Recent changes to feature-requests

getBoolean into DataSet

Anonymous — Thu, 23 Jun 2011 09:26:02 -0000

Hi guys :)

Perhaps it could be usefull, for me at least, you would implement a "getBoolean" method into DataSet interface in order to parser row fields with '0'/'1' or 'Y'/'N' values and API return a boolean value.

Thank you! ;)

Flatpack dependency on old version slf4j library.

Anonymous — Mon, 10 Jan 2011 13:25:21 -0000

I'm wondering if it would be possible to upgrade the dependency on slf4j to a more recent version of slf4j for the flatpack library.
Currently I have some dependency problems using both flatpack and DBunit in the same project.
I tried to exclude the old version in maven but it's not compatible with the recent api of slf4j.

Thx
Luc

Expose the current record type

Yuri Schimke — Mon, 27 Jul 2009 09:01:16 -0000

When parsing the DataSet, the record type must obviously be known internally. However it is not exposed externally, so the client code must check a set of record types manually. See the Camel Flatpack code below which is hard coded to check "header" and "trailer". This would be simplified, if it the record type was exposed.

http://fisheye6.atlassian.com/browse/camel/trunk/components/camel-flatpack/src/main/java/org/apache/camel/component/flatpack/FlatpackConverter.java?r=HEAD

boolean header = dataSet.isRecordID(FlatpackComponent.HEADER_ID);
boolean trailer = dataSet.isRecordID(FlatpackComponent.TRAILER_ID);

// the columns can vary depending on header, body or trailer
String[] columns;
if (header) {
columns = dataSet.getColumns(FlatpackComponent.HEADER_ID);
} else if (trailer) {
columns = dataSet.getColumns(FlatpackComponent.TRAILER_ID);
} else {
columns = dataSet.getColumns();
}

Maven2 Support

Roland Klein — Wed, 09 Jan 2008 12:29:17 -0000

Hi Paul,

some time ago i sent You several enhancements requests and You integrated most or all except a correct Maven2 integration.
http://sourceforge.net/tracker/index.php?func=detail&aid=1402993&group_id=127761&atid=709366

But there are some questions left:

Why do You build Your own Maven repo, when there is an global open own ( http://repo1.maven.org/maven2/ ) ?
Why don't You distribute the Maven2 necessary file "pom.xml", otherwise there is no dependency resolving possible?

Don't get me wrong, the developing from pzfilereader to FlatPack is great, but the Maven2 support isn't there. The distributed files in http://objectlabkit.sourceforge.net/m2-repo/net/sf/flatpack/flatpack/3.1.1/ are useless because of the absence of the pom.xml and they seem to be just a copy of the Maven1 versions with a different directory structure.
Is there any planning to solve this issues? Or should i distribute a pom.xml?

Sorry for blaming, but up to now we can't use this new versions.

Roland

Option To Not Trim Fixed Text

Paul Zepernick — Sat, 03 Nov 2007 02:03:53 -0000

Hi,

I'm using FP to read fixed length text and I would like to keep a blank value blank and not have it trimmed, like it's actually done in FixedWidthParserUtils.splitFixedText(List, String), line 70 : splitResult.add(tempValue.trim());

Could you add such an option, like {set,get}TrimBlankStrings(boolean) on the net.sf.flatpack.Parser interface ?

Regards,

Jérôme

Retrieve Actual Line Data

Paul Zepernick — Wed, 21 Feb 2007 20:19:21 -0000

Look into adding the ability of returning the data as it was in the file back to the user. See the following post:

http://sourceforge.net/forum/forum.php?thread_id=1677920&forum_id=434272

Simple validation

Anonymous — Tue, 13 Feb 2007 17:47:35 -0000

I would like to see more validation config in XML mapping if possible like
<?xml version='1.0'?>
<!DOCTYPE PZMAP SYSTEM
"http://pzfilereader.sourceforge.net/pzfilereader.dtd" >
<PZMAP>
<COLUMN name="FIRSTNAME" require="true" maxlenght="15"/>
<COLUMN name="LASTNAME" require="true" maxlenght="15"/>
<COLUMN name="ADDRESS" require="true" maxlenght="40"/>
<COLUMN name="CITY" require="true" maxlenght="20"/>
<COLUMN name="STATE" require="true" maxlenght="2"/>
<COLUMN name="ZIP" require="true" maxlenght="9"/>

</PZMAP>

TSV

Anonymous — Mon, 12 Feb 2007 14:45:54 -0000

I want to support TSV(Tab separate value) File.

2 other metadata options...

Benoit Xhenseval — Tue, 03 Oct 2006 11:02:14 -0000

2 options that, IMHO, would be useful in the XML
description for fixed width records

1/ One could want the parser to TRIM the strings
2/ One could want the parsed to nullify empty strings
(after trim event).

What do you think?

Kind regards

Benoit

Add metadata for a column

Benoit Xhenseval — Tue, 03 Oct 2006 09:54:38 -0000

I was looking for something to parse a flat file with
fixed width records... This seems to be perfect!

I noticed the ability to getDate/getDouble etc but I
think that it would be great if the XML file could
define a couple of attribute to provide some basic
transformation, eg:

<!ATTLIST COLUMN name CDATA #REQUIRED
length CDATA #IMPLIED
targetClass CDATA #IMPLIED
format CDATA #IMPLIED
>

Where targetClass would be say a java.util.Date and the
format would describe a SimpleDateFormat.

When I request DataSet.getObject("columnName"); I would
get the object i.e. the date if it was a date, a Double
if it was a double, etc

Support for BigDecimal would also be great for me
(getBigDecimal())

What do you think?

Many thanks for this great tool!

Benoit