Problem with column sort option

I have a two page form application. Page 1 shows a list of entered estimates for a sales consultant. Page 2 is the detailed estimation page. When I delete the record on the detail-page and branch back to page 1 I get the follwoing error. "failed to parse SQL query: ORA-06502: PL/SQL: numeric or value error: NULL index table key value". This happens only when I remove the last estimation record AND I have at least one column flagged as "sort". If I take the sort option of the column(s) it is working fine. If I exit the page and come back-in again everything is fine as well. I have specified to clear the cache when I branch from page 2 to page 1. I hope this is enough detail.

I think I get really good at answering my own questions.
It seems when you select the sort option on a column you must also set the "Sort Sequence".

Similar Messages

Problem with Column Sorting in Request Table View

We've enabled column sorting on the table view in an Answers request but it doesn't work when more than around 400 rows are returned. It works Ok when fewer than 400 rows are returned. Does anyone know if there is a specific limit to the number of rows where column sorting works?
Thanks,
Mike

I've dug into the query log a little more. When this request returns more than around 400 records and you click on a column heading to sort the table view, the query from the log is not changing to reflect the new sort order. It stays as whatever the default sort order was for the request.
Column heading sorting on other requests in different subject areas on this same server works fine. I've tested this other request with up to 6400 rows returning and it's sorting works.
All caching is turned off for the subject area.

Problem with column sort in myFaces dataTable

Hello,
In my dataTable I am trying to do a column sort on each column, When i click on the column header link nothing is happening. Could somebody please help. Please let me know if I am missing something.
Here is the code.
The jsf is as follows:
<h:panelGrid columns="1" styleClass="tableHeaderAction">
<t:dataTable id="fooS" value="#{supervisorsDto.tableRows}"
var="tableRow" rows"30" styleClass="dataTable"
headerClass="tableHeader" first="0"
columnClasses="string,string,string,string"
rowClasses="odd,even"
sortAscending="#{supervisorsDto.ascending}"
sortColumn="#{supervisorsDto.sort}"
preserveSort="true" >
<t:column>
<f:facet name="header">
<t:commandSortHeader columnName="lastName" arrow="false">
<t:outputText value="#{screen.lastName}" />
</t:commandSortHeader>
</f:facet>
<t:outputText value="#{tableRow.lastName}" />
</t:column>
<t:column>
<f:facet name="header">
<t:commandSortHeader columnName="firstName" arrow="false">
<t:outputText value="#{screen.firstName}" />
</t:commandSortHeader>
</f:facet>
<t:outputText value="#{tableRow.firstName}" />
</t:column>
<t:column>
<f:facet name="header">
<t:commandSortHeader columnName="loginName" arrow="false">
<t:outputText value="#{screen.loginName}" />
</t:commandSortHeader>
</f:facet>
<t:outputText value="#{tableRow.loginName}" />
</t:column>
<t:column>
<f:facet name="header">
<t:commandSortHeader columnName="team" arrow="false">
<t:outputText value="#{screen.team}" />
</t:commandSortHeader>
</f:facet>
<t:outputText value="#{tableRow.team}" />
<h:inputHidden id="key" value="#{tableRow.empID}" />
</t:column>
</t:dataTable>
</h:panelGrid>
The sort method in SupervisorsDto is as follows:
public class SupervisorsDto implements Serializable {
private String sort="lastName";
private boolean ascending= true;
@SuppressWarnings("unchecked")
public void sort(final String column,final boolean ascending)
Comparator comparator = new Comparator()
public int compare(Object o1, Object o2)
Supervisors c1 = (Supervisors)o1;
Supervisors c2 = (Supervisors)o2;
if (column == null)
return 0;
if (column.equals("lastName"))
return ascending ? c1.getLastName().compareTo(c2.getLastName()) : c2.getLastName().compareTo(c1.getLastName());
else if (column.equals("firstName"))
return ascending ? c1.getFirstName().compareTo(c2.getFirstName()) : c2.getFirstName().compareTo(c1.getFirstName());
else if (column.equals("loginName"))
return ascending ? c1.getLoginName().compareTo(c2.getLoginName()) : c2.getLoginName().compareTo(c1.getLoginName());
else if (column.equals("team"))
return ascending ? c1.getTeam().compareTo(c2.getTeam()) : c2.getTeam().compareTo(c1.getTeam());
else return 0;
Collections.sort(tableRows, comparator);
public void sort(String sortColumn)
if (sortColumn == null)
throw new IllegalArgumentException("Argument sortColumn must not be null.");
if (sort.equals(sortColumn))
//current sort equals new sortColumn -> reverse sort order
ascending = !ascending;
else
//sort new column in default direction
sort = sortColumn;
ascending = isDefaultAscending(sort);
sort(sort, ascending);
protected boolean isDefaultAscending(String sortColumn)
return true;
public String getSort() {
return sort;
public void setSort(String sort) {
this.sort = sort;
public boolean isAscending() {
return ascending;
public void setAscending(boolean ascending) {
if(ascending != this.ascending)
this.ascending = ascending;
Thanks!

