpaired surrogates early > > > Key: OAK-5506 > URL: https://issues.apache.org/jira/browse/OAK-5506 > Project: Jackrabbit Oak > Issue Type: Wish > Components: core, jcr, segment-tar >Affects Versions: 1.5.18 >Reporter: Julian Reschke >Priority: Minor > Fix For: 1.10 > > Attachments: OAK-5506-01.patch, OAK-5506-02.patch, OAK-5506-4.diff, > OAK-5506-bench.diff, OAK-5506-name-conversion.diff, OAK-5506-segment.diff, > OAK-5506-segment2.diff, OAK-5506-segment3.diff, OAK-5506.diff, > ValidNamesTest.java > > > Apparently, the following node name is accepted: >{{"foo\ud800"}} > but a subsequent {{getPath()}} call fails: > {noformat} > javax.jcr.InvalidItemStateException: This item [/test_node/foo?] does not > exist anymore > at > org.apache.jackrabbit.oak.jcr.delegate.ItemDelegate.checkAlive(ItemDelegate.java:86) > at > org.apache.jackrabbit.oak.jcr.session.operation.ItemOperation.checkPreconditions(ItemOperation.java:34) > at > org.apache.jackrabbit.oak.jcr.delegate.SessionDelegate.prePerform(SessionDelegate.java:615) > at > org.apache.jackrabbit.oak.jcr.delegate.SessionDelegate.perform(SessionDelegate.java:205) > at > org.apache.jackrabbit.oak.jcr.session.ItemImpl.perform(ItemImpl.java:112) > at > org.apache.jackrabbit.oak.jcr.session.ItemImpl.getPath(ItemImpl.java:140) > at > org.apache.jackrabbit.oak.jcr.session.NodeImpl.getPath(NodeImpl.java:106) > at > org.apache.jackrabbit.oak.jcr.ValidNamesTest.nameTest(ValidNamesTest.java:271) > at > org.apache.jackrabbit.oak.jcr.ValidNamesTest.testUnpairedSurrogate(ValidNamesTest.java:259) > at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method) > at sun.reflect.NativeMethodAccessorImpl.invoke(Unknown Source){noformat} > (test case follows) -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Commented] (OAK-5506) reject item names with unpaired surrogates early
[ https://issues.apache.org/jira/browse/OAK-5506?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16353763#comment-16353763 ] Julian Reschke commented on OAK-5506: - a I can take care of the unit test. b) Not sure whether the utils should be configurable - caller will need to be aware of it anyway, right? Feedback appreciated. > reject item names with unpaired surrogates early > > > Key: OAK-5506 > URL: https://issues.apache.org/jira/browse/OAK-5506 > Project: Jackrabbit Oak > Issue Type: Wish > Components: core, jcr, segment-tar >Affects Versions: 1.5.18 >Reporter: Julian Reschke >Priority: Minor > Fix For: 1.10 > > Attachments: OAK-5506-01.patch, OAK-5506-02.patch, OAK-5506-4.diff, > OAK-5506-bench.diff, OAK-5506-name-conversion.diff, OAK-5506-segment.diff, > OAK-5506-segment2.diff, OAK-5506-segment3.diff, OAK-5506.diff, > ValidNamesTest.java > > > Apparently, the following node name is accepted: >{{"foo\ud800"}} > but a subsequent {{getPath()}} call fails: > {noformat} > javax.jcr.InvalidItemStateException: This item [/test_node/foo?] does not > exist anymore > at > org.apache.jackrabbit.oak.jcr.delegate.ItemDelegate.checkAlive(ItemDelegate.java:86) > at > org.apache.jackrabbit.oak.jcr.session.operation.ItemOperation.checkPreconditions(ItemOperation.java:34) > at > org.apache.jackrabbit.oak.jcr.delegate.SessionDelegate.prePerform(SessionDelegate.java:615) > at > org.apache.jackrabbit.oak.jcr.delegate.SessionDelegate.perform(SessionDelegate.java:205) > at > org.apache.jackrabbit.oak.jcr.session.ItemImpl.perform(ItemImpl.java:112) > at > org.apache.jackrabbit.oak.jcr.session.ItemImpl.getPath(ItemImpl.java:140) > at > org.apache.jackrabbit.oak.jcr.session.NodeImpl.getPath(NodeImpl.java:106) > at > org.apache.jackrabbit.oak.jcr.ValidNamesTest.nameTest(ValidNamesTest.java:271) > at > org.apache.jackrabbit.oak.jcr.ValidNamesTest.testUnpairedSurrogate(ValidNamesTest.java:259) > at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method) > at sun.reflect.NativeMethodAccessorImpl.invoke(Unknown Source){noformat} > (test case follows) -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Commented] (OAK-5506) reject item names with unpaired surrogates early
[ https://issues.apache.org/jira/browse/OAK-5506?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16353759#comment-16353759 ] Francesco Mari commented on OAK-5506: - [~mduerig], it sounds reasonable to me. > reject item names with unpaired surrogates early > > > Key: OAK-5506 > URL: https://issues.apache.org/jira/browse/OAK-5506 > Project: Jackrabbit Oak > Issue Type: Wish > Components: core, jcr, segment-tar >Affects Versions: 1.5.18 >Reporter: Julian Reschke >Priority: Minor > Fix For: 1.10 > > Attachments: OAK-5506-01.patch, OAK-5506-02.patch, OAK-5506-4.diff, > OAK-5506-bench.diff, OAK-5506-name-conversion.diff, OAK-5506-segment.diff, > OAK-5506-segment2.diff, OAK-5506-segment3.diff, OAK-5506.diff, > ValidNamesTest.java > > > Apparently, the following node name is accepted: >{{"foo\ud800"}} > but a subsequent {{getPath()}} call fails: > {noformat} > javax.jcr.InvalidItemStateException: This item [/test_node/foo?] does not > exist anymore > at > org.apache.jackrabbit.oak.jcr.delegate.ItemDelegate.checkAlive(ItemDelegate.java:86) > at > org.apache.jackrabbit.oak.jcr.session.operation.ItemOperation.checkPreconditions(ItemOperation.java:34) > at > org.apache.jackrabbit.oak.jcr.delegate.SessionDelegate.prePerform(SessionDelegate.java:615) > at > org.apache.jackrabbit.oak.jcr.delegate.SessionDelegate.perform(SessionDelegate.java:205) > at > org.apache.jackrabbit.oak.jcr.session.ItemImpl.perform(ItemImpl.java:112) > at > org.apache.jackrabbit.oak.jcr.session.ItemImpl.getPath(ItemImpl.java:140) > at > org.apache.jackrabbit.oak.jcr.session.NodeImpl.getPath(NodeImpl.java:106) > at > org.apache.jackrabbit.oak.jcr.ValidNamesTest.nameTest(ValidNamesTest.java:271) > at > org.apache.jackrabbit.oak.jcr.ValidNamesTest.testUnpairedSurrogate(ValidNamesTest.java:259) > at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method) > at sun.reflect.NativeMethodAccessorImpl.invoke(Unknown Source){noformat} > (test case follows) -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Commented] (OAK-5506) reject item names with unpaired surrogates early
