CFind results switch for xml

[sfc-gh-jyost]

I am able to make connections to PACS and use send_c_find to parse the database for meta data. The results come back in this format...
(0008, 0005) Specific Character Set CS: 'ISO_IR 192',
0008, 0052) Query/Retrieve Level CS: 'PATIENT',
0008, 0054) Retrieve AE Title AE: 'ANY-SCP',
(0010, 0010) Patient's Name PN: 'Test^Bart'"

Is there a way to force results to come back in xml format or something easier to parse?

[mrbean-bremen]

send_c_find returns a dataset, not a string - you can handle the dataset items any way you want by iterating over them. There is also the possibility to convert the dataset to json.

[scaramallion]

It's never in XML format, either...

[sfc-gh-jyost]

Thanks for getting back to fast! I will try the convert to json. My issue is that I can iterate through it and put each record in a single column in a db table. But I then need to transform that dataset to key value pairs into another more readable table. Here is an example..
"(0008, 0005) Specific Character Set CS: 'ISO_IR 192'\n(0008, 0052) Query/Retrieve Level CS: 'PATIENT'\n(0008, 0054) Retrieve AE Title AE: 'ANY-SCP'\n(0010, 0010) Patient's Name PN: 'Prescott^Bart'"

thats all in one column in my db. I want to break that out to a tag column and value columns. JSON would be easier to work with, I'll try that before I put it in the db. I was hoping the findscu had an option to dump into xml like dcm4che does or offis. thanks for the confirmation that it doesn't :)

[mrbean-bremen]

Can't you just iterate the returned dataset directly instead of relying on some string output? Who writes that output - isn't this your own code?

[sfc-gh-jyost]

findscu returns a dataset like this...
[(0008, 0005) Specific Character Set CS: ''
(0008, 0052) Query/Retrieve Level CS: 'PATIENT'
(0008, 0054) Retrieve AE Title AE: 'ANY-SCP'
(0010, 0010) Patient's Name PN: 'Doe^John',
(0008, 0005) Specific Character Set CS: ''
(0008, 0052) Query/Retrieve Level CS: 'PATIENT'
(0008, 0054) Retrieve AE Title AE: 'ANY-SCP'
(0010, 0010) Patient's Name PN: 'Kirk^James T.'
]

So I can iterate through that and take each record above and parse it in python, but was hoping there were options on the output that findscu returns or pydicom had a method to parse it.

[mrbean-bremen]

I'm still not sure I understand your problem. You get a Dataset as a result, and iterating over it you get DataElement instances, which you query for the tag, VM, multiplicity and value, and you can feed the needed values to your database. There is nothing to parse here, unless I don't understand something in your workflow.

[sfc-gh-jyost]

Nevermind. I figured it out. I clearly did not understand the dataset and how to extract the elements.
https://towardsdatascience.com/introducing-pydicom-its-classes-methods-and-attributes-518c1d71162
this solved my problem. Instead of doing a dcmread(somedicomimage.dcm) I already had the dataset and the iteration of it in results. so simply
patient_name = results.PatientName
st.text(patient_name)

Appreciate all the comments and questions. They pushed me to better understand what I was working with. thanks!!