[jira] [Commented] (LUCENE-6013) Remove IndexableFieldType.indexed()
[ https://issues.apache.org/jira/browse/LUCENE-6013?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14178088#comment-14178088 ] ASF subversion and git services commented on LUCENE-6013: - Commit 1633296 from [~mikemccand] in branch 'dev/trunk' [ https://svn.apache.org/r1633296 ] LUCENE-6013: remove IndexableFieldType.indexed and FieldInfo.indexed (it's redundant with IndexOptions != null) Remove IndexableFieldType.indexed() --- Key: LUCENE-6013 URL: https://issues.apache.org/jira/browse/LUCENE-6013 Project: Lucene - Core Issue Type: Improvement Reporter: Michael McCandless Assignee: Michael McCandless Fix For: 5.0, Trunk Attachments: LUCENE-6013.patch, LUCENE-6013.patch, LUCENE-6013.patch, LUCENE-6013.patch Like LUCENE-6006, here's another pre-cursor for LUCENE-6005 ... because I think it's important to nail down Lucene's low-schema (FieldType/FieldInfos) semantics before adding a high-schema. IndexableFieldType.indexed() is redundant with IndexableFieldType.indexOptions() != null, so we should remove it, codecs shouldn't have to write/read it, high-schema should not configure it, etc. Similarly, the FieldInfo.indexed bit is redundant, so I removed it, but I left the sugar API (FieldInfo.isIndexed) and implement it as just checking IndexOptions != null. -- This message was sent by Atlassian JIRA (v6.3.4#6332) - To unsubscribe, e-mail: dev-unsubscr...@lucene.apache.org For additional commands, e-mail: dev-h...@lucene.apache.org
[jira] [Commented] (LUCENE-6013) Remove IndexableFieldType.indexed()
[ https://issues.apache.org/jira/browse/LUCENE-6013?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14178188#comment-14178188 ] ASF subversion and git services commented on LUCENE-6013: - Commit 1633328 from [~mikemccand] in branch 'dev/branches/branch_5x' [ https://svn.apache.org/r1633328 ] LUCENE-6013: remove IndexableFieldType.indexed and FieldInfo.indexed (it's redundant with IndexOptions != null) Remove IndexableFieldType.indexed() --- Key: LUCENE-6013 URL: https://issues.apache.org/jira/browse/LUCENE-6013 Project: Lucene - Core Issue Type: Improvement Reporter: Michael McCandless Assignee: Michael McCandless Fix For: 5.0, Trunk Attachments: LUCENE-6013.patch, LUCENE-6013.patch, LUCENE-6013.patch, LUCENE-6013.patch Like LUCENE-6006, here's another pre-cursor for LUCENE-6005 ... because I think it's important to nail down Lucene's low-schema (FieldType/FieldInfos) semantics before adding a high-schema. IndexableFieldType.indexed() is redundant with IndexableFieldType.indexOptions() != null, so we should remove it, codecs shouldn't have to write/read it, high-schema should not configure it, etc. Similarly, the FieldInfo.indexed bit is redundant, so I removed it, but I left the sugar API (FieldInfo.isIndexed) and implement it as just checking IndexOptions != null. -- This message was sent by Atlassian JIRA (v6.3.4#6332) - To unsubscribe, e-mail: dev-unsubscr...@lucene.apache.org For additional commands, e-mail: dev-h...@lucene.apache.org
[jira] [Commented] (LUCENE-6013) Remove IndexableFieldType.indexed()
[ https://issues.apache.org/jira/browse/LUCENE-6013?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14176811#comment-14176811 ] Robert Muir commented on LUCENE-6013: - Patch looks good. I'm afraid about FieldInfosReader/Writer, can you please look this over one more time just to ensure its ok? Since it was writing a bit for this before, this is absolutely the most likely place for bugs in the patch... and the worse place for my brain to hit TooManyUnrelatedStyleChangesException when reviewing it :( Remove IndexableFieldType.indexed() --- Key: LUCENE-6013 URL: https://issues.apache.org/jira/browse/LUCENE-6013 Project: Lucene - Core Issue Type: Improvement Reporter: Michael McCandless Assignee: Michael McCandless Fix For: 5.0, Trunk Attachments: LUCENE-6013.patch Like LUCENE-6006, here's another pre-cursor for LUCENE-6005 ... because I think it's important to nail down Lucene's low-schema (FieldType/FieldInfos) semantics before adding a high-schema. IndexableFieldType.indexed() is redundant with IndexableFieldType.indexOptions() != null, so we should remove it, codecs shouldn't have to write/read it, high-schema should not configure it, etc. Similarly, the FieldInfo.indexed bit is redundant, so I removed it, but I left the sugar API (FieldInfo.isIndexed) and implement it as just checking IndexOptions != null. -- This message was sent by Atlassian JIRA (v6.3.4#6332) - To unsubscribe, e-mail: dev-unsubscr...