How to split, re-arrange in order HL7 message

I receive an HL7 message which looks like:
MSH|fdsas43423|fdsfs|432423|fdfads|423423|
PID|43243|fdfds|654|HGDF|76554|HGDS
PV1|5434|gfsd|543534|hgdfz|
ORC||||F|RE
OBR|20140709|gffsd|425643|hgd|543
OBX|65465|GFFSD|7664|aHGFF|4654
NTE|GHFDSGF|543|GFSFD|654|HGDF
OBX|423|FDSAD|432423|FDAS|423423
ORC||||p|RE
OBR|20140706|gffsd|425643|hgd|543
OBX|65465|GFFSD|7664|aHGFF|4654
NTE|GHFDSGF|543|GFSFD|654|HGDF
OBX|423|FDSAD|432423|FDAS|423423
ORC||||F|RE
OBR|20140710|gffsd|425643|hgd|543
OBX|65465|GFFSD|7664|aHGFF|4654
NTE|GHFDSGF|543|GFSFD|654|HGDF
OBX|423|FDSAD|432423|FDAS|423423
ORC||||F|RE
OBR|20140711|gffsd|425643|hgd|543
OBX|65465|GFFSD|7664|aHGFF|4654
NTE|GHFDSGF|543|GFSFD|654|HGDF
OBX|423|FDSAD|432423|FDAS|423423

I need to split the message into sections which start at ORC and end at the next one.  the field next to OBR is a timestamp which is alwasy unique and once the message is split into sections it needs to be rebuilt in order of the time stamp from latest to earliest.  

So it should look like:
MSH|fdsas43423|fdsfs|432423|fdfads|423423|
PID|43243|fdfds|654|HGDF|76554|HGDS
PV1|5434|gfsd|543534|hgdfz|
ORC||||p|RE
OBR|20140706|gffsd|425643|hgd|543
OBX|65465|GFFSD|7664|aHGFF|4654
NTE|GHFDSGF|543|GFSFD|654|HGDF
OBX|423|FDSAD|432423|FDAS|423423
ORC||||F|RE
OBR|20140709|gffsd|425643|hgd|543
OBX|65465|GFFSD|7664|aHGFF|4654
NTE|GHFDSGF|543|GFSFD|654|HGDF
OBX|423|FDSAD|432423|FDAS|423423
ORC||||F|RE
OBR|20140710|gffsd|425643|hgd|543
OBX|65465|GFFSD|7664|aHGFF|4654
NTE|GHFDSGF|543|GFSFD|654|HGDF
OBX|423|FDSAD|432423|FDAS|423423
ORC||||F|RE
OBR|20140711|gffsd|425643|hgd|543
OBX|65465|GFFSD|7664|aHGFF|4654
NTE|GHFDSGF|543|GFSFD|654|HGDF
OBX|423|FDSAD|432423|FDAS|423423

Thanks.
Nick MaloneAsked:
Who is Participating?
 
aikimarkConnect With a Mentor Commented:
It bothered me that I was reading the file twice, so I streamlined the code.
import os

def main():
    #!/usr/local/bin/python2.7
    os.chdir( "\\users\\aikimark\\downloads")
    f = open('HL7input.txt', 'r')
    linelist = f.readlines()
    f.close

    groups = []
    agroup=[]
    groupkey=''
    recnum=0

    for line in linelist:
      a=line.split('|')
      if a[0]=='ORC':
          if len(agroup) == 1:
            #add to groups and reset
            agroup.append(recnum-1)
            agroup.append(groupkey)
            groups.append(agroup)
            agroup=[]
            groupkey=''
            agroup.append(recnum)
          else:
            agroup.append(recnum)

      elif a[0]=='OBR':
          #retain group key
          groupkey=a[1]
      recnum=recnum+1

    #add last group to the groups list
    agroup.append(recnum-1)
    agroup.append(groupkey)
    groups.append(agroup)
    firstgroupstart = groups[0][0]

