mirror of
https://github.com/bherbruck/solidcraft.git
synced 2026-11-15 13:10:45 +00:00
Brennen Herbruck
916b91e8d0
solidcore's main at c0d3cc0124, rewritten to one commit per same-author run per PR (2,800 commits, author credit kept; tree identical). Full original history: tangency-labs/solidcore archive/full-history. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> git-subtree-dir: solidcore git-subtree-mainline:44525da669git-subtree-split:c440e5f892
65 lines
2.4 KiB
Bash
Executable File
Vendored
65 lines
2.4 KiB
Bash
Executable File
Vendored
#!/usr/bin/env bash
|
|
# Fetch the STEP Tools sample STEP files (AP203 and AP214 sets,
|
|
# https://steptools.com/docs/stpfiles/) into corpus/steptools/ (gitignored).
|
|
#
|
|
# The files are "All Rights Reserved" by STEP Tools, Inc.; the site permits only
|
|
# informational, non-commercial copying. NEVER commit them, and never copy them into
|
|
# fixtures/. This script only downloads them for local corpus runs.
|
|
#
|
|
# Every file is pinned by SHA-256 in scripts/steptools-corpus.sha256; a download whose
|
|
# hash differs fails the script (nothing is kept). Idempotent: files already present
|
|
# with the pinned hash are not downloaded again; a present file with another hash is an
|
|
# error (remove it and re-run).
|
|
#
|
|
# The corpus is data only: goldens --step-corpus parses it as untrusted input and never
|
|
# executes anything from it.
|
|
#
|
|
# Usage: scripts/fetch-steptools-corpus.sh [dest] (default <repo>/corpus/steptools)
|
|
# Used by: cargo run --profile goldens -p goldens -- --step-corpus
|
|
set -euo pipefail
|
|
|
|
BASE="https://steptools.com/docs/stpfiles"
|
|
|
|
HERE="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)"
|
|
ROOT="$(cd "$HERE/.." && pwd)"
|
|
MANIFEST="$HERE/steptools-corpus.sha256"
|
|
DEST="${1:-$ROOT/corpus/steptools}"
|
|
|
|
command -v sha256sum >/dev/null || { echo "fetch-steptools-corpus: sha256sum not found" >&2; exit 1; }
|
|
command -v curl >/dev/null || { echo "fetch-steptools-corpus: curl not found" >&2; exit 1; }
|
|
|
|
hash_of() { sha256sum "$1" | cut -d' ' -f1; }
|
|
|
|
mkdir -p "$DEST"
|
|
fetched=0
|
|
present=0
|
|
while read -r sha path; do
|
|
case "$sha" in ''|'#'*) continue ;; esac
|
|
# Paths are fixed in the manifest; refuse anything that could leave DEST.
|
|
case "$path" in
|
|
*/../*|../*|/*|*/..) echo "fetch-steptools-corpus: bad path '$path'" >&2; exit 1 ;;
|
|
esac
|
|
out="$DEST/$path"
|
|
if [ -e "$out" ]; then
|
|
if [ "$(hash_of "$out")" = "$sha" ]; then
|
|
present=$((present + 1))
|
|
continue
|
|
fi
|
|
echo "fetch-steptools-corpus: $out exists with another hash; remove it and re-run" >&2
|
|
exit 1
|
|
fi
|
|
mkdir -p "$(dirname "$out")"
|
|
tmp="$out.part"
|
|
rm -f "$tmp"
|
|
curl --fail --silent --show-error --location --max-time 300 -o "$tmp" "$BASE/$path"
|
|
got="$(hash_of "$tmp")"
|
|
if [ "$got" != "$sha" ]; then
|
|
rm -f "$tmp"
|
|
echo "fetch-steptools-corpus: $path: sha256 $got, pinned $sha" >&2
|
|
exit 1
|
|
fi
|
|
mv "$tmp" "$out"
|
|
fetched=$((fetched + 1))
|
|
done < "$MANIFEST"
|
|
echo "steptools corpus in $DEST: $fetched fetched, $present already present"
|