I went through the Cloudera docs.
It doesn't mention how to add a column to an existing view. Is it because there is no way to achieve that?
If I use DESCRIBE FORMATTED and use the original view definition of select * from xyz(an example is shown here) to drop the existing view and recreate the view with the same userId, will it work exactly as it did before?
You are right - DROP-CREATE should work. Even ALTER VIEW view_name AS SELECT ..., new_col FROM table should work. May be you can share what query you are trying and what is the error you are facing.
Related
In FSCM I am looking to modify the Search view on Add/Update PO page (Main Menu--> Purchasing--> Purchase Orders--> Add/Update POs) to display the Requisition ID associated with the PO in the search results page. The only table I have found that has both PO_ID and REQ_ID is PS_PO_LINE_DISTRIB however unless I use a SELECT DISTINCT clause I will get multiple PO_ID rows when there are more than 1 line on a PO.
Within Purchase Order Inquiry you can see the related Requisition ID's related to a PO by clicking on Document Status link inside the Purchase Order inquiry details page.
I started looking at the PeopleCode within the the Purchase Order Inquiry to see how they are linking the PO to a Requisition and it appears to use work tables with related PeopleCode function libraries, but I wasn't able to figure our how they get linked. I am hoping someone else may know the answer to this. Thank you.
I'm on an old version of PeopleSoft (SCM 8.80, Tools 8.51), so your mileage may vary. I'm assuming you're familiar with App Designer. If not, comment below and I'll add some details about what I'm clicking on.
Find the name of the Add/Update PO component.
Open the PURCHASE_ORDER component in App Designer. Now let's find the name of the search record. Note that there is a different record for the Add Search Record, so if you want to change that too, do all of this for that record as well.
Open the PO_SRCH record, and add the REQ_ID field to it. Make sure you mark the field as a key. You should consider saving your modified PO_SRCH under a new name in case you want to be able to revert to vanilla PeopleSoft. If you do, change the Search Record in the component to your new record name.
We can see that PO_SRCH is a view. So let's modify the view to pull in REQ_ID from PO_LINE_DISTRIB. As you mentioned above, there doesn't appear to be another table with both PO_ID and REQ_ID, so you'll have to do a SELECT DISTINCT.
We should do a LEFT OUTER JOIN instead of a standard join because if you do a standard join and you enter a purchase order with no lines and save it, then you'll never be able to retrieve that purchase order in this window. Since REQ_ID is a key field, we can't have a null, so we have to do the CASE.
One odd thing that I ran into here was building the view now gave me an error about selecting fewer columns in the SQL than I had in my record definition. I solved it by modifying the view for SQL Server. I've never had to do that before and I don't know why I had to do it for this specific record. But anyway, I entered the same SQL under the record's "Microsoft SQL Server" definition.
In the properties of PO_SRCH, we can see that it has a related language record. If you're only using one language, you can probably get away without changing this, but I'll do it for completeness. Open PO_SRCHLN. Now add REQ_ID to it (mark it as a key field like you did above), and save it as PO_SRCHLN2 (I'm saving it under a new name so I don't break anything else that may be using PO_SRCHLN).
Edit the SQL the same was as you did above. Note: I didn't have to also change the Microsoft SQL Server definition like I did above. I have no idea why.
Now build PO_SRCHLN2.
Go back to PO_SRCH and change its related language record to PO_SRCHLN2.
Now build PO_SRCH.
Hopefully you didn't get any errors and your search page has the requisition ID in it now. My system doesn't use requisitions so they're all blank in the example below, but the new field is there.
i have done a calculated sql with a custom query.
Implemented a table from that data model.
Until here we are ok.
My problem is the next step:
In my custom query i have a field concatAll(that is all fields concatenated).
In my page i want to put a search box to do a "contains" search in my field concatAll and present those results in my table.
Anyone knows how to do this?
Tried multiple approaches, but could not have the result.
Thanks
As #Pavel Shkleinik send to me this is possible using :params in the query of calculated_sql.
Example:
or (c.phone like concat(CONCAT("%", ifnull(:concatAll,''), "%")))
and you put concatAll as a string parameter on datasource page.
Regards
I created an app with the sxc module.
Now I have like 500 empty rows which I want to delete.
I searched for them in the database to delete them all but I cannot seem to find them and I think it is a waste of time to delete them all 1 by 1.
