Files
TimeTracker/docker/entrypoint_fixed.sh
Dries Peeters 86f7946120 fix: resolve SQLAlchemy reserved word conflict in ClientNotification model
- Rename 'metadata' column to 'extra_data' in ClientNotification model
  to avoid SQLAlchemy Declarative API reserved word conflict
- Update ClientNotificationService to use 'extra_data' instead of 'metadata'
- Maintain API compatibility by returning 'metadata' key in to_dict() method
- Update migration to create 'extra_data' column instead of 'metadata'
- Improve migration idempotency and SQLite compatibility with proper checks
- Enhance backup directory handling with configurable BACKUP_FOLDER support
- Update admin routes to use centralized backup directory function

This fixes the application startup error:
  sqlalchemy.exc.InvalidRequestError: Attribute name 'metadata' is reserved
  when using the Declarative API.

The migration is now idempotent and handles both PostgreSQL and SQLite
databases safely.
2026-01-02 07:50:04 +01:00

1354 lines
46 KiB
Bash

#!/bin/bash
# TimeTracker Docker Entrypoint with Automatic Migration Detection
# This script automatically detects database state and chooses the correct migration strategy
# Don't exit on errors - let the script continue and show what's happening
# set -e
echo "=== TimeTracker Docker Container Starting ==="
echo "Timestamp: $(date)"
echo "Container ID: $(hostname)"
echo "Python version: $(python --version 2>/dev/null || echo 'Python not available')"
echo "Flask version: $(flask --version 2>/dev/null || echo 'Flask CLI not available')"
echo "Current directory: $(pwd)"
echo "User: $(whoami)"
echo
# Function to log messages with timestamp
log() {
echo "[$(date '+%Y-%m-%d %H:%M:%S')] $1"
}
# Function to check if a command exists
command_exists() {
command -v "$1" >/dev/null 2>&1
}
# Function to wait for database
wait_for_database() {
local db_url="$1"
local max_retries=60 # Increased retries
local retry_delay=3 # Increased delay
log "Waiting for database to be available..."
log "Database URL: $db_url"
for attempt in $(seq 1 $max_retries); do
log "Attempt $attempt/$max_retries to connect to database..."
if [[ "$db_url" == postgresql* ]]; then
log "Testing PostgreSQL connection..."
# Test 1: Try psql if available
if command_exists psql; then
log "Testing with psql..."
if psql "$db_url" -c "SELECT 1" >/dev/null 2>&1; then
log "✓ PostgreSQL database is available (via psql)"
return 0
else
log "psql connection failed"
fi
else
log "psql not available, skipping psql test"
fi
# Test 2: Always try Python connection
log "Testing with Python psycopg2..."
if python -c "
import psycopg2
import sys
try:
# Parse connection string to remove +psycopg2 if present
conn_str = '$db_url'.replace('+psycopg2://', '://')
print(f'Attempting connection to: {conn_str}')
conn = psycopg2.connect(conn_str)
conn.close()
print('Connection successful')
sys.exit(0)
except Exception as e:
print(f'Connection failed: {e}')
sys.exit(1)
" 2>/dev/null; then
log "✓ PostgreSQL database is available (via Python)"
return 0
else
log "Python connection failed"
fi
elif [[ "$db_url" == sqlite://* ]]; then
log "Testing SQLite connection..."
local db_file="${db_url#sqlite://}"
if [[ -f "$db_file" ]] || [[ -w "$(dirname "$db_file")" ]]; then
log "✓ SQLite database is available"
return 0
else
log "SQLite file not accessible"
fi
else
log "Unknown database URL format: $db_url"
fi
log "Database not ready (attempt $attempt/$max_retries)"
if [[ $attempt -lt $max_retries ]]; then
log "Waiting $retry_delay seconds before next attempt..."
