josh-sync/lib/sync.sh

#!/usr/bin/env bash
# lib/sync.sh — Sync algorithms: forward, reverse, import, reset
#
# All four operations:
#   1. Create a temp work dir with cleanup trap
#   2. Perform git operations
#   3. Return a status string via stdout
#
# Requires: lib/core.sh, lib/config.sh, lib/auth.sh, lib/state.sh sourced
# Expects: SYNC_BRANCH_MONO, SYNC_BRANCH_SUBREPO, JOSH_FILTER, etc.

# ─── Forward Sync: Monorepo → Subrepo ──────────────────────────────
#
# Returns: fresh | skip | clean | lease-rejected | conflict | unrelated

forward_sync() {
  local mono_branch="$SYNC_BRANCH_MONO"
  local subrepo_branch="$SYNC_BRANCH_SUBREPO"
  local work_dir
  work_dir=$(mktemp -d)
  # shellcheck disable=SC2064  # Intentional early expansion — work_dir is local
  trap "rm -rf '$work_dir'" EXIT

  log "INFO" "=== Forward sync: mono/${mono_branch} → subrepo/${subrepo_branch} ==="

  # 1. Clone the monorepo subfolder through josh (filtered view)
  log "INFO" "Cloning filtered monorepo via josh-proxy..."
  git clone "$(josh_auth_url)" \
    --branch "$mono_branch" --single-branch \
    "${work_dir}/filtered" || die "Failed to clone through josh-proxy"

  cd "${work_dir}/filtered" || exit
  git config user.name "$BOT_NAME"
  git config user.email "$BOT_EMAIL"

  local mono_head
  mono_head=$(git rev-parse HEAD)
  log "INFO" "Mono filtered HEAD: ${mono_head:0:12}"

  # 2. Record subrepo HEAD before any operations (the "lease")
  local subrepo_sha
  subrepo_sha=$(subrepo_ls_remote "$subrepo_branch")
  log "INFO" "Subrepo HEAD (lease): ${subrepo_sha:-(empty)}"

  # 3. Handle fresh push (subrepo branch doesn't exist)
  if [ -z "$subrepo_sha" ]; then
    log "INFO" "Subrepo branch doesn't exist — doing fresh push"
    git push "$(subrepo_auth_url)" "HEAD:refs/heads/${subrepo_branch}" \
      || die "Failed to push to subrepo"
    echo "fresh"
    return
  fi

  # 4. Fetch subrepo for comparison
  git remote add subrepo "$(subrepo_auth_url)"
  git fetch subrepo "$subrepo_branch"

  # 5. Compare trees — skip if identical
  local mono_tree subrepo_tree
  mono_tree=$(git rev-parse 'HEAD^{tree}')
  # shellcheck disable=SC1083  # {tree} is git syntax, not shell brace expansion
  subrepo_tree=$(git rev-parse "subrepo/${subrepo_branch}^{tree}" 2>/dev/null || echo "none")

  if [ "$mono_tree" = "$subrepo_tree" ]; then
    log "INFO" "Trees identical — nothing to sync"
    echo "skip"
    return
  fi

  # 6. Attempt merge: start from subrepo state, merge mono changes
  git checkout -B sync-attempt "subrepo/${subrepo_branch}" >&2

  if git merge --no-commit --no-ff "$mono_head" >&2 2>&1; then
    # Clean merge
    if git diff --cached --quiet; then
      log "INFO" "Merge empty — skip"
      git merge --abort 2>/dev/null || true
      echo "skip"
      return
    fi

    git commit -m "Sync from monorepo $(date -u +%Y-%m-%dT%H:%M:%SZ)

${BOT_TRAILER}: forward/${mono_branch}/$(date -u +%Y-%m-%dT%H:%M:%SZ)" >&2

    # 7. Push with force-with-lease (explicit SHA)
    if git push \
      --force-with-lease="refs/heads/${subrepo_branch}:${subrepo_sha}" \
      "$(subrepo_auth_url)" \
      "HEAD:refs/heads/${subrepo_branch}"; then

      log "INFO" "Forward sync complete"
      echo "clean"
    else
      log "WARN" "Force-with-lease rejected — subrepo changed during sync"
      echo "lease-rejected"
    fi

  else
    # Check: unrelated histories (filter change) vs normal merge conflict
    if ! git merge-base "subrepo/${subrepo_branch}" "$mono_head" >/dev/null 2>&1; then
      log "INFO" "No common ancestor — histories are unrelated (filter change?)"
      echo "unrelated"
      return
    fi

    # Normal merge conflict
    local conflicted
    conflicted=$(git diff --name-only --diff-filter=U 2>/dev/null || echo "(unknown)")
    git merge --abort

    log "WARN" "Merge conflict on: ${conflicted}"