[ https://issues.apache.org/jira/browse/OAK-5506?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16353754#comment-16353754 ] Michael Dürig commented on OAK-5506: +1 for the patch from my side provided we follow up with * A unit test for {{CharsetEncodingUtils}} ([~reschke], can you take care of this?) * A way to fall back to the previous behaviour for tests and via feature flag ([~reschke], can {{CharsetEncodingUtils}} be parametrized to that respect or would it be better to leave this to the caller side?) * A migration test ensuring repositories written before these change can still be safely read with this patch (I can take this up). [~frm] WDYT? > reject item names with unpaired surrogates early > > > Key: OAK-5506 > URL: https://issues.apache.org/jira/browse/OAK-5506 > Project: Jackrabbit Oak > Issue Type: Wish > Components: core, jcr, segment-tar >Affects Versions: 1.5.18 >Reporter: Julian Reschke >Priority: Minor > Fix For: 1.10 > > Attachments: OAK-5506-01.patch, OAK-5506-02.patch, OAK-5506-4.diff, > OAK-5506-bench.diff, OAK-5506-name-conversion.diff, OAK-5506-segment.diff, > OAK-5506-segment2.diff, OAK-5506-segment3.diff, OAK-5506.diff, > ValidNamesTest.java > > > Apparently, the following node name is accepted: >{{"foo\ud800"}} > but a subsequent {{getPath()}} call fails: > {noformat} > javax.jcr.InvalidItemStateException: This item [/test_node/foo?] does not > exist anymore > at > org.apache.jackrabbit.oak.jcr.delegate.ItemDelegate.checkAlive(ItemDelegate.java:86) > at > org.apache.jackrabbit.oak.jcr.session.operation.ItemOperation.checkPreconditions(ItemOperation.java:34) > at > org.apache.jackrabbit.oak.jcr.delegate.SessionDelegate.prePerform(SessionDelegate.java:615) > at > org.apache.jackrabbit.oak.jcr.delegate.SessionDelegate.perform(SessionDelegate.java:205) > at > org.apache.jackrabbit.oak.jcr.session.ItemImpl.perform(ItemImpl.java:112) > at > org.apache.jackrabbit.oak.jcr.session.ItemImpl.getPath(ItemImpl.java:140) > at > org.apache.jackrabbit.oak.jcr.session.NodeImpl.getPath(NodeImpl.java:106) > at > org.apache.jackrabbit.oak.jcr.ValidNamesTest.nameTest(ValidNamesTest.java:271) > at > org.apache.jackrabbit.oak.jcr.ValidNamesTest.testUnpairedSurrogate(ValidNamesTest.java:259) > at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method) > at sun.reflect.NativeMethodAccessorImpl.invoke(Unknown Source){noformat} > (test case follows) -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Commented] (OAK-5506) reject item names with unpaired surrogates early
[ https://issues.apache.org/jira/browse/OAK-5506?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16352457#comment-16352457 ] Julian Reschke commented on OAK-5506: - Moved the utility class into segment-tar for now: https://issues.apache.org/jira/secure/attachment/12909233/OAK-5506-segment3.diff > reject item names with unpaired surrogates early > > > Key: OAK-5506 > URL: https://issues.apache.org/jira/browse/OAK-5506 > Project: Jackrabbit Oak > Issue Type: Wish > Components: core, jcr, segment-tar >Affects Versions: 1.5.18 >Reporter: Julian Reschke >Priority: Minor > Fix For: 1.10 > > Attachments: OAK-5506-01.patch, OAK-5506-02.patch, OAK-5506-4.diff, > OAK-5506-bench.diff, OAK-5506-name-conversion.diff, OAK-5506-segment.diff, > OAK-5506-segment2.diff, OAK-5506-segment3.diff, OAK-5506.diff, > ValidNamesTest.java > > > Apparently, the following node name is accepted: >{{"foo\ud800"}} > but a subsequent {{getPath()}} call fails: > {noformat} > javax.jcr.InvalidItemStateException: This item [/test_node/foo?] does not > exist anymore > at > org.apache.jackrabbit.oak.jcr.delegate.ItemDelegate.checkAlive(ItemDelegate.java:86) > at > org.apache.jackrabbit.oak.jcr.session.operation.ItemOperation.checkPreconditions(ItemOperation.java:34) > at > org.apache.jackrabbit.oak.jcr.delegate.SessionDelegate.prePerform(SessionDelegate.java:615) > at > org.apache.jackrabbit.oak.jcr.delegate.SessionDelegate.perform(SessionDelegate.java:205) > at > org.apache.jackrabbit.oak.jcr.session.ItemImpl.perform(ItemImpl.java:112) > at > org.apache.jackrabbit.oak.jcr.session.ItemImpl.getPath(ItemImpl.java:140) > at > org.apache.jackrabbit.oak.jcr.session.NodeImpl.getPath(NodeImpl.java:106) > at > org.apache.jackrabbit.oak.jcr.ValidNamesTest.nameTest(ValidNamesTest.java:271) > at > org.apache.jackrabbit.oak.jcr.ValidNamesTest.testUnpairedSurrogate(ValidNamesTest.java:259) > at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method) > at sun.reflect.NativeMethodAccessorImpl.invoke(Unknown Source){noformat} > (test case follows) -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Commented] (OAK-5506) reject item names with unpaired surrogates early
[ https://issues.apache.org/jira/browse/OAK-5506?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16350590#comment-16350590 ] Julian Reschke commented on OAK-5506: - OK. One thing I'd like to change is to move the utility class into segment-tar for now. Let's get this working there first - later we can think of the right API for use in other OAK modules... > reject item names with unpaired surrogates early > > > Key: OAK-5506 > URL: https://issues.apache.org/jira/browse/OAK-5506 > Project: Jackrabbit Oak > Issue Type: Wish > Components: core, jcr, segment-tar >Affects Versions: 1.5.18 >Reporter: Julian Reschke >Priority: Minor > Fix For: 1.10 > > Attachments: OAK-5506-01.patch, OAK-5506-02.patch, OAK-5506-4.diff, > OAK-5506-bench.diff, OAK-5506-name-conversion.diff, OAK-5506-segment.diff, > OAK-5506-segment2.diff, OAK-5506.diff, ValidNamesTest.java > > > Apparently, the following node name is accepted: >{{"foo\ud800"}} > but a subsequent {{getPath()}} call fails: > {noformat} > javax.jcr.InvalidItemStateException: This item [/test_node/foo?] does not > exist anymore > at > org.apache.jackrabbit.oak.jcr.delegate.ItemDelegate.checkAlive(ItemDelegate.java:86) > at > org.apache.jackrabbit.oak.jcr.session.operation.ItemOperation.checkPreconditions(ItemOperation.java:34) > at > org.apache.jackrabbit.oak.jcr.delegate.SessionDelegate.prePerform(SessionDelegate.java:615) > at > org.apache.jackrabbit.oak.jcr.delegate.SessionDelegate.perform(SessionDelegate.java:205) > at > org.apache.jackrabbit.oak.jcr.session.ItemImpl.perform(ItemImpl.java:112) > at > org.apache.jackrabbit.oak.jcr.session.ItemImpl.getPath(ItemImpl.java:140) > at > org.apache.jackrabbit.oak.jcr.session.NodeImpl.getPath(NodeImpl.java:106) > at > org.apache.jackrabbit.oak.jcr.ValidNamesTest.nameTest(ValidNamesTest.java:271) > at > org.apache.jackrabbit.oak.jcr.ValidNamesTest.testUnpairedSurrogate(ValidNamesTest.java:259) > at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method) > at sun.reflect.NativeMethodAccessorImpl.invoke(Unknown Source){noformat} > (test case follows) -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Commented] (OAK-5506) reject item names with unpaired surrogates early
