#85 IFS=

filters/01-delete-non-java-files.sh

@@ -37,7 +37,7 @@ fi
 
 mkdir -p "$(dirname "${list}")"
 find "${home}" -type f -not -name '*.java' -print > "${list}"
-while read -r f; do
+while IFS= read -r f; do
     rm -f "${f}"
 done < "${list}"
 

filters/02-delete-package-info.sh

@@ -34,7 +34,7 @@ fi
 
 mkdir -p "$(dirname "${list}")"
 find "${home}" -type f -a -name 'package-info.java' -print > "${list}"
-while read -r f; do
+while IFS= read -r f; do
     rm -f "${f}"
 done < "${list}"
 

filters/03-delete-tests.sh

@@ -42,7 +42,7 @@ mkdir -p "$(dirname "${list}")"
     find "${home}" -type f -name '*Tests.java' -print;
     find "${home}" -type f -path '**/src/test/java/**/*.java' -print
 } > "${list}"
-while read -r f; do
+while IFS= read -r f; do
     rm -f "${f}"
 done < "${list}"
 

filters/04-delete-unparseable.sh

@@ -44,7 +44,7 @@ candidates=${temp}/files-to-parse.txt
 mkdir -p "$(dirname "${candidates}")"
 find "${home}" -type f -name '*.java' -print > "${candidates}"
 py=${LOCAL}/filters/delete-unparseable.py
-while read -r f; do
+while IFS= read -r f; do
     printf "python3 %s %s %s\n" "${py@Q}" "${f@Q}" "${list@Q}" >> "${jobs}"
 done < "${candidates}"
 "${LOCAL}/help/parallel.sh" "${jobs}"

filters/05-delete-long-lines.sh

@@ -39,7 +39,7 @@ touch "${list}"
 
 candidates=${temp}/files-to-check-line-lengths.txt
 find "${home}" -type f -name '*.java' -print > "${candidates}"
-while read -r f; do
+while IFS= read -r f; do
     length=$(LC_ALL=C awk '{ print length($0) }' < "${f}" | sort -n | tail -1)
     if [ -z "${length}" ]; then continue; fi
     if [ "${length}" -gt "${max}" ]; then

filters/06-delete-non-classes.sh

@@ -44,7 +44,7 @@ candidates=${temp}/classes-to-challenge.txt
 mkdir -p "$(dirname "${candidates}")"
 find "${home}" -type f -name '*.java' -print > "${candidates}"
 py=${LOCAL}/filters/delete-non-classes.py
-while read -r f; do
+while IFS= read -r f; do
     printf "python3 %s %s %s\n" "${py@Q}" "${f@Q}" "${list@Q}" >> "${jobs}"
 done < "${candidates}"
 "${LOCAL}/help/parallel.sh" "${jobs}"

filters/07-delete-invalid-files.sh

@@ -44,7 +44,7 @@ candidates=${temp}/classes-to-filter.txt
 mkdir -p "$(dirname "${candidates}")"
 find "${home}" -type f -name '*.java' -print > "${candidates}"
 py=${LOCAL}/filters/delete-invalid-files.py
-while read -r f; do
+while IFS= read -r f; do
     printf "python3 %s %s %s\n" "${py@Q}" "${f@Q}" "${list@Q}" >> "${jobs}"
 done < "${candidates}"
 "${LOCAL}/help/parallel.sh" "${jobs}"

steps/aggregate-repo.sh

@@ -36,8 +36,8 @@ if [ -e "${ddir}" ]; then
     exit
 fi
 
-find "${dir}" -name '*.m' | while read -r m; do
-    ls "${m}".* | while read -r v; do
+find "${dir}" -name '*.m' | while IFS= read -r m; do
+    ls "${m}".* | while IFS= read -r v; do
         java=$(echo "${v}" | sed "s|${dir}||" | sed "s|\.m\..*$||")
         metric=$(echo "${v}" | sed "s|${dir}${java}.m.||")
         csv=${ddir}/${metric}.csv

