Extract string followed by specific word/symbol

Question

I have two lines as shown below in my input file input.txt and I need to extract claimStartDate from first line and claimEndDate from second line.

<ProfessionalClaim paymentIndicator="P" claimProcessedDateTime="20180409120000102" claimEndDate="2018-04-02" claimStartDate="2018-04-02" sourceSystemId="abcd" claimActionCode="00"> <ProfessionalClaim paymentIndicator="P" claimProcessedDateTime="20180430120000281" claimEndDate="2018-04-17" claimStartDate="2018-04-17" sourceSystemId="abcd" claimActionCode="00"> rm input.txt awk '/<ProfessionalClaim/' test.xml | head -1 > input.txt awk '/<ProfessionalClaim/' test.xml | tail -1 >> input.txt awk '{match($0, "claimStartDate=\"([^\"]+)\"", start); print start[1]} \ {match($0, "claimEndDate=\"([^\"]+)\"", end); print end[1]}' input.txt 

Answer 1

$ awk '/F_LINE/ {match($0, "claimStartDate=\"([^\"]+)\"", start); print start[1]} \ /L_LINE/ {match($0, "claimEndDate=\"([^\"]+)\"", end); print end[1]}' input.txt 2018-04-02 2018-04-17 

---

EDIT due to your new information:

$ awk 'NR==1 {match($0, "claimStartDate=\"([^\"]+)\"", start); print start[1]} \ NR==2 {match($0, "claimEndDate=\"([^\"]+)\"", end); print end[1]}' input.txt 2018-04-02 2018-04-17 

You can also do this all in one run:

$ grep "<ProfessionalClaim" text.xml \ | sed -n '1p;$p' \ | $ awk 'NR==1 {match($0, "claimStartDate=\"([^\"]+)\"", start); print start[1]} \ NR==2 {match($0, "claimEndDate=\"([^\"]+)\"", end); print end[1]}' 

• grep find all line with <ProfessionalClaim in text.xml
• sed truncate the lines to the first and the last onyl
• awk will print the claimStartDate for the first line and ClaimEndDate for the second line

Answer 2

Assuming some XML input document like the following:

<?xml version="1.0"?> <root> <ProfessionalClaim paymentIndicator="P" claimProcessedDateTime="20180409120000102" claimEndDate="2018-04-02" claimStartDate="2018-04-02" sourceSystemId="abcd" claimActionCode="00"/> <ProfessionalClaim paymentIndicator="P" claimProcessedDateTime="20180430120000281" claimEndDate="2018-04-17" claimStartDate="2018-04-17" sourceSystemId="abcd" claimActionCode="00"/> <ProfessionalClaim paymentIndicator="P" claimProcessedDateTime="20180430120000281" claimEndDate="2018-04-18" claimStartDate="2018-04-18" sourceSystemId="abcd" claimActionCode="00"/> <ProfessionalClaim paymentIndicator="P" claimProcessedDateTime="20180430120000281" claimEndDate="2018-04-19" claimStartDate="2018-04-19" sourceSystemId="abcd" claimActionCode="00"/> </root> 

... we may use xmlstarlet to extract the claimStartDate attribute's value from each ProfessionalClaim node that has another ProfessionalClaim node following it, together with that next ProfessionalClaim node's claimEndDate attribute's value:

xmlstarlet select --template \ --match '//ProfessionalClaim[following-sibling::ProfessionalClaim/@claimEndDate]' \ --value-of 'concat(@claimStartDate, " ", following-sibling::ProfessionalClaim/@claimEndDate)' \ -nl input.txt 

This first matches each ProfessionalClaim node that is followed by another ProfessionalClaim node.

For each such node, the value of the claimStartDate attribute is concatenated with the value of the claimEndDate attribute of the following ProfessionalClaim node, with a single space character as delimiter.

Given my example document above, this would generate

2018-04-02 2018-04-17 2018-04-17 2018-04-18 2018-04-18 2018-04-19