dep/anitomy: port to use UCS-4 rather than wide strings rationale: wide strings are not the same on every platform, and might not even be Unicode. (while they usually are, its possible that they are not) I was *going* to change StringToInt to use a string stream, but outputting to an integer doesn't seem to work at all with UCS-4, even though it ought to, so I just rolled my own that uses the arabic digits only.

#!/bin/bash -eu
# This script is meant to be run by
# https://github.com/google/oss-fuzz/blob/master/projects/utf8proc/Dockerfile

mkdir build
cd build
cmake .. -DUTF8PROC_ENABLE_TESTING=ON -DLIB_FUZZING_ENGINE="$LIB_FUZZING_ENGINE"
make -j$(nproc)

cp $SRC/utf8proc/build/fuzzer $OUT/utf8proc_fuzzer

find $SRC/utf8proc/test -name "*.txt" | \
     xargs zip $OUT/utf8proc_fuzzer_seed_corpus.zip