[ https://issues.apache.org/jira/browse/OAK-5506?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16350535#comment-16350535 ] Michael Dürig commented on OAK-5506: I can take over. Optimally after you committed the first bits from your end. > reject item names with unpaired surrogates early > > > Key: OAK-5506 > URL: https://issues.apache.org/jira/browse/OAK-5506 > Project: Jackrabbit Oak > Issue Type: Wish > Components: core, jcr, segment-tar >Affects Versions: 1.5.18 >Reporter: Julian Reschke >Priority: Minor > Fix For: 1.10 > > Attachments: OAK-5506-01.patch, OAK-5506-02.patch, OAK-5506-4.diff, > OAK-5506-bench.diff, OAK-5506-name-conversion.diff, OAK-5506-segment.diff, > OAK-5506-segment2.diff, OAK-5506.diff, ValidNamesTest.java > > > Apparently, the following node name is accepted: >{{"foo\ud800"}} > but a subsequent {{getPath()}} call fails: > {noformat} > javax.jcr.InvalidItemStateException: This item [/test_node/foo?] does not > exist anymore > at > org.apache.jackrabbit.oak.jcr.delegate.ItemDelegate.checkAlive(ItemDelegate.java:86) > at > org.apache.jackrabbit.oak.jcr.session.operation.ItemOperation.checkPreconditions(ItemOperation.java:34) > at > org.apache.jackrabbit.oak.jcr.delegate.SessionDelegate.prePerform(SessionDelegate.java:615) > at > org.apache.jackrabbit.oak.jcr.delegate.SessionDelegate.perform(SessionDelegate.java:205) > at > org.apache.jackrabbit.oak.jcr.session.ItemImpl.perform(ItemImpl.java:112) > at > org.apache.jackrabbit.oak.jcr.session.ItemImpl.getPath(ItemImpl.java:140) > at > org.apache.jackrabbit.oak.jcr.session.NodeImpl.getPath(NodeImpl.java:106) > at > org.apache.jackrabbit.oak.jcr.ValidNamesTest.nameTest(ValidNamesTest.java:271) > at > org.apache.jackrabbit.oak.jcr.ValidNamesTest.testUnpairedSurrogate(ValidNamesTest.java:259) > at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method) > at sun.reflect.NativeMethodAccessorImpl.invoke(Unknown Source){noformat} > (test case follows) -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Commented] (OAK-5506) reject item names with unpaired surrogates early
[ https://issues.apache.org/jira/browse/OAK-5506?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16350518#comment-16350518 ] Julian Reschke commented on OAK-5506: - [~mduerig] - can you point me to an example test class, from where I could steal? Or do you want to take over? > reject item names with unpaired surrogates early > > > Key: OAK-5506 > URL: https://issues.apache.org/jira/browse/OAK-5506 > Project: Jackrabbit Oak > Issue Type: Wish > Components: core, jcr, segment-tar >Affects Versions: 1.5.18 >Reporter: Julian Reschke >Priority: Minor > Fix For: 1.10 > > Attachments: OAK-5506-01.patch, OAK-5506-02.patch, OAK-5506-4.diff, > OAK-5506-bench.diff, OAK-5506-name-conversion.diff, OAK-5506-segment.diff, > OAK-5506-segment2.diff, OAK-5506.diff, ValidNamesTest.java > > > Apparently, the following node name is accepted: >{{"foo\ud800"}} > but a subsequent {{getPath()}} call fails: > {noformat} > javax.jcr.InvalidItemStateException: This item [/test_node/foo?] does not > exist anymore > at > org.apache.jackrabbit.oak.jcr.delegate.ItemDelegate.checkAlive(ItemDelegate.java:86) > at > org.apache.jackrabbit.oak.jcr.session.operation.ItemOperation.checkPreconditions(ItemOperation.java:34) > at > org.apache.jackrabbit.oak.jcr.delegate.SessionDelegate.prePerform(SessionDelegate.java:615) > at > org.apache.jackrabbit.oak.jcr.delegate.SessionDelegate.perform(SessionDelegate.java:205) > at > org.apache.jackrabbit.oak.jcr.session.ItemImpl.perform(ItemImpl.java:112) > at > org.apache.jackrabbit.oak.jcr.session.ItemImpl.getPath(ItemImpl.java:140) > at > org.apache.jackrabbit.oak.jcr.session.NodeImpl.getPath(NodeImpl.java:106) > at > org.apache.jackrabbit.oak.jcr.ValidNamesTest.nameTest(ValidNamesTest.java:271) > at > org.apache.jackrabbit.oak.jcr.ValidNamesTest.testUnpairedSurrogate(ValidNamesTest.java:259) > at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method) > at sun.reflect.NativeMethodAccessorImpl.invoke(Unknown Source){noformat} > (test case follows) -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Commented] (OAK-5506) reject item names with unpaired surrogates early
[ https://issues.apache.org/jira/browse/OAK-5506?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16350149#comment-16350149 ] Michael Dürig commented on OAK-5506: Performance looks good AFICS. Re. backwards compatibility test: I think we should add a way to disable the extra check for testing in addition to the static feature flag property. The test could be on the FileStore level, which likely simplifies things. > reject item names with unpaired surrogates early > > > Key: OAK-5506 > URL: https://issues.apache.org/jira/browse/OAK-5506 > Project: Jackrabbit Oak > Issue Type: Wish > Components: core, jcr, segment-tar >Affects Versions: 1.5.18 >Reporter: Julian Reschke >Priority: Minor > Fix For: 1.10 > > Attachments: OAK-5506-01.patch, OAK-5506-02.patch, OAK-5506-4.diff, > OAK-5506-bench.diff, OAK-5506-name-conversion.diff, OAK-5506-segment.diff, > OAK-5506-segment2.diff, OAK-5506.diff, ValidNamesTest.java > > > Apparently, the following node name is accepted: >{{"foo\ud800"}} > but a subsequent {{getPath()}} call fails: > {noformat} > javax.jcr.InvalidItemStateException: This item [/test_node/foo?] does not > exist anymore > at > org.apache.jackrabbit.oak.jcr.delegate.ItemDelegate.checkAlive(ItemDelegate.java:86) > at > org.apache.jackrabbit.oak.jcr.session.operation.ItemOperation.checkPreconditions(ItemOperation.java:34) > at > org.apache.jackrabbit.oak.jcr.delegate.SessionDelegate.prePerform(SessionDelegate.java:615) > at > org.apache.jackrabbit.oak.jcr.delegate.SessionDelegate.perform(SessionDelegate.java:205) > at > org.apache.jackrabbit.oak.jcr.session.ItemImpl.perform(ItemImpl.java:112) > at > org.apache.jackrabbit.oak.jcr.session.ItemImpl.getPath(ItemImpl.java:140) > at > org.apache.jackrabbit.oak.jcr.session.NodeImpl.getPath(NodeImpl.java:106) > at > org.apache.jackrabbit.oak.jcr.ValidNamesTest.nameTest(ValidNamesTest.java:271) > at > org.apache.jackrabbit.oak.jcr.ValidNamesTest.testUnpairedSurrogate(ValidNamesTest.java:259) > at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method) > at sun.reflect.NativeMethodAccessorImpl.invoke(Unknown Source){noformat} > (test case follows) -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Commented] (OAK-5506) reject item names with unpaired surrogates early