    # Push mono state as a conflict branch for PR
    local ts
    ts=$(date +%Y%m%d-%H%M%S)
    local conflict_branch="auto-sync/mono-${mono_branch}-${ts}"
    git checkout -B "$conflict_branch" "$mono_head" >&2
    git push "$(subrepo_auth_url)" "${conflict_branch}"

    # Create PR on subrepo
    local pr_body conflicted_list
    # shellcheck disable=SC2001
    conflicted_list=$(echo "$conflicted" | sed 's/^/- /')
    pr_body="## Sync Conflict\n\nMonorepo \`${mono_branch}\` has changes that conflict with \`${subrepo_branch}\`.\n\n**Conflicted files:**\n${conflicted_list}\n\nPlease resolve and merge this PR to complete the sync."

    create_pr "${SUBREPO_API}" "${SUBREPO_TOKEN}" \
      "$subrepo_branch" "$conflict_branch" \
      "[Sync Conflict] mono/${mono_branch} → ${subrepo_branch}" \
      "$pr_body" \
      || die "Failed to create conflict PR on subrepo (check SUBREPO_TOKEN)"

    log "INFO" "Conflict PR created on subrepo"
    echo "conflict"
  fi
}

# ─── Filter Change Reconciliation ─────────────────────────────────
# When the josh filter changes (e.g., exclude patterns added/removed),
# josh-proxy recomputes filtered history with new SHAs. This creates a
# merge commit on the subrepo that connects old and new histories,
# re-establishing shared ancestry without a destructive force-push.
# Returns: reconciled | lease-rejected

reconcile_filter_change() {
  local mono_branch="$SYNC_BRANCH_MONO"
  local subrepo_branch="$SYNC_BRANCH_SUBREPO"
  local work_dir
  work_dir=$(mktemp -d)
  # shellcheck disable=SC2064  # Intentional early expansion — work_dir is local
  trap "rm -rf '$work_dir'" EXIT

  log "INFO" "=== Filter change reconciliation: ${mono_branch} ==="

  # 1. Clone subrepo
  git clone "$(subrepo_auth_url)" \
    --branch "$subrepo_branch" --single-branch \
    "${work_dir}/subrepo" || die "Failed to clone subrepo"

  cd "${work_dir}/subrepo" || exit
  git config user.name "$BOT_NAME"
  git config user.email "$BOT_EMAIL"

  local subrepo_head
  subrepo_head=$(git rev-parse HEAD)
  log "INFO" "Subrepo HEAD: ${subrepo_head:0:12}"

  # 2. Fetch josh-proxy filtered view (new filter)
  git remote add josh-filtered "$(josh_auth_url)"
  git fetch josh-filtered "$mono_branch" || die "Failed to fetch from josh-proxy"

  local josh_head josh_tree
  josh_head=$(git rev-parse "josh-filtered/${mono_branch}")
  # shellcheck disable=SC1083  # {tree} is git syntax, not shell brace expansion
  josh_tree=$(git rev-parse "josh-filtered/${mono_branch}^{tree}")
  log "INFO" "Josh-proxy HEAD (new filter): ${josh_head:0:12}"

  # 3. Check if trees are already identical (filter change had no effect)
  local subrepo_tree
  # shellcheck disable=SC1083
  subrepo_tree=$(git rev-parse "HEAD^{tree}")
  if [ "$josh_tree" = "$subrepo_tree" ]; then
    log "INFO" "Trees identical after filter change — no reconciliation needed"
    echo "skip"
    return
  fi

  # 4. Create merge commit: josh-proxy HEAD (first parent) + subrepo HEAD, with josh-proxy's tree
  # Josh follows first-parent traversal — josh-filtered MUST be first so josh can map
  # the history back to the monorepo. Old subrepo history hangs off parent 2.
  local merge_commit
  merge_commit=$(git commit-tree "$josh_tree" \
    -p "$josh_head" \
    -p "$subrepo_head" \
    -m "Sync: filter configuration updated

${BOT_TRAILER}: filter-change/${mono_branch}/$(date -u +%Y-%m-%dT%H:%M:%SZ)")

  git reset --hard "$merge_commit" >&2
  log "INFO" "Created reconciliation merge: ${merge_commit:0:12}"

  # 5. Record lease and push
  local subrepo_sha
  subrepo_sha=$(subrepo_ls_remote "$subrepo_branch")

  if git push \
    --force-with-lease="refs/heads/${subrepo_branch}:${subrepo_sha}" \
    "$(subrepo_auth_url)" \
    "HEAD:refs/heads/${subrepo_branch}"; then

    log "INFO" "Filter change reconciled — shared ancestry re-established"
    echo "reconciled"
  else
    log "WARN" "Force-with-lease rejected — subrepo changed during reconciliation"
    echo "lease-rejected"
  fi
}

