Files
omarchy-pkgs/bin/sync-repo
T
David Heinemeier HanssonandClaude Opus 5 dbe1db4b03 Read the remote database without depending on bsdtar
The partial-tree guard parsed omarchy.db with bsdtar, which is not installed on
the repository host. Every sync there aborted with "the remote database exists
but could not be read" -- a guard meant to catch a partial tree instead blocked
a complete one, stopping a publish after sign, promote and update had already
succeeded.

GNU tar reads the database fine when it is a seekable file; the pipe was what
defeated it originally, and that is already downloaded to a temp file. tar now
leads, with bsdtar as a fallback for a tar too old to detect zstd.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
2026-08-12 04:59:22 -07:00

244 lines
8.5 KiB
Bash
Executable File

#!/bin/bash
# Source common functions
BUILD_ROOT=$(realpath "${BASH_SOURCE[0]%/*}/..")
source "$BUILD_ROOT/helpers/message-helpers.sh"
source "$BUILD_ROOT/helpers/paths.sh"
# Default remote (production)
DEFAULT_REMOTE="pkgs.omarchy.org:omarchy-pkgs"
REMOTE="$DEFAULT_REMOTE"
SKIP_PROD_CHECK=false
PRUNE=false
# Print header
print_header "Sync Repository to Remote"
# This script has no `set -e`, so a `shift 2` past the end of the argument list
# fails without consuming anything and the loop spins forever. Check first.
require_value() {
if [[ $# -lt 2 || -z "$2" ]]; then
print_error "Option $1 requires a value"
exit 1
fi
}
# Parse arguments
while [[ $# -gt 0 ]]; do
case $1 in
--arch)
require_value "$@"
ARCH="$2"
update_arch_paths
shift 2
;;
--mirror)
require_value "$@"
MIRROR="$2"
update_arch_paths
shift 2
;;
--remote)
require_value "$@"
REMOTE="$2"
shift 2
;;
--skip-prod-check)
SKIP_PROD_CHECK=true
shift
;;
--prune)
PRUNE=true
shift
;;
-h | --help)
echo "Usage: $0 [OPTIONS]"
echo ""
echo "Options:"
echo " --arch <arch> Target architecture (x86_64 or aarch64, default: x86_64)"
echo " --mirror <mirror> Mirror to use (edge or stable, default: edge)"
echo " --remote <remote> Rclone remote destination (default: $DEFAULT_REMOTE)"
echo " --skip-prod-check Skip production sync confirmation"
echo " --prune Also delete remote packages missing locally"
echo " -h, --help Show this help message"
echo ""
echo "Uploads are additive by default. Removing packages from the remote"
echo "requires --prune, which only makes sense from a complete local tree."
exit 0
;;
*)
print_error "Unknown option: $1"
exit 1
;;
esac
done
print_info "Mirror: $MIRROR"
print_info "Architecture: $ARCH"
print_info "Local Repository: $REPO_DIR"
print_info "Remote: $REMOTE"
echo ""
# Check if local repo exists
if [[ ! -d "$REPO_DIR" ]]; then
print_error "Local repository directory not found: $REPO_DIR"
exit 1
fi
DESTINATION_DIRECTORY="$MIRROR/$ARCH"
# Check if syncing to production and warn user
if [[ "$REMOTE" == "$DEFAULT_REMOTE" ]] && [[ "$SKIP_PROD_CHECK" != true ]]; then
print_warning "You are about to sync to PRODUCTION ($REMOTE/$DESTINATION_DIRECTORY)"
echo ""
read -p "Are you sure you want to sync to production? (y/N) " -n 1 -r
echo
if [[ ! $REPLY =~ ^[Yy]$ ]]; then
print_info "Sync cancelled"
exit 0
fi
fi
print_info "Syncing to: $REMOTE/$DESTINATION_DIRECTORY"
# The database is what users actually resolve against, and repo-add builds it
# from this tree alone. Publishing one built from a partial tree hides every
# package it does not know about, even though the files are still on the remote.
# Refuse to shrink the package list unless that is the stated intent.
#
# Compare package names, not file counts: this tree keeps several versions of
# each package (bin/repo clean --keep 2) while the database carries one entry per
# name, so counting files would compare unrelated quantities and let a partial
# tree through whenever its spare versions made up the difference.
strip_version() { sed -E 's/-[^-]+-[^-]+-[^-]+\.pkg\.tar\.[^.]+$//'; }
LOCAL_NAMES=$(ls -1 "$REPO_DIR" 2>/dev/null | grep -v '\.sig$' | grep '\.pkg\.tar\.' |
strip_version | sort -u)
# Distinguish "no repository there yet" from "cannot read the repository". Only
# the first is safe to treat as an empty remote; failing open on a credential or
# network error is how a partial database reaches production.
REMOTE_LISTING=$(rclone lsf "$REMOTE/$DESTINATION_DIRECTORY/" --s3-no-head 2>&1)
RCLONE_STATUS=$?
# rclone exit 3 is "directory not found", which is what a mirror that has never
# been published looks like. Every other failure means the remote could not be
# read, and an unread remote must not be mistaken for an empty one.
if [[ $RCLONE_STATUS -eq 3 ]]; then
