Deprecate the LoadXML feature

pandajedi/jedicore/JediTaskBuffer.py

@@ -6,7 +6,7 @@
 from pandacommon.pandalogger.PandaLogger import PandaLogger
 
 from pandajedi.jediconfig import jedi_config
-from pandaserver.taskbuffer import ParseJobXML, TaskBuffer
+from pandaserver.taskbuffer import TaskBuffer
 from pandaserver.taskbuffer.InputChunk import InputChunk
 from pandaserver.taskbuffer.JediCacheSpec import JediCacheSpec
 from pandaserver.taskbuffer.JediDatasetSpec import JediDatasetSpec
@@ -72,7 +72,6 @@ def insertFilesForDataset_JEDI(
         nChunksForScout: Any,
         includePatt: list[str],
         excludePatt: list[str],
-        xmlConfig: Any,
         noWaitParent: bool,
         parent_tid: int | None,
         pid: str | None,
@@ -111,7 +110,6 @@ def insertFilesForDataset_JEDI(
                 nChunksForScout,
                 includePatt,
                 excludePatt,
-                xmlConfig,
                 noWaitParent,
                 parent_tid,
                 pid,
@@ -353,7 +351,6 @@ def getOutputFiles_JEDI(
         instantiatedSite: str | None = None,
         isUnMerging: bool = False,
         isPrePro: bool = False,
-        xmlConfigJob: ParseJobXML.dom_job | None = None,
         siteDsMap: dict[Any, Any] | None = None,
         middleName: str = "",
         registerDatasets: bool = False,
@@ -372,7 +369,6 @@ def getOutputFiles_JEDI(
                 instantiatedSite,
                 isUnMerging,
                 isPrePro,
-                xmlConfigJob,
                 siteDsMap,
                 middleName,
                 registerDatasets,

pandajedi/jediorder/ContentsFeeder.py

@@ -188,11 +188,7 @@ def feed_contents_to_tasks(self, task_ds_list: list[tuple[int, list[JediDatasetS
             nFilesPerJob = taskSpec.getNumFilesPerJob()
             # the number of chunks used by scout
             nChunksForScout = 10
-            # load XML
-            if taskSpec.useLoadXML():
-                xmlConfig = taskParamMap["loadXML"]
-            else:
-                xmlConfig = None
+
             # skip files used by another task
             if "skipFilesUsedBy" in taskParamMap:
                 skipFilesUsedBy = taskParamMap["skipFilesUsedBy"]
@@ -616,7 +612,6 @@ def feed_contents_to_tasks(self, task_ds_list: list[tuple[int, list[JediDatasetS
                                 nChunksForScout,
                                 includePatt,
                                 excludePatt,
-                                xmlConfig,
                                 noWaitParent,
                                 taskSpec.parent_tid,
                                 self.pid,

