I believe that is by design, and it's a pattern that you'll see in many different data access scenarios (for example, the MySQL driver on .NET). Since iBATIS.NET doesn't have an actual MySQL, T-SQL etc parser, it doesn't know in what order to apply the parameters based on the column attribute alone. It would be nice if the property style syntax (#Name#) were supported, but I don't think that it is; perhaps someone else could give an insightful reason as to why. The column attribute is, I believe, only required for OUT parameters when calling stored procedures.
I remember making the same assumption and also being astonished, but it does make some degree of sense when you think about it. To iBATIS, the resulting SQL is still "just a string." Hope that helps. V/R, Nicholas Piasecki From: tech fan [mailto:[email protected]] Sent: Wednesday, May 13, 2009 5:20 AM To: [email protected] Subject: parameterMaps based on sequance instead of names?! Hi, I can not believe this, that the parameterMaps are based on sequance instead of names. I'm using 1.6.2 and oracle, it seems to be so. I have a definination here: <parameterMaps> <parameterMap id ="UpdateMap" class ="Process"> <parameter property ="Id" column="ID"/> <parameter property="Name" column="NAME" /> <parameter property="Owner" column="OWNER" /> <parameter property="EMail" column="EMAIL"/> <parameter property="Comments" column="COMMENTS"/> <parameter property="Workflow.Id" column="PARENTID"/> <parameter property="IsDeleted" column="ISDELETED"/> </parameterMap> </parameterMaps> and sql is this <insert id="Insert" parameterClass="Process" parameterMap="UpdateMap"> INSERT INTO process (ID, NAME, OWNER, EMAIL, COMMENTS, OVERLAYS, PARENTID, ISDELETED) VALUES (?, ?, ?, ?, ?, ?, ?, ?) </insert> if I switch the sequance of ISDELETE and PARENTID, in either sql or the parameterMap, then it always generates error. (ISDELETED is varchar2(1)) Anyone encountered this? Regards, T.M.
