goblin/vendor/github.com/pelletier/go-toml/v2/ci.sh

281 lines
6.7 KiB
Bash

#!/usr/bin/env bash
stderr() {
echo "$@" 1>&2
}
usage() {
b=$(basename "$0")
echo $b: ERROR: "$@" 1>&2
cat 1>&2 <<EOF
DESCRIPTION
$(basename "$0") is the script to run continuous integration commands for
go-toml on unix.
Requires Go and Git to be available in the PATH. Expects to be ran from the
root of go-toml's Git repository.
USAGE
$b COMMAND [OPTIONS...]
COMMANDS
benchmark [OPTIONS...] [BRANCH]
Run benchmarks.
ARGUMENTS
BRANCH Optional. Defines which Git branch to use when running
benchmarks.
OPTIONS
-d Compare benchmarks of HEAD with BRANCH using benchstats. In
this form the BRANCH argument is required.
-a Compare benchmarks of HEAD against go-toml v1 and
BurntSushi/toml.
-html When used with -a, emits the output as HTML, ready to be
embedded in the README.
coverage [OPTIONS...] [BRANCH]
Generates code coverage.
ARGUMENTS
BRANCH Optional. Defines which Git branch to use when reporting
coverage. Defaults to HEAD.
OPTIONS
-d Compare coverage of HEAD with the one of BRANCH. In this form,
the BRANCH argument is required. Exit code is non-zero when
coverage percentage decreased.
EOF
exit 1
}
cover() {
branch="${1}"
dir="$(mktemp -d)"
stderr "Executing coverage for ${branch} at ${dir}"
if [ "${branch}" = "HEAD" ]; then
cp -r . "${dir}/"
else
git worktree add "$dir" "$branch"
fi
pushd "$dir"
go test -covermode=atomic -coverpkg=./... -coverprofile=coverage.out.tmp ./...
cat coverage.out.tmp | grep -v fuzz | grep -v testsuite | grep -v tomltestgen | grep -v gotoml-test-decoder > coverage.out
go tool cover -func=coverage.out
echo "Coverage profile for ${branch}: ${dir}/coverage.out" >&2
popd
if [ "${branch}" != "HEAD" ]; then
git worktree remove --force "$dir"
fi
}
coverage() {
case "$1" in
-d)
shift
target="${1?Need to provide a target branch argument}"
output_dir="$(mktemp -d)"
target_out="${output_dir}/target.txt"
head_out="${output_dir}/head.txt"
cover "${target}" > "${target_out}"
cover "HEAD" > "${head_out}"
cat "${target_out}"
cat "${head_out}"
echo ""
target_pct="$(tail -n2 ${target_out} | head -n1 | sed -E 's/.*total.*\t([0-9.]+)%.*/\1/')"
head_pct="$(tail -n2 ${head_out} | head -n1 | sed -E 's/.*total.*\t([0-9.]+)%/\1/')"
echo "Results: ${target} ${target_pct}% HEAD ${head_pct}%"
delta_pct=$(echo "$head_pct - $target_pct" | bc -l)
echo "Delta: ${delta_pct}"
if [[ $delta_pct = \-* ]]; then
echo "Regression!";
target_diff="${output_dir}/target.diff.txt"
head_diff="${output_dir}/head.diff.txt"
cat "${target_out}" | grep -E '^github.com/pelletier/go-toml' | tr -s "\t " | cut -f 2,3 | sort > "${target_diff}"
cat "${head_out}" | grep -E '^github.com/pelletier/go-toml' | tr -s "\t " | cut -f 2,3 | sort > "${head_diff}"
diff --side-by-side --suppress-common-lines "${target_diff}" "${head_diff}"
return 1
fi
return 0
;;
esac
cover "${1-HEAD}"
}
bench() {
branch="${1}"
out="${2}"
replace="${3}"
dir="$(mktemp -d)"
stderr "Executing benchmark for ${branch} at ${dir}"
if [ "${branch}" = "HEAD" ]; then
cp -r . "${dir}/"
else
git worktree add "$dir" "$branch"
fi
pushd "$dir"
if [ "${replace}" != "" ]; then
find ./benchmark/ -iname '*.go' -exec sed -i -E "s|github.com/pelletier/go-toml/v2|${replace}|g" {} \;
go get "${replace}"
fi
export GOMAXPROCS=2
nice -n -19 taskset --cpu-list 0,1 go test '-bench=^Benchmark(Un)?[mM]arshal' -count=5 -run=Nothing ./... | tee "${out}"
popd
if [ "${branch}" != "HEAD" ]; then
git worktree remove --force "$dir"
fi
}
fmktemp() {
if mktemp --version|grep GNU >/dev/null; then
mktemp --suffix=-$1;
else
mktemp -t $1;
fi
}
benchstathtml() {
python3 - $1 <<'EOF'
import sys
lines = []
stop = False
with open(sys.argv[1]) as f:
for line in f.readlines():
line = line.strip()
if line == "":
stop = True
if not stop:
lines.append(line.split(','))
results = []
for line in reversed(lines[1:]):
v2 = float(line[1])
results.append([
line[0].replace("-32", ""),
"%.1fx" % (float(line[3])/v2), # v1
"%.1fx" % (float(line[5])/v2), # bs
])
# move geomean to the end
results.append(results[0])
del results[0]
def printtable(data):
print("""
<table>
<thead>
<tr><th>Benchmark</th><th>go-toml v1</th><th>BurntSushi/toml</th></tr>
</thead>
<tbody>""")
for r in data:
print(" <tr><td>{}</td><td>{}</td><td>{}</td></tr>".format(*r))
print(""" </tbody>
</table>""")
def match(x):
return "ReferenceFile" in x[0] or "HugoFrontMatter" in x[0]
above = [x for x in results if match(x)]
below = [x for x in results if not match(x)]
printtable(above)
print("<details><summary>See more</summary>")
print("""<p>The table above has the results of the most common use-cases. The table below
contains the results of all benchmarks, including unrealistic ones. It is
provided for completeness.</p>""")
printtable(below)
print('<p>This table can be generated with <code>./ci.sh benchmark -a -html</code>.</p>')
print("</details>")
EOF
}
benchmark() {
case "$1" in
-d)
shift
target="${1?Need to provide a target branch argument}"
old=`fmktemp ${target}`
bench "${target}" "${old}"
new=`fmktemp HEAD`
bench HEAD "${new}"
benchstat "${old}" "${new}"
return 0
;;
-a)
shift
v2stats=`fmktemp go-toml-v2`
bench HEAD "${v2stats}" "github.com/pelletier/go-toml/v2"
v1stats=`fmktemp go-toml-v1`
bench HEAD "${v1stats}" "github.com/pelletier/go-toml"
bsstats=`fmktemp bs-toml`
bench HEAD "${bsstats}" "github.com/BurntSushi/toml"
cp "${v2stats}" go-toml-v2.txt
cp "${v1stats}" go-toml-v1.txt
cp "${bsstats}" bs-toml.txt
if [ "$1" = "-html" ]; then
tmpcsv=`fmktemp csv`
benchstat -csv -geomean go-toml-v2.txt go-toml-v1.txt bs-toml.txt > $tmpcsv
benchstathtml $tmpcsv
else
benchstat -geomean go-toml-v2.txt go-toml-v1.txt bs-toml.txt
fi
rm -f go-toml-v2.txt go-toml-v1.txt bs-toml.txt
return $?
esac
bench "${1-HEAD}" `mktemp`
}
case "$1" in
coverage) shift; coverage $@;;
benchmark) shift; benchmark $@;;
*) usage "bad argument $1";;
esac