[ https://issues.apache.org/jira/browse/OAK-5506?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16348865#comment-16348865 ] Julian Reschke commented on OAK-5506: - Tests rerun with {{-DStringWriteTest.propertyCount=1000}}. Old: {noformat} Apache Jackrabbit Oak 1.10-SNAPSHOT # StringWriteTest C min 10% 50% 90% max N Oak-Segment-Tar1 15 15 31 32 79 2171 Oak-Segment-Tar1 11 15 31 32 78 2088 Oak-Segment-Tar1 12 15 31 32 69 2153 Oak-Segment-Tar1 10 15 31 32 79 2024 {noformat} New: {noformat} Apache Jackrabbit Oak 1.10-SNAPSHOT # StringWriteTest C min 10% 50% 90% max N Oak-Segment-Tar1 9 15 31 32 78 2073 Oak-Segment-Tar1 15 15 31 32 63 2113 Oak-Segment-Tar1 13 15 31 32 78 2070 Oak-Segment-Tar1 13 15 31 35 79 2026 {noformat} > reject item names with unpaired surrogates early > > > Key: OAK-5506 > URL: https://issues.apache.org/jira/browse/OAK-5506 > Project: Jackrabbit Oak > Issue Type: Wish > Components: core, jcr, segment-tar >Affects Versions: 1.5.18 >Reporter: Julian Reschke >Priority: Minor > Fix For: 1.10 > > Attachments: OAK-5506-01.patch, OAK-5506-02.patch, OAK-5506-4.diff, > OAK-5506-bench.diff, OAK-5506-name-conversion.diff, OAK-5506-segment.diff, > OAK-5506-segment2.diff, OAK-5506.diff, ValidNamesTest.java > > > Apparently, the following node name is accepted: >{{"foo\ud800"}} > but a subsequent {{getPath()}} call fails: > {noformat} > javax.jcr.InvalidItemStateException: This item [/test_node/foo?] does not > exist anymore > at > org.apache.jackrabbit.oak.jcr.delegate.ItemDelegate.checkAlive(ItemDelegate.java:86) > at > org.apache.jackrabbit.oak.jcr.session.operation.ItemOperation.checkPreconditions(ItemOperation.java:34) > at > org.apache.jackrabbit.oak.jcr.delegate.SessionDelegate.prePerform(SessionDelegate.java:615) > at > org.apache.jackrabbit.oak.jcr.delegate.SessionDelegate.perform(SessionDelegate.java:205) > at > org.apache.jackrabbit.oak.jcr.session.ItemImpl.perform(ItemImpl.java:112) > at > org.apache.jackrabbit.oak.jcr.session.ItemImpl.getPath(ItemImpl.java:140) > at > org.apache.jackrabbit.oak.jcr.session.NodeImpl.getPath(NodeImpl.java:106) > at > org.apache.jackrabbit.oak.jcr.ValidNamesTest.nameTest(ValidNamesTest.java:271) > at > org.apache.jackrabbit.oak.jcr.ValidNamesTest.testUnpairedSurrogate(ValidNamesTest.java:259) > at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method) > at sun.reflect.NativeMethodAccessorImpl.invoke(Unknown Source){noformat} > (test case follows) -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Commented] (OAK-5506) reject item names with unpaired surrogates early
[ https://issues.apache.org/jira/browse/OAK-5506?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16348577#comment-16348577 ] Julian Reschke commented on OAK-5506: - bq. But I think we need to come up with a test case writing the old way and reading the new way. This should be doable if there is a way to disable the new behaviour I guess. [~mduerig] - do you have something specific in mind? I could imagine a system prop, but if we want to make it static we'd have to do major hacking in the test case... > reject item names with unpaired surrogates early > > > Key: OAK-5506 > URL: https://issues.apache.org/jira/browse/OAK-5506 > Project: Jackrabbit Oak > Issue Type: Wish > Components: core, jcr, segment-tar >Affects Versions: 1.5.18 >Reporter: Julian Reschke >Assignee: Francesco Mari >Priority: Minor > Fix For: 1.10 > > Attachments: OAK-5506-01.patch, OAK-5506-02.patch, OAK-5506-4.diff, > OAK-5506-bench.diff, OAK-5506-name-conversion.diff, OAK-5506-segment.diff, > OAK-5506-segment2.diff, OAK-5506.diff, ValidNamesTest.java > > > Apparently, the following node name is accepted: >{{"foo\ud800"}} > but a subsequent {{getPath()}} call fails: > {noformat} > javax.jcr.InvalidItemStateException: This item [/test_node/foo?] does not > exist anymore > at > org.apache.jackrabbit.oak.jcr.delegate.ItemDelegate.checkAlive(ItemDelegate.java:86) > at > org.apache.jackrabbit.oak.jcr.session.operation.ItemOperation.checkPreconditions(ItemOperation.java:34) > at > org.apache.jackrabbit.oak.jcr.delegate.SessionDelegate.prePerform(SessionDelegate.java:615) > at > org.apache.jackrabbit.oak.jcr.delegate.SessionDelegate.perform(SessionDelegate.java:205) > at > org.apache.jackrabbit.oak.jcr.session.ItemImpl.perform(ItemImpl.java:112) > at > org.apache.jackrabbit.oak.jcr.session.ItemImpl.getPath(ItemImpl.java:140) > at > org.apache.jackrabbit.oak.jcr.session.NodeImpl.getPath(NodeImpl.java:106) > at > org.apache.jackrabbit.oak.jcr.ValidNamesTest.nameTest(ValidNamesTest.java:271) > at > org.apache.jackrabbit.oak.jcr.ValidNamesTest.testUnpairedSurrogate(ValidNamesTest.java:259) > at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method) > at sun.reflect.NativeMethodAccessorImpl.invoke(Unknown Source){noformat} > (test case follows) -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Commented] (OAK-5506) reject item names with unpaired surrogates early
[ https://issues.apache.org/jira/browse/OAK-5506?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16348565#comment-16348565 ] Julian Reschke commented on OAK-5506: - trunk: [r1822875|http://svn.apache.org/r1822875] - documents that system behavior for broken strings as "undefined" for now > reject item names with unpaired surrogates early > > > Key: OAK-5506 > URL: https://issues.apache.org/jira/browse/OAK-5506 > Project: Jackrabbit Oak > Issue Type: Wish > Components: core, jcr, segment-tar >Affects Versions: 1.5.18 >Reporter: Julian Reschke >Assignee: Francesco Mari >Priority: Minor > Fix For: 1.10 > > Attachments: OAK-5506-01.patch, OAK-5506-02.patch, OAK-5506-4.diff, > OAK-5506-bench.diff, OAK-5506-name-conversion.diff, OAK-5506-segment.diff, > OAK-5506-segment2.diff, OAK-5506.diff, ValidNamesTest.java > > > Apparently, the following node name is accepted: >{{"foo\ud800"}} > but a subsequent {{getPath()}} call fails: > {noformat} > javax.jcr.InvalidItemStateException: This item [/test_node/foo?] does not > exist anymore > at > org.apache.jackrabbit.oak.jcr.delegate.ItemDelegate.checkAlive(ItemDelegate.java:86) > at > org.apache.jackrabbit.oak.jcr.session.operation.ItemOperation.checkPreconditions(ItemOperation.java:34) > at > org.apache.jackrabbit.oak.jcr.delegate.SessionDelegate.prePerform(SessionDelegate.java:615) > at > org.apache.jackrabbit.oak.jcr.delegate.SessionDelegate.perform(SessionDelegate.java:205) > at > org.apache.jackrabbit.oak.jcr.session.ItemImpl.perform(ItemImpl.java:112) > at > org.apache.jackrabbit.oak.jcr.session.ItemImpl.getPath(ItemImpl.java:140) > at > org.apache.jackrabbit.oak.jcr.session.NodeImpl.getPath(NodeImpl.java:106) > at > org.apache.jackrabbit.oak.jcr.ValidNamesTest.nameTest(ValidNamesTest.java:271) > at > org.apache.jackrabbit.oak.jcr.ValidNamesTest.testUnpairedSurrogate(ValidNamesTest.java:259) > at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method) > at sun.reflect.NativeMethodAccessorImpl.invoke(Unknown Source){noformat} > (test case follows) -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Commented] (OAK-5506) reject item names with unpaired surrogates early
