finished

Dockerfile

@@ -4,6 +4,7 @@ RUN apt update -y && apt install -y make
 RUN apt install -y git
 RUN apt install -y gcc
 RUN apt install -y gawk
+RUN apt install -y python
 RUN gcc --version
 RUN apt install -y build-essential
 RUN git clone https://github.com/usnistgov/SCTK.git

Jenkinsfile

@@ -7,9 +7,10 @@ pipeline {
                 checkout([$class: 'GitSCM', branches: [[name: '*/master']], doGenerateSubmoduleConfigurations: false, extensions: [], submoduleCfg: [], userRemoteConfigs: [[credentialsId: '8cdbf19a-e62d-4807-9112-bc7a8d225074', url: 'https://git.wmi.amu.edu.pl/s416094/s416094-mlworkshops']]])
                 sh label: '', script: 'chmod 777 counter.sh'
                 sh label: '', script: './counter.sh'
-                archiveArtifacts 'hypothesis.trn'
-                archiveArtifacts 'reference.trn'
-                archiveArtifacts 'results.txt'
+                archiveArtifacts 'wikiniews_results.tsv'
+                archiveArtifacts 'wer_avg.txt'
+                archiveArtifacts 'srr.txt'
+                archiveArtifacts 'wer_srr.txt'
             }
         }
     }

counter.sh

@@ -1,4 +1,6 @@
 
 awk 'BEGIN { FS="\t" } {print gensub("^\\*\\*", "\\\\**", "g", $2) "\t(sp1_"NR")"}'  wikiniews_results.tsv >  reference.trn
 awk 'BEGIN { FS="\t" } {print gensub("^\\*\\*", "\\\\**", "g", $3) "\t(sp1_"NR")"}'  wikiniews_results.tsv >  hypothesis.trn
-sclite -f 0 -r reference.trn trn -h hypothesis.trn trn -e utf-8 -i rm -o all stdout >> results.txt
+sclite -f 0 -r reference.trn trn -h hypothesis.trn trn -e utf-8 -i rm -o all stdout >> results.txt
+python extract.py
+mv results.tsv wikiniews_results.tsv

extract.py

@@ -0,0 +1,55 @@
+import re
+import fileinput
+import csv
+wars = []
+corrects=0
+with open("results.txt", "r", encoding='utf-8') as ins:
+    for line in ins:
+        if line.startswith('Scores'):
+            correct = int(re.findall('\d+', line)[0])
+            substitutions = int(re.findall('\d+', line)[1])
+            deletes = int(re.findall('\d+', line)[2])
+            inserts = int(re.findall('\d+', line)[3])
+            war = (substitutions+deletes+inserts)/(substitutions+deletes+correct)
+            print(correct, substitutions, deletes, inserts)
+            wars.append(war)
+            if substitutions==0 and deletes==0 and inserts==0:
+                corrects = corrects + 1
+
+srr = corrects / sum(wars)
+print(srr)
+
+war_avg = sum(wars) / len(wars)
+print(war_avg)
+with open("wer_avg.txt", "w") as wer_avg_file:
+    wer_avg_file.write("WER avg: ")
+    wer_avg_file.write(str(war_avg))
+with open("srr.txt", "w") as srr_file:
+    srr_file.write("SRR: ")
+    srr_file.write(str(srr))
+
+with open("wer_srr.txt", "w") as wer_srr:
+    wer_srr.write("SRR: ")
+    wer_srr.write(str(srr))
+    wer_srr.write("\n")
+    wer_srr.write("WER avg: ")
+    wer_srr.write(str(war_avg))
+
+
+counter = 0;
+with open("wikiniews_results.tsv",'r', encoding="utf-8") as wik_in:
+    with open("result.tsv",'w', encoding="utf-8") as f_out:
+        writer = csv.writer(f_out, delimiter='\t', lineterminator='\n')
+        reader = csv.reader(wik_in, delimiter='\t')
+
+        result = []
+
+        for row in reader:
+            # add new column values
+            row.append(wars[counter])
+            result.append(row)
+            counter=counter+1
+
+        writer.writerows(result)
+
+