sleep $retry_delay
fi
done
log "✗ Database is not available after maximum retries"
return 1
}
# Function to detect database state
detect_database_state() {
local db_url="$1"
if [[ "$db_url" == postgresql* ]]; then
# Simple direct Python execution without temp files
python -c "
import psycopg2
try:
conn_str = '$db_url'.replace('+psycopg2://', '://')
conn = psycopg2.connect(conn_str)
cursor = conn.cursor()
cursor.execute(\"\"\"
SELECT EXISTS (
SELECT FROM information_schema.tables
WHERE table_name = 'alembic_version'
)
\"\"\")
has_alembic = cursor.fetchone()[0]
# Check for actual application tables (not just alembic_version)
cursor.execute(\"\"\"
SELECT table_name
FROM information_schema.tables
WHERE table_schema = 'public'
AND table_name NOT IN ('alembic_version', 'spatial_ref_sys')
ORDER BY table_name
\"\"\")
existing_tables = [row[0] for row in cursor.fetchall()]
# Count core application tables
core_tables = ['users', 'projects', 'time_entries', 'settings', 'clients']
core_table_count = len([t for t in existing_tables if t in core_tables])
conn.close()
# If alembic_version exists but no core tables, treat as fresh (stale alembic_version)
if has_alembic and core_table_count == 0:
print('fresh')
elif has_alembic and core_table_count >= 3:
print('migrated')
elif existing_tables:
print('legacy')
else:
print('fresh')
except Exception as e:
print('unknown')
" 2>/dev/null
elif [[ "$db_url" == sqlite://* ]]; then
local db_file="${db_url#sqlite://}"
if [[ ! -f "$db_file" ]]; then
echo "fresh"
return
fi
python -c "
import sqlite3
try:
conn = sqlite3.connect('$db_file')
cursor = conn.cursor()
cursor.execute('SELECT name FROM sqlite_master WHERE type=\"table\" AND name=\"alembic_version\"')
has_alembic = cursor.fetchone() is not None
# Get all tables except system tables
cursor.execute('SELECT name FROM sqlite_master WHERE type=\"table\" AND name NOT IN (\"sqlite_sequence\")')
all_tables = [row[0] for row in cursor.fetchall()]
# Filter out alembic_version to get actual application tables
existing_tables = [t for t in all_tables if t != 'alembic_version']
# Count core application tables
core_tables = ['users', 'projects', 'time_entries', 'settings', 'clients']
core_table_count = len([t for t in existing_tables if t in core_tables])
conn.close()
# If alembic_version exists but no core tables, treat as fresh (stale alembic_version)
if has_alembic and core_table_count == 0:
print('fresh')
elif has_alembic and core_table_count >= 3:
print('migrated')
elif existing_tables:
print('legacy')
else:
print('fresh')
except Exception as e:
print('unknown')
" 2>/dev/null
else
echo "unknown"
fi
}
# Function to choose migration strategy
choose_migration_strategy() {
local db_state="$1"
case "$db_state" in
"fresh")
echo "fresh_init"
;;
"migrated")
echo "check_migrations"
;;
"legacy")
echo "comprehensive_migration"
;;
*)
echo "comprehensive_migration"
;;
esac
}
# Function to execute migration strategy
execute_migration_strategy() {
local strategy="$1"
local db_url="$2"
log "Executing migration strategy: '$strategy'"
case "$strategy" in
"fresh_init")
log "Executing fresh_init strategy..."
execute_fresh_init "$db_url"
;;
"check_migrations")
log "Executing check_migrations strategy..."
execute_check_migrations "$db_url"
;;
"comprehensive_migration")
log "Executing comprehensive_migration strategy..."
execute_comprehensive_migration "$db_url"
;;
*)
log "✗ Unknown migration strategy: '$strategy'"
return 1
;;
esac
}
# Compile translations (.po -> .mo) if needed
compile_translations() {
log "Compiling translation catalogs (if needed)..."
# Try pybabel if available
if command_exists pybabel; then
# Ensure writable permissions before compiling
chmod -R u+rwX,g+rwX /app/translations 2>/dev/null || true
if pybabel compile -d /app/translations >/dev/null 2>&1; then
log "✓ Translations compiled via pybabel"
return 0
else
log "⚠ pybabel compile failed or no catalogs to compile"
fi
else
log "pybabel not available; trying Python fallback"
fi
# Python fallback using app.utils.i18n
if python - <<'PY'
try:
import os
from app.utils.i18n import ensure_translations_compiled
try:
import pathlib
p = pathlib.Path('/app/translations')
for sub in p.glob('**/LC_MESSAGES'):
try:
os.chmod(str(sub), 0o775)
except Exception:
pass
except Exception:
pass
ensure_translations_compiled('/app/translations')
print('Python fallback: ensure_translations_compiled executed')
except Exception as e:
print(f'Python fallback failed: {e}')
PY
then
log "✓ Translations compiled via Python fallback"
return 0
else
log "⚠ Could not compile translations (will fallback to msgid)"
return 1
fi
}
# Function to execute fresh database initialization
execute_fresh_init() {
local db_url="$1"
log "Executing fresh database initialization..."