##    print 'before sort'
##    for g in groups:
##      print g

    #sort on third item in tuple -- the date of the group in yyyymmdd format string
    sgroups=sorted(groups, key=lambda g: g[2])

##    print 'after sort'
##    for g in sgroups:
##      print g
    f = open('HL7output.txt', 'w')

    #print 'sorted file'
    print range(firstgroupstart)
    for l in range(firstgroupstart):
      f.write( linelist[l])
    for g in sgroups:
      for outline in range(g[0],g[1]+1):
        f.write( linelist[outline])
    f.close

Open in new window

0
 
aikimarkCommented:
It might be more helpful if you posted an example with different data in each 'section'.
0
 
CEHJCommented:
Don't use homebrewed techniques - there are parsers available,  e.g. http://hl7api.sourceforge.net/
0
The new generation of project management tools

With monday.com’s project management tool, you can see what everyone on your team is working in a single glance. Its intuitive dashboards are customizable, so you can create systems that work for you.

 
aikimarkCommented:
Since you are using Python, you might want to consider a Python-centric parser, such as hl7apy.
http://sourceforge.net/projects/hl7apy/
0
 
Nick MaloneAuthor Commented:
I've requested that this question be deleted for the following reason:

no answers.
0
 
aikimarkCommented:
You have seen two different potential solutions and received a request to clarify your data.  Your "no answers" assertion is bogus.  I object to your deletion request.  Please work with the participating experts.
0
 
CEHJCommented:
Your "no answers" assertion is bogus.  I object to your deletion request.

.. as do I
0
 
aikimarkCommented:
The following seems to work in my Windows environment using your posted data.  Be aware that if your posted data does not strongly resemble your actual/production data, this code will not work.
import os

def main():
    #!/usr/local/bin/python2.7
    os.chdir( "\\users\\aikimark\\downloads")
    f = open('HL7input.txt', 'r')

    groups = []
    agroup=[]
    groupkey=''
    recnum=0
    for line in f:
      a=line.split('|')
      if a[0]=='ORC':
          if len(agroup) == 1:
            #add to groups and reset
            agroup.append(recnum-1)
            agroup.append(groupkey)
            groups.append(agroup)
            agroup=[]
            groupkey=''
            agroup.append(recnum)
          else:
            agroup.append(recnum)

      elif a[0]=='OBR':
          #retain group key
          groupkey=a[1]
      recnum=recnum+1

    #add last group to the groups list
    agroup.append(recnum-1)
    agroup.append(groupkey)
    groups.append(agroup)
    firstgroupstart = groups[0][0]

##    print 'before sort'
##    for g in groups:
##      print g

    #sort on third item in tuple -- the date of the group in yyyymmdd format string
    sgroups=sorted(groups, key=lambda g: g[2])

##    print 'after sort'
##    for g in sgroups:
##      print g
    f.close
    f = open('HL7input.txt', 'r')
    linelist = f.readlines()
    f.close
    f = open('HL7output.txt', 'w')

    #print 'sorted file'
    print range(firstgroupstart)
    for l in range(firstgroupstart):
      f.write( linelist[l],)
    for g in sgroups:
      for outline in range(g[0],g[1]+1):
        f.write( linelist[outline],)
    f.close

if __name__ == '__main__':
    main()

Open in new window

0
 
aikimarkCommented:
accept http:#a40237568 as the solution
0
 
CEHJCommented:
Points splits
0
 
aikimarkCommented:
@CEHJ

Your proposed solution only went so far as the parsing.  It didn't address the sorting.
0
 
CEHJCommented:
a. that's why a points split
b. use of a proper API means that regular Java Collection sorting is trivial
0
 
aikimarkCommented:
a. although we both proposed a parsing library via a link to different sourceforge projects, I actually posted Python code that did the parsing without any extra parsing libraries.  My comment is a complete solution.
b. Java collection sorting is not a consideration with this Python problem.
0
All Courses

From novice to tech pro — start learning today.