Files
libnovel/scripts/pb-init.sh
Admin 22b6ee824e
All checks were successful
CI / Scraper / Test (pull_request) Successful in 14s
CI / UI / Build (pull_request) Successful in 17s
CI / UI / Docker Push (pull_request) Has been skipped
CI / Scraper / Lint (pull_request) Successful in 21s
CI / Scraper / Docker Push (pull_request) Has been skipped
iOS CI / Build (pull_request) Successful in 8m22s
iOS CI / Test (pull_request) Successful in 12m35s
fix(pb-init): use python3 for JSON parsing in ensure_field; add heartbeat_at fields
The sed-based collection id and fields extraction was greedy and broke on
collections with multiple fields (grabbed the last field id instead of the
top-level collection id → PATCH to wrong URL → 404).

Rewrite ensure_field to use python3 for reliable JSON parsing. Also adds the
missing heartbeat_at (date) field to scraping_tasks and audio_jobs which was
never applied on the initial deploy because the bug prevented the PATCH.
2026-03-15 21:53:54 +05:00

346 lines
12 KiB
Bash
Executable File

#!/bin/sh
# pb-init.sh — idempotent PocketBase collection bootstrap
#
# Creates all collections required by libnovel. Safe to re-run: POST returns
# 400/422 when a collection already exists; both are treated as success.
#
# Required env vars (with defaults):
# POCKETBASE_URL http://pocketbase:8090
# POCKETBASE_ADMIN_EMAIL admin@libnovel.local
# POCKETBASE_ADMIN_PASSWORD changeme123
set -e
PB_URL="${POCKETBASE_URL:-http://pocketbase:8090}"
PB_EMAIL="${POCKETBASE_ADMIN_EMAIL:-admin@libnovel.local}"
PB_PASSWORD="${POCKETBASE_ADMIN_PASSWORD:-changeme123}"
log() { echo "[pb-init] $*"; }
# ─── 0. Ensure curl is available ─────────────────────────────────────────────
if ! command -v curl > /dev/null 2>&1; then
apk add --no-cache curl > /dev/null 2>&1
fi
# ─── 1. Wait for PocketBase to be ready ──────────────────────────────────────
log "waiting for PocketBase at $PB_URL ..."
until curl -sf "$PB_URL/api/health" > /dev/null 2>&1; do
sleep 2
done
log "PocketBase is up"
# ─── 2. Ensure the superuser exists, then authenticate ───────────────────────
#
# The muchobien/pocketbase image does NOT auto-create a superuser from env vars.
# On a fresh install PocketBase exposes a one-time install JWT in its log output
# at /pb_data/logs/ — but we can't read that from here.
#
# Strategy:
# a) Try to auth normally (works on subsequent runs once the account exists).
# b) If that returns 400/401, PocketBase is fresh. Use the install token
# obtained from the /_/ redirect Location header (PocketBase v0.23+).
log "ensuring superuser $PB_EMAIL exists ..."
# Try to get the install token from the /_/ redirect Location header.
LOCATION=$(curl -sf -o /dev/null -w "%{redirect_url}" "$PB_URL/_/" 2>/dev/null || true)
if echo "$LOCATION" | grep -q "pbinstal/"; then
INSTALL_TOKEN=$(echo "$LOCATION" | sed 's|.*pbinstal/||' | tr -d ' \r\n')
log "install token found — creating superuser via install endpoint"
curl -sf -X POST "$PB_URL/api/collections/_superusers/records" \
-H "Content-Type: application/json" \
-H "Authorization: Bearer $INSTALL_TOKEN" \
-d "{\"email\":\"$PB_EMAIL\",\"password\":\"$PB_PASSWORD\",\"passwordConfirm\":\"$PB_PASSWORD\"}" \
> /dev/null 2>&1 || true
log "superuser create attempted (may already exist)"
fi
# ─── 3. Authenticate and obtain a superuser token ────────────────────────────
log "authenticating as $PB_EMAIL ..."