# ─── Reverse Sync: Subrepo → Monorepo ──────────────────────────────
#
# Always creates a PR on the monorepo — never pushes directly.
# Returns: skip | pr-created | josh-rejected

reverse_sync() {
  local mono_branch="$SYNC_BRANCH_MONO"
  local subrepo_branch="$SYNC_BRANCH_SUBREPO"
  local work_dir
  work_dir=$(mktemp -d)
  # shellcheck disable=SC2064  # Intentional early expansion — work_dir is local
  trap "rm -rf '$work_dir'" EXIT

  log "INFO" "=== Reverse sync: subrepo/${subrepo_branch} → mono/${mono_branch} ==="

  # 1. Clone subrepo
  git clone "$(subrepo_auth_url)" \
    --branch "$subrepo_branch" --single-branch \
    "${work_dir}/subrepo" || die "Failed to clone subrepo"

  cd "${work_dir}/subrepo" || exit
  git config user.name "$BOT_NAME"
  git config user.email "$BOT_EMAIL"

  # 2. Fetch monorepo's filtered view for comparison
  git remote add mono-filtered "$(josh_auth_url)"
  git fetch mono-filtered "$mono_branch" || die "Failed to fetch from josh-proxy"

  # 3. Compare trees — skip if subrepo matches josh-filtered view
  local subrepo_tree josh_tree
  # shellcheck disable=SC1083  # {tree} is git syntax, not shell brace expansion
  subrepo_tree=$(git rev-parse "HEAD^{tree}")
  # shellcheck disable=SC1083
  josh_tree=$(git rev-parse "mono-filtered/${mono_branch}^{tree}")

  if [ "$subrepo_tree" = "$josh_tree" ]; then
    log "INFO" "Subrepo tree matches josh-filtered view — nothing to sync"
    echo "skip"
    return
  fi

  # 4. Find new human commits (excludes bot commits from forward sync)
  local human_commits
  human_commits=$(git log --ancestry-path "mono-filtered/${mono_branch}..HEAD" \
    --oneline --invert-grep --grep="^${BOT_TRAILER}:" 2>/dev/null || echo "")

  if [ -z "$human_commits" ]; then
    log "INFO" "No new human commits in subrepo — nothing to sync"
    echo "skip"
    return
  fi

  log "INFO" "New human commits to sync:"
  echo "$human_commits" >&2

  # 4. Push through josh to a staging branch
  local ts
  ts=$(date +%Y%m%d-%H%M%S)
  local staging_branch="auto-sync/subrepo-${subrepo_branch}-${ts}"

  if git push -o "base=${mono_branch}" "$(josh_auth_url)" "HEAD:refs/heads/${staging_branch}"; then
    log "INFO" "Pushed to staging branch via josh: ${staging_branch}"

    # 5. Create PR on monorepo (NEVER direct push)
    local pr_body
    pr_body="## Subrepo changes\n\nNew commits from subrepo \`${subrepo_branch}\`:\n\n\`\`\`\n${human_commits}\n\`\`\`\n\n**Review checklist:**\n- [ ] Changes scoped to synced subfolder\n- [ ] No leaked credentials or environment-specific config\n- [ ] CI passes"

    create_pr "${MONOREPO_API}" "${GITEA_TOKEN}" \
      "$mono_branch" "$staging_branch" \
      "[Subrepo Sync] ${subrepo_branch} → ${mono_branch}" \
      "$pr_body" \
      || die "Failed to create PR on monorepo (check GITEA_TOKEN)"

    log "INFO" "Reverse sync PR created on monorepo"
    echo "pr-created"
  else
    log "ERROR" "Josh rejected push — check josh-proxy logs"
    echo "josh-rejected"
  fi
}

# ─── Initial Import: Subrepo → Monorepo (first time) ───────────────
#
# Used when a subrepo already has content and you're adding it to the
# monorepo for the first time. Creates a PR.
# Usage: initial_import [clone_url_override]
#   clone_url_override — if set, clone from this URL instead of subrepo_auth_url()
#                        (used by onboard to clone from the archived repo)
# Returns: skip | pr-created

initial_import() {
  local clone_url="${1:-$(subrepo_auth_url)}"
  local mono_branch="$SYNC_BRANCH_MONO"
  local subrepo_branch="$SYNC_BRANCH_SUBREPO"
  local subfolder
  subfolder=$(echo "$JOSH_SYNC_TARGETS" | jq -r --arg n "$JOSH_SYNC_TARGET_NAME" \
    '.[] | select(.name == $n) | .subfolder')
  local work_dir
  work_dir=$(mktemp -d)
  # shellcheck disable=SC2064  # Intentional early expansion — work_dir is local
  trap "rm -rf '$work_dir'" EXIT

  log "INFO" "=== Initial import: subrepo/${subrepo_branch} → mono/${mono_branch}:${subfolder}/ ==="

