span8
span4
span8
span4
Idea by marko · · sql serverattributesattribute handling
Writing to SQL Server can be frustrating if you don't pay attention to the attribute types. Generally Varchar is needed as any whitespace may cause issues with styling data in Arcmap or other GIS programs. The default character type should change to reflect the best types for any given writer.
radochudy commented ·
I have small recomendation. If you plan to use Geoserver for data publishing I realy recommend to use nvarchar instead varchar. Varchar makes strange things in geoserver,mostly using in SLD or CQL filters...
svasa76 commented ·
Yay, finally got 2016.1 two hours ago! There is varchar datatype, right, but why can't I delete the numer in the width field? Does it give the max number of characters? And what can I do if I don't want to give a max length? Thanks for your help...
markatsafe commented ·
This has been addressed in FME 2016.1. By default mapping will be to varchar whereever possible
daleatsafe commented ·
Thanks for pointing out a very old default that is hardcoded in workbench. WIth time we are trying to make Workbench much smarter with data types in general, but in the very short term (2016.1) we will change this default to varchar which will be so much smarter on most modern systems.
erik_jan commented ·
I would suggest in every location that FME now decides what datatype to use. So if the datatype is not predefined in the database (new writer or new featuretype) the default should be used.
The precision could be defined by the data, but I have experienced quite often that the precision is defined by the first x number of rows and is not sufficient for the complete dataset.
ravenkopelman commented ·
What context do you see this default in? When generating a new workspace that translates from format X to SQL Server, or when manually adding writer feature types?
erik_jan commented ·
The nicest way to do this is using the FME options and being able to set up the default datatype per Writer for character and numeric attributes. You could even add a default precision per datatype (e.g Oracle character: Varchar2(20), numeric: Number(38,0) ).
erik_jan commented ·
I second this.
For Oracle too the Varchar2 type is highly recommended over the default Char type as the Char type appends spaces to the actual value.
Share your great idea, or help out by voting for other people's ideas.
Reader should read only required attributes
'Case Sensitive' option when testing attribute values
Set attributes in Creator and randomize geometries
Add option to immediately rename exposed attributes
Expose List Attributes From a Custom Transformer Published Parameter
Filter Advanced Value Handling in AttributeCreator
Option to use @Value(AttributeName) in the AttributeRenamer for New Attribute Names
Provide Copy-Past functionality for attribute management tools/functions
© 2019 Safe Software Inc | Legal