Liran Zelkha has posted comments on this change. Change subject: core: Improve Dynamic Query SQL generator ......................................................................
Patch Set 12: (10 comments) http://gerrit.ovirt.org/#/c/28135/12/backend/manager/modules/bll/src/test/java/org/ovirt/engine/core/bll/SearchQueryTest.java File backend/manager/modules/bll/src/test/java/org/ovirt/engine/core/bll/SearchQueryTest.java: Line 66: final StoragePoolDAO storagePoolDAO = Mockito.mock(StoragePoolDAO.class); Line 67: final GlusterVolumeDao glusterVolumeDao = Mockito.mock(GlusterVolumeDao.class); Line 68: final NetworkViewDao networkViewDao = Mockito.mock(NetworkViewDao.class); Line 69: final DbEngineDialect dbEngineDialect = Mockito.mock(DbEngineDialect.class); Line 70: final DbFacade facadeMock = new DbFacade() { > build is broken in this file and this line. Done Line 71: @Override Line 72: public DiskDao getDiskDao() { Line 73: return diskDao; Line 74: } http://gerrit.ovirt.org/#/c/28135/12/backend/manager/modules/searchbackend/src/main/java/org/ovirt/engine/core/searchbackend/SearchObjectAutoCompleter.java File backend/manager/modules/searchbackend/src/main/java/org/ovirt/engine/core/searchbackend/SearchObjectAutoCompleter.java: Line 80: // vds - storage domain Line 81: addJoin(SearchObjects.VDS_OBJ_NAME, Line 82: "storage_pool_id", Line 83: SearchObjects.VDC_STORAGE_DOMAIN_OBJ_NAME, Line 84: "storage_pool_id"); > Just curios, is that a bug-fix, or an optimization? Bug fix for the optimization. There's some mixup in the views between storage_id and storage_pool_id. Line 85: Line 86: // cluster - storage domain Line 87: addJoin(SearchObjects.VDC_CLUSTER_OBJ_NAME, "storage_id", SearchObjects.VDC_STORAGE_DOMAIN_OBJ_NAME, "id"); Line 88: Line 387: } Line 388: Line 389: public String getInnerJoin(String searchObj, String crossRefObj, boolean useTags) { Line 390: final String[] joinKey = mJoinDictionary.get(StringFormat.format("%1$s.%2$s", searchObj, crossRefObj)); Line 391: final String crossRefTable = getRelatedTableName(crossRefObj, true); > useTags isn't relevant here? it should always be true? Done Line 392: final String searchObjTable = getRelatedTableName(searchObj, useTags); Line 393: Line 394: return StringFormat.format(" LEFT OUTER JOIN %3$s ON %1$s.%2$s=%3$s.%4$s ", searchObjTable, joinKey[0], Line 395: crossRefTable, joinKey[1]); Line 414: if (useTags) { Line 415: return getRelatedTableName(obj); Line 416: } Line 417: Line 418: return getRelatedTableNameWithOutTags(obj); > s/WithOut/Without Done Line 419: } Line 420: Line 421: private String getRelatedTableNameWithOutTags(String obj) { Line 422: if (obj == null) { http://gerrit.ovirt.org/#/c/28135/12/backend/manager/modules/searchbackend/src/main/java/org/ovirt/engine/core/searchbackend/StorageDomainFieldAutoCompleter.java File backend/manager/modules/searchbackend/src/main/java/org/ovirt/engine/core/searchbackend/StorageDomainFieldAutoCompleter.java: Line 43: Line 44: // building the ColumnName Dict Line 45: columnNameDict.put(NAME, "storage_name"); Line 46: columnNameDict.put(STATUS, "storage_domain_shared_status"); Line 47: columnNameDict.put(DATACENTER, "storage_pool_name::text"); > why did you specify here ::text? due to casting? For some reason in the database it is defined as "unknown", hence we need the cast :-( Line 48: columnNameDict.put(TYPE, "storage_type"); Line 49: columnNameDict.put(SIZE, "available_disk_size"); Line 50: columnNameDict.put(USED, "used_disk_size"); Line 51: columnNameDict.put(COMMITTED, "commited_disk_size"); http://gerrit.ovirt.org/#/c/28135/12/backend/manager/modules/searchbackend/src/main/java/org/ovirt/engine/core/searchbackend/SyntaxChecker.java File backend/manager/modules/searchbackend/src/main/java/org/ovirt/engine/core/searchbackend/SyntaxChecker.java: Line 860: Line 861: // Add special handling (not deleted) to event searches Line 862: if (searchObjStr.equalsIgnoreCase("EVENT")) { Line 863: whereBuilder.add("not deleted"); Line 864: } > not sure I understand this special treatment. Done Line 865: Line 866: // adding WHERE if required and All implicit AND Line 867: StringBuilder wherePhrase = new StringBuilder(); Line 868: if (whereBuilder.size() > 0) { Line 860: Line 861: // Add special handling (not deleted) to event searches Line 862: if (searchObjStr.equalsIgnoreCase("EVENT")) { Line 863: whereBuilder.add("not deleted"); Line 864: } > Oved, we keep in the table of events events that are also deleted, but UI s >From DB perspective, will be the same. And then it's just one more view to >maintain. I prefer to keep it this way. We have a view inflation as it is. Line 865: Line 866: // adding WHERE if required and All implicit AND Line 867: StringBuilder wherePhrase = new StringBuilder(); Line 868: if (whereBuilder.size() > 0) { Line 1111: Line 1112: String tableName; Line 1113: Line 1114: // We will take the table with tags for all subtables Line 1115: // TODO: Optimize this > what about this TODO, is should be done or omitted from the final patch Only in a future patch, so I thought to leave it here for the next patch. It's much more tedious and dangerous change Line 1116: if (conditionType == ConditionType.ConditionwithSpesificObj) { Line 1117: tableName = mSearchObjectAC.getRelatedTableName(objName, true); Line 1118: } else { Line 1119: tableName = mSearchObjectAC.getRelatedTableName(objName, fieldName); Line 1113: Line 1114: // We will take the table with tags for all subtables Line 1115: // TODO: Optimize this Line 1116: if (conditionType == ConditionType.ConditionwithSpesificObj) { Line 1117: tableName = mSearchObjectAC.getRelatedTableName(objName, true); > same here. is it always true? Yes - this is the optimization TODO remark Line 1118: } else { Line 1119: tableName = mSearchObjectAC.getRelatedTableName(objName, fieldName); Line 1120: } Line 1121: if (customizedRelation.equalsIgnoreCase("LIKE") || customizedRelation.equalsIgnoreCase("ILIKE")) { http://gerrit.ovirt.org/#/c/28135/12/backend/manager/modules/searchbackend/src/main/java/org/ovirt/engine/core/searchbackend/SyntaxContainer.java File backend/manager/modules/searchbackend/src/main/java/org/ovirt/engine/core/searchbackend/SyntaxContainer.java: Line 19: private int privateMaxCount; Line 20: private long searchFrom = 0; Line 21: private boolean caseSensitive=true; Line 22: Line 23: public boolean isTags() { > Please rename to something more meaningful Done Line 24: return mOrigText.contains("tag") Line 25: || (getSearchObjectStr() != null && (getSearchObjectStr().equals(SearchObjects.VDC_USER_OBJ_NAME))) Line 26: || (getCrossRefObjList().contains(SearchObjects.VDC_STORAGE_DOMAIN_OBJ_NAME) && (getSearchObjectStr() == null || !getSearchObjectStr().equals(SearchObjects.VDS_OBJ_NAME))); Line 27: } -- To view, visit http://gerrit.ovirt.org/28135 To unsubscribe, visit http://gerrit.ovirt.org/settings Gerrit-MessageType: comment Gerrit-Change-Id: I862873171c6753f8c8863c10c336e981e39dc8cb Gerrit-PatchSet: 12 Gerrit-Project: ovirt-engine Gerrit-Branch: master Gerrit-Owner: Liran Zelkha <lzel...@redhat.com> Gerrit-Reviewer: Barak Azulay <bazu...@redhat.com> Gerrit-Reviewer: Eli Mesika <emes...@redhat.com> Gerrit-Reviewer: Liran Zelkha <lzel...@redhat.com> Gerrit-Reviewer: Oved Ourfali <oourf...@redhat.com> Gerrit-Reviewer: Yair Zaslavsky <yzasl...@redhat.com> Gerrit-Reviewer: automat...@ovirt.org Gerrit-Reviewer: oVirt Jenkins CI Server Gerrit-HasComments: Yes _______________________________________________ Engine-patches mailing list Engine-patches@ovirt.org http://lists.ovirt.org/mailman/listinfo/engine-patches