I think I get really good at answering my own questions.
It seems when you select the sort option on a column you must also set the "Sort Sequence".

Problem with column sorting in af:table

Hi,
I am populating an af:table using programmatic view object. One of the columns of the table has its sorting set to true. However, when i click on the sort arrow in the column, no data is loaded, and it displays a message "No data to display."
Kindly, help me in resolving the issue. Thanks in advance for yours assistance.
Regards,
Max.

Hi all,
I have an exact scenario what you explained. Let me explain my scenario and the way i implemented.
Mine is also custom java data source implementation where i get data from a service. Followed the implementation details given in document.
How the framework works:
1. code to retrieve data from service is written inside executeQueryForCollection.
2. hasNextForCollection returns true or false based on the no. of data fetched. setFetchComplete(true) is set when it reaches the end of retrieved data.
3. createRowFromResultSet will be called until setFetchComplete is set to true.
Worked fine and i am able to get the data in my view object.
1st Issue:
Since the code to retrieve data from service is written inside executeQueryForCollection, everytime it is getting executed whenever executeQuery is called (eg. in-memory filtering/sorting, table filtering, table sortine), which is a performance issue. I don't want to call my service often since i've retrieved all the data already.
Tried:
I added an if condition inside executeQueryForCollection to skip service call (skipped only service call but not hasNextForCollection() and super.executeQueryForCollection()) by adding a service call flag. Default flag value is false. Whenever we need service call, we need to set the flag (i.e. it is actually a property with setter and getter) to true, followed by executeQuery().
2nd issue:
when user performs table filtering/sorting, since i skipped the service call, since there is no rows to populate, createRowFromResultSet() will not be called and so i couldn't see any data in table at all. (I don't know why it expect me to give data.)
Tried:
To solve this, i added a local variable to keep the retrived data from service. Whenever i perform service call, i'll keep the data in a local variable. By doing this, whenever executeQueryForCollection is called, i'll check the service call flag. It will be false. In the else part i set the local variable data as user data for collection. The only difference in this implementation is that instead of serving the data from service, i serve data from the local variable. By doing this, filtering/sorting is working fine and i am able to avoid the service call too (but i am not able to avoid the createRowFromResultSet().)

Problem with columns in the middle of my book

I am writing my second book with Pages 2.02 and have run into a problem with columns. I want to add a long list of items as a list with check boxes for readers to check off. If I paste in enough text for three or four pages of two columns it gets messed up in various ways - the text flows into the following text if I don't use section breaks at the end instead of layout breaks, if I use section breaks it solves that one, but if I add anything to the list it changes all the columns across the pages so that one or several of the pages has way less in the right column than in the left. So far the left column fills the entire page all the time, but the right column has lots of spaces. If I switch back and forth from one column to two, it fixes it. If I add text before the list section it messes everything up again and because there is a section break I end up with a little bit of text on the last page and a huge gap before the single column part that follows.
Does anyone know how to get reliable results that will not mess me up when adding content before the list or in the list itself?
PowerBook G4 Mac OS X (10.4.8)

