[
https://issues.jboss.org/browse/TEIID-4883?page=com.atlassian.jira.plugin...
]
Ramesh Reddy commented on TEIID-4883:
-------------------------------------
[~dvilaverde] In that case, this is bug. What I see is Teiid is not accounting for blank
cells in header, thus using empty column name and considering the next cell as duplicate.
I was thinking your usecase is much like
||heading 1||heading 2|| ||heading 3||
|col A1|col A2| |col A3|
|col B1|col B2| |col B3|
Where 3 should not be part of the table. Even this is good usecase.
So, for the bug, we would need to make sure headerCell is not blank, if it is blank then
either skip the cell or end the the table creation at that column based on the
{{getAllowEmptyCells}} proeprty you created. If you want to submit patch I will review it
again.
Teiid Excel Translator should filter blank cells from header row
----------------------------------------------------------------
Key: TEIID-4883
URL:
https://issues.jboss.org/browse/TEIID-4883
Project: Teiid
Issue Type: Enhancement
Components: Misc. Connectors
Reporter: David Vilaverde
Assignee: Steven Hawkins
Priority: Minor
Fix For: Open To Community
Attachments: fr-demo1-sample-data.xls
The teiid translator for excel will throw a duplicate column name exception when using a
spreadsheet that has 2 or more columns where the header row doesn't specify a header
name. In the attached spreadsheet it's difficult to see that rows E-H should be
deleted from the dataset.
The proposed enhancement is to allow the translator for excel to exclude any trailing
columns where there is no column name defined.
I took a shot at making this enhancement and created a PR at
https://github.com/teiid/teiid/pull/951
If this PR is not going to be accepted, can the methods in the ExcelMetadataProcessor me
changed from private to protected so that the processor can be extended easily?
--
This message was sent by Atlassian JIRA
(v7.2.3#72005)