Weasis error for studies with Umlaut like "ä"

jjoerg · Jun 7th 2016

Hello,

surely a german-speaking problem: studies containing an umlaut (ä,ö,ü) fail to load in the weasis viewer. weasis opens smoothly but i Recieve an error:
error weasis:Error on loading the XML Manifest from ...
When I change the study description on a 1.4.19beta instance from "MR Schädel" to "MR Schaedel" everything runs fine. Is there a solution or a workaround? "Conquest does not modify or interpret any field data", this will not work.

A bit strange, with conquest 1.4.17d and Apache 2.2 On Windows Small Business Server 2008 it works, on Server 2003R2 not.
Any ideas or hints?

Kind regards an thanks in advance!

Jörg

marcelvanherk · Jun 7th 2016

Hi Joerg,

can you inspect the weasisseriesxml file for any obvious errors? It may be that a string needs to be escaped. The code that generates it is in weasisseriesxml.lua. You can for instance try to add a line after 5 saying

patid=''

This may circumvent the issue.

Marcel

jjoerg · Jun 7th 2016

Hi Marcel,

thanks for your reply, your workaround doesn't work. The issue concerns all study-descriptions and patient names with umlauts.
I try to follow this direction implementing in line 37
r[0].StudyDescription = 'test'
it runs fine.

I try to follow further and implement a kind of stringreplacement in the weasisseriesxml, or do you have better ideas?

Kind regards

Jörg

jjoerg · Jun 7th 2016

Hi Marcel,

quick and dirty with success:
incorporated FixUTF8 to remove Umlaute

http://notebook.kulchenko.com/…ing-malformed-utf8-in-lua

Trying to perform the rigth replacement ä=ae and so on

Kind regards

Jörg

marcelvanherk · Jun 7th 2016

Great Joerg,

once the code is complete, I will integrate it in the release.

Marcel

jjoerg · Jun 8th 2016

Hi Marcel,

this was my first contact to lua. So I stick to the "quick and dirty" solution.

As far as I understand: the strings I want to manipulate are mor or less information strings for the WADO request, the information displayed by the weasis viewers is picked from the dicom images retrieved. In my final solution the non alphanumeric characters are replaced by an "_". Modification marked as ascii-compliance insert

weasisseries.xml after line 22:

Code

-- updated to pass compress mode, mvh 20140126print('Content-Type: application/xml\n')local patid = string.gsub(series2, ':.*$', '')local seriesuid = string.gsub(series2, '^.*:', '')local q = DicomObject:new()q.QueryRetrieveLevel = 'IMAGE'q.PatientID = patidq.SeriesInstanceUID = seriesuidq.SOPInstanceUID = ''q.PatientBirthDate = ''q.PatientName = ''q.StudyInstanceUID = ''q.StudyDescription = ''q.StudyDate = ''q.StudyTime = ''q.SeriesDescription = ''q.SeriesNumber = ''q.Modality = ''q.ImageNumber = ''r = dicomquery(servercommand('get_param:MyACRNema'), 'IMAGE', q)-- ascii compliance insert: remove  non alphanumeric characters 06/2016 r[0].PatientName =r[0].PatientName:gsub('%W','_')r[0].StudyDescription =r[0].StudyDescription:gsub('%W','_')r[0].SeriesDescription = r[0].SeriesDescription:gsub('%W','_')-- end ascii compliance insert wt = ""if compress=='J3' or compress=='j3' then wt = ' WadoTransferSyntaxUID="1.2.840.10008.1.2.4.50"' endif compress=='J4' or compress=='j4' then wt = ' WadoTransferSyntaxUID="1.2.840.10008.1.2.4.51"' endif compress=='J5' or compress=='j5' then wt = ' WadoTransferSyntaxUID="1.2.840.10008.1.2.4.53"' endif compress=='J1' or compress=='j1' then wt = ' WadoTransferSyntaxUID="1.2.840.10008.1.2.4.55"' endif compress=='J2' or compress=='j2' then wt = ' WadoTransferSyntaxUID="1.2.840.10008.1.2.4.57"' endif compress=='J1' or compress=='j1' then wt = ' WadoTransferSyntaxUID="1.2.840.10008.1.2.4.70"' endif compress=='JK' or compress=='jk' then wt = ' WadoTransferSyntaxUID="1.2.840.10008.1.2.4.90"' endif compress=='JL' or compress=='jl' then wt = ' WadoTransferSyntaxUID="1.2.840.10008.1.2.4.91"' endlocal s = DicomObject:new()print([[<?xml version="1.0" encoding="utf-8" ?><wado_query xmlns= "http://www.weasis.org/xsd" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" wadoURL="]]..webscriptadress..[[" requireOnlySOPInstanceUID="false" overrideDicomTagsList="0x00000000" >  <Patient PatientID="]]..patid..[[" PatientName="]]..r[0].PatientName..[[" PatientBirthDate="]]..r[0].PatientBirthDate..[[" >    <Study StudyInstanceUID="]]..r[0].StudyInstanceUID..[[" StudyDescription="]]..r[0].StudyDescription..[[" StudyDate="]]..r[0].StudyDate..[[" StudyTime="]]..r[0].StudyTime..[[" >      <Series SeriesInstanceUID="]]..r[0].SeriesInstanceUID..[[" SeriesDescription="]]..r[0].SeriesDescription..[[" SeriesNumber="]]..r[0].SeriesNumber..[[" Modality="]]..r[0].Modality..[["]]..wt..[[ >]])for i=0, #r-1 do  print([[<Instance SOPInstanceUID="]]..r[i].SOPInstanceUID..[[" InstanceNumber="]]..i..[[" />]])endprint([[      </Series>    </Study>  </Patient></wado_query>]])

