[AINode] Integrate python code standardization #15593
Merged
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Codecov ReportAll modified and coverable lines are covered by tests ✅
Additional details and impacted files@@ Coverage Diff @@
## master #15593 +/- ##
=========================================
Coverage 39.05% 39.05%
Complexity 198 198
=========================================
Files 4740 4740
Lines 307786 307786
Branches 38478 38478
=========================================
+ Hits 120203 120211 +8
+ Misses 187583 187575 -8 ☔ View full report in Codecov by Sentry. 🚀 New features to boost your workflow:
|
CRZbulabula
changed the title
CRZbulabula
changed the title
ycycse
approved these changes
May 29, 2025
|
Is there a command to automatically format all the code? |
Definitely! Run |
This should be merged quickly
|
jt2594838
added a commit
that referenced
this pull request
Jul 24, 2025
* Pipe: Fixed the bug that disruptor queue close may block forever by not waiting real-time tsFile close (#15464) * Pipe: Added max / min function to aggregate-processor (#15459) * fix: jline history (#15463) * Interrupt wal-delete thread when WALManager is closed (#15442) * Load: Fixed the index update logic of tree-model (#15474) * avoid external memory space supplied when wal mode is disabled (#15476) * Pipe: Enable flushing storage engine after last terminate event reported (#15465) Co-authored-by: Steve Yurong Su <[email protected]> * Support LEFT and RIGHT JOIN in TableModel * Pipe: Added metrics for batch & handler size & compression ratio & compression time for iotdb connector (#15397) Co-authored-by: Steve Yurong Su <[email protected]> * fix npe (#15477) * [AINode] Introduce built-in timer_xl model for forecasting (#15468) * latest timerxl code * add dependency of einops in ainode/pyproject.toml * Update pyproject.toml --------- Co-authored-by: OswinGuai <[email protected]> Co-authored-by: YangCaiyin <[email protected]> * fix: Boundary checks for maxStandardError in APPROX_COUNT_DISTINCT. * Subscription IT: Refactor how IT is categorized for intro table model tests (#15485) * Subscription: tree/table model isolation for topic/consumer/subscription entities (#15484) * Feature/cppclient table redirection (#15439) * Implement high availability CPP client under table model * Implement high availability CPP client under table model; Refactored the CPP client codebase * Implement high availability CPP client under table model;Refactored the CPP client codebase * Refactored the CPP client codebase * Refactored the CPP client codebase * fix compilation error * Extended data type support to cover STRING, BLOB, DATE, and TIMESTAMP; Fixed memory leaks in test cases (Valgrind verified) * fix compilation error on ubuntu14.04/boost1.60 * Implement high availability CPP client under table&tree model;Refactored the CPP client codebase;Extended data type support to cover STRING, BLOB, DATE, and TIMESTAMP * Support built-in forecast table function for table model * Fixed the reserved word introduced by tree view in tree model SQL & Enhanced the tree view IT & Handled the cases in field / measurement names & Temporarily banned source measurements with the same name (#15475) * Async Load: Set the database name null when appropriate & Enable auto-creation of the database by reading the isAutoCreateSchemaEnabled flag from config (#15489) * Use reference time position for PartitionTableAutoCleaner * add it & fix bugs for TreeDeviceViewScan * [AINode] Support built-in inference of table/tree model for time_xl * Fixed the issue of abnormal output when the input length of timerxl is not a multiple of 96 (#15495) Co-authored-by: OswinGuai <[email protected]> * docker-compose file:fix a slip of the pen and add the new way to set memory with environment variables (#15496) * docker: support configuring JVM memory environment variables in docker-compose file:IOTDB_JMX_OPTS for datanode,CONFIGNODE_JMX_OPTS for confignode * recover the place of setting ON_HEAP_MEMORY and OFF_HEAP_MEMORY to fix atmos * docker-compose file:fix a slip of the pen and add the new way to set memory with environment variables * fix thread leak (#15497) * Implememt auto release technique for Blob Allocator (#15491) * add auto release logic * add license * enhance Signed-off-by: OneSizeFitQuorum <[email protected]> * Refine API Signed-off-by: OneSizeFitQuorum <[email protected]> * fix typo Signed-off-by: OneSizeFitQuorum <[email protected]> --------- Signed-off-by: OneSizeFitQuorum <[email protected]> Co-authored-by: MrQuansy <[email protected]> * [Table Model Subquery] Support correlated scalar subquery * Fix auto cleaning partition table bug * Support correlated quantified comparison * Applied the new functional definition to tree to table view * perf: Optimize the code structure and add comments for CaseWhen * Remove SessionPool sensitive info (#15506) * IoTV2: Close receiver's executor to prevent request stuck when stopping dn (#15505) * close receiver's executor * typo * [AINode] Support multi-column as input in built-in model inference * Pipe: Fixed the bug that sorter may use wrong value when there are duplicated timestamps & Refactor (#15488) * Load: fix memory leak when failed in 2nd phase (#15503) * Fixed the bug that update / delete devices with no effect may end up changing all the devices * Datanode cannot start when metaDataCacheEnable=false * Pipe: implement external sources strategy and MQTT extractor (#15275) Co-authored-by: Steve Yurong Su <[email protected]> * Fix bugs in accumulator of UDAF * Make foreacast tvf's table relation default order by timecol desc * Fix case of process empty block in Streaming(Hash)AggregationOperator * Fix IAuthorCache never update * Subscription IT: add IoTDBSubscriptionTopicIT for table model (#15525) * Load: fix memory leak when failed to send chunk data in first phase (#15518) This commit addresses a memory leak issue in the load process by ensuring that memory usage is reduced even when the dispatch of a piece node fails. The key changes include: - Storing the result of the dispatch call in a variable (isDispatchSuccess) before reducing memory usage. - Deducting the memory usage prior to checking the dispatch result to avoid leaks. - Returning false immediately after the reduction when the dispatch fails. * Load: fix data loss when loading different files with same name in the same directory (#15527) * Subscription: fully managed tsfile parsing process for tsfile format topic (#15524) * Subscription: support drop subscription from session & intro allTopicMessagesHaveBeenConsumed for pull consumer (#15486) - Introducing a new DropSubscriptionTask and corresponding updates across executor interfaces, cluster executors, and managers. - Updating the SQL parser to recognize the DROP SUBSCRIPTION statement. - Enhancing client session APIs and internal data models to support subscription deletion. * Implement ASOF LEFT Join * Fix TVF return unexpected result when passing ORIGIN argument * forbid changing encrypt key again (#15509) * forbid changing encrypt key again * code review and forbid changing encrypt key totally * delete parameter encrypt_flag (#15514) * delete parameter encrypt flag * update tsfile version * Enable the kernel to support passing topics to MQTT custom plugins (#15523) * Fix cases of AggregationTableScan used with Join * Set correct operator type for LastQueryAggTableScanOperator * Fix DataNodeWrapper didn't set Timezone * Pipe: Fixed the bug that the insertion for newer tsFile in one region may report progress beyond the older tsFile when it is not flushed (#15515) * Fix null pointer exception casued by measurements values and measureSchemas values are not consistent (#15532) * fix the problem of null pointer below: NullPointerException: Cannot invoke "org.apache.tsfile.write.schema.MeasurementSchema.serializeTo(java.io.OutputStream)" because "this.measurementSchemas[i]" is null * fix the problem of null pointer for table model below: NullPointerException: Cannot invoke "org.apache.tsfile.write.schema.MeasurementSchema.serializeTo(java.io.OutputStream)" because "this.measurementSchemas[i]" is null * Enable unsetting of the current database when the database is dropped * Fix getColumns in JDBC DataBaseMetaData (#15533) * Fix getColumns in jdbc DataBaseMetaData * Fix getColumns in jdbc DataBaseMetaData +1 * Fix getColumns in jdbc DataBaseMetaData +2 --------- Co-authored-by: 2b3c511 <[email protected]> * Pipe: enable passing topic names to customized formatter in MQTT extractor (#15545) * Pipe: refine memory cost estimation of Binary types for more precise memory control (#15540) * Pipe: Added skip-if semantic to iotdb-connector (#15530) * NonAlignedTreeDeviceViewScan may return limit + offset lines result when the offset is pushed down * Modify the exception and message of wrong 'setConfiguration' Statement in table model (#15541) * Modify the exception message of wrong setConfigurationStatement in table model * modify Exception * Pipe: Optimized the hybrid switching algorithm (#15528) Co-authored-by: Steve Yurong Su <[email protected]> * Added privilege for information_schema & Removed the privilege requirement for show pipePlugins * Pipe: Fix Active Load not getting DataBaseName correctly (#15551) * add patch (#15499) * Subscription: decrease reference count for managed parsed tsfile event (#15555) * reset metric (#15556) * Push down offset into TableScanNode when there is only one data region * Fixed the bug that the first query may be rejected when read only in table model for simple consensus * Remove all kinds of ProcedureException's subclass & Fix a security problem (#15563) * Fix default paramters cannot stop ainode (#15531) * Feature/cpp client tsblock (#15511) * cpp client tsblock * rename Common.cc to Common.cpp * Fixed some unit test issues * Modified some Exception message prompts * Change Error Code from 301 to 719 in case of sort memory not enough * [AINode] Return ainodeClient in clientPool * Add a judgment in order to avoid occur NPE when one value of measurementSchemas is null (#15574) * Change Error Code from 301 and 305 to NO_AVAILABLE_REPLICA(1012) while all replicas are unavailable * Cancel query which contains FI sent to shutdown datanode by mistake * Pipe: Enable synonym detection for OpcUaConnector & Refactor remaining ids -> tags (#15573) * Pipe: add mark-as-general-write-request parameter in pipe to force forwarding event (#15572) * Load: Support "database" substitution for "database-name" (#15562) * Perfect error message of ASOF JOIN * Optimize sort properties of JoinNode in case of JOIN USING * Correct error messages when encountered IllegalPathException (#15559) * Addresses an issue where the error log message is inaccurate when an IllegalPathException occurs. Addresses an issue where the error log message is inaccurate when an IllegalPathException occurs cause by a insert sql is executed. A create sql is added in order to validate if the exception message is legal when IllegalPathExeption occurs. * format code * Fix the issue it scripts can not be past. * Feat: Add APPROX_MOST_FREQUENT Aggregation Function * IoTV2: Make consensus event retry forever. (#15565) * IoTV2: Refine receiver and sender transfer logic to prevent stuck (#15569) * Make it easier for users to extend ClientRPCServiceImpl to export more rpc services * Fix 15569 and 15565 compile error (#15591) * Pipe/Load: Assign distinct progress indexes for loading tsfiles in time partitions to reduce pipe reprocessing after restart & Decrease pipe heartbeat interval (#15583) * Pipe: Optimize Batch and WAL memory allocation algorithms (#15534) * Pipe: Optimize Drop Pipe high priority tasks cannot obtain SubTask object lock (#15404) Co-authored-by: Steve Yurong Su <[email protected]> * Subscription: retain tsfile events in tsfile batch to avoid premature commit (#15598) Consider the historical data export snapshot scenario: 1. The events delivered upstream are, in order, the tsfile event and the termination event. 2. The tsfile event is parsed into multiple tablet events, and then the reference count of the tsfile event is set to 0 (should report as false). 3. Assuming that for some reason the tablet events were not sent to the peer in time, the reference count of the transfer termination event is set to 0 (should report as true). 4. At this point, because the tablet events were not enriched with a commit id (see Subscription: fully managed tsfile parsing process for tsfile format topic #15524), the termination event successfully marks the corresponding DR complete, which in turn leads to data loss. * Subscription: support unsubscribe from completed topics under client heartbeat thread (#15595) * [AINode] Add Holt-Winters algorithm and fix related bugs * Added nodes/config_nodes/data_nodes table for information_schema & Fixed the header of show ainodes * Change keep_input to preserve_input * Changed the error code of create table view in tree dialect to 701 & Fixed the bug that table device is treated as a device with template * Reorder the show databases / tables dataset by database/table name * Subscription: apply IoTConsensusV2 as cluster mode for integration test (#15546) * IoTV2: Fix the issue of resend lots of replicated files when DN restarts. (#15600) IoTV2 use RecoverProgressIndex to represent the replicate progress of its consensus pipe. the RecoverProgressIndex is consist of binary tuple, where key represents the source data node id of the writing data and value represents the operation sequence for one leader, naming nodeA, it may contain both local write data(RecoverProgress's key is local data node id) and data replicated from other old leader.