AUTH_RESPONSE=$(curl -sf -X POST "$PB_URL/api/collections/_superusers/auth-with-password" \
-H "Content-Type: application/json" \
-d "{\"identity\":\"$PB_EMAIL\",\"password\":\"$PB_PASSWORD\"}")
TOKEN=$(echo "$AUTH_RESPONSE" | sed 's/.*"token":"\([^"]*\)".*/\1/')
if [ -z "$TOKEN" ] || [ "$TOKEN" = "$AUTH_RESPONSE" ]; then
log "ERROR: failed to obtain auth token. Response: $AUTH_RESPONSE"
exit 1
fi
log "auth token obtained"
# ─── 4. Helpers ───────────────────────────────────────────────────────────────
create_collection() {
NAME="$1"
BODY="$2"
STATUS=$(curl -s -o /dev/null -w "%{http_code}" \
-X POST "$PB_URL/api/collections" \
-H "Content-Type: application/json" \
-H "Authorization: Bearer $TOKEN" \
-d "$BODY")
case "$STATUS" in
200|201) log "created collection: $NAME" ;;
400|422) log "collection already exists (skipped): $NAME" ;;
*) log "WARNING: unexpected status $STATUS for collection: $NAME" ;;
esac
}
# ensure_field COLLECTION FIELD_NAME FIELD_TYPE
#
# Checks whether FIELD_NAME exists in COLLECTION's schema. If it is missing,
# sends a PATCH with the full current fields list plus the new field appended.
ensure_field() {
COLL="$1"
FIELD_NAME="$2"
FIELD_TYPE="$3"
SCHEMA=$(curl -sf \
-H "Authorization: Bearer $TOKEN" \
"$PB_URL/api/collections/$COLL" 2>/dev/null)
# Use python3 to reliably parse the JSON schema.
PARSED=$(echo "$SCHEMA" | python3 -c "
import sys, json
try:
d = json.load(sys.stdin)
fields = d.get('fields', [])
exists = any(f.get('name') == '$FIELD_NAME' for f in fields)
print('exists=' + str(exists))
print('id=' + d.get('id', ''))
if not exists:
fields.append({'name': '$FIELD_NAME', 'type': '$FIELD_TYPE'})
print('fields=' + json.dumps(fields))
except Exception as e:
print('error=' + str(e))
" 2>/dev/null)
if echo "$PARSED" | grep -q "^exists=True"; then
log "field $COLL.$FIELD_NAME already exists — skipping"
return
fi
COLLECTION_ID=$(echo "$PARSED" | grep "^id=" | sed 's/^id=//')
if [ -z "$COLLECTION_ID" ]; then
log "WARNING: could not get id for collection $COLL — skipping ensure_field"
return
fi
NEW_FIELDS=$(echo "$PARSED" | grep "^fields=" | sed 's/^fields=//')
PATCH_BODY="{\"fields\":${NEW_FIELDS}}"
STATUS=$(curl -s -o /dev/null -w "%{http_code}" \
-X PATCH "$PB_URL/api/collections/$COLLECTION_ID" \
-H "Content-Type: application/json" \
-H "Authorization: Bearer $TOKEN" \
-d "$PATCH_BODY")
case "$STATUS" in
200|201) log "patched $COLL — added field: $FIELD_NAME ($FIELD_TYPE)" ;;
*) log "WARNING: patch returned $STATUS when adding $FIELD_NAME to $COLL" ;;
esac
}
# ─── 5. Create collections (idempotent — skips if already exist) ─────────────
create_collection "books" '{
"name": "books",
"type": "base",
"fields": [
{"name": "slug", "type": "text", "required": true},
{"name": "title", "type": "text", "required": true},
{"name": "author", "type": "text"},
{"name": "cover", "type": "text"},
{"name": "status", "type": "text"},
{"name": "genres", "type": "json"},
{"name": "summary", "type": "text"},
{"name": "total_chapters", "type": "number"},
{"name": "source_url", "type": "text"},
{"name": "ranking", "type": "number"},
{"name": "meta_updated", "type": "date"}
]
}'
create_collection "chapters_idx" '{
"name": "chapters_idx",
"type": "base",
"fields": [
{"name": "slug", "type": "text", "required": true},
{"name": "number", "type": "number", "required": true},
{"name": "title", "type": "text"},
{"name": "date_label", "type": "text"}
]
}'
create_collection "ranking" '{
"name": "ranking",
"type": "base",
"fields": [
{"name": "rank", "type": "number", "required": true},
{"name": "slug", "type": "text", "required": true},
{"name": "title", "type": "text"},
{"name": "author", "type": "text"},
{"name": "cover", "type": "text"},
{"name": "status", "type": "text"},
{"name": "genres", "type": "json"},
{"name": "source_url", "type": "text"},
{"name": "updated", "type": "date"}
]
}'
create_collection "progress" '{
"name": "progress",
"type": "base",
"fields": [
{"name": "session_id", "type": "text", "required": true},
{"name": "user_id", "type": "text"},
{"name": "slug", "type": "text", "required": true},
{"name": "chapter", "type": "number"},
{"name": "updated", "type": "date"}
]
}'
create_collection "audio_cache" '{
"name": "audio_cache",
"type": "base",
"fields": [
{"name": "cache_key", "type": "text", "required": true},
{"name": "filename", "type": "text"},
{"name": "updated", "type": "date"}
]
}'
create_collection "app_users" '{
"name": "app_users",
"type": "base",
"fields": [
{"name": "username", "type": "text", "required": true},
{"name": "password_hash", "type": "text", "required": true},
{"name": "role", "type": "text"},
{"name": "created", "type": "date"},
{"name": "avatar_url", "type": "text"}
]
}'
create_collection "user_settings" '{
"name": "user_settings",
"type": "base",
"fields": [
{"name": "session_id", "type": "text", "required": true},
{"name": "user_id", "type": "text"},
{"name": "auto_next", "type": "bool"},
{"name": "voice", "type": "text"},
{"name": "speed", "type": "number"},
{"name": "updated", "type": "date"}
]
}'
