From 8e9616a5cfd2aa9d8443d5fde4a29c5dc00c9c6a Mon Sep 17 00:00:00 2001
From: Conor Schaefer <conor@freedom.press>
Date: Tue, 4 Aug 2020 16:07:43 -0700
Subject: [PATCH] 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.
---
 scripts/build-debianpackage | 43 ++++++++++++++++++++++++++++++++-----
 1 file changed, 38 insertions(+), 5 deletions(-)

diff --git a/scripts/build-debianpackage b/scripts/build-debianpackage
index c5723f64..b9f4bcd7 100755
--- a/scripts/build-debianpackage
+++ b/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 +59,46 @@ 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 $prod_fingerprint"
+}
+
+# Dynamically generate a tarball, from the Python source code,
+# that is byte-for-byte reproducible. Infers timestamp
+# from the changelog, same as for the deb package.
 function build_source_tarball() {
     repo_url="https://github.com/freedomofpress/${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
-    (cd "$build_dir" && python setup.py sdist 1>&2)
-    find "${build_dir}/dist/" | grep -P '\.tar.gz$' | head -n1
+
+    # Verify tag, using only the prod key
+    verify_git_tag "$build_dir" "$PKG_VERSION"
+
+    # Tag is verified, proceed with checkout
+    git -C "$build_dir" checkout "$PKG_VERSION"
+    (cd "$build_dir" && LC_ALL="C.UTF-8" python setup.py sdist)
+
+    # Initial tarball will contain timestamps from NOW, let's repack
+    # with timestamps from the changelog, which is static.
+    raw_tarball="$(find "${build_dir}/dist/" | grep -P '\.tar.gz$' | head -n1)"
+    dch_time="$(date "+%Y-%m-%d %H:%M:%S %z" -d@$(dpkg-parsechangelog --file $PKG_NAME/debian/changelog-$PLATFORM -STimestamp)) "
+    (cd "$build_dir" && tar -xzf "dist/$(basename $raw_tarball)")
+    tarball_basename="$(basename "$raw_tarball")"
+    # Repack with tar only, so env vars are respected
+    (cd "$build_dir" && tar -cf "${tarball_basename%.gz}" --mode=go=rX,u+rw,a-s --mtime="$dch_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 "${tarball_basename%.gz}")
+    (cd "$build_dir" && mv "$tarball_basename" dist/)
+    echo "$raw_tarball"
 }
 
 # If the package is contained in the list, it should be a python package. In
@@ -77,7 +109,8 @@ if [[ "${PKG_NAME}" =~ ^(securedrop-client|securedrop-proxy|securedrop-export|se
     if [[ -z "${PKG_PATH:-}" ]]; then
         # Build from source
         echo "PKG_PATH not set, building from source (version $PKG_VERSION)..."
-        candidate_pkg_path="$(build_source_tarball)"
+        build_source_tarball
+        candidate_pkg_path="$(find /tmp/$PKG_NAME/dist -type f -iname '*.tar.gz')"
         if [[ -f "$candidate_pkg_path" ]]; then
             PKG_PATH="$candidate_pkg_path"
             echo "Found tarball at $PKG_PATH, override with PKG_PATH..."