Hi Cymru,
Thank you for thinking about my problems. I guess I s half asleep when I posted and it didn't make sense.
I have so far 77 pages of the book divided into several chapters via section breaks. Inside the 5th section there is almost two pages of paragraphs. Those paragraphs could become 3.75 pages or who knows how long before the list of items. The list of items is 226 items separated by carriage returns and go from one word with 5 characters to 7 words and 34 characters. It needs to be a list with check boxes before them so that the user can check off the items that apply. It is a workbook about turning bipolar into an advantage and is currently 8.5 x 11 and printed on my laser printer for now, but when completed will be more like 8 x 10, so it needs to resize without major hassles.
I have tried every combination of breaks to get the list to flow in two columns, but perhaps have not done the correct combination. I show invisibles and layout. At the end of my paragraphs I do a paragraph return. I then do either a section break or a layout break, followed by a paragraph return and another layout or section break. I then put my curser before the paragraph break, but after the first section/paragraph break, and set columns to 2, evenly divided by default with 3.2026" columns and .3375" gutter, but I want them to change with page size in the future.
Once set up, I copy the text out of BBEdit and paste into the left column. It takes up several pages because it is formatted to do double spaces, so I use one of my styles to make it bold and single spaces. I then apply list style with image bullets and one of the default box images that came with Pages. There are 5 items in each column on the first page because it is close to the bottom, but flows into the next two pages. (with text edits it might end up 20 items in each column on the first page) The second page has 39 items in each column and the last page has 8 items in the left column only, but if with section breaks take up the whole page and the next paragraphs start on the next page. If I then add items to the list, every page gets messed up and the second page ends up with 39 on the left column and less items in the right column for each added item.
If I use layout breaks instead of section breaks, the paragraphs following my lists (now with only one column) end up under the list. What that means is I can see my bold items, but beneath them are my paragraphs that belong after the list. I can see the layout break at the end of the list, but the columns that are shown as part of the 'view layout' are visible right over the single column text also for the rest of the page. The only way to solve that is to go with section breaks and put up with the gap from the end of the list to the following paragraphs.
I hope this makes more sense.
Tom
PowerBook G4 Mac OS X (10.4.8)

Problem with column heading sorting

I have a problem with a classic report.
I created a simple report. We want to sort the report, when we click the column heading. But when we cklick on the column heading, nothing happens. In firefox we get the error "apex.jQuery.datepicker is undefined".
When we created another simple report and connect against another database (but the same web-server), we can sort the report, when we cklick a column heading. On both databases we installed the same apex version (=4.0.1.00.03).
on "apex.oracle.com" (workspace:ama / user:demo / pwd:demo) I created an example.
what is wrong?? Can anyone help me ??
Thanks
Robert

Hi Jari,
yes, you are right. I changed my query. I attempted to find a solution and so I changed the query several times => I used other tables, with or without date columns, with or without columns in the select and so on.
You said, the problem is the authentication scheme. It crossed my mind that I do another change:
during the creation of my application apex created a login page => page 101. For my example application I don't need a login page, and so I droped the page. After starting the application I saw the error message "page 101 not found ". So I searched for a property where apex definies the "start page" respectively "login page". But I don't found anything. Then I looked into the export file and searched for "101". I found the package procedure "wwv_flow_api.create_auth_setup" and the parameter "p_invalid_session_page", which receive the value "101". In apex the property "Session Not Valid Page" was blank. So I inserted the page number "1" (=the first and sole page in my application).
After the explained changes I could start my report.
I don't know, whether these informations help to find the problem!
Regards
Robert

Problem with the Photomail option within " How would you like to share ?"

I have a PC with Windows 7 and Elements 10 installed. For the past couple of days, when I attempt to share photos using the Photomail option under "How would you like to share ?" column , after selecting the recipient, I get an error message dialogue box that says : "Elements 10 Organizer has stopped working - a problem caused the program to stop working correctly. Windows will close the program and notify you if a solution is available". Then the Organizer stays closed down until I reopen it. However, I am able to share photos using the "Email Attachment" option.
What could be wrong with the "Photomail" option ?
Thanks

Hi,
I have seen this sort of failure caused by a permissions problem.
Try this:
Close elements
Create a shortcut on your desktop which goes direct to the organizer
Right click on that shortcut and select "Run as Administrator" (even if you are an administrator)
Now try your Photomail again
If it does work, you only need to do this once, it seems to correct itself after that.
Good luck
Brian

Problem with ALV Filter Option

Hi
In 4.6C version when i click on a column for Filter option, i am able to key only 20 Chars( Visible length is 16 ). Whereas in 4.7 version, i am able to key in 40 Chars in the Input field of the filter option. Is this a problem with 4.6C version. Please Help...