[ https://issues.apache.org/jira/browse/OAK-5506?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16347006#comment-16347006 ] Michael Dürig commented on OAK-5506: Agree wrt. performance. In other places we already switched to byte buffers. Wrt. compatibility, this sounds reasonable. But I think we need to come up with a test case writing the old way and reading the new way. This should be doable if there is a way to disable the new behaviour I guess. > reject item names with unpaired surrogates early > > > Key: OAK-5506 > URL: https://issues.apache.org/jira/browse/OAK-5506 > Project: Jackrabbit Oak > Issue Type: Wish > Components: core, jcr, segment-tar >Affects Versions: 1.5.18 >Reporter: Julian Reschke >Assignee: Francesco Mari >Priority: Minor > Fix For: 1.10 > > Attachments: OAK-5506-01.patch, OAK-5506-02.patch, OAK-5506-4.diff, > OAK-5506-bench.diff, OAK-5506-name-conversion.diff, OAK-5506-segment.diff, > OAK-5506-segment2.diff, OAK-5506.diff, ValidNamesTest.java > > > Apparently, the following node name is accepted: >{{"foo\ud800"}} > but a subsequent {{getPath()}} call fails: > {noformat} > javax.jcr.InvalidItemStateException: This item [/test_node/foo?] does not > exist anymore > at > org.apache.jackrabbit.oak.jcr.delegate.ItemDelegate.checkAlive(ItemDelegate.java:86) > at > org.apache.jackrabbit.oak.jcr.session.operation.ItemOperation.checkPreconditions(ItemOperation.java:34) > at > org.apache.jackrabbit.oak.jcr.delegate.SessionDelegate.prePerform(SessionDelegate.java:615) > at > org.apache.jackrabbit.oak.jcr.delegate.SessionDelegate.perform(SessionDelegate.java:205) > at > org.apache.jackrabbit.oak.jcr.session.ItemImpl.perform(ItemImpl.java:112) > at > org.apache.jackrabbit.oak.jcr.session.ItemImpl.getPath(ItemImpl.java:140) > at > org.apache.jackrabbit.oak.jcr.session.NodeImpl.getPath(NodeImpl.java:106) > at > org.apache.jackrabbit.oak.jcr.ValidNamesTest.nameTest(ValidNamesTest.java:271) > at > org.apache.jackrabbit.oak.jcr.ValidNamesTest.testUnpairedSurrogate(ValidNamesTest.java:259) > at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method) > at sun.reflect.NativeMethodAccessorImpl.invoke(Unknown Source){noformat} > (test case follows) -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Commented] (OAK-5506) reject item names with unpaired surrogates early
[ https://issues.apache.org/jira/browse/OAK-5506?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16346957#comment-16346957 ] Julian Reschke commented on OAK-5506: - With respect to compatibility with segment-tar storage written by previous versions: my understanding is that any string value (be it an item name or a string property) persisted to segment-tar will have gone through {{String.getBytes(UTF8)}}. In the case of a string that doesn't roundtrip through UTF-8, this means that those characters will have been replaced by a replacement character (here: "?"). So the persisted state *does* represent valid UTF-8, and when read back, will just use that replacement character. The proposed patch only affects writing of strings that are invalid, thus previously written garbage shouldn't be an issue. > reject item names with unpaired surrogates early > > > Key: OAK-5506 > URL: https://issues.apache.org/jira/browse/OAK-5506 > Project: Jackrabbit Oak > Issue Type: Wish > Components: core, jcr, segment-tar >Affects Versions: 1.5.18 >Reporter: Julian Reschke >Assignee: Francesco Mari >Priority: Minor > Fix For: 1.10 > > Attachments: OAK-5506-01.patch, OAK-5506-02.patch, OAK-5506-4.diff, > OAK-5506-bench.diff, OAK-5506-name-conversion.diff, OAK-5506-segment.diff, > OAK-5506-segment2.diff, OAK-5506.diff, ValidNamesTest.java > > > Apparently, the following node name is accepted: >{{"foo\ud800"}} > but a subsequent {{getPath()}} call fails: > {noformat} > javax.jcr.InvalidItemStateException: This item [/test_node/foo?] does not > exist anymore > at > org.apache.jackrabbit.oak.jcr.delegate.ItemDelegate.checkAlive(ItemDelegate.java:86) > at > org.apache.jackrabbit.oak.jcr.session.operation.ItemOperation.checkPreconditions(ItemOperation.java:34) > at > org.apache.jackrabbit.oak.jcr.delegate.SessionDelegate.prePerform(SessionDelegate.java:615) > at > org.apache.jackrabbit.oak.jcr.delegate.SessionDelegate.perform(SessionDelegate.java:205) > at > org.apache.jackrabbit.oak.jcr.session.ItemImpl.perform(ItemImpl.java:112) > at > org.apache.jackrabbit.oak.jcr.session.ItemImpl.getPath(ItemImpl.java:140) > at > org.apache.jackrabbit.oak.jcr.session.NodeImpl.getPath(NodeImpl.java:106) > at > org.apache.jackrabbit.oak.jcr.ValidNamesTest.nameTest(ValidNamesTest.java:271) > at > org.apache.jackrabbit.oak.jcr.ValidNamesTest.testUnpairedSurrogate(ValidNamesTest.java:259) > at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method) > at sun.reflect.NativeMethodAccessorImpl.invoke(Unknown Source){noformat} > (test case follows) -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Commented] (OAK-5506) reject item names with unpaired surrogates early
[ https://issues.apache.org/jira/browse/OAK-5506?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16346962#comment-16346962 ] Julian Reschke commented on OAK-5506: - With respect to performance: this code change opens the door to serialize to {{ByteBuffer}}s instead of {{byte[]}}, potentially avoiding array copies. Not sure whether segment-tar could take advantage of that... > reject item names with unpaired surrogates early > > > Key: OAK-5506 > URL: https://issues.apache.org/jira/browse/OAK-5506 > Project: Jackrabbit Oak > Issue Type: Wish > Components: core, jcr, segment-tar >Affects Versions: 1.5.18 >Reporter: Julian Reschke >Assignee: Francesco Mari >Priority: Minor > Fix For: 1.10 > > Attachments: OAK-5506-01.patch, OAK-5506-02.patch, OAK-5506-4.diff, > OAK-5506-bench.diff, OAK-5506-name-conversion.diff, OAK-5506-segment.diff, > OAK-5506-segment2.diff, OAK-5506.diff, ValidNamesTest.java > > > Apparently, the following node name is accepted: >{{"foo\ud800"}} > but a subsequent {{getPath()}} call fails: > {noformat} > javax.jcr.InvalidItemStateException: This item [/test_node/foo?] does not > exist anymore > at > org.apache.jackrabbit.oak.jcr.delegate.ItemDelegate.checkAlive(ItemDelegate.java:86) > at > org.apache.jackrabbit.oak.jcr.session.operation.ItemOperation.checkPreconditions(ItemOperation.java:34) > at > org.apache.jackrabbit.oak.jcr.delegate.SessionDelegate.prePerform(SessionDelegate.java:615) > at > org.apache.jackrabbit.oak.jcr.delegate.SessionDelegate.perform(SessionDelegate.java:205) > at > org.apache.jackrabbit.oak.jcr.session.ItemImpl.perform(ItemImpl.java:112) > at > org.apache.jackrabbit.oak.jcr.session.ItemImpl.getPath(ItemImpl.java:140) > at > org.apache.jackrabbit.oak.jcr.session.NodeImpl.getPath(NodeImpl.java:106) > at > org.apache.jackrabbit.oak.jcr.ValidNamesTest.nameTest(ValidNamesTest.java:271) > at > org.apache.jackrabbit.oak.jcr.ValidNamesTest.testUnpairedSurrogate(ValidNamesTest.java:259) > at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method) > at sun.reflect.NativeMethodAccessorImpl.invoke(Unknown Source){noformat} > (test case follows) -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Commented] (OAK-5506) reject item names with unpaired surrogates early