It's data in "Manage content / data" table.
Let me know please.
I have another question:
If I edit an item. The title of the module gets changed with the first items 'name' field. How to avoid that? Is it an bug?
Thanks in advance.
Basically JKings answer is correct - this kind of bulk-operation can easily be done using export/import, because on re-import you can tell 2sxc to delete all items not found in the import. This ensures that 2sxc can take care of data integrity etc. Instructions https://2sxc.org/en/Learn/Content-Export-and-Import
So the correct steps are:
export the list
open in notepad, xml editor (or use excel, as shown in the link)
Remove all those you don't want
Re-import, but choose the option to "Remove all entities not found in import"
You're set :)
I'm using the Feeds module to import lots of Feed Item nodes. Due to a malformed feed file, I'm getting lots of duplicates. I'm using a View to display these nodes, and need to be able to add a DISTINCT filter on the "Node: Post Date" field, so I only get 1 result for each post-date.
I will also look into tackling the problem at the source so to speak (I don't want to have all those duplicates in the first place), but this is an interesting issue in itself - I can't find a way to add a DISTINCT filter on a field other than the Node ID (which has it's own option in the View's Basic Settings box).
I found a great article on a good way to alter the SQL queries that are generated from views before they get executed: http://echodittolabs.org/blog/2010/06/group-views. I used this to basically suffix a GROUP BY clause to the end of the query (in a really nice, clean and versatile way).
As an aside, I also found a way to tackle the issue of importing lots of duplicate feed items, the details of which are here: http://drupal.org/node/661314#comment-3667228. It adopts quite an extreme approach (deleting all items before each update), but this is the only solution for some nasty malformed feeds.
I was holding out for some undiscovered feature of Views that let you do this, but I don't think there is one - maybe in the next version ;)
There are two option to solve this:\
apply this patch
OR
hook_views_query_alter => just paste
$query->distinct = 1;
$query->no_distinct = 'views_groupby';
I guess you have two options: either put some logic in the view template file to skip the duplicate items or implement hook_views_query_alter() to change the query used by the view, adding the DISTINCT clause.
We found this issue in drupal 6.x view - had 7 of 150 items duplicated one or twice. No idea why. Issue only appeared for anonymous users. Luckily, views 6.x.2.16 provides a 'distinct' setting under the basic settings, I set it to Yes and got rid of the duplicates.
I have two views that I would like to combine into one.
The first view shows all items of X where company ID = Y. This is to give preferential sort to the client first, and then everyone else.
So I created a second view, all items of X, where company ID != Y.
I created it as an Attachment to attach to the first view, but I don't think I got the intended result.
How can I combine these views so the first view results are listed first, and then the second view is too, using the same pager, filters, and arguments?
Is there any way of achieving this without programming it?
From a MySQL point of view, the order-by-field syntax would be the appropriate way to handle this. Example:
SELECT * FROM tickets ORDER BY FIELD(priority, 'High', 'Normal', 'Low');
It would be great if there is a module that add this kind of functionality to Views, but AFAIK, it does not exist.
If you want to solve this without programming, I think you can use the rules module to automatically set the 'sticky' checkbox on nodes where company ID = Y. With that in place, you can order the View on the sticky value.
Along the lines of the 'sticky' idea, if you didn't want to override that, maybe you could add a checkbox field to the company type -- isClient. Make it false for everyone except the client, and sort by that.
I haven't done this, but maybe you need to create both versions as attachments, and attach them both to another display...?
for drupal 5 there was views union. Someone started something for D6, but I don't know how far they got.
http://drupal.org/node/275162
Create the second view as an attachment and attach it to first.
Set all Inherit arguments, Inherit exposed filters and Inherit pager to Yes.
how is the client parameter passed to the view? as an url argument? if so, you can create your second view like i outline here and then select the exclude the argument option on the appropriate location.
usually the easiest way to achieve this is with a small hook_query_alter, but that requires a small amount of programming.
A little bit later... but I've found a better solution using only the Views module:
Create a Block View with that shows the first list that you need ("all items of X where company ID = Y")
Create another View that must be a "Page view" with the second list (all items of X, where company ID != Y)
In the "HEADER" settings of this second view, click "Add" and select "Global: View area".
In the "View to insert" list, select the first you have created (and check "Inherid contextual filters" if you are using it)
And that's it!