Skip to main content
  • EN

I am just starting to test out 2022.2. Using a ODBC 3.x connection with SQL executor and getting errors I have never seen before.

 

.....was skipped as it was not a valid SELECT statement

BADNEWS: Attribute name 'fme_type' removed from schema feature as it doesn't have a type. Current value is 'fme_no_geom'

(C:\\code\\fme\\foundation\\framework\\control\\metamap.cpp:1872) class StatusInfo *__cdecl STFMetaMapper::mapFeature(class STFFeature &,bool,bool,bool)

BADNEWS: Cannot convert null string into an integer

(C:\\code\\fme\\foundation\\framework\\util\\stf\\stfutil.cpp:347) class StatusInfo *__cdecl STF_stringToUInt(const class ObsoleteString &,unsigned int &)

... Last 2 lines repeated 40 times ...

 

The query still runs and I get results. The query works fine in 2021 with no problems.

The job runs multiple statements and I don't want 1000+ errors every time the job runs.

If I switch to only data features i still get the error:

BADNEWS: Cannot convert null string into an integer

(C:\\code\\fme\\foundation\\framework\\util\\stf\\stfutil.cpp:347) class StatusInfo *__cdecl STF_stringToUInt(const class ObsoleteString &,unsigned int &)

... Last 2 lines repeated 3736 times ...

 

So what's up?

 

 

 

Hi @robotix​ 

It appears like you have LOG_DEBUG turned on.

You can turn off the message by navigating

Go-To Tools --> FME Options --> Translation --> Log Message Filter --> Log Debug 

and Uncheck the box.

That doesn't explain the errors. I want to know why the errors exist, not how to ignore them.

Having log debug on will sometimes show errors and warnings which you can't control and which have no effect on the translation. It's best to leave it off unless the regular log messages aren't sufficient.

Maybe a solution for the "BADNEWS: Cannot convert null string into an integer" is to create the SQL statement in an AttributeCreator outside of the SQLExecutor. Is there an empty parameter which needs to be an integer?

Having log debug on will sometimes show errors and warnings which you can't control and which have no effect on the translation. It's best to leave it off unless the regular log messages aren't sufficient.

Maybe a solution for the "BADNEWS: Cannot convert null string into an integer" is to create the SQL statement in an AttributeCreator outside of the SQLExecutor. Is there an empty parameter which needs to be an integer?

Nah, tried that.

The error happens right after a row is selected and FME tries to add the result to a spatial index, even though I am using a non-spatial MSSQL connection.

 

FFS writer: Writing 1 feature(s) to file 'C:\\Users\\..............1 15 fo 1 RESULT 4 fc94bf7630bb5f4d8d355d9b35174e53c59dbdb3.fts'

Spatial Index: Adding 1 features

BADNEWS: Cannot convert null string into an integer........

 

It's just annoying to sift through a bunch of useless errors when trying to debug or tune. I wouldn't have debug logging on during normal operation.

 

Looks like i can avoid the "BADNEWS: Attribute name 'fme_type' removed from schema feature as it doesn't have a type. Current value is 'fme_no_geom'"

------if I change the executor to 'Data Features' only.

That removed about 3/4ths of the errors.

Not sure why FME all of a sudden assumes there is supposed to be an fme_type attribute...

 

Reply


Community Stats

30,126
Posts
114,351
Replies
37,758
Members
Terms & ConditionsCookie settings
A product of
Safe Software respectfully acknowledges that we live, learn and work on the traditional and unceded territories of the Kwantlen, Katzie, and Semiahmoo First Nations.