Hi,

i am importing some extracts that have 300+ columns, and their column name are really bad, with all sorts of non-alpha numeric characters.  such column names are like these:

column A (has a comment)
column B contrains %
column C can have & as well

so when Talend reads these files header - we get the default names as:

_column_A__has_a_comment__
_column_B_contrains__
_column_C_can_have___as_well_

which is fine, but i though i could change these names to avoid double underscores and so on
thus i created a job to process the Talend metadata XML.

* here is the question *

i can not get the job to write an attribute named 'default' even though it exists in the original XML.

have a look at the attached screenshots.

does any one know how to work around this?

i have manually added this attribute, but surely i am missing something there.

i am reading Talend metadata using an XML input, a tJava to alter what i need and writing out with a tAdvanceFileOutputXML.

here is a sample of the output:

<?xml version="1.0" encoding="ISO-8859-15"?>
  <column comment="" key="false" label="IDENTIFIER" length="15" nullable="true" originalDbColumnName="" pattern="" precision="0" talendType="id_String"/>
  <column comment="" key="false" label="REASON_FOR_ACCESS" length="8" nullable="true" originalDbColumnName="" pattern="" precision="0" talendType="id_String"/>
...
</schema>


when it should be something like:
...
  <column default="unknown" comment="" key="false" label="IDENTIFIER" length="15" nullable="true" originalDbColumnName="" pattern="" precision="0" talendType="id_String"/>
...



please let me know if you need more info.


thanks,

Nicolas

Last edited by nicolasdiogo (2010-06-03 19:42:32)