# ─── 6. Schema migrations (idempotent field additions) ───────────────────────
# Ensures fields added after initial deploy are present in existing instances.
ensure_field "progress" "user_id" "text"
ensure_field "progress" "audio_time" "number"
ensure_field "user_settings" "user_id" "text"
ensure_field "app_users" "avatar_url" "text"
create_collection "book_comments" '{
"name": "book_comments",
"type": "base",
"fields": [
{"name": "slug", "type": "text", "required": true},
{"name": "user_id", "type": "text"},
{"name": "username", "type": "text"},
{"name": "body", "type": "text", "required": true},
{"name": "upvotes", "type": "number"},
{"name": "downvotes", "type": "number"},
{"name": "created", "type": "date"},
{"name": "parent_id", "type": "text"}
]
}'
create_collection "comment_votes" '{
"name": "comment_votes",
"type": "base",
"fields": [
{"name": "comment_id", "type": "text", "required": true},
{"name": "user_id", "type": "text"},
{"name": "session_id", "type": "text", "required": true},
{"name": "vote", "type": "text", "required": true}
]
}'
create_collection "scraping_tasks" '{
"name": "scraping_tasks",
"type": "base",
"fields": [
{"name": "kind", "type": "text"},
{"name": "target_url", "type": "text"},
{"name": "from_chapter", "type": "number"},
{"name": "to_chapter", "type": "number"},
{"name": "worker_id", "type": "text"},
{"name": "status", "type": "text", "required": true},
{"name": "books_found", "type": "number"},
{"name": "chapters_scraped", "type": "number"},
{"name": "chapters_skipped", "type": "number"},
{"name": "errors", "type": "number"},
{"name": "error_message", "type": "text"},
{"name": "started", "type": "date"},
{"name": "finished", "type": "date"}
]
}'
create_collection "audio_jobs" '{
"name": "audio_jobs",
"type": "base",
"fields": [
{"name": "cache_key", "type": "text", "required": true},
{"name": "slug", "type": "text", "required": true},
{"name": "chapter", "type": "number", "required": true},
{"name": "voice", "type": "text"},
{"name": "worker_id", "type": "text"},
{"name": "status", "type": "text", "required": true},
{"name": "error_message", "type": "text"},
{"name": "started", "type": "date"},
{"name": "finished", "type": "date"}
]
}'
create_collection "user_library" '{
"name": "user_library",
"type": "base",
"fields": [
{"name": "session_id", "type": "text", "required": true},
{"name": "user_id", "type": "text"},
{"name": "slug", "type": "text", "required": true},
{"name": "saved_at", "type": "date"}
]
}'
create_collection "user_sessions" '{
"name": "user_sessions",
"type": "base",
"fields": [
{"name": "user_id", "type": "text", "required": true},
{"name": "session_id", "type": "text", "required": true},
{"name": "user_agent", "type": "text"},
{"name": "ip", "type": "text"},
{"name": "created_at", "type": "date"},
{"name": "last_seen", "type": "date"}
]
}'
create_collection "user_subscriptions" '{
"name": "user_subscriptions",
"type": "base",
"fields": [
{"name": "follower_id", "type": "text", "required": true},
{"name": "followee_id", "type": "text", "required": true},
{"name": "created", "type": "date"}
]
}'
# ─── 7. Post-initial-deploy field additions ───────────────────────────────────
# heartbeat_at is used by the backend runner to detect stale tasks.
ensure_field "scraping_tasks" "heartbeat_at" "date"
ensure_field "audio_jobs" "heartbeat_at" "date"
log "all collections ready"