# Set FLASK_APP if not already set
if [[ -z "$FLASK_APP" ]]; then
log "⚠ FLASK_APP not set, setting it to app.py"
export FLASK_APP=app.py
fi
# Clean up any stale alembic_version table if database is truly fresh (no core tables)
log "Checking for stale alembic_version table..."
if [[ "$db_url" == sqlite://* ]]; then
local db_file="${db_url#sqlite://}"
if [[ -f "$db_file" ]]; then
local core_table_count=$(python -c "
import sqlite3
try:
conn = sqlite3.connect('$db_file')
cursor = conn.cursor()
cursor.execute('SELECT name FROM sqlite_master WHERE type=\"table\" AND name IN (\"users\", \"projects\", \"time_entries\", \"settings\", \"clients\")')
core_tables = cursor.fetchall()
conn.close()
print(len(core_tables))
except:
print(0)
" 2>/dev/null || echo "0")
if [[ "$core_table_count" -eq 0 ]]; then
log "No core tables found, cleaning up stale alembic_version if present..."
python -c "
import sqlite3
try:
conn = sqlite3.connect('$db_file')
cursor = conn.cursor()
cursor.execute('DROP TABLE IF EXISTS alembic_version')
conn.commit()
conn.close()
except:
pass
" 2>/dev/null || true
fi
fi
elif [[ "$db_url" == postgresql* ]]; then
local core_table_count=$(python -c "
import psycopg2
try:
conn_str = '$db_url'.replace('+psycopg2://', '://')
conn = psycopg2.connect(conn_str)
cursor = conn.cursor()
cursor.execute(\"\"\"
SELECT COUNT(*)
FROM information_schema.tables
WHERE table_schema = 'public'
AND table_name IN ('users', 'projects', 'time_entries', 'settings', 'clients')
\"\"\")
count = cursor.fetchone()[0]
conn.close()
print(count)
except:
print(0)
" 2>/dev/null || echo "0")
if [[ "$core_table_count" -eq 0 ]]; then
log "No core tables found, cleaning up stale alembic_version if present..."
python -c "
import psycopg2
try:
conn_str = '$db_url'.replace('+psycopg2://', '://')
conn = psycopg2.connect(conn_str)
cursor = conn.cursor()
cursor.execute('DROP TABLE IF EXISTS alembic_version')
conn.commit()
conn.close()
except:
pass
" 2>/dev/null || true
fi
fi
# Check if migrations directory already exists
if [[ -d "/app/migrations" ]]; then
log "⚠ Migrations directory already exists, checking if it's properly configured..."
# Check if we have the required files
if [[ -f "/app/migrations/env.py" && -f "/app/migrations/alembic.ini" ]]; then
log "✓ Migrations directory is properly configured, skipping init"
log "Checking if we need to create initial migration..."
# Check if we have any migration versions
if [[ ! -d "/app/migrations/versions" ]] || [[ -z "$(ls -A /app/migrations/versions 2>/dev/null)" ]]; then
log "No migration versions found, creating initial migration..."
if ! flask db migrate -m "Initial database schema"; then
log "✗ Initial migration creation failed"
log "Error details:"
flask db migrate -m "Initial database schema" 2>&1 | head -20
return 1
fi
log "✓ Initial migration created"
else
log "✓ Migration versions already exist"
fi
# Check current migration status
log "Checking current migration status..."
local current_revision=$(flask db current 2>/dev/null | tr -d '\n' || echo "none")
log "Current migration revision: $current_revision"
# If we have alembic_version but no tables, check what revision is stored
if [[ "$current_revision" != "none" ]] && [[ -n "$current_revision" ]]; then
log "Database has alembic_version with revision: $current_revision"
# Check if we actually have tables
if [[ "$db_url" == sqlite://* ]]; then
local db_file="${db_url#sqlite://}"
if [[ -f "$db_file" ]]; then
local table_count=$(python -c "
import sqlite3
try:
conn = sqlite3.connect('$db_file')
cursor = conn.cursor()
cursor.execute('SELECT name FROM sqlite_master WHERE type=\"table\" AND name != \"sqlite_sequence\" AND name != \"alembic_version\"')
tables = cursor.fetchall()
conn.close()
print(len(tables))
except:
print(0)
" 2>/dev/null || echo "0")
if [[ "$table_count" -eq 0 ]]; then
log "⚠ Database has alembic_version but no application tables"
log "This indicates a failed migration. Clearing alembic_version to start fresh..."