  # 1. Clone monorepo (directly, not through josh — we need the real repo)
  local mono_auth_url
  local api_host_path
  api_host_path=$(echo "$MONOREPO_API" | sed 's|https://||; s|/api/v1/repos/|/|')
  mono_auth_url="https://${BOT_USER}:${GITEA_TOKEN}@${api_host_path}.git"

  git clone "$mono_auth_url" \
    --branch "$mono_branch" --single-branch \
    "${work_dir}/monorepo" || die "Failed to clone monorepo"

  # 2. Clone subrepo (or archived repo when clone_url is overridden)
  git clone "$clone_url" \
    --branch "$subrepo_branch" --single-branch \
    "${work_dir}/subrepo" || die "Failed to clone subrepo"

  local file_count
  file_count=$(find "${work_dir}/subrepo" -not -path '*/.git/*' -not -path '*/.git' -type f | wc -l | tr -d ' ')
  log "INFO" "Subrepo has ${file_count} files"

  # 3. Copy subrepo content into monorepo subfolder
  cd "${work_dir}/monorepo" || exit
  git config user.name "$BOT_NAME"
  git config user.email "$BOT_EMAIL"

  local ts
  ts=$(date +%Y%m%d-%H%M%S)
  local staging_branch="auto-sync/import-${JOSH_SYNC_TARGET_NAME}-${ts}"
  git checkout -B "$staging_branch" >&2

  mkdir -p "$subfolder"
  rsync -a --exclude='.git' "${work_dir}/subrepo/" "${subfolder}/"
  git add "$subfolder"

  if git diff --cached --quiet; then
    log "INFO" "No changes — subfolder already matches subrepo"
    echo "skip"
    return
  fi

  git commit -m "Import ${JOSH_SYNC_TARGET_NAME} from subrepo/${subrepo_branch}

${BOT_TRAILER}: import/${JOSH_SYNC_TARGET_NAME}/${ts}" >&2

  # 4. Push branch
  git push origin "$staging_branch" || die "Failed to push import branch"
  log "INFO" "Pushed import branch: ${staging_branch}"

  # 5. Create PR on monorepo
  local pr_body
  pr_body="## Initial import\n\nImporting existing subrepo \`${subrepo_branch}\` (${file_count} files) into \`${subfolder}/\`.\n\n**Review checklist:**\n- [ ] Content looks correct\n- [ ] No leaked credentials or environment-specific config\n- [ ] CI passes"

  create_pr "${MONOREPO_API}" "${GITEA_TOKEN}" \
    "$mono_branch" "$staging_branch" \
    "[Import] ${JOSH_SYNC_TARGET_NAME}: ${subrepo_branch}" \
    "$pr_body" \
    || die "Failed to create PR on monorepo (check GITEA_TOKEN)"

  log "INFO" "Import PR created on monorepo"
  echo "pr-created"
}

# ─── Subrepo Reset: Force-push josh-filtered view to subrepo ───────
# Run this AFTER merging an import PR to establish shared josh history.
# Returns: reset

subrepo_reset() {
  local mono_branch="$SYNC_BRANCH_MONO"
  local subrepo_branch="$SYNC_BRANCH_SUBREPO"
  local work_dir
  work_dir=$(mktemp -d)
  # shellcheck disable=SC2064  # Intentional early expansion — work_dir is local
  trap "rm -rf '$work_dir'" EXIT

  log "INFO" "=== Subrepo reset: josh-filtered mono/${mono_branch} → subrepo/${subrepo_branch} ==="

  # 1. Clone monorepo through josh (filtered view — this IS the shared history)
  log "INFO" "Cloning filtered monorepo via josh-proxy..."
  git clone "$(josh_auth_url)" \
    --branch "$mono_branch" --single-branch \
    "${work_dir}/filtered" || die "Failed to clone through josh-proxy"

  cd "${work_dir}/filtered" || exit

  local head_sha
  head_sha=$(git rev-parse HEAD)
  log "INFO" "Josh-filtered HEAD: ${head_sha:0:12}"

  # 2. Force-push to subrepo (replaces subrepo history with josh-filtered history)
  log "WARN" "Force-pushing to subrepo/${subrepo_branch} — this replaces subrepo history"
  git push --force "$(subrepo_auth_url)" "HEAD:refs/heads/${subrepo_branch}" \
    || die "Failed to force-push to subrepo"

  log "INFO" "Subrepo reset complete — histories are now linked through josh"
  echo "reset"
}