[
https://issues.jboss.org/browse/TEIID-2896?page=com.atlassian.jira.plugin...
]
Steven Hawkins commented on TEIID-2896:
---------------------------------------
To build off what you are saying, we have three scenarios:
1. standard sv or fixed width such that all rows have the same characteristics (should
already be covered by designer)
2. sv file with multiple row types - here both the row selector and column selector can be
used. This functionality existed before this issue. I'm not sure if designer had any
support. This allows the user to create denormalized rows, such as:
type,firstName,lastName
p,John,Smith
c,Sue
c,Billy
p,Mildred,Jones
c,Tobias
With the texttable:
TEXTTABLE (file SELECTOR 'c' COLUMNS firstName string, lastName string SELECTOR
'p' 3 HEADER)
Would produce:
Sue,Smith
Billy,Smith
Tobias,Jones
3. A fixed width file with multiple row types. Because of this issue it doesn't
always matter that the rows conform to a particular expected data length, or by TEIID-2898
you can selectively only parse the rows matching the row selector.
Add TEXTTABLE function capabilities to interpret fixed width data
file content that is shorter than the defined line length
---------------------------------------------------------------------------------------------------------------------------
Key: TEIID-2896
URL:
https://issues.jboss.org/browse/TEIID-2896
Project: Teiid
Issue Type: Enhancement
Components: Query Engine
Reporter: Van Halbert
Assignee: Steven Hawkins
Fix For: 8.7
Attachments: sample, TEIID-2896.patch
Current behavior is that data lines too short to fit a fixed width data interpretation
produces an error such as "Error: Text parse error: Fixed width line width 35 is
smaller than the expected 60 on text line 1 in file:/tmp/file.csv.". The
customer's use case is interpreting data files from another application which bundles
multiple types of data along with additional descriptive info/metadata about the data
contained within the file. As the error is thrown if any lines after the first interpreted
line are shorter than the summed value of the fixed width columns, they are unable to
query the file. The customer wishes for the ERROR to be suppressed and the query to return
the data independent of whether it fits the fixed width criteria and potentially to be
able to filter the data being retrieved.
--
This message was sent by Atlassian JIRA
(v6.2.3#6260)