REMOTE_LISTING=""
elif [[ $RCLONE_STATUS -ne 0 ]]; then
print_error "Cannot read the remote repository (rclone exit $RCLONE_STATUS)"
echo "$REMOTE_LISTING"
echo ""
echo "Refusing to sync: an unreadable remote cannot be checked for packages"
echo "this tree would hide."
exit 1
fi
if grep -qx 'omarchy\.db' <<<"$REMOTE_LISTING"; then
# bsdtar, not tar: the database is compressed and GNU tar will not detect that
# on a pipe. repo-add has used both gzip and zstd, so let libarchive decide.
REMOTE_DB_FILE=$(mktemp)
trap 'rm -f "$REMOTE_DB_FILE"' EXIT
rclone cat "$REMOTE/$DESTINATION_DIRECTORY/omarchy.db" --s3-no-head >"$REMOTE_DB_FILE" 2>/dev/null
# Download to a file rather than piping: GNU tar detects the compression from
# a seekable archive but not from a pipe, and the database has been both gzip
# and zstd over its life. bsdtar reads either too, but it is not present on
# every host, so tar leads and bsdtar covers a tar too old for zstd.
REMOTE_ENTRIES=$(tar -tf "$REMOTE_DB_FILE" 2>/dev/null)
if [[ -z "$REMOTE_ENTRIES" ]] && command -v bsdtar >/dev/null 2>&1; then
REMOTE_ENTRIES=$(bsdtar -tf "$REMOTE_DB_FILE" 2>/dev/null)
fi
REMOTE_NAMES=$(printf '%s\n' "$REMOTE_ENTRIES" | sed 's|/.*||' |
sed -E 's/-[^-]+-[^-]+$//' | sort -u | grep -v '^$')
if [[ -z "$REMOTE_NAMES" ]]; then
print_error "The remote database exists but could not be read"
echo ""
echo "Refusing to sync rather than assume the remote is empty. Check that"
echo "omarchy.db is not corrupt: rclone cat $REMOTE/$DESTINATION_DIRECTORY/omarchy.db | tar -t"
exit 1
fi
HIDDEN=$(comm -23 <(echo "$REMOTE_NAMES") <(echo "$LOCAL_NAMES"))
HIDDEN_COUNT=$(grep -c '' <<<"$HIDDEN")
[[ -z "$HIDDEN" ]] && HIDDEN_COUNT=0
if [[ "$HIDDEN_COUNT" -gt 0 && "$PRUNE" != true ]]; then
print_error "$HIDDEN_COUNT package(s) in the remote database are missing from this tree"
echo ""
echo "$HIDDEN" | head -10 | sed 's/^/ /'
[[ "$HIDDEN_COUNT" -gt 10 ]] && echo " ... and $((HIDDEN_COUNT - 10)) more"
echo ""
echo "Publishing a database built here would hide them, even though their"
echo "files remain on the mirror."
echo ""
echo "To publish packages built on this machine, push them to the build host,"
echo "which holds the complete repository:"
echo " bin/repo push --mirror $MIRROR --arch $ARCH"
echo ""
echo "If shrinking the repository is genuinely what you want, pass --prune."
exit 1
fi
fi
# Upload packages first, database last, so the remote never advertises a package
# it does not yet have.
#
# This is `copy`, not `sync`: a local tree is not authoritative about what should
# exist on the remote. Packages built on another machine live only in that
# machine's build-output, and pkgs.omarchy.org/ is gitignored, so any checkout
# that has not run a full release is missing nearly everything. `sync` would read
# those absences as deletions and empty the repository. --ignore-existing keeps
# versions already published from being overwritten.
if [[ "$PRUNE" == true ]]; then
print_warning "Pruning: remote packages missing from $REPO_DIR will be DELETED"
if [[ "$SKIP_PROD_CHECK" != true ]]; then
read -p "Prune the remote to match this tree? (y/N) " -n 1 -r
echo
if [[ ! $REPLY =~ ^[Yy]$ ]]; then
print_info "Sync cancelled"
exit 0
fi
fi
print_info "Syncing packages (with prune)..."
if ! rclone sync "$REPO_DIR" "$REMOTE/$DESTINATION_DIRECTORY" \
--s3-no-head \
--exclude "omarchy.db*" \
--exclude "omarchy.files*" \
--ignore-existing \
--copy-links --delete-after -v; then
print_error "Package sync failed — not publishing the database"
exit 1
fi
else
print_info "Syncing packages..."
if ! rclone copy "$REPO_DIR" "$REMOTE/$DESTINATION_DIRECTORY" \
--s3-no-head \
--exclude "omarchy.db*" \
--exclude "omarchy.files*" \
--ignore-existing \
--copy-links -v; then
print_error "Package upload failed — not publishing the database"
print_warning "The remote database still describes the previous contents, so"
print_warning "the repository is unchanged and consistent."
exit 1
fi
fi
# Then sync database files last to ensure repository integrity
print_info "Updating repository database..."
if ! rclone copy "$REPO_DIR" "$REMOTE/$DESTINATION_DIRECTORY" \
--s3-no-head \
--include "omarchy.*" \
--checksum --copy-links -v; then
print_error "Database upload failed"
print_warning "Packages were uploaded but the database still describes the"
print_warning "previous contents. Re-run sync to finish publishing them."
exit 1
fi
print_success "Sync complete!"