python -c "
import sqlite3
try:
conn = sqlite3.connect('$db_file')
cursor = conn.cursor()
cursor.execute('DROP TABLE IF EXISTS alembic_version')
conn.commit()
conn.close()
print('Cleared stale alembic_version table')
except Exception as e:
print(f'Error: {e}')
" 2>/dev/null || true
current_revision="none"
fi
fi
fi
fi
# Check if database has any tables (to determine if it's truly fresh)
local has_tables=false
if [[ "$db_url" == sqlite://* ]]; then
local db_file="${db_url#sqlite://}"
if [[ -f "$db_file" ]]; then
local table_count=$(python -c "
import sqlite3
try:
conn = sqlite3.connect('$db_file')
cursor = conn.cursor()
cursor.execute('SELECT name FROM sqlite_master WHERE type=\"table\" AND name != \"sqlite_sequence\"')
tables = cursor.fetchall()
conn.close()
print(len(tables))
except:
print(0)
" 2>/dev/null || echo "0")
if [[ "$table_count" -gt 0 ]]; then
has_tables=true
fi
fi
elif [[ "$db_url" == postgresql* ]]; then
local table_count=$(python -c "
import psycopg2
try:
conn_str = '$db_url'.replace('+psycopg2://', '://')
conn = psycopg2.connect(conn_str)
cursor = conn.cursor()
cursor.execute(\"\"\"
SELECT COUNT(*)
FROM information_schema.tables
WHERE table_schema = 'public'
\"\"\")
count = cursor.fetchone()[0]
conn.close()
print(count)
except:
print(0)
" 2>/dev/null || echo "0")
if [[ "$table_count" -gt 0 ]]; then
has_tables=true
fi
fi
# Only stamp if database has tables but no alembic_version (legacy database)
# For truly fresh databases, we should apply migrations from the beginning
if [[ "$current_revision" == "none" ]] && [[ "$has_tables" == "true" ]]; then
log "Database has tables but no alembic_version, stamping with current revision..."
local head_revision=$(flask db heads 2>/dev/null | tr -d '\n' || echo "")
if [[ -n "$head_revision" ]]; then
if ! flask db stamp "$head_revision"; then
log "✗ Database stamping failed"
log "Error details:"
flask db stamp "$head_revision" 2>&1 | head -20
return 1
fi
log "✓ Database stamped with revision: $head_revision"
fi
elif [[ "$current_revision" == "none" ]] && [[ "$has_tables" == "false" ]]; then
log "Fresh database detected (no tables), will apply migrations from beginning..."
fi
# Apply any pending migrations
log "Applying pending migrations..."
# Ensure we're in the right directory and FLASK_APP is set
cd /app
export FLASK_APP=${FLASK_APP:-app.py}
# Run migration with proper error capture
# Store output in a file and also show it in real-time for debugging
MIGRATION_OUTPUT=$(mktemp)
set +e # Don't exit on error immediately
# Run migration with Python traceback enabled for better error visibility
# Use 'heads' instead of 'head' to handle multiple migration branches
PYTHONUNBUFFERED=1 python -u -c "
import sys
import traceback
import os
# Set up environment
os.environ['PYTHONUNBUFFERED'] = '1'
try:
from flask import Flask
from flask_migrate import upgrade
from app import create_app
app = create_app()
with app.app_context():
try:
# Use 'heads' to upgrade all migration heads (handles branching)
upgrade(revision='heads')
sys.exit(0)
except Exception as e:
print('ERROR: Migration failed!', file=sys.stderr)
print(f'Error type: {type(e).__name__}', file=sys.stderr)
print(f'Error message: {e}', file=sys.stderr)
print('\\nFull traceback:', file=sys.stderr)
traceback.print_exc(file=sys.stderr)
sys.exit(1)
except Exception as e:
print('ERROR: Failed to initialize Flask app for migration!', file=sys.stderr)
print(f'Error type: {type(e).__name__}', file=sys.stderr)
print(f'Error message: {e}', file=sys.stderr)
print('\\nFull traceback:', file=sys.stderr)
traceback.print_exc(file=sys.stderr)
sys.exit(1)
" 2>&1 | tee "$MIGRATION_OUTPUT"
MIGRATION_EXIT_CODE=${PIPESTATUS[0]}
set -e # Re-enable exit on error
if [[ $MIGRATION_EXIT_CODE -ne 0 ]]; then
log "✗ Migration application failed (exit code: $MIGRATION_