steps/aggregate.sh

@@ -38,7 +38,7 @@ touch "${jobs}"
 
 declare -i repo=0
 sh="$(dirname "$0")/aggregate-repo.sh"
-echo "${repos}" | while read -r r; do
+echo "${repos}" | while IFS= read -r r; do
     repo=$((repo+1))
     printf "%s %s %s %s\n" "${sh@Q}" "${r@Q}" "${repo@Q}" "${total@Q}" >> "${jobs}"
 done
@@ -48,15 +48,15 @@ wait
 mkdir -p "${TARGET}/data"
 rm -rf "${TARGET}/data/*.csv"
 printf "repository,file" >> "${TARGET}/data/all.csv"
-echo "${all}" | while read -r a; do
+echo "${all}" | while IFS= read -r a; do
     printf ',%s' "${a}" >> "${TARGET}/data/all.csv"
 done
 printf "\n" >> "${TARGET}/data/all.csv"
 
-find "${TARGET}/data" -maxdepth 2 -mindepth 2 -type d -print | while read -r d; do
+find "${TARGET}/data" -maxdepth 2 -mindepth 2 -type d -print | while IFS= read -r d; do
     r=$(realpath --relative-to="${TARGET}/data" "$d" )
-    find "${d}" -name '*.csv' -maxdepth 1 -exec basename {} \; | while read -r csv; do
-        while read -r t; do
+    find "${d}" -name '*.csv' -maxdepth 1 -exec basename {} \; | while IFS= read -r csv; do
+        while IFS= read -r t; do
             echo "${r},${t}" >> "${TARGET}/data/${csv}"
         done < "${d}/${csv}"
     done

steps/filter.sh

@@ -24,7 +24,7 @@ set -e
 set -o pipefail
 
 mkdir -p "${TARGET}/temp/reports"
-find "${LOCAL}/filters" -name '*.sh' -exec realpath --relative-to="${LOCAL}/filters" {} \; | sort | while read -r filter; do
+find "${LOCAL}/filters" -name '*.sh' -exec realpath --relative-to="${LOCAL}/filters" {} \; | sort | while IFS= read -r filter; do
     tex=${TARGET}/temp/reports/${filter}.tex
     if [ -e "${tex}" ]; then
         echo "The ${filter} filter was already completed earlier, see report in '${tex}'"
@@ -45,7 +45,7 @@ and published its results to ${TARGET}/temp/reports/${filter}.tex "
 done
 
 
-find "${TARGET}/temp/reports" -type f -exec basename {} \; | while read -r f; do
+find "${TARGET}/temp/reports" -type f -exec basename {} \; | while IFS= read -r f; do
     echo "${f}:"
     cat "${TARGET}/temp/reports/${f}"
     echo ""

steps/jpeek-repo.sh

@@ -98,7 +98,7 @@ for type in all cvc; do
         echo "No files generated by jpeek in ${dir}"
         continue
     fi
-    find "${dir}" -type f -maxdepth 1 -print | while read -r report; do
+    find "${dir}" -type f -maxdepth 1 -print | while IFS= read -r report; do
         echo "${report}" >> "${files}"
         if echo "${report}" | grep -q "${accept}"; then
             packages=$(xmlstarlet sel -t -v 'count(/metric/app/package/@id)' "${report}")

steps/measure-file.sh

@@ -32,14 +32,14 @@ start=$(date +%s%N)
 
