Update BSO with external data

vinsnash

I'm not sure I understand the User Guide on this where it states an "*" can be used instead of an attribute ID.

I have a BSO that has instance address attributes we want to update. We don't want to add new instances. Sample CSV file:

"Unit.UIC","Unit.Address1","Unit.Address2","Unit.City","Unit.State","Unit.ZIP"
"031LMD","NOSC","15 Sea St.","Battleship Cove","MA","01209"

The BSO contains instances with the Unit.UIC but none of the other attributes have values at present. How to use the "*" to update the address attributes is the question.

Does it go with the header "Unit.Code"," or with each data line "031LMD","?

aware_support

The system will need to find a UNIQUE instance of the object that you want to modify for each CSV row. This means that your data must have one or more attributes that uniquely identify the instance. You need to mark such attributes with the "*" symbol in the header row, for example, *Unit.Code

vinsnash

When I use "*UnitTable.UnitCode" in the CSV header the server reports an error:

"XML returned from server does not have root element"

The server log shows:

2011-05-11 13:10:46,529 ERROR USNSCC#admin -Execution Exception after starting system process message Invalid description of data column ??"/*/U/n/i/t/T/a/b/l/e/./U/n/i/t/C/o/d/e/" It must start with UnitCode.

"/" represents undisplayable characters in the log entry that can't be copied.

I've experienced the same error when inadvertantly ommitting the header from an import .CSV File. I hope theres a fix for this because we will have to do a lot of this type of BO update.

vinsnash

I made a typo in the previous. It should read "It must start with UnitTable"

tford

Is your first character " or *?

It should be *

vinsnash

Do you mean "UnitTable.UnitCode" rather than "UnitTable.UnitCode" ? I'll try it but I think I already did.

Vins

tford

It should be:
*UnitTable.UnitCode

You will find some files use in the sample applications in our \samples\CRM folder of your AwareIM installation.

vinsnash

*"UnitTable.UnitCode","UnitTable.DrillAddress1","UnitTable.DrillAddress2","UnitTable.DrillCity","UnitTable.DrillState","UnitTable.DrillZIP","UnitTable.UnitURL"
"011BMD","Gloucester High School","32 Leslie O Johnson Road","Gloucester","MA","01930-2500","http://www.bostonminutemendivision.com"

These are the first two lines of the CSV - header and 1st line of data. I've tried "*UnitTable.UnitCode" and *"UnitTable.UnitCode" with the same result. "Must begin with UnitTable".

There must be something not apparent in the documentation. There is a shortage of explicit examples and an assumption of knowledge that just is out here. Another "Dummies" topic.

Vins Nash

vinsnash

The samples I've used and work for import that adds have headers that use quotes for the columns i.e.

"ObjA.Attr1","ObjA.Attr2"," etc.

Are we saying here the header doesn't use quotes UnitTable.UnitCode,UnitTable.Address1, etc. rather than "UnitTable.UnitCode","UnitTable.Address1","......?

I'm confused.

Vins Nash

aware_support

You don't need quotes in the header row. Please remove them. Just look at the sample CSV data in the sample applications - it should be quite obvious what you need to do

vinsnash

Removed quotes and got the same error. This time:

2011-05-11 13:10:46,529 ERROR USNSCC#admin -Execution Exception after starting system process message Invalid description of data column ???*/U/n/i/t/T/a/b/l/e/./U/n/i/t/C/o/d/e It must start with UnitTable.

Using no quotes here. Your process evidently doesn't like *UnitTable.UnitCode either.

Sample applications are one thing, documentation is something else altogether. Expecting users to search through samples for answers to problems with no index is a bit of a stretch. I'm sure your samples have plenty of samples but your documentation should contain the examples or point the user to sample(s) that contain the solution. At least point a user to the specific sample when answering a question.

Vins Nash

aware_support

Documentation is available in the file AwareIM/Docs/ExportImport.txt

As far as your problem is concerned - look at the error message. It is quite clear that you have some strange characters in your header row. Don't know what editor you are using to create the file, but try opening it with the Notepad and see if you have suspicious characters there

vinsnash

The file was keyed in/created with Notepad for testing purposes. It's not an export from another data base yet. The header line never will be anything other than keyed in using Notepad. It mirrors exactly syntax I've been using all along for import to other objects. The two line sample in the previous post is copied and pasted from Notepad.

I assumed the "strange" characters were part of whatever scripting is going on with the server. It manages to pick up the correct spelling in the last part of the error message - "must start with UnitTable". I'm at a total loss now.

I'll try keying the whole file over again and see what happens.

Vins Nash

vinsnash

After several test scenarios I've determined that the problem is with the .CSV file structure. I used Notepad to key in a header and a couple of lines of data and that worked. I did the same with a couple of other objects which worked. I could not, however, copy and past the header that worked into the file that was causing the problem. Copying and pasting lines of data from the problem file to a file with the header that worked caused the same error to appear.

I'll be getting another copy of the data exported from the same data base next week and will try again. If it fails again I'll have to come up with a plan "B". I'll have to get some sort of hex editor because nothing is apparent opening the file with Notepad, Wordpad, or Word. The file may be using some code that Notepad reads but not the AwareIM import process. Except for the undisplayable characters in the error message all the characters that make up the header element display.

Vins Nash