[jira] [Commented] (CAMEL-22458) flatpack converter accesses invalid column names for header and trailer record value retrieval of fixed width files

["Claus Ibsen (Jira)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/CAMEL-22458?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=18023475#comment-18023475 ] 

Claus Ibsen commented on CAMEL-22458:
-------------------------------------

You are welcome to work on a patch and send a PR

> flatpack converter accesses invalid column names for header and trailer record value retrieval of fixed width files
> -------------------------------------------------------------------------------------------------------------------
>
>                 Key: CAMEL-22458
>                 URL: https://issues.apache.org/jira/browse/CAMEL-22458
>             Project: Camel
>          Issue Type: Bug
>    Affects Versions: 4.14.0
>         Environment: * probably all
>  * discovered on camel 4 quarkus (after migrating from camel 2 fuse)
>  ** quarkus-camel-bom, quarkus-bom version: 3.16.2
>            Reporter: Gandrinor
>            Priority: Minor
>              Labels: flatpack
>
> *Errror Szenario*
> When org.apache.camel.component.flatpack.FlatpackConverter toMap function is called, for example while using the org.apache.camel.component.flatpack.DataSetList iterator, the putValues(Map<String, Object> map, Record recordObj) is called.
> In this method the recordObj.getColumns() is called to retrieve the available columns of the record. The columns are iterated over to retrieve the respective values.
> When a fixed with file with the following example.pzmap.xml scheme was parsed, 
>  
> {code:java}
> <?xml version='1.0'?>
> <!-- DTD can be pulled from the Jar or over the web-->
> <!DOCTYPE PZMAP SYSTEM	"flatpack.dtd" >
> <!--<!DOCTYPE PZMAP SYSTEM	"http://flatpack.sourceforge.net/flatpack.dtd" >-->
> <PZMAP>
>     <RECORD id="header" indicator="0" elementNumber="1">
>         <COLUMN name="RECORDINDICATOR" length="5" />
>         <COLUMN name="HEADERDATA" length="20" />
>     </RECORD> 
>         <COLUMN name="FIRSTNAME" length="35" />
> 	<COLUMN name="LASTNAME" length="35" />
> 	<COLUMN name="ADDRESS" length="100" />
> 	<COLUMN name="CITY" length="100" />
> 	<COLUMN name="STATE" length="2" />
> 	<COLUMN name="ZIP" length="5" />
>     <RECORD id="trailer" indicator="9" elementNumber="1">
>         <COLUMN name="RECORDINDICATOR" length="5" />
>         <COLUMN name="TRAILERDATA" length="15" />
>     </RECORD>
> 		
> </PZMAP>  {code}
> this leads to the following error being thrown: 
>  
> {code:java}
> java.util.NoSuchElementException: Column [FIRSTNAME] does not exist, check case/spelling. key:[header] at net.sf.flatpack.util.ParserUtils.getColumnIndex(ParserUtils.java:618) at net.sf.flatpack.RowRecord.getStringValue(RowRecord.java:269) at net.sf.flatpack.RowRecord.getString(RowRecord.java:315) at org.apache.camel.component.flatpack.FlatpackConverter.putValues(FlatpackConverter.java:117) at org.apache.camel.component.flatpack.FlatpackConverter.toMap(FlatpackConverter.java:48) at org.apache.camel.component.flatpack.DataSetList$1.next(DataSetList.java:90) at org.apache.camel.component.flatpack.DataSetList$1.next(DataSetList.java:68) {code}
>  
>  
> *Error explanation*
> The first record is the header record, which differs from the rows below. But the recordObj.getColumns() method retrieves not the header records columns (RECORDINDICATOR, HEADERDATA), but the other rows columns (FIRSTNAME, LASTNAME, ADDRESS, CITY, STATE, ZIP), which do not match the record structure. This leads to error.
>  
> *Solution Suggestion*
>  * use the local getColumns(DataSet dataSet) method instead of recordObj.getColumns() to retrieve the correct colums
>  * change the getColumns(DataSet dataSet) methods signature to getColumns(Record dataSet)
>  
> {code:java}
> private static void putValues(Map<String, Object> map, net.sf.flatpack.Record recordObj) {
>     String[] columns = getColumns(recordObj); // !!!modified!!!
>     for (String column : columns) {
>         String value = recordObj.getString(column);
>         map.put(column, value);
>     }
> }
> private static String[] getColumns(Record dataSet) { // !!!modified!!!
>     // the columns can vary depending on header, body or trailer
>     if (dataSet.isRecordID(FlatpackComponent.HEADER_ID)) {
>         return dataSet.getColumns(FlatpackComponent.HEADER_ID);
>     } else if (dataSet.isRecordID(FlatpackComponent.TRAILER_ID)) {
>         return dataSet.getColumns(FlatpackComponent.TRAILER_ID);
>     } else {
>         return dataSet.getColumns();
>     }
> } {code}
>  
>  
>  



--
This message was sent by Atlassian Jira
(v8.20.10#820010)