Replaces explicit listing of fuzzer target and corpus files with generic (#2455)

detection of these files using wildcards.

projects/icu/build.sh

@@ -36,34 +36,23 @@ make -j$(nproc)
 $CXX $CXXFLAGS -std=c++11 -c $SRC/icu/icu4c/source/test/fuzzer/locale_util.cc \
      -I$SRC/icu4c/source/test/fuzzer
 
-FUZZERS="break_iterator_fuzzer \
+FUZZER_PATH=$SRC/icu/icu4c/source/test/fuzzer
-  converter_fuzzer \
+# Assumes that all fuzzers files end with'_fuzzer.cc'.
-  locale_fuzzer \
+FUZZERS=$FUZZER_PATH/*_fuzzer.cc
-  number_format_fuzzer \
+
-  ucasemap_fuzzer \
-  uloc_canonicalize_fuzzer \
-  uloc_for_language_tag_fuzzer \
-  uloc_get_name_fuzzer \
-  uloc_is_right_to_left_fuzzer \
-  uloc_open_keywords_fuzzer \
-  unicode_string_codepage_create_fuzzer \
-  uregex_open_fuzzer
-  "
 for fuzzer in $FUZZERS; do
+  file=${fuzzer:${#FUZZER_PATH}+1}
   $CXX $CXXFLAGS -std=c++11 \
-    $SRC/icu/icu4c/source/test/fuzzer/$fuzzer.cc -o $OUT/$fuzzer locale_util.o \
+    $fuzzer -o $OUT/${file/.cc/} locale_util.o \
     -I$SRC/icu/icu4c/source/common -I$SRC/icu/icu4c/source/i18n -L$WORK/icu/lib \
     $LIB_FUZZING_ENGINE -licui18n -licuuc -licutu -licudata
 done
 
-CORPUS="uloc_canonicalize_fuzzer_seed_corpus \
+# Assumes that all seed files end with '*_fuzzer_seed_corpus.txt'.
-  uloc_for_language_tag_fuzzer_seed_corpus \
+CORPUS=$SRC/icu/icu4c/source/test/fuzzer/*_fuzzer_seed_corpus.txt
-  uloc_get_name_fuzzer_seed_corpus \
-  uloc_is_right_to_left_fuzzer_seed_corpus \
-  uloc_open_keywords_fuzzer_seed_corpus
-  "
 for corpus in $CORPUS; do
-  zip $OUT/$corpus.zip $SRC/icu/icu4c/source/test/fuzzer/$corpus.txt
+    zipfile=${corpus:${#FUZZER_PATH}+1}
+    zip $OUT/${zipfile/.txt/.zip} $corpus
 done
 
 cp $SRC/icu/icu4c/source/test/fuzzer/*.dict  $OUT/