Parse Data output from a remote STAF command - regex

Question

What would the easiest way to parse the Data section from this STAF command be?
Cannot find a STAF parameter which I can pass to the command to automatically do this,
so looks like parsing/regular expression might be best option?
Note: I do not want to use any external libraries.
[root#source ~]# STAF target PROCESS START SHELL COMMAND "ls" WAIT RETURNSTDOUT
Response
--------
{
Return Code: 0
Key : <None>
Files : [
{
Return Code: 0
Data : myFile.txt
myFile2.txt
myFile3.txt
}
]
}
Instead I would like the output/result to be formated like ..
[root#source ~]# STAF target PROCESS START SHELL COMMAND "ls" WAIT RETURNSTDOUT
myFile.txt
myFile2.txt
myFile3.txt

Answer 1

Best way to this is Create a XML file and use python script to access the data part of STAFResult since STAF Return data in Marshalled form as "CONTENT" and python can be use to grab that.
I will try to explain it with simple example, Its an HTTP request to server.
<stafcmd>
<location>'%s'%machineName</location>
<service>'http'</service>
<request>'DOGET URL %s?phno=%s&shortCode=%s&query=%s' % (url, phno, shortCode, escapeQuery)</request>
</stafcmd>
<if expr="RC == 0">
<sequence>
<call function="'func_Script'"></call>
<if expr="rc == 0"> <!-- Pass At First Query -->
<sequence>
<message>'PASS#Fisrt HTTPRequest: Keyword = %s,\nRequired Response = %s,\ncontent=%s' %(query, response, content)</message>
<tcstatus result="'pass'">'Pass:' </tcstatus>
</sequence>
<else> <!-- Check For MORE -->
<call function="'Validate_QueryMore'"> </call>
</else>
</if>
</sequence>
<else>
<message>'ERROR: HTTPRequest QUERY : RC = %s Result= %s' %(rc,STAFResult)</message>
</else>
</if>
<function name="func_Script">
<script>
import re
content = STAFResult['content'].lower()
response = response.lower()
test = content.find(response)
if test != -1:
rc = 0
else:
rc = 1
</script>
</function>
Hope It will give you some Help.

Answer 2

You can pipe the output of your command through an sed script that will filter out only the filenames for you. Here's a first cut:
sed -ne '/^[a-z]/p;/Data/s/[^:]*: \(.*\)/\1/p'
The idea is: If a line starts with a lower-case letter, that's a file name (expression up to the first semicolon). If the string "Data" is on the line, take everything that comes after the first colon in that line (expression after the semicolon). Everything else is ignored.
You might want to be more specific than just expecting a lower-case letter at the beginning (this would filter out the "Response" line at the beginning, but if your filename might start with an upper-case letter, that won't work). Also, just looking for the string "Data" might be a bit too general -- that string might occur in the filename as well. But hopefully you get the idea. To use this, run your command like this:
STAF ... | sed -ne ...