@lucene.apache.org For additional commands, e-mail: dev-h...@lucene.apache.org
[jira] [Commented] (LUCENE-6013) Remove IndexableFieldType.indexed()
[ https://issues.apache.org/jira/browse/LUCENE-6013?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14176921#comment-14176921 ] Michael McCandless commented on LUCENE-6013: Thanks Rob, sorry about the noise in Lucene50FIS, I'll double check: there is a REAL change here, I completely rewrote how we encode IndexOptions. I found it really confusing how it was OMIT_THIS, STORE_THAT, OMIT_THIS_OTHER_THING before. Remove IndexableFieldType.indexed() --- Key: LUCENE-6013 URL: https://issues.apache.org/jira/browse/LUCENE-6013 Project: Lucene - Core Issue Type: Improvement Reporter: Michael McCandless Assignee: Michael McCandless Fix For: 5.0, Trunk Attachments: LUCENE-6013.patch Like LUCENE-6006, here's another pre-cursor for LUCENE-6005 ... because I think it's important to nail down Lucene's low-schema (FieldType/FieldInfos) semantics before adding a high-schema. IndexableFieldType.indexed() is redundant with IndexableFieldType.indexOptions() != null, so we should remove it, codecs shouldn't have to write/read it, high-schema should not configure it, etc. Similarly, the FieldInfo.indexed bit is redundant, so I removed it, but I left the sugar API (FieldInfo.isIndexed) and implement it as just checking IndexOptions != null. -- This message was sent by Atlassian JIRA (v6.3.4#6332) - To unsubscribe, e-mail: dev-unsubscr...@lucene.apache.org For additional commands, e-mail: dev-h...@lucene.apache.org
[jira] [Commented] (LUCENE-6013) Remove IndexableFieldType.indexed()
[ https://issues.apache.org/jira/browse/LUCENE-6013?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14176926#comment-14176926 ] Robert Muir commented on LUCENE-6013: - No, its good to fix the style, its just scary in a 'nuke fieldinfos boolean change'. I also feel the encoding of this thing is complex in packing norms/docvalues bit flags? Remove IndexableFieldType.indexed() --- Key: LUCENE-6013 URL: https://issues.apache.org/jira/browse/LUCENE-6013 Project: Lucene - Core Issue Type: Improvement Reporter: Michael McCandless Assignee: Michael McCandless Fix For: 5.0, Trunk Attachments: LUCENE-6013.patch Like LUCENE-6006, here's another pre-cursor for LUCENE-6005 ... because I think it's important to nail down Lucene's low-schema (FieldType/FieldInfos) semantics before adding a high-schema. IndexableFieldType.indexed() is redundant with IndexableFieldType.indexOptions() != null, so we should remove it, codecs shouldn't have to write/read it, high-schema should not configure it, etc. Similarly, the FieldInfo.indexed bit is redundant, so I removed it, but I left the sugar API (FieldInfo.isIndexed) and implement it as just checking IndexOptions != null. -- This message was sent by Atlassian JIRA (v6.3.4#6332) - To unsubscribe, e-mail: dev-unsubscr...@lucene.apache.org For additional commands, e-mail: dev-h...@lucene.apache.org
[jira] [Commented] (LUCENE-6013) Remove IndexableFieldType.indexed()