Hi
Can you check report BCALV_TEST_FULLSCREEN_FILTER in 46C. When you put a filter on C60 ( 60 Char field ) the screen display of select-option is small but when you select the multiple options, you can key in upto 60 chars and the filter works well. Check if thsi rpogram helps you.
Cheers.

IGrid 11.5 column sort options

A simple question with probably a complex answer.
I have a iGrid based off a Xacute time-based query. The user would like the ability to click on a column header to resort the iGrid. That part is easy using the columnselectionevent and passing the columnName back to the transaction. This method executes the transaction on each header click.
I do not want the transaction to run on each header click though. QueryCaching I don't think is a option as this is a time-based query.
Any ideas out there.?
Regards,
Roger DeWeese

Roger,
We missed you and Tom down in Nashville last week.
You are correct - QueryCaching is ignored in a Time Based query since it would lead to excessive cache objects, which would probably never be matched again by a subsequent user request.
Column sorting with Ctrl+Click on the Column Heading has been added to the iGrid in upcoming version 12.1, but of course that doesn't help you today.
Without running the Transaction again you could however leverage the /Illuminator/StyleSheets/GenericSortFilter.xsl transform and enable QueryCaching. If you map your Transaction date properties in the Param.x location and it will not know that it is time based, but you of course will lose the SD/ED buttons and VCR control buttons in the toolbar and potentially take charge of the date logic in the BLS too.
Then from the ColumnSelectionEvent if you set one of the XParamValue.x settings it will use the cached query results and just reapply the sorting xsl. (Look at the params at the top of the xsl file to see the parameter name settings needed to map into your Xacute Query template below the Transform)
Regards,
Jeremy

Problem with H and Option key on a macbook keyboard

I have a problem with my macbook keyboard, any time i tried to use the "h" or "option" keys i get the TM symbol.
All other keys works perfect
i don't know what happend?
Thanks

To access NTFS Windows from your MAC OS via Bootcamp, use this software:
http://www.paragon-software.com/home/ntfs-mac/
To access Mac from your Windows via Bootcamp, use this software:
http://www.mediafour.com/products/macdrive/bootcamp/

Ipod Touch 5th Generation ios7 problem with Genre sorting

Hi. I upgraded my 5th gen ipod touch to ios7. But I can no longer sort my music by genre. For example, I have Bruce Springsteen divided up into Rock and 80's. But when I go to "80s" genre and click "Bruce Springsteen" it shows ALL songs by Bruce Springsteen. The ones I have labeled as 80s AND the ones labeled as Rock. And vice versa. If I go to Rock and Bruce Springsteen, I get the Rock songs and the 80s songs as well.
Basically, if I go to ANY genre and pick an artist that has songs in multiple genres, it will list ALL songs and albums by that artist and not just the songs in that particular genre. So it makes sorting by genre completely useless.
I've seen this on every device with ios7 installed. I didn't have this problem with ios6 and I prefer to sort my music by genre. I've spent a lot of time organizing it that way because that's how I want it.
Does anybody know any way around this? Thanks!

You say you have iOS 7. If you are a developer you have access to the private OS 7 developer's forum
iOS 7 Developer forum- Apple Developer
Otherwise you are on your own
PSA: iOS 7 If you're not a developer, don't install developer betas on your iPhone | TUAW - The Unofficial Apple Weblog

Problem with column selection in GUI_DOWLOAD

Hi everyone !
I want to save data from an internal table to a DAT file using GUI_DOWLOAD. In this internal table, there are 2 colums that won't be downloaded so I call the FM this way :
    CALL FUNCTION 'GUI_DOWNLOAD'
         EXPORTING
              FILENAME                = L_FILE
              FILETYPE                = 'DAT'
              COL_SELECT              = 'X'
              COL_SELECT_MASK          = V_COLSELECT
         TABLES
              DATA_TAB                = TAB_OUT
              FIELDNAMES              = TAB_FIELDNAMES.
V_COLSELECT contains an 'X' for each column I need to be downloaded and a space for the 2 other columns.
In the result file, the columns that I didn't select don't appear BUT there are 2 columns with no header and filled with zeros at the right edge of the file. I don't know where these 2 columns come from. Maybe a problem with fieldnames but I'm not sure (in TAB_OUT, all columns are present but in TAB_FIELDNAMES there are only headers for the selected columns).
If anybody have an idea, please just tell me.
Points will be rewarded for helpful answers.
Thanks.
Nicolas.