In weasisstudy.xml only patient name and study name can be treated at the same place, the series name has to be modified in the loop iterating through the different series.

Code

-- updated to pass compress mode, mvh 20140126


print('Content-Type: application/xml\n')




local patid = string.gsub(study2, ':.*$', '')
local studyuid = string.gsub(study2, '^.*:', '')
local q = DicomObject:new()
q.QueryRetrieveLevel = 'SERIES'
q.PatientID = patid
q.StudyInstanceUID = studyuid
q.PatientBirthDate = ''
q.PatientName = ''
q.StudyDescription = ''
q.StudyDate = ''
q.StudyTime = ''
q.SeriesInstanceUID = ''
q.SeriesDescription = ''
q.SeriesNumber = ''
q.Modality = ''
r = dicomquery(servercommand('get_param:MyACRNema'), 'SERIES', q)


wt = ""
if compress=='J3' or compress=='j3' then wt = ' WadoTransferSyntaxUID="1.2.840.10008.1.2.4.50"' end
if compress=='J4' or compress=='j4' then wt = ' WadoTransferSyntaxUID="1.2.840.10008.1.2.4.51"' end
if compress=='J5' or compress=='j5' then wt = ' WadoTransferSyntaxUID="1.2.840.10008.1.2.4.53"' end
if compress=='J1' or compress=='j1' then wt = ' WadoTransferSyntaxUID="1.2.840.10008.1.2.4.55"' end
if compress=='J2' or compress=='j2' then wt = ' WadoTransferSyntaxUID="1.2.840.10008.1.2.4.57"' end
if compress=='J1' or compress=='j1' then wt = ' WadoTransferSyntaxUID="1.2.840.10008.1.2.4.70"' end
if compress=='JK' or compress=='jk' then wt = ' WadoTransferSyntaxUID="1.2.840.10008.1.2.4.90"' end
if compress=='JL' or compress=='jl' then wt = ' WadoTransferSyntaxUID="1.2.840.10008.1.2.4.91"' end


local s = DicomObject:new()


-- ascii compliance insert: remove  non alphanumeric characters 06/2016
r[0].PatientName =r[0].PatientName:gsub('%W','_')
r[0].StudyDescription =r[0].StudyDescription:gsub('%W','_')
-- end ascii compliance insert 


print([[
<?xml version="1.0" encoding="utf-8" ?>
<wado_query xmlns= "http://www.weasis.org/xsd" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" wadoURL="]]..webscriptadress..[[" requireOnlySOPInstanceUID="false" overrideDicomTagsList="0x00000000">
  <Patient PatientID="]]..patid..[[" PatientName="]]..r[0].PatientName..[[" PatientBirthDate="]]..r[0].PatientBirthDate..[[" >
    <Study StudyInstanceUID="]]..r[0].StudyInstanceUID..[[" StudyDescription="]]..r[0].StudyDescription..[[" StudyDate="]]..r[0].StudyDate..[[" StudyTime="]]..r[0].StudyTime..[[" >
]])


for i=0, #r-1 do


-- ascii compliance insert: remove  non alphanumeric characters 06/2016
r[i].SeriesDescription = r[i].SeriesDescription:gsub('%W','_')
-- end ascii compliance insert


  print([[
      <Series SeriesInstanceUID="]]..r[i].SeriesInstanceUID..[[" SeriesDescription="]]..r[i].SeriesDescription..[[" SeriesNumber="]]..r[i].SeriesNumber..[[" Modality="]]..r[i].Modality..[["]]..wt..[[ >
  ]])


  s.QueryRetrieveLevel = 'IMAGE'
  s.SOPInstanceUID = ''
  s.ImageNumber = ''
  s.SeriesInstanceUID = r[i].SeriesInstanceUID
  t = dicomquery(servercommand('get_param:MyACRNema'), 'IMAGE', s)
  for j=0, #t-1 do
    print([[<Instance SOPInstanceUID="]]..t[j].SOPInstanceUID..[[" InstanceNumber="]]..j..[[" />]])
  end


  print([[
      </Series>
]])
end


print([[
    </Study>
  </Patient>
</wado_query>
]])

Display More

Kind regards

Jörg

marcelvanherk · Jun 8th 2016

Thanks for that!

The _ trick is a good start. It may also be that such characters can be replaced with their % equivalents.

regards,

Marcel

Weasis error for studies with Umlaut like "ä"

Participate now!

Share