pandajedi/jediorder/JobGenerator.py

@@ -28,7 +28,7 @@
 from pandaserver.dataservice import DataServiceUtils
 from pandaserver.dataservice.DataServiceUtils import select_scope
 from pandaserver.srvcore import CoreUtils
-from pandaserver.taskbuffer import EventServiceUtils, JobUtils, ParseJobXML
+from pandaserver.taskbuffer import EventServiceUtils, JobUtils
 from pandaserver.taskbuffer.FileSpec import FileSpec
 from pandaserver.taskbuffer.InputChunk import InputChunk
 from pandaserver.taskbuffer.JediDatasetSpec import JediDatasetSpec
@@ -1019,11 +1019,6 @@ def doGenerate(
         # register datasets
         registerDatasets = taskSpec.toRegisterDatasets()
         try:
-            # load XML
-            xmlConfig = None
-            if taskSpec.useLoadXML():
-                loadXML = taskParamMap["loadXML"]
-                xmlConfig = ParseJobXML.dom_parser(xmlStr=loadXML)
             # using grouping with boundaryID
             useBoundary = taskSpec.useGroupWithBoundaryID()
             # loop over all sub chunks
@@ -1096,12 +1091,7 @@ def doGenerate(
             fetched_out_sub_chunks = {}
             fetched_serial_numbers = {}
             fetched_parallel_out_map = {}
-            if (
-                xmlConfig is None
-                and (useBoundary is None or not useBoundary["outMap"])
-                and not taskSpec.on_site_merging()
-                and taskSpec.getFieldNumToLFN() is None
-            ):
+            if (useBoundary is None or not useBoundary["outMap"]) and not taskSpec.on_site_merging() and taskSpec.getFieldNumToLFN() is None:
                 for site_name in n_jobs_per_site:
                     # set site name if the chunks instantiate template datasets
                     if to_produce_outputs_merged_later or (taskSpec.instantiateTmpl() and taskSpec.instantiateTmplSite()):
@@ -1123,7 +1113,6 @@ def doGenerate(
                         site_to_instantiate,
                         to_produce_outputs_merged_later,
                         False,
-                        None,
                         siteDsMap,
                         "",
                         registerDatasets,
@@ -1699,17 +1688,7 @@ def doGenerate(
                     except Exception:
                         jobSpec.resource_type = "Undefined"
                         tmpLog.error(f"set resource_type excepted with {traceback.format_exc()}")
-                    # XML config
-                    xmlConfigJob = None
-                    if xmlConfig is not None:
-                        if boundaryID is None:
-                            tmpLog.error(f"failed to get XML config for N={boundaryID}")
-                            return failedRet
-                        try:
-                            xmlConfigJob = xmlConfig.jobs[boundaryID]
-                        except Exception:
-                            tmpLog.error(f"failed to get XML config for N={boundaryID}")
-                            return failedRet
+
                     # num of output files per job
                     if taskSpec.on_site_merging():
                         readStat, taskParamMap = self.readTaskParams(taskSpec, taskParamMap, tmpLog)
@@ -1740,7 +1719,6 @@ def doGenerate(
                             instantiatedSite,
                             isUnMerging,
                             False,
-                            xmlConfigJob,
                             siteDsMap,
                             middleName,
                             registerDatasets,
@@ -1822,10 +1800,7 @@ def doGenerate(
                         tmpBuildFileSpec = copy.copy(buildFileSpec)
                         jobSpec.addFile(tmpBuildFileSpec)
                         paramList.append(("LIB", buildFileSpec.lfn))
-                    # placeholders for XML config
-                    if xmlConfigJob is not None:
-                        paramList.append(("XML_OUTMAP", xmlConfigJob.get_outmap_str(outSubChunk)))
-                        paramList.append(("XML_EXESTR", xmlConfigJob.exec_string_enc()))
+
                     # middle name
                     paramList.append(("MIDDLENAME", middleName))
                     # segment ID

pandajedi/jediorder/JobSplitter.py

@@ -52,7 +52,7 @@ def doSplit(
             nEventsPerJob = taskSpec.getNumEventsPerJob()
             # number of files per job if defined
             nFilesPerJob = taskSpec.getNumFilesPerJob()
-            if nFilesPerJob is None and nEventsPerJob is None and inputChunk.useScout() and not taskSpec.useLoadXML() and not taskSpec.respectSplitRule():
+            if nFilesPerJob is None and nEventsPerJob is None and inputChunk.useScout() and not taskSpec.respectSplitRule():
                 nFilesPerJob = 1
             # grouping with boundaryID
             useBoundary = taskSpec.useGroupWithBoundaryID()

pandajedi/jedirefine/TaskRefinerBase.py

@@ -341,9 +341,6 @@ def extractCommon(
                 self.setSplitRule(taskParamMap, "maxJumboPerSite", JediTaskSpec.splitRuleToken["maxJumboPerSite"])
         if "minCpuEfficiency" in taskParamMap:
             self.setSplitRule(taskParamMap, "minCpuEfficiency", JediTaskSpec.splitRuleToken["minCpuEfficiency"])
-        if "loadXML" in taskParamMap:
-            self.setSplitRule(None, 3, JediTaskSpec.splitRuleToken["loadXML"])
-            self.setSplitRule(None, 4, JediTaskSpec.splitRuleToken["groupBoundaryID"])
         if "pfnList" in taskParamMap:
             self.setSplitRule(None, 1, JediTaskSpec.splitRuleToken["pfnList"])
         if "noWaitParent" in taskParamMap and taskParamMap["noWaitParent"] is True:

pandaserver/taskbuffer/JediTaskSpec.py

@@ -905,10 +905,6 @@ def setErrDiag(self, diag: str | None, append: bool | None = False, prepend: boo
         else:
             self.errorDialog = diag
 
-    # use loadXML
-    def useLoadXML(self) -> bool:
-        return self.check_split_rule("loadXML")
-
     # make VOMS FQANs
     def makeFQANs(self) -> list[str]:
         # no working group