lili/OCCT
1
0
Fork 0
mirror of https://github.com/Open-Cascade-SAS/OCCT.git synced 2026-09-06 04:37:55 +08:00
Code Issues Packages Projects Releases Wiki Activity

0029830: STEPCAFControl_Reader poor performance - quadratic dependence

Browse Source 
Various performance improvements in STEP read/write algorithms:
- Search for the  label of a shape or component shape is improved using map mechanism instead of brute force iteration.
- Invariant FindEntities() is moved out of the loop in the method getStyledItem in STEPCAFControl/STEPCAFControl_Writer.cxx.
- A pointer to the end of binders chain is added in Transfer_Binder class to speed up adding a binder to the chain.
- Small fixes are added  to eliminate excess copying of handles, calls of handle DownCasts and so on.

Stack overflow is removed during destruction of STEP model with long chains of Transfer_Binder.
It is possible to use the Draw commands ReadStep and WriteStep to read/write from the session without accessing the disk file (use '.' for the file name).

Performance test cases for STEP reading/writing have been added.
This commit is contained in:
msv
2018-06-01 14:38:39 +03:00
committed by bugmaster
parent 0a96e0bbc4
commit 63cdf48ec1
17 changed files with 349 additions and 185 deletions
Download Patch File Download Diff File Expand all files Collapse all files
src/STEPControl/STEPControl_ActorRead.cxx
+13 -10
View File
@@ -658,7 +658,7 @@ static void getSDR(const Handle(StepRepr_ProductDefinitionShape)& PDS,
Handle(TransferBRep_ShapeBinder) STEPControl_ActorRead::TransferEntity(const Handle(StepRepr_NextAssemblyUsageOccurrence)& NAUO ,
const Handle(Transfer_TransientProcess)& TP)
{
Handle(TransferBRep_ShapeBinder) shbinder;
Handle(TransferBRep_ShapeBinder) shbinder;
Handle(StepBasic_ProductDefinition) PD;
const Interface_Graph& graph = TP->Graph();
gp_Trsf Trsf;
@@ -685,21 +685,24 @@ Handle(TransferBRep_ShapeBinder) STEPControl_ActorRead::TransferEntity(const Han
// find real ProductDefinition used rep
Interface_EntityIterator subs3 = TP->Graph().Sharings(rep);
for (subs3.Start(); subs3.More(); subs3.Next()) {
if ( subs3.Value()->IsKind(STANDARD_TYPE(StepShape_ShapeDefinitionRepresentation))) {
DeclareAndCast(StepShape_ShapeDefinitionRepresentation,SDR,subs3.Value());
const Handle(Standard_Transient)& aSubs3Val = subs3.Value();
if (Handle(StepShape_ShapeDefinitionRepresentation) SDR =
Handle(StepShape_ShapeDefinitionRepresentation)::DownCast (aSubs3Val))
{
Handle(StepRepr_ProductDefinitionShape) PDS1 =
Handle(StepRepr_ProductDefinitionShape)::DownCast(SDR->Definition().PropertyDefinition());
if(PDS1.IsNull()) continue;
Interface_EntityIterator subs4 = graph.Shareds(PDS1);
for (subs4.Start(); subs4.More(); subs4.Next()) {
Handle(StepBasic_ProductDefinition) PD1 =
Handle(StepBasic_ProductDefinition)::DownCast(subs4.Value());
if(PD1.IsNull()) continue;
PD=PD1;
for (subs4.Start(); PD.IsNull() && subs4.More(); subs4.Next())
{
PD = Handle(StepBasic_ProductDefinition)::DownCast(subs4.Value());
}
}
else if(subs3.Value()->IsKind(STANDARD_TYPE(StepRepr_ShapeRepresentationRelationship))) {
SRR = Handle(StepRepr_ShapeRepresentationRelationship)::DownCast(subs3.Value());
else if (aSubs3Val->IsKind (STANDARD_TYPE(StepRepr_ShapeRepresentationRelationship)))
{
// NB: C cast is used instead of DownCast() to improve performance on some cases.
// This saves ~10% of elapsed time on "testgrid perf de bug29* -parallel 0".
SRR = (StepRepr_ShapeRepresentationRelationship*)(aSubs3Val.get());
}
}
}