Hi all
We've found a bug in mdlPattern_addAssociative (and in mdlPattern_addAssociative2 as well). If you use code like this:
PatternParams patPar; memset(&patPar, 0, sizeof(patPar)); patPar.modifiers |= PATMOD_SPACE1; patPar.modifiers |= PATMOD_ANGLE1; patPar.modifiers |= PATMOD_COLOR; patPar.space1 = 5000; patPar.angle1 = fc_pi / 4; patPar.color = 3; rc = mdlPattern_addAssociative(edPP, 0, 0, &patPar, NULL, NULL, PATTERN_HATCH, mdlModelRef_getActive());
the function returns SUCCESS, BUT at first glance it seems like no pattern have been added to the shape (edPP). But if you use the "Show Pattern Attributes" tool you will se that PA (pattern angle) is zero and PD (pattern distance) is zero as well.
In V8i SS3 this works just fine...
Regards, Evan
I know what edPP is, I'm trying to see exactly what you are doing with it (because the issue almost certainly lies somewhere in your code not mdlPattern_addAssociative, even if you somehow got away with it in a previous version).
For example, something like the following would be bad:
void func1() { MSElementDescrP edP; mdlElmdscr_read(&edP, ...); func2(edP); mdlElmdscr_add(edP); mdlElmdscr_freeAll(&edP); } void func2(MSElementDescrP edP) { func3(&edP); } void func3(MSElementDescrH edPP) { mdlPattern_addAssociative(edPP, ...); }
Hi Brien
I've now created a very small and simple test application that illustrates the problem.
Private void setSymb(MSElementDescr **edPP, long color) { PatternParams patPar; long rc; memset(&patPar, 0, sizeof(patPar)); patPar.modifiers |= PATMOD_SPACE1; patPar.modifiers |= PATMOD_ANGLE1; patPar.modifiers |= PATMOD_COLOR; patPar.space1 = 2000; patPar.angle1 = fc_pi / 4; patPar.color = color; rc = mdlPattern_addAssociative(edPP, 0, 0, &patPar, NULL, NULL, PATTERN_HATCH, mdlModelRef_getActive()); } extern "C" DLLEXPORT int MdlMain (int argc, char **argv) { UShort typeMask[6]; ScanCriteria *scanCriteria; int scanStatus, scanSize; UInt32 scanPosition; ULong scanBuf[1024], *lP; memset(typeMask, 0, sizeof (typeMask)); typeMask[0] = TMSK0_CMPLX_SHAPE | TMSK0_SHAPE | TMSK0_CELL_HEADER; scanCriteria = mdlScanCriteria_create(); mdlScanCriteria_setModel(scanCriteria, mdlModelRef_getActive()); mdlScanCriteria_setReturnType(scanCriteria, MSSCANCRIT_RETURN_FILEPOS, FALSE, FALSE); mdlScanCriteria_setElementTypeTest(scanCriteria, typeMask, sizeof(typeMask)); do { scanSize = sizeof(scanBuf) / sizeof(short); scanStatus = mdlScanCriteria_scan(scanCriteria, scanBuf, &scanSize, &scanPosition); for (lP = scanBuf; lP < scanBuf + scanSize / 2; ++lP) { MSElementDescr *edP; mdlElmdscr_read(&edP, *lP, mdlModelRef_getActive(), FALSE, NULL); setSymb(&edP, 3); mdlElmdscr_add(edP); mdlElmdscr_freeAll(&edP); } } while (scanStatus == BUFF_FULL); if (scanCriteria) mdlScanCriteria_free(scanCriteria); return(0); }
I've now modified the setSymb function to test the suggestion from Jon and the extracted parameters:
rc = mdlPattern_addAssociative(edPP, -1, -1, &patPar, NULL, NULL, PATTERN_HATCH, mdlModelRef_getActive()); PatternParams extractedPatPar; char msg[80]; mdlPattern_extractAssociative(&extractedPatPar, NULL, &((*edPP)->el), mdlModelRef_getActive(), 0); sprintf(msg, "space1 = %f, angle1 = %f", extractedPatPar.space1, extractedPatPar.angle1); mdlDialog_dmsgsPrint(msg);
The result is exactly the same as if I used the "Show Pattern Attributes" tool: Pattern angle and distance is zero. In SS3 (8.11.09.752) it works just fine. There are no element->file callback registered.
Hi EvanH,
I copied your code as-is and was able to place hatches successfully.
SHOW PATTERN reports correctly “Hatching: PA=45.0000,0.0000, PD=2.0000,0.0000”.
The only side issue I noticed was the call to mdlElmdscr_add (), which is meant for creating a clone of the original shape. This is not a problem, just an observation. The hatch was created on the cloned shape successfully.
Replacing it with mdlElmdscr_rewrite () created the hatch on the original shape successfully too.
Please let me know what I am missing. When you run your code, does it create a hatch at all? Are you trying to update an existing hatch?
Can you send me a DGN file containing a hatch created by your tool in MicroStation V8i Ss3?
Unknown said:I copied your code as-is and was able to place hatches successfully
In which version of MicroStation? This problem seems to be strongly related to version.
Regards, Jon Summers LA Solutions
Hi Sunand,
I'm VERY surprised that you can attach the pattern with the code listed. As Jon points out the problem is strongly related to version. Please check the code on the released version.
The reason for using mdlElmdscr_add instead of mdlElmdscr_rewrite is that the original shape might come from a reference file (when using my original application)
I've attached to .dgn files PatternTest_SS3.dgn and PatternTest_SS4.dgn showing the results when running the test in the two versions.
PatternTest_SS3.dgnPatternTest_SS4.dgn