[ https://issues.apache.org/jira/browse/OAK-5506?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16346875#comment-16346875 ] Julian Reschke commented on OAK-5506: - ...was just looking at the last columns. Will rerun later with more iterations. > reject item names with unpaired surrogates early > > > Key: OAK-5506 > URL: https://issues.apache.org/jira/browse/OAK-5506 > Project: Jackrabbit Oak > Issue Type: Wish > Components: core, jcr, segment-tar >Affects Versions: 1.5.18 >Reporter: Julian Reschke >Assignee: Francesco Mari >Priority: Minor > Fix For: 1.10 > > Attachments: OAK-5506-01.patch, OAK-5506-02.patch, OAK-5506-4.diff, > OAK-5506-bench.diff, OAK-5506-name-conversion.diff, OAK-5506-segment.diff, > OAK-5506-segment2.diff, OAK-5506.diff, ValidNamesTest.java > > > Apparently, the following node name is accepted: >{{"foo\ud800"}} > but a subsequent {{getPath()}} call fails: > {noformat} > javax.jcr.InvalidItemStateException: This item [/test_node/foo?] does not > exist anymore > at > org.apache.jackrabbit.oak.jcr.delegate.ItemDelegate.checkAlive(ItemDelegate.java:86) > at > org.apache.jackrabbit.oak.jcr.session.operation.ItemOperation.checkPreconditions(ItemOperation.java:34) > at > org.apache.jackrabbit.oak.jcr.delegate.SessionDelegate.prePerform(SessionDelegate.java:615) > at > org.apache.jackrabbit.oak.jcr.delegate.SessionDelegate.perform(SessionDelegate.java:205) > at > org.apache.jackrabbit.oak.jcr.session.ItemImpl.perform(ItemImpl.java:112) > at > org.apache.jackrabbit.oak.jcr.session.ItemImpl.getPath(ItemImpl.java:140) > at > org.apache.jackrabbit.oak.jcr.session.NodeImpl.getPath(NodeImpl.java:106) > at > org.apache.jackrabbit.oak.jcr.ValidNamesTest.nameTest(ValidNamesTest.java:271) > at > org.apache.jackrabbit.oak.jcr.ValidNamesTest.testUnpairedSurrogate(ValidNamesTest.java:259) > at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method) > at sun.reflect.NativeMethodAccessorImpl.invoke(Unknown Source){noformat} > (test case follows) -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Commented] (OAK-5506) reject item names with unpaired surrogates early
[ https://issues.apache.org/jira/browse/OAK-5506?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16346786#comment-16346786 ] Marcel Reutegger commented on OAK-5506: --- The result doesn't look that useful. The measured time is zero up to the 50 percentile. > reject item names with unpaired surrogates early > > > Key: OAK-5506 > URL: https://issues.apache.org/jira/browse/OAK-5506 > Project: Jackrabbit Oak > Issue Type: Wish > Components: core, jcr, segment-tar >Affects Versions: 1.5.18 >Reporter: Julian Reschke >Assignee: Francesco Mari >Priority: Minor > Fix For: 1.10 > > Attachments: OAK-5506-01.patch, OAK-5506-02.patch, OAK-5506-4.diff, > OAK-5506-bench.diff, OAK-5506-name-conversion.diff, OAK-5506-segment.diff, > OAK-5506-segment2.diff, OAK-5506.diff, ValidNamesTest.java > > > Apparently, the following node name is accepted: >{{"foo\ud800"}} > but a subsequent {{getPath()}} call fails: > {noformat} > javax.jcr.InvalidItemStateException: This item [/test_node/foo?] does not > exist anymore > at > org.apache.jackrabbit.oak.jcr.delegate.ItemDelegate.checkAlive(ItemDelegate.java:86) > at > org.apache.jackrabbit.oak.jcr.session.operation.ItemOperation.checkPreconditions(ItemOperation.java:34) > at > org.apache.jackrabbit.oak.jcr.delegate.SessionDelegate.prePerform(SessionDelegate.java:615) > at > org.apache.jackrabbit.oak.jcr.delegate.SessionDelegate.perform(SessionDelegate.java:205) > at > org.apache.jackrabbit.oak.jcr.session.ItemImpl.perform(ItemImpl.java:112) > at > org.apache.jackrabbit.oak.jcr.session.ItemImpl.getPath(ItemImpl.java:140) > at > org.apache.jackrabbit.oak.jcr.session.NodeImpl.getPath(NodeImpl.java:106) > at > org.apache.jackrabbit.oak.jcr.ValidNamesTest.nameTest(ValidNamesTest.java:271) > at > org.apache.jackrabbit.oak.jcr.ValidNamesTest.testUnpairedSurrogate(ValidNamesTest.java:259) > at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method) > at sun.reflect.NativeMethodAccessorImpl.invoke(Unknown Source){noformat} > (test case follows) -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Commented] (OAK-5506) reject item names with unpaired surrogates early
[ https://issues.apache.org/jira/browse/OAK-5506?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16346638#comment-16346638 ] Julian Reschke commented on OAK-5506: - results from {{java -jar target/oak-benchmarks-1.10-SNAPSHOT.jar benchmark Oak-Segment-Tar StringWriteTest}}: new: {noformat} Apache Jackrabbit Oak 1.10-SNAPSHOT # StringWriteTest C min 10% 50% 90% max N Oak-Segment-Tar1 0 0 0 16 33 16726 Oak-Segment-Tar1 0 0 0 16 47 17000 Oak-Segment-Tar1 0 0 0 16 47 17739 Oak-Segment-Tar1 0 0 0 15 46 17635 {noformat} old: {noformat} Apache Jackrabbit Oak 1.10-SNAPSHOT # StringWriteTest C min 10% 50% 90% max N Oak-Segment-Tar1 0 0 0 16 35 16790 Oak-Segment-Tar1 0 0 0 16 47 17092 Oak-Segment-Tar1 0 0 0 16 46 16937 Oak-Segment-Tar1 0 0 0 16 47 16851 {noformat} > reject item names with unpaired surrogates early > > > Key: OAK-5506 > URL: https://issues.apache.org/jira/browse/OAK-5506 > Project: Jackrabbit Oak > Issue Type: Wish > Components: core, jcr, segment-tar >Affects Versions: 1.5.18 >Reporter: Julian Reschke >Assignee: Francesco Mari >Priority: Minor > Fix For: 1.10 > > Attachments: OAK-5506-01.patch, OAK-5506-02.patch, OAK-5506-4.diff, > OAK-5506-bench.diff, OAK-5506-name-conversion.diff, OAK-5506-segment.diff, > OAK-5506-segment2.diff, OAK-5506.diff, ValidNamesTest.java > > > Apparently, the following node name is accepted: >{{"foo\ud800"}} > but a subsequent {{getPath()}} call fails: > {noformat} > javax.jcr.InvalidItemStateException: This item [/test_node/foo?] does not > exist anymore > at > org.apache.jackrabbit.oak.jcr.delegate.ItemDelegate.checkAlive(ItemDelegate.java:86) > at > org.apache.jackrabbit.oak.jcr.session.operation.ItemOperation.checkPreconditions(ItemOperation.java:34) > at > org.apache.jackrabbit.oak.jcr.delegate.SessionDelegate.prePerform(SessionDelegate.java:615) > at > org.apache.jackrabbit.oak.jcr.delegate.SessionDelegate.perform(SessionDelegate.java:205) > at > org.apache.jackrabbit.oak.jcr.session.ItemImpl.perform(ItemImpl.java:112) > at > org.apache.jackrabbit.oak.jcr.session.ItemImpl.getPath(ItemImpl.java:140) > at > org.apache.jackrabbit.oak.jcr.session.NodeImpl.getPath(NodeImpl.java:106) > at > org.apache.jackrabbit.oak.jcr.ValidNamesTest.nameTest(ValidNamesTest.java:271) > at > org.apache.jackrabbit.oak.jcr.ValidNamesTest.testUnpairedSurrogate(ValidNamesTest.java:259) > at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method) > at sun.reflect.NativeMethodAccessorImpl.invoke(Unknown Source){noformat} > (test case follows) -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Commented] (OAK-5506) reject item names with unpaired surrogates early