 mkdir -p "$(dirname "${javam}")"
 metrics=$(find "${LOCAL}/metrics/" -type f -exec basename {} \;)
-echo "${metrics}" | while read -r m; do
+echo "${metrics}" | while IFS= read -r m; do
     # @todo #74 This metric doesn't work, because when we run it,
     #  the directory already doesn't have the ".git" folder --- it was filtered out
     #  by one of the filters in the "filters/" directory. Probably, we should avoid shallow clones
     #  and then get an opportunity to have more Git-based metrics.
     if [ "${m}" = "authors.sh" ]; then continue; fi
     if timeout 30m "metrics/${m}" "${java}" "${javam}"; then
-        while read -r t; do
+        while IFS= read -r t; do
             IFS=' ' read -r -ra M <<< "${t}"
             echo "${M[1]}" > "${javam}.${M[0]}"
         done < "${javam}"

steps/measure.sh

@@ -38,7 +38,7 @@ touch "${jobs}"
 
 declare -i file=0
 sh="$(dirname "$0")/measure-file.sh"
-echo "${javas}" | while read -r java; do
+echo "${javas}" | while IFS= read -r java; do
     file=$((file+1))
     rel=$(realpath --relative-to="${TARGET}/github" "${java}")
     javam=${TARGET}/measurements/${rel}.m

steps/polish.sh

@@ -30,7 +30,7 @@ if [ ! -e "${dir}" ]; then
 fi
 
 repos=$(find "${dir}" -maxdepth 2 -mindepth 2 -type d -exec realpath --relative-to="${dir}" {} \;)
-echo "${repos}" | while read -r repo; do
+echo "${repos}" | while IFS= read -r repo; do
     if [ -z "${repo}" ]; then continue; fi
     if grep -Fxq "${repo}" "${TARGET}/repositories.csv"; then
         echo "Directory of ${repo} is already here"
@@ -42,7 +42,7 @@ done
 echo "All $(echo "${repos}" | wc -l | xargs) repo directories inside ${dir} look good"
 
 orgs=$(find "${dir}" -maxdepth 1 -mindepth 1 -type d -exec realpath --relative-to="${dir}" {} \;)
-echo "${orgs}" | while read -r org; do
+echo "${orgs}" | while IFS= read -r org; do
     if [ "$(find "${dir}/${org}" | wc -l | xargs)" == '0' ]; then
         rm -rf "${dir:?}/${org}"
         echo "Organization ${org} is empty and was deleted"

steps/report.sh

@@ -27,7 +27,7 @@ rm -f "${TARGET}/temp/list-of-metrics.tex"
 
 java=${TARGET}/temp/Foo.java
 mkdir -p "$(dirname "${java}")"
-find "${LOCAL}/metrics" -type f -exec basename {} \; | while read -r m; do
+find "${LOCAL}/metrics" -type f -exec basename {} \; | while IFS= read -r m; do
     echo "class Foo {}" > "${java}"
     metric=${TARGET}/temp/Foo.${m}.m
     rm -f "${metric}"

steps/tests.sh

@@ -26,7 +26,7 @@ set -o pipefail
 temp=${LOCAL}/test-zone
 mkdir -p "${temp}"
 
-find "${LOCAL}/tests" -name '*.sh' | sort | while read -r test; do
+find "${LOCAL}/tests" -name '*.sh' | sort | while IFS= read -r test; do
     name=$(realpath --relative-to="${LOCAL}/tests" "${test}")
     if [ -n "${TEST}" ] && [ ! "${TEST}" = "${name}" ]; then
         continue

tests/steps/test-aggregate.sh

@@ -25,12 +25,12 @@ set -o pipefail
 
 temp=$1
 
-repo="foo/bar test"
+repo="foo/bar test ; "
 dir="${TARGET}/measurements/${repo}/a"
 mkdir -p "${dir}"
 touch "${dir}/Foo.java.m"
 echo "42" > "${dir}/Foo.java.m.loc"
-"${LOCAL}/steps/aggregate.sh"
+"${LOCAL}/steps/aggregate.sh" >/dev/null
 test -e "${TARGET}/data/${repo}/all.csv"
 test -e "${TARGET}/data/${repo}/loc.csv"
 cat "${TARGET}/data/${repo}/loc.csv" | grep ",42" >/dev/null