Remove hardcoded local filesystem path

[wtwhite]

Removes the last "important" case of /home/wtwhite appearing in a script.

---

Remaining local filesystem paths

• coverage-debug/. Not included in cross-build-diff-test.zip as it's not inside find-diff-classes. Safe to ignore.
• Poms and compiled .class files below compile/. Can be ignored since they're deleted by make clean-compile.
• 4 specific log files. I checked one, and it is just warnings about multiple SLF4J bindings. Safe to ignore.

find-diff-classes/run/gaoss/org/apache/commons/commons-configuration2/2.7/org.apache.commons.configuration2.DataConfiguration_ESTest.err
find-diff-classes/run/gaoss/org/apache/commons/commons-configuration2/2.8.0/org.apache.commons.configuration2.DataConfiguration_ESTest.err
find-diff-classes/run/mvnc/org/apache/commons/commons-configuration2/2.7/org.apache.commons.configuration2.DataConfiguration_ESTest.err
find-diff-classes/run/mvnc/org/apache/commons/commons-configuration2/2.8.0/org.apache.commons.configuration2.DataConfiguration_ESTest.err

• Surefire XML reports like run-mvn/gaoss/com/github/oshi/oshi-core/6.4.0/TEST-oshi.driver.windows.registry.HkeyPerformanceDataUtil_ESTest.xml (these are obsolete and can be ignored)
• EvoSuite test generation logs (can be ignored)
• Generated ..._ESTest_scaffolding.java source files:

  public static void setSystemProperties() {

    java.lang.System.setProperties((java.util.Properties) defaultProperties.clone());
    java.lang.System.setProperty("java.io.tmpdir", "/tmp");
    java.lang.System.setProperty("user.home", "/home/wtwhite");
    java.lang.System.setProperty("user.dir", "/home/wtwhite/code/regression-test-generation/find-diff-classes/testgen/mvnc/com/github/oshi/oshi-core/6.4.0");
  }

• Given that these don't seem to affect results, I'll leave them as-is. In theory they could affect results, however, so I'll make an issue to record the fact.

Testing

Fully tested on a fresh machine, using the instructions in the README to compile and run already-generated tests, including un-exporting $DISPLAY (from the Common Issues). Took ~2 hours to produce identical results to the paper:

bineq@gangbusters1:~/code/regression-test-generation-NEW-ZIP/find-diff-classes$ time make
--snip--
sqlite3 -tabs "/home/bineq/code/regression-test-generation-NEW-ZIP/find-diff-classes/db.sqlite" "SELECT * FROM differing_failed_test_methods ORDER BY 1, 2, 3, 4, 5, 6, 7;" > different_test_outcomes.tsv
sqlite3 -tabs "/home/bineq/code/regression-test-generation-NEW-ZIP/find-diff-classes/db.sqlite" "select provider1, replace(groupId, '/', '.') || ':' || artifactId || ':' || version as artifact, testClass as 'test class', count(*) as 'tests' from differing_failed_test_methods group by 1, 2, 3 order by 1, 2, 3;" | perl -lpe 's/\t/ & /g; $_ .= " \\\\"; s/_/\\_/g' > data-summary-table-body.tex
( /usr/bin/echo '\begin{tabular}{p{0.5cm}p{5cm}p{7.5cm}r} ' ; /usr/bin/echo '\toprule ' ; /usr/bin/echo '\textbf{build} & \textbf{artifact} & \textbf{test class} & \textbf{tests} \\ \midrule' ; cat data-summary-table-body.tex ; /usr/bin/echo ' \bottomrule \end{tabular}' ) > data-summary-table-tabular.tex

real    149m46.652s
user    139m18.401s
sys     7m6.919s
bineq@gangbusters1:~/code/regression-test-generation-NEW-ZIP/find-diff-classes$ echo $?
0
bineq@gangbusters1:~/code/regression-test-generation-NEW-ZIP/find-diff-classes$ wc -l different_test_outcomes.tsv
60 different_test_outcomes.tsv
bineq@gangbusters1:~/code/regression-test-generation-NEW-ZIP/find-diff-classes$ md5sum ../paper/generated/data-summary-table-tabular.tex data-summary-table-tabular.tex
a84c7bf57649ec7165852cbfd3199447  ../paper/generated/data-summary-table-tabular.tex
a84c7bf57649ec7165852cbfd3199447  data-summary-table-tabular.tex