[ https://issues.apache.org/jira/browse/LUCENE-6013?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14177154#comment-14177154 ] Michael McCandless commented on LUCENE-6013: I'll update the javadocs for the format changes ... and I'll compact the current 3 bit flags. Remove IndexableFieldType.indexed() --- Key: LUCENE-6013 URL: https://issues.apache.org/jira/browse/LUCENE-6013 Project: Lucene - Core Issue Type: Improvement Reporter: Michael McCandless Assignee: Michael McCandless Fix For: 5.0, Trunk Attachments: LUCENE-6013.patch, LUCENE-6013.patch Like LUCENE-6006, here's another pre-cursor for LUCENE-6005 ... because I think it's important to nail down Lucene's low-schema (FieldType/FieldInfos) semantics before adding a high-schema. IndexableFieldType.indexed() is redundant with IndexableFieldType.indexOptions() != null, so we should remove it, codecs shouldn't have to write/read it, high-schema should not configure it, etc. Similarly, the FieldInfo.indexed bit is redundant, so I removed it, but I left the sugar API (FieldInfo.isIndexed) and implement it as just checking IndexOptions != null. -- This message was sent by Atlassian JIRA (v6.3.4#6332) - To unsubscribe, e-mail: dev-unsubscr...@lucene.apache.org For additional commands, e-mail: dev-h...@lucene.apache.org
[jira] [Commented] (LUCENE-6013) Remove IndexableFieldType.indexed()
[ https://issues.apache.org/jira/browse/LUCENE-6013?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14177189#comment-14177189 ] Robert Muir commented on LUCENE-6013: - Looks much simpler! I guess the one thing is the duplication of indexed constants (e.g. INDEXED_DOCS_ONLY). I dont think we should make anything sophisticated, lets keep it simple, but right after the last constant we could add something like static assert IndexOptions.values().length == 4 or something? Remove IndexableFieldType.indexed() --- Key: LUCENE-6013 URL: https://issues.apache.org/jira/browse/LUCENE-6013 Project: Lucene - Core Issue Type: Improvement Reporter: Michael McCandless Assignee: Michael McCandless Fix For: 5.0, Trunk Attachments: LUCENE-6013.patch, LUCENE-6013.patch, LUCENE-6013.patch Like LUCENE-6006, here's another pre-cursor for LUCENE-6005 ... because I think it's important to nail down Lucene's low-schema (FieldType/FieldInfos) semantics before adding a high-schema. IndexableFieldType.indexed() is redundant with IndexableFieldType.indexOptions() != null, so we should remove it, codecs shouldn't have to write/read it, high-schema should not configure it, etc. Similarly, the FieldInfo.indexed bit is redundant, so I removed it, but I left the sugar API (FieldInfo.isIndexed) and implement it as just checking IndexOptions != null. -- This message was sent by Atlassian JIRA (v6.3.4#6332) - To unsubscribe, e-mail: dev-unsubscr...@lucene.apache.org For additional commands, e-mail: dev-h...@lucene.apache.org
[jira] [Commented] (LUCENE-6013) Remove IndexableFieldType.indexed()
[ https://issues.apache.org/jira/browse/LUCENE-6013?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14177446#comment-14177446 ] Robert Muir commented on LUCENE-6013: - +1 We may want to followup with some modification of the name indexOptions. It would be nice in the future to maybe just call it 'indexed', just have e.g. setIndexed(DOCS_ONLY) or whatever. But we should remove the boolean here first! Remove IndexableFieldType.indexed() --- Key: LUCENE-6013 URL: https://issues.apache.org/jira/browse/LUCENE-6013 Project: Lucene - Core Issue Type: Improvement Reporter: Michael McCandless Assignee: Michael McCandless Fix For: 5.0, Trunk Attachments: LUCENE-6013.patch, LUCENE-6013.patch, LUCENE-6013.patch, LUCENE-6013.patch Like LUCENE-6006, here's another pre-cursor for LUCENE-6005 ... because I think it's important to nail down Lucene's low-schema (FieldType/FieldInfos) semantics before adding a high-schema. IndexableFieldType.indexed() is redundant with IndexableFieldType.indexOptions() != null, so we should remove it, codecs shouldn't have to write/read it, high-schema should not configure it, etc. Similarly, the FieldInfo.indexed bit is redundant, so I removed it, but I left the sugar API (FieldInfo.isIndexed) and implement it as just checking IndexOptions != null. -- This message was sent by Atlassian JIRA (v6.3.4#6332) - To unsubscribe, e-mail: dev-unsubscr...@lucene.apache.org For additional commands, e-mail: dev-h...@lucene.apache.org