(RecoverProgress's key is old leader's data node id) if nodeA is restarted, consensus pipe will try to flush and collect all tsFiles on it. Those tsFiles may be high-level which are compacted from local write 0-level tsFile and replicated 0-level tsFile. In that case, Those high-level tsFiles' RecoverProgress may contain both local data node id key and old leader's data node id key. IoTV2 will only transfer tsFiles with progressIndex greater than startIndex, for those tsFiles are considered to contain unprocessed data and are necessary to be resent. The problem is, the comparison of RecoverProgressIndex is quite strict (for more details please refer to RecoverProgressIndex source code.), causing some high-level tsFiles that don't contain local writing data are still considered necessary to be resent. So for IoTV2, we only need to consider local write data and its dedicated progressIndex for deciding whether tsFiles contain unprocessed data. * Pipe: Fixed the problem of null pointer and incorrect Tablet format being modified during TSFile construction (#15602) * [AINode] Integrate Sundial as built-in model (#15586) * [AINode] Remove TimerXL from the builtin list (#15616) * [AINode] Integrate python code standardization (#15593) * Introduce row pattern recognition * Fixed the bug that the tree model database deletion may clear the table model cache * Pipe: Unbinded the default meter for remaining insertion event count for hybrid degrading & Changed some default values of related properties (#15615) * Fixed the bug that the view from tree can be written * Perfect errMsg of unsupported join criteria in JoinNode * Update last cache in load (#15604) * add load update last cache strategies * add cacheLastValuesForLoad * add cache last values for load * add memory control when caching last values * fix * fix UPDATE_NO_BLOB does not take effect * fix comments * spotless * ignore perf test * Pipe: Fixed the issue that the DropPipe Connector subtask process waits too long for the callback thread to end (#15605) * Pipe: Fixed the bug that config receiver cannot detect runtime password change (#15621) * Load: Introduce thread pool for tablet conversion and insertion (#15611) * Load: Check whether TsFile is a tree model when DataBase is not specified (#15592) * Pipe: Add retry when TsFile parsing failed to avoid race among processor threads (#15624) * Ignore mem chunks whose data type is not matched * Finish window function query planning stage * Update release infomation for v2.0.3 (#15629) * Avoid the NPE problem occur that tsTable is null (#15618) * Avoid the NPE problem occur that tsTable is null, and ensure if table is not exist and catch this exception to tell leader datanode for retry or do nothing. * Avoid the NPE problem occur that tsTable is null via retry to get the tsTable. * Add license description * [Py-client] Fix test_dialect failed (#15630) * Revert "Update dockerfile to use the latest ubuntu 24 (#14903)" (#15632) This reverts commit 8e4fef3. * fix datanode's memory configuration (#15634) * Fix ConfigNodePropertiesTest (#15635) * fix (#15633) * Skip generate distributed plan when treeDBName is null * Ignore not exist device when storage engine is not ready * Bump version to 2.0.5-SNAPSHOT (#15642) * resolve the NPE problem (#15646) * Add cached device number info for last query * Pipe: Fix the problem that the initialization of Pipe consumer memory module caused DN writing to get stuck (#15645) * Pipe: Fix the problem that the initialization of Pipe consumer memory module caused DN writing to get stuck * fix * fix * fix * Fix tree ttl in table view scan * When the queries aligned series has some inconsistent data types in memtable, the other column will also be ignored * fix compaction path util npe (#15653) * Pipe: Add retry when TsFile parsing failed to avoid race among processor threads (follow up #15624) (#15644) * Pipe: Add retry when TsFile parsing failed to avoid race among processor threads * refactor * refactor * refactor * [remove datanode] Accelerate GCR load balancing implement (#15535) * Fix SessionConnection opens too much transport without closing (#15649) * [AINode] Enable log record by default (#15654) * Pipe: Changed the default value about batch delay configuration (#15657) * Fix * Fix * Refactor batch * Update