[ https://issues.apache.org/jira/browse/OAK-5506?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16345730#comment-16345730 ] Julian Reschke commented on OAK-5506: - updated potential patch for segment-tar: https://issues.apache.org/jira/secure/attachment/12908423/OAK-5506-segment2.diff > reject item names with unpaired surrogates early > > > Key: OAK-5506 > URL: https://issues.apache.org/jira/browse/OAK-5506 > Project: Jackrabbit Oak > Issue Type: Wish > Components: core, jcr, segment-tar >Affects Versions: 1.5.18 >Reporter: Julian Reschke >Assignee: Francesco Mari >Priority: Minor > Fix For: 1.10 > > Attachments: OAK-5506-01.patch, OAK-5506-02.patch, OAK-5506-4.diff, > OAK-5506-bench.diff, OAK-5506-name-conversion.diff, OAK-5506-segment.diff, > OAK-5506-segment2.diff, OAK-5506.diff, ValidNamesTest.java > > > Apparently, the following node name is accepted: >{{"foo\ud800"}} > but a subsequent {{getPath()}} call fails: > {noformat} > javax.jcr.InvalidItemStateException: This item [/test_node/foo?] does not > exist anymore > at > org.apache.jackrabbit.oak.jcr.delegate.ItemDelegate.checkAlive(ItemDelegate.java:86) > at > org.apache.jackrabbit.oak.jcr.session.operation.ItemOperation.checkPreconditions(ItemOperation.java:34) > at > org.apache.jackrabbit.oak.jcr.delegate.SessionDelegate.prePerform(SessionDelegate.java:615) > at > org.apache.jackrabbit.oak.jcr.delegate.SessionDelegate.perform(SessionDelegate.java:205) > at > org.apache.jackrabbit.oak.jcr.session.ItemImpl.perform(ItemImpl.java:112) > at > org.apache.jackrabbit.oak.jcr.session.ItemImpl.getPath(ItemImpl.java:140) > at > org.apache.jackrabbit.oak.jcr.session.NodeImpl.getPath(NodeImpl.java:106) > at > org.apache.jackrabbit.oak.jcr.ValidNamesTest.nameTest(ValidNamesTest.java:271) > at > org.apache.jackrabbit.oak.jcr.ValidNamesTest.testUnpairedSurrogate(ValidNamesTest.java:259) > at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method) > at sun.reflect.NativeMethodAccessorImpl.invoke(Unknown Source){noformat} > (test case follows) -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Commented] (OAK-5506) reject item names with unpaired surrogates early
[ https://issues.apache.org/jira/browse/OAK-5506?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16343227#comment-16343227 ] Michael Dürig commented on OAK-5506: bq. is it worth the effort? I don't know. But finding out is part of this effort. > reject item names with unpaired surrogates early > > > Key: OAK-5506 > URL: https://issues.apache.org/jira/browse/OAK-5506 > Project: Jackrabbit Oak > Issue Type: Wish > Components: core, jcr, segment-tar >Affects Versions: 1.5.18 >Reporter: Julian Reschke >Assignee: Francesco Mari >Priority: Minor > Fix For: 1.10 > > Attachments: OAK-5506-01.patch, OAK-5506-02.patch, OAK-5506-4.diff, > OAK-5506-bench.diff, OAK-5506-name-conversion.diff, OAK-5506-segment.diff, > OAK-5506.diff, ValidNamesTest.java > > > Apparently, the following node name is accepted: >{{"foo\ud800"}} > but a subsequent {{getPath()}} call fails: > {noformat} > javax.jcr.InvalidItemStateException: This item [/test_node/foo?] does not > exist anymore > at > org.apache.jackrabbit.oak.jcr.delegate.ItemDelegate.checkAlive(ItemDelegate.java:86) > at > org.apache.jackrabbit.oak.jcr.session.operation.ItemOperation.checkPreconditions(ItemOperation.java:34) > at > org.apache.jackrabbit.oak.jcr.delegate.SessionDelegate.prePerform(SessionDelegate.java:615) > at > org.apache.jackrabbit.oak.jcr.delegate.SessionDelegate.perform(SessionDelegate.java:205) > at > org.apache.jackrabbit.oak.jcr.session.ItemImpl.perform(ItemImpl.java:112) > at > org.apache.jackrabbit.oak.jcr.session.ItemImpl.getPath(ItemImpl.java:140) > at > org.apache.jackrabbit.oak.jcr.session.NodeImpl.getPath(NodeImpl.java:106) > at > org.apache.jackrabbit.oak.jcr.ValidNamesTest.nameTest(ValidNamesTest.java:271) > at > org.apache.jackrabbit.oak.jcr.ValidNamesTest.testUnpairedSurrogate(ValidNamesTest.java:259) > at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method) > at sun.reflect.NativeMethodAccessorImpl.invoke(Unknown Source){noformat} > (test case follows) -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Commented] (OAK-5506) reject item names with unpaired surrogates early
[ https://issues.apache.org/jira/browse/OAK-5506?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16343117#comment-16343117 ] Julian Reschke commented on OAK-5506: - [~mduerig] - in earlier comments you said that the current behavior is ok because it "garbage in, garbage out" - why do you care how garbage is handled then? FWIW, we certainly can do more work than having an emergency system variable, the question is: is it worth the effort? > reject item names with unpaired surrogates early > > > Key: OAK-5506 > URL: https://issues.apache.org/jira/browse/OAK-5506 > Project: Jackrabbit Oak > Issue Type: Wish > Components: core, jcr, segment-tar >Affects Versions: 1.5.18 >Reporter: Julian Reschke >Assignee: Francesco Mari >Priority: Minor > Fix For: 1.10 > > Attachments: OAK-5506-01.patch, OAK-5506-02.patch, OAK-5506-4.diff, > OAK-5506-bench.diff, OAK-5506-name-conversion.diff, OAK-5506-segment.diff, > OAK-5506.diff, ValidNamesTest.java > > > Apparently, the following node name is accepted: >{{"foo\ud800"}} > but a subsequent {{getPath()}} call fails: > {noformat} > javax.jcr.InvalidItemStateException: This item [/test_node/foo?] does not > exist anymore > at > org.apache.jackrabbit.oak.jcr.delegate.ItemDelegate.checkAlive(ItemDelegate.java:86) > at > org.apache.jackrabbit.oak.jcr.session.operation.ItemOperation.checkPreconditions(ItemOperation.java:34) > at > org.apache.jackrabbit.oak.jcr.delegate.SessionDelegate.prePerform(SessionDelegate.java:615) > at > org.apache.jackrabbit.oak.jcr.delegate.SessionDelegate.perform(SessionDelegate.java:205) > at > org.apache.jackrabbit.oak.jcr.session.ItemImpl.perform(ItemImpl.java:112) > at > org.apache.jackrabbit.oak.jcr.session.ItemImpl.getPath(ItemImpl.java:140) > at > org.apache.jackrabbit.oak.jcr.session.NodeImpl.getPath(NodeImpl.java:106) > at > org.apache.jackrabbit.oak.jcr.ValidNamesTest.nameTest(ValidNamesTest.java:271) > at > org.apache.jackrabbit.oak.jcr.ValidNamesTest.testUnpairedSurrogate(ValidNamesTest.java:259) > at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method) > at sun.reflect.NativeMethodAccessorImpl.invoke(Unknown Source){noformat} > (test case follows) -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Commented] (OAK-5506) reject item names with unpaired surrogates early