Hi,
Try to keep fields same in TAB_FIELDNAMES and TAB_OUT. because anyway you are using COL_SELECT_MASK.
I think you are using COL_SELECT_MASK properly ie
if you have 6 field in the TAB_OUT out of that you need to donwload 1,3,6 then your
COL_SELECT_MASK will be X X X
aRs

Problem with file download option in ALV via ITS. Please help!

Hi Experts,
I have a ALV report which I have web enabled using ITS service. In the ALV there is a delivered functionality for file download. It works fine when the report is run in the SE80. But when I run the ITS service for the report and choose file download option it displays a page saying "this page has been diaplayed due to technical reasons" with title "File properties". In addition it looks for Java runtime in the browser. After all this it opens a popup window for selecting path and file name.
This is not user friendly and many customers have rasied their concerns on it.
My question is that,
Q1: What is the need of this "File properties page"? Why ITS calls this while ABAP editor does not?
Q2: Is there any way we can suppress this so that it directly opens a popup window for selecting path and file name?
Please help me.
Thanks
Gopal

Hello Klaus,
i have also the problem with these intermediate screens!
We have in between installed patch 22 for ITS 6.20 and SP18 at SAP Netweaver 2004!
But they are still not processed invisible for the users.
Are there any additional parameters, which controls this behaviour(~webgui_updown_make_visible is not used!)?
How can i check, if i have really a.ex. the actual ws.jar-file or other related files of the patch?
We are using the standalone ITS with an own (modified) IAC with an own CI-design for the webpages. The intermediate screens are displayed with the "normal" SAPGUI-design because of the parameter ~generatedynpro. Without this it does not work, because the template SAPLSIT_0100 is missing!
Or do you have an example for such a intermediate template with the applet for an upload of a file?
Thanks in advance
Felix

Problem with meta key option in Terminal app!

I used Terminal for awhile now with and without meta-key option but I like the meta-key to be turned on cause im familair with tabbing that way. But one thing I can't do in like irssi, is to make symbols with altkey1234567890.. I can do that without the meya-key option.. Anyone know how to solve this problem?

This seems a [long-standing|http://www.entropy.ch/blog/MacOSX/2009/01/27/Mac-OS-X-Problem-Unsolved-for-Years-Terminal-Meta-Key-and-Internati onal-Keyboards.html|Mac OS X Problem Unsolved for Years: Terminal Meta Key and International Keyboards] issue.
A possible solution could be to find how to specify an arbitrary alternate key as the META key, for example the Caps Lock or Fn key.
Another solution could be to tweak the shell to print the requested characters, even with the "use Option as the META key" option enabled. For what I understand, this route for the Bash shell is explained [here|http://anders.janmyr.com/2009/01/making-mac-os-x-terminal-hum.html|Making the Mac OS X Terminal Hum]. However, I don't know enough of *nix and shells to really understand what's written in that article… moreover, I use the tcsh shell and those instructions seems to me very shell-specific.
Any idea of what needs to be tweaked in a tcsh shell environment?

Apache FOP Problem with column width

We are using Apache FOP to generate simple pdf documents out of reports.
My problem is that the table in the pdf is as standard filled in with columns of same length.
Further on the content of some table fields is to large or sometimes there is to much room.
is there any possiblity to let fop react like html?
The smallest table field size should be like the content. I don't want to adjust all my reports, cause there are really many reports with pdf printing.
anyone an idea how to solve this with FOP?
i have no BI pubisher or other reporting engines for this problem
thx so far

Oliver,
I think what you're looking for is having your column width automatically adjusted based on the content of your column values. That is not possible with the built-in, generic XSL-FO template. What you can do is manually adjusting the column width to appropriate values on the print attributes page.
Regards,
Marc

Problem with column sort option

Similar Messages

Maybe you are looking for