Ensures source tarballs are reproducible

When building tarballs dynamically, let's take the time to ensure that
they're fully reproducible. We still run 'python setup.py sdist', but
since that tool doesn't (yet) support SOURCE_DATE_EPOCH, we'll manually
repack the archive with native tar & gzip, forcing predictable
timestamps from the git info, resulting in a deterministic build.

Author: Conor Schaefer

scripts/build-debianpackage

@@ -37,6 +37,7 @@ if [[ -z "${PKG_NAME:-}" ]]; then
 fi
 
 
+# Look up most recent release from GitHub repo
 function find_latest_version() {
     repo_url="https://github.com/freedomofpress/${PKG_NAME}/releases"
     curl -s "$repo_url" \
@@ -58,15 +59,45 @@ fi
 # Copy over the debian directory (including new changelog) from repo
 cp -r "$CUR_DIR/$PKG_NAME/" "$TOP_BUILDDIR/"
 
+# Ensures that a given git tag is signed with the prod release key
+# If "rc" is in the tag name, this will fail.
+function verify_git_tag() {
+    local d
+    local t
+    d="$1"
+    t="$2"
+    prod_fingerprint="22245C81E3BAEB4138B36061310F561200F4AD77"
+    git -C "$build_dir" tag --verify "$PKG_VERSION" 2>&1 \
+        | grep -q -F "using RSA key $prod_fingerprint"
+}
+
+# Dynamically generate a tarball, from the Python source code,
+# that is byte-for-byte reproducible. Use timestamps from the git tag.
 function build_source_tarball() {
     repo_url="https://github.com/freedomofpress/${PKG_NAME}"
     build_dir="/tmp/${PKG_NAME}"
     rm -rf "$build_dir"
     git clone "$repo_url" "$build_dir"
-    git -C "$build_dir" tag --verify "$PKG_VERSION" 1>&2
-    git -C "$build_dir" checkout "$PKG_VERSION" 1>&2
+
+    # Verify tag, using only the prod key
+    verify_git_tag "$build_dir" "$PKG_VERSION" || exit 1
+
+    # Tag is verified, proceed with checkout
+    git -C "$build_dir" checkout "$PKG_VERSION" 1>&2 || exit 1
     (cd "$build_dir" && python setup.py sdist 1>&2)
-    find "${build_dir}/dist/" | grep -P '\.tar.gz$' | head -n1
+
+    # Initial tarball will contain timestamps from NOW, let's repack
+    # with timestamps from the signed tag.
+    raw_tarball="$(find "${build_dir}/dist/" | grep -P '\.tar.gz$' | head -n1)"
+    tag_time="$(git -C "$build_dir" log --format="%ai" --no-patch -n1 "$PKG_VERSION")"
+    (cd "$build_dir" && tar -xzf "dist/$(basename $raw_tarball)" 1>&2)
+    tarball_basename="$(basename "$raw_tarball")"
+    # Repack with tar only, so env vars are respected
+    (cd "$build_dir" && tar -cf "${tarball_basename%.gz}" --mode=go=rX,u+rw,a-s --mtime="$tag_time" --sort=name --owner=root:0 --group=root:0 "${tarball_basename%.tar.gz}" 1>&2)
+    # Then gzip it separately, so we can pass args
+    (cd "$build_dir" && gzip --no-name "${tarball_basename%.gz}" 1>&2)
+    (cd "$build_dir" && mv "$tarball_basename" dist/ 1>&2)
+    echo "$raw_tarball"
 }
 
 # If the package is contained in the list, it should be a python package. In