[ https://issues.apache.org/jira/browse/OAK-5506?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16343106#comment-16343106 ] Michael Dürig commented on OAK-5506: TBH I'm not in favour of this change unless we have better ways for dealing with upgrades than an emergency button. See my [comment|https://issues.apache.org/jira/browse/OAK-7208?focusedCommentId=16343101&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-16343101] on OAK-7280, which applies here respectively. > reject item names with unpaired surrogates early > > > Key: OAK-5506 > URL: https://issues.apache.org/jira/browse/OAK-5506 > Project: Jackrabbit Oak > Issue Type: Wish > Components: core, jcr, segment-tar >Affects Versions: 1.5.18 >Reporter: Julian Reschke >Assignee: Francesco Mari >Priority: Minor > Fix For: 1.10 > > Attachments: OAK-5506-01.patch, OAK-5506-02.patch, OAK-5506-4.diff, > OAK-5506-bench.diff, OAK-5506-name-conversion.diff, OAK-5506-segment.diff, > OAK-5506.diff, ValidNamesTest.java > > > Apparently, the following node name is accepted: >{{"foo\ud800"}} > but a subsequent {{getPath()}} call fails: > {noformat} > javax.jcr.InvalidItemStateException: This item [/test_node/foo?] does not > exist anymore > at > org.apache.jackrabbit.oak.jcr.delegate.ItemDelegate.checkAlive(ItemDelegate.java:86) > at > org.apache.jackrabbit.oak.jcr.session.operation.ItemOperation.checkPreconditions(ItemOperation.java:34) > at > org.apache.jackrabbit.oak.jcr.delegate.SessionDelegate.prePerform(SessionDelegate.java:615) > at > org.apache.jackrabbit.oak.jcr.delegate.SessionDelegate.perform(SessionDelegate.java:205) > at > org.apache.jackrabbit.oak.jcr.session.ItemImpl.perform(ItemImpl.java:112) > at > org.apache.jackrabbit.oak.jcr.session.ItemImpl.getPath(ItemImpl.java:140) > at > org.apache.jackrabbit.oak.jcr.session.NodeImpl.getPath(NodeImpl.java:106) > at > org.apache.jackrabbit.oak.jcr.ValidNamesTest.nameTest(ValidNamesTest.java:271) > at > org.apache.jackrabbit.oak.jcr.ValidNamesTest.testUnpairedSurrogate(ValidNamesTest.java:259) > at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method) > at sun.reflect.NativeMethodAccessorImpl.invoke(Unknown Source){noformat} > (test case follows) -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Commented] (OAK-5506) reject item names with unpaired surrogates early
[ https://issues.apache.org/jira/browse/OAK-5506?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16343072#comment-16343072 ] Julian Reschke commented on OAK-5506: - For segment-tar, it would mean that it would stop loosing information, but instead throw exceptions. Otherwise, the answer is "no", that is, we may want to have a system variable to turn this off in emergencies. > reject item names with unpaired surrogates early > > > Key: OAK-5506 > URL: https://issues.apache.org/jira/browse/OAK-5506 > Project: Jackrabbit Oak > Issue Type: Wish > Components: core, jcr, segment-tar >Affects Versions: 1.5.18 >Reporter: Julian Reschke >Assignee: Francesco Mari >Priority: Minor > Fix For: 1.10 > > Attachments: OAK-5506-01.patch, OAK-5506-02.patch, OAK-5506-4.diff, > OAK-5506-bench.diff, OAK-5506-name-conversion.diff, OAK-5506-segment.diff, > OAK-5506.diff, ValidNamesTest.java > > > Apparently, the following node name is accepted: >{{"foo\ud800"}} > but a subsequent {{getPath()}} call fails: > {noformat} > javax.jcr.InvalidItemStateException: This item [/test_node/foo?] does not > exist anymore > at > org.apache.jackrabbit.oak.jcr.delegate.ItemDelegate.checkAlive(ItemDelegate.java:86) > at > org.apache.jackrabbit.oak.jcr.session.operation.ItemOperation.checkPreconditions(ItemOperation.java:34) > at > org.apache.jackrabbit.oak.jcr.delegate.SessionDelegate.prePerform(SessionDelegate.java:615) > at > org.apache.jackrabbit.oak.jcr.delegate.SessionDelegate.perform(SessionDelegate.java:205) > at > org.apache.jackrabbit.oak.jcr.session.ItemImpl.perform(ItemImpl.java:112) > at > org.apache.jackrabbit.oak.jcr.session.ItemImpl.getPath(ItemImpl.java:140) > at > org.apache.jackrabbit.oak.jcr.session.NodeImpl.getPath(NodeImpl.java:106) > at > org.apache.jackrabbit.oak.jcr.ValidNamesTest.nameTest(ValidNamesTest.java:271) > at > org.apache.jackrabbit.oak.jcr.ValidNamesTest.testUnpairedSurrogate(ValidNamesTest.java:259) > at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method) > at sun.reflect.NativeMethodAccessorImpl.invoke(Unknown Source){noformat} > (test case follows) -- This message was sent by Atlassian JIRA (v7.6.3#76005)
[jira] [Commented] (OAK-5506) reject item names with unpaired surrogates early
[ https://issues.apache.org/jira/browse/OAK-5506?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16343061#comment-16343061 ] Michael Dürig commented on OAK-5506: [~reschke], do we understand the backward compatibility concerns the latest patch introduces? Do we have a migration path for existing repositories that might contain such characters now disallowed? Do we have test coverage for such migration cases? > reject item names with unpaired surrogates early > > > Key: OAK-5506 > URL: https://issues.apache.org/jira/browse/OAK-5506 > Project: Jackrabbit Oak > Issue Type: Wish > Components: core, jcr, segment-tar >Affects Versions: 1.5.18 >Reporter: Julian Reschke >Assignee: Francesco Mari >Priority: Minor > Fix For: 1.10 > > Attachments: OAK-5506-01.patch, OAK-5506-02.patch, OAK-5506-4.diff, > OAK-5506-bench.diff, OAK-5506-name-conversion.diff, OAK-5506-segment.diff, > OAK-5506.diff, ValidNamesTest.java > > > Apparently, the following node name is accepted: >{{"foo\ud800"}} > but a subsequent {{getPath()}} call fails: > {noformat} > javax.jcr.InvalidItemStateException: This item [/test_node/foo?] does not > exist anymore > at > org.apache.jackrabbit.oak.jcr.delegate.ItemDelegate.checkAlive(ItemDelegate.java:86) > at > org.apache.jackrabbit.oak.jcr.session.operation.ItemOperation.checkPreconditions(ItemOperation.java:34) > at > org.apache.jackrabbit.oak.jcr.delegate.SessionDelegate.prePerform(SessionDelegate.java:615) > at > org.apache.jackrabbit.oak.jcr.delegate.SessionDelegate.perform(SessionDelegate.java:205) > at > org.apache.jackrabbit.oak.jcr.session.ItemImpl.perform(ItemImpl.java:112) > at > org.apache.jackrabbit.oak.jcr.session.ItemImpl.getPath(ItemImpl.java:140) > at > org.apache.jackrabbit.oak.jcr.session.NodeImpl.getPath(NodeImpl.java:106) > at > org.apache.jackrabbit.oak.jcr.ValidNamesTest.nameTest(ValidNamesTest.java:271) > at > org.apache.jackrabbit.oak.jcr.ValidNamesTest.testUnpairedSurrogate(ValidNamesTest.java:259) > at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method) > at sun.reflect.NativeMethodAccessorImpl.invoke(Unknown Source){noformat} > (test case follows) -- This message was sent by Atlassian JIRA (v7.6.3#76005)