# Editing a submission from the KoBoToolbox server does not allow me to save upon completion

Hi @Kal_Lam
I am still getting some submissions, collectd via ODK android app, that need editing on the koboToolbox dashboard but today I am getting an error when trying to submit after editing a text response answer.

I am not sure why i am getting the error in red shown in the screenshot. firstly the date is 20th July whereas the form was entered on 16th July. Secondly the time with me is 11:42 am whereas the time on form is 08:12 am therefore not in the future, and finally i was not editing this response but a text response below new phone number.
Any help is appreciated
Catherine

Hi @cp622,

I assume it’s a constraint (used while building your survey project) that is hindering you to save the changes made. Would you mind checking the constraint. It should solve your issue.

Have a great day!

Hi @Kal_Lam
Yes there is a constraint in the form which works well on data collection. I am wondering why it flags up while editing the response?
constraint= . <=now() constraint_message= The call attempt time cannot be in the future. relevant= (${call_attempt}>= ‘2’) and (${participant_status1} = ‘2’ or \${participant_status1} = ‘4’) required= yes

Hi @cp622,

Delete the constraint and you should be able to edit without any issue. But kindly please be informed that if the data collection is still on it should affect the data that is yet to be collected.

Or, maybe you could make edits once the data collection is over.

Have a great day!

Hi @Kal_Lam
Unfortunately neither of those solutions are possible, data collection is ongoing so I can’t edit the xlfForm to delete the constraint. I cannot wait until data collection is over because when I download the data everyday this submission splits over 2 rows.
I guess I will need to ask the enumerator to re-enter data and delete the problematic submission on the dashboard
Catherine

1 Like

Hello @Kal_Lam,

1. Every submissions is linked to the form version used for (first) data entry?
2. In KoBoCollect, on.local device, re-edit can only done with this version (before sent).
3. On server level, edit/save is done with the newest deployed version?
4. Data export from the server is also done based on the newest deployed version? (But enhanced by data and labels from all previous versions if the include option is selected)?

This means for example:

• Data cleaning on server can use changed constraints, relevants. Even additional calculations?
• There is no possibility to change the used form on local device after data entry/save?

Hi @wroos, cc: @cp622,

This issue is related with the form design. If i am not mistaken it’s associated with the following post which controls the same (i.e. enumerators are not allowed to manipulate the date during data entry):

Else, if it would have been a normal form a user could easily edit a submission without any issues form the KoBoToobox server.

Have a great day!

Hi @Kal_Lam and @wroos
The main issue is that the csv export from KoboToolbox dashboard for some submissions data is split across 2 rows, I think this is due to the enumerator using carriage returns or extra spaces in text box responses (though i am not certain).

When I download the data and try to process it in SAS the code crashes because the first variable is meant to be a date. however instead the text reponse from the row above splits and so text comes into the date column.

I previously submitted a query to the forum when @Kal_lam said I could edit responses on the dashboard. This works well in most of the cases, however in today’s download I was not able to edit the submission. I was previously able to edit so I am not sure what the difference between today’s submission and all the other ones I have been able to edit.

If you need me to clarify any points please let me know

Catherine

Hi Catherine,

This is definitely an issue with how the csv is being pursed by the data conversion programme. When you look at the CSV in its raw format on Excel, are the rows also split? If the rows are split then this is definitely a system issue where continuous texts are not being processed within open quotes to denote them. If this is the case, kindly send the name of your project and account name (only) on private message and I will find sometime to look at it.

I would first start by just confirming an obvious to clear any doubts; do you own this form or it was shared with you?

1. If shared to you you, kindly ask the person to re-share the same to you.
2. If this is your own form, could you share a screenshot of what you are seeing when you try to edit the form?

Stephane

1 Like

the split line can be seen in rows 4601 & 4602

I work with my colleague who is the study coordinator and she set up the project and uploaded the xlsForm, however I have her permission to have the username and password. I am the data manager for the project.

Best wishes
Catherine

Hi Catherine,
This file seems to have been pushed as semi colon separated values as opposed to comma separated values. What happens is that if one or more of the responses there had a semi-colon and an open or close quotation, the system may have read it as a variable.

In this case the errors should be in the rows you have highlighted. Edit the form before you download the csv.

Stephane

1 Like

Hello @stephanealoo and @cp622 ,
Can we try to find out exactly, please, what characters/elements in text fields may cause this error for exported (XLSX or csv) data. So we could suggest a standard regex on all text field to avoid such problems, also for the rest of the community.
Kind regards

Yes @wroos
Even though regex would only be a form specific fix and working for the future instances of data collection on that form.

I will be abke to try and see behaviour properly and then ask developers to check if this a true bug.

Stephane

Hello @cp622 and @stephanealoo,
You may also try if your recent problem can be solved by using Excel >> create Empty Workbook, Then >> Data >> From Text. There you can configure how delimiter etc are to be treated.
You could also try XLS export.

2 Likes

Dear @stephanealoo & @wroos
So far i have tested a few scenarios on both the ODK collect app and on the enketo webform. Please see the screenshot

i have used the data\text to columns in excel for ease of reading.
StudyID 99999901 pressed carriage return before entering the free text and then I did StudyID 99999903 and pressed return after entering the free text in enketo
StudyID 99999902 and 99999910 was done via ODK collect app, with the same method of carriage return before entering free text, then after enter entering the free text. I also tested a 2 word text answer with a capital letter to distinguish the two words, this also caused a split in the two lines MaigaminDilaxin.
When i replaced with an underscore there was no issue.

I tried to edit yesterdays response via the dashboard, as i had been able to do previously, on my laptop and the error of time cannot be in future kept appearing. I am not sure why it worked on other occasions and didn’t work yesterday.

So in my opinion there are 2 issues, the underlying issue of data splitting across 2 lines and the solution for this which only seems to work some of the times.

Best wishes

Catherine

Hi Catherine,
If I get you right, you deliberately introduced data entry errors which were captured

On a separate note, I still strongly believe that the settings you are applying for data conversion on excel could be causing this issue too

At this level, I suggest you send me your account name and the name of project on private message so that I go through it later in the day today.

Stephane

1 Like

Hi @stephanealoo
The issue appeared some weeks ago with data splitting across 2 lines. I entered data deliberately to test what was causing this issue, i have given a summary of these tests in my last reply.
I am using the csv XML values and headers download option, i am not doing any conversion, if i open the data in excel or in textpad without any edits i see that there is a studyID who’s data has appeared on 2 rows.
In my opinion the issue is with the csv output. I downloaded it by XLS values and headers and the issue of the studyID’s data over 2 rows does not exist. However if i now decide to download the data using XLS export type I will beed to re-write my SAS import script. I wanted to avoid this. I could enter some regex constraints, however I am not sure of all the characters that cause the issue.

Can you instruct me how to send a private message to you, i previously tried with @Kal_lam and ended up flaggng posts as inappropriate

Best wishes

Catherine

Hi
Send me that data privately I look at it.

Stephane

Hi @cp622,

Kindly please have a look on how to send a private message:

Have a great day!

I am just wondering whether there is any update on either of the queries raised in this thread?

Regards

Catherine