CommonConfig.java * Update PipeConnectorSubtask.java * Update CommonConfig.java * Update CommonConfig.java * [AINode] Refactor the built-in TimerXL (#15655) * Revert "Fix process of last query if intermediate state of last cache is read when query concurrently" * Pipe: Fix TSFile transfer blocking InsertNode sending (#15666) * Pipe: Fix TSFile transfer blocking InsertNode sending * fix * fix * fix * fix * fix * fix * fix * Pipe/IoTV2: Try to persist progressIndex in local for resend event more efficently (#15599) * feat: progressIndex persistent in local process * fix ci and support cancel persist task * pipeMeta in cn doesn't need to persist progressIndex * remove duplicate * clean progressIndex background service * fix and add conf * Update PipeConfig.java * Update PipeMetaDeSerTest.java * cp-gap * Update PipeTaskMeta.java * partial * Update PipeRuntimeMeta.java * Update PipeRuntimeMeta.java * Partial * Update PipeTaskMeta.java * Added config * clean * Change default * Update PipeConnectorConstant.java * Partial * Update PipeTaskMeta.java * Update PipeTaskMeta.java --------- Co-authored-by: Caideyipi <[email protected]> * Pipe: Add InsertNode & tsFile transmission time metrics (#15668) * Pipe: Add InsertNode transmission time * Pipe: Add InsertNode transmission time * update * update * fix * fix * fix * Pipe: Fixed the metric NPE caused by batch disable (#15672) * Add print-wal tool and logs when failures of reading wal files (#15416) * Add print-wal tool and logs when failures of reading wal files * fix position printed * Subscription: unsubscribe completed topics when remove consumer config (#15660) * Subscription: optimize prefetching queue report state overhead (#15664) * setup * add comments * Pipe: Fixed multiple bugs (#15674) (#15683) 1. Doubled the TCP connection num 2. Made the memory for wal and batch fixed 3. Changed the default batch time to 20ms 4. Disabled the batch load for wal cache 5. Separated the tsFile and insertNode client --------- Co-authored-by: luoluoyuyu <[email protected]> * Fix Session reconnection increase connection number (#15677) * Fix Session reconnection increase connection number * Fix Session reconnection increase connection number * Optimize the implement * Pipe: Fix the problem that the Pipe memory block fails to call the Wait function (#15687) * Support dataset module for iotdb in ainode (#15686) * [AINode] Add a model ckpt path retrieve interface (#15689) * Implemented fast last query for tree model with prefix path (#15678) * Fast last query on local * Remove debug settings 2 * Bug fix * Update SessionExample.java * Update SessionConnection.java * Fix * logger * Update DualKeyCacheImpl.java * Update DualKeyCacheImpl.java * Pipe: Changed the default batch memory request size to avoid CI failure (#15690) * Fix * Update PipeConsensusTransferBatchReqBuilder.java * Update IoTDBConfig.java * Update PipeConnectorConstant.java * Update PipeConnectorConstant.java * Fix * Update IoTDBConfig.java * Pipe: Fixed the problem of not being able to write normally due to insufficient memory (#15701) * Pipe: Fixed the problem of not being able to write normally due to insufficient memory * fix * fix * fix * [AINode] Enhance the robustness of AINode (#15695) * add debug entry (#15694) * [AINode] Update python dependencies (#15712) * more accurate mermory size (#15713) * IoTV2: Remove events that do not retry in leader's transfer buffer & fix concurrency bug of retryQueue & fix retry interval & refine log (#15684) * remove events that do not retry * refine log & fix retry interval & fix concurrency issue of retryQueue * fix cpp client processing null value (#15700) * User password encryption upgraded from MD5 to SHA256. (#15680) * User password encryption upgraded from MD5 to SHA256. Existing users' passwords will be upgraded upon next login. * fix some issues * fix login4pipe * Fix arrary index out-of-bounds in agg query with align by device when cross region * Subscription IT: intro retry rule for flaky tests (#15698) * Pipe: Add retry for tablet batch req to avoid retransmission when memory is insufficient (#15715) * Load: Convert to tablets when node is read-only (#15693) * Pipe: Fix the inconsistency between schema and values columns in the process of building tsfile (#15625) * use environment variable and PBKDF to generate main encrypt key (#15711) * use environment variable and PBKDF to generate main encrypt key * modify the error information * Put last query IT only in 1c1d * Support order by time for Agg align by device query with cross region data * Skip update-cache logic when loading TsFile in IoTV2Consensus (#15729) * fix datanode check (#15730) * Enhance "remove region" sql to handle the case of DataNode not exist (#15728) * [remove datanode] Fix ArrayIndexOutOfBoundsException in computeInitialDbLoad (#15718) * Fixed the bug that the characters surrounded by "" cannot be parsed in tree view path * Using join to query views of non-aligned series may throw NullPointerException * Fixed the bug that the old view's cache is not cleared when altering its name * Fix timeout didn't pass to client bug * Subscription: fix drop subscription from session under table model (#15725) * Return empty result set for non aligned series when no filed columns can be obtained from view * fix exception msg in RPR * Fixed the potential failure for table view query with pattern beyond databases * Conf Template: Fix wrong symbol (#15742) * Pipe: Set WAL to uncompressed when using real-time sync (#15733) * Pipe: Fixed the insert node EMA logic for degrading (#15741) * Fix RPROperator blocking behavior * fix tsfile checker leak (#15710) * Avoiding error log by removing ThreadLocal for Mqtt Client * Fix TypeProvider and AggregationDescriptor serde in Templated Query * Reset global time filter for each device in table view scan * Add more tests for removing datanode (#15749) * Fix stuck when stopping a DataNode with large unremovable WAL files (#15727) * Fix stuck when stopping a DataNode with large unremovable WAL files * spotless * add shutdown hook watcher * Fix logDispatcher stuck * add re-interrupt * Fixed the device NPE for table view in view device & Added semantic check to prohibit the last data query for prefix paths with "*" & Refactored metric log of tag number * Pipe: Delete the extra tags created by the PipeWALInsertNodeCacheMetrics (#15761) * Optimize group by query in ClientRPCServiceImpl to reduce cpu usage * Fix thread leak in session while connected DN is down * Pipe: Fixed the NPE when closing sync connector (#15772) * Subscription: return immediately when unsubscribe empty complete topics (#15770) * Pipe: Fixed the NPE caused by the compression timer (#15782) * Pipe/IoTV2: Persist progress index locally before shutdown to accurate recovery after restart (#15779) * Opt error msg while time column is empty string * Improve argument validation for approx_most_frequent function * [Remove DataNode] Increase the waiting time for removing DN test (#15769) * Fix multiple bugs in window function * [AINode] Unify .safetensors model loading through huggingface interfaces (#15756) * update go:package_prefix to v2 (#15795) * Fix the returning state of NotifyRegionMigrationProcedure (#15797) * Export TsFile: manual commit instead of auto commit (#15798) * Pipe: Add upper bound check for Pipe request decompression buffer (#15699) Co-authored-by: Steve Yurong Su <[email protected]> * Fixed the issue with deeply nested Or Filter in NonAlignedTreeViewScan * Prevent the contents of log_datanode_sample_query.log and log_explain_ananlyze.log from being output to log_datanode_all.log (#15800) * Avoid NPE in fetch FI Info * Add new parameter [ignoreNull] and supported data type for VARIATION Function (#15628) * Parallel disptach read fragment instance * Throw CANNOT_FETCH_FI_STATE(722) instead of 301/305 while DN restarting * Fix sort elimination when table join self in TableModel * Pause remove DN tests * Clarify the exception message when using a closed session (#15811) * Avoid 301/305 code about fetch schema failed * Make timecol be case-insensitive * [AINode] Update the class fields of IoTDB dataset (#15799) * Rest Service add FastLastQuery method * Enhance it for cluster-restart tests (#15794) * Fix stuck when stopping a DataNode with large unremovable WAL files * spotless * add shutdown hook watcher * Fix logDispatcher stuck * add re-interrupt * Enhance IT frame to run tests that restart the cluster * code refinement * modify test configuration * fix test * IoTV2: Refactor replicate index so that it is shared at the pipe task level & Add some log for delete local peer (#15815) * use custom replicate index for each consensus pipe * test conf * fix * Revert "test conf" This reverts commit f0f13af. * fix review * Fix elimination rule of TopK when generate DistributionPlan * Correct status code from 301 to 719 while memory not enough & Record release resource blocked too long by driver (#15828) * [AINode] Add TimerXL to the built-in list (#15832) * [AINode] Modify dataset module for table-model training (#15816) * Add WAL entry number for one TsFile metric (#15818) * [AINode] Support customize session config in AINode (#15834) * [AINode] Refactor the result columns of "show models" SQL (#15831) * column refactored and model can be auto downloaded * Update AINodeBasicIT.java --------- Co-authored-by: Yongzao <[email protected]> * [AINode] Simplify the CREATE MODEL SQL for model training (#15840) * [AINode] Quick bug fix patch (#15841) * Fixed the issue that columns may not be pruned when querying non-field columns in tree view scan * Subscription: add config to enable or disable subscription (#15820) * setup * minor improve * disable subscription executor * fix ExportTsFileTestIT * fix IoTDBDatabaseIT * Resolve IndexOutOfPartitionBound and symbol not found in window functions. * fix a problem about data type of field was displayed wrong. * delete unuseful code. * The logic that the data type of alignedChunkMetadata is different of the data type of measurement after changed which it can be considered is an defect in the FastAlignedSeriesCompactionExecutor. * To Aligned Series: Resolve the problem that only the first data type of chunkMetadata is adjusted. Add the logic that the statistics of chunkMetadata need be analyzed with new data type. In addition, the data type of valueChunkMetadata need be updated. To NonAligned Series: Avoid the problem that chunk in the memory need be still rewrited when data type don't be changed. * Adjust the chunk object need be merged statistics. Don't need rewrite chunk data when data type have not been changed. * Because this problem is not exist as below: "The logic that the data type of alignedChunkMetadata is different of the data type of measurement after changed which it can be considered is an defect in the FastAlignedSeriesCompactionExecutor." So delete unuseful code. * Fix some problems after merge code from master branch. * Resolve the problem can not find symbol that variable ColumnCategory. * Resolve the problem the query result is null from MemTable when don't execute the flush operation. * Resolve the problem CI can not be passed sucessfully. This is reason that the problem cause: When type of timeseries is the align, the chunkMetadata of chunkMetadataElement contain two inner ChunkMetadata struct, only change their value with reference, don't need change value with the "set" function. --------- Signed-off-by: OneSizeFitQuorum <[email protected]> Co-authored-by: Caideyipi <[email protected]> Co-authored-by: CritasWang <[email protected]> Co-authored-by: Jiang Tian <[email protected]> Co-authored-by: Steve Yurong Su <[email protected]> Co-authored-by: Weihao Li <[email protected]> Co-authored-by: Peng Junzhi <[email protected]> Co-authored-by: jtmer <[email protected]> Co-authored-by: OswinGuai <[email protected]> Co-authored-by: YangCaiyin <[email protected]> Co-authored-by: FearfulTomcat27 <[email protected]> Co-authored-by: VGalaxies <[email protected]> Co-authored-by: Hongzhi Gao <[email protected]> Co-authored-by: Jackie Tien <[email protected]> Co-authored-by: Zhenyu Luo <[email protected]> Co-authored-by: Yongzao <[email protected]> Co-authored-by: shuwenwei <[email protected]> Co-authored-by: YangCaiyin <[email protected]> Co-authored-by: changxue2022 <[email protected]> Co-authored-by: Potato <[email protected]> Co-authored-by: MrQuansy <[email protected]> Co-authored-by: Liao Lanyu <[email protected]> Co-authored-by: Haonan <[email protected]> Co-authored-by: Zikun Ma <[email protected]> Co-authored-by: nanxiang xia <[email protected]> Co-authored-by: Chen YZ <[email protected]> Co-authored-by: jintao zhu <[email protected]> Co-authored-by: wenyanshi-123 <[email protected]> Co-authored-by: Summer <[email protected]> Co-authored-by: 2b3c511 <[email protected]> Co-authored-by: William Song <[email protected]> Co-authored-by: Li Yu Heng <[email protected]> Co-authored-by: 张正明 <[email protected]> Co-authored-by: Le Yang <[email protected]> Co-authored-by: Zhihao Shen <[email protected]> Co-authored-by: Xiangpeng Hu <[email protected]> Co-authored-by: luoluoyuyu <[email protected]> Co-authored-by: Le Yang <[email protected]> Co-authored-by: Zeyu Zhang <[email protected]> Co-authored-by: CloudWise-Lukemiao <[email protected]> Co-authored-by: Yongzao <[email protected]>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
To format all the python code in this project, we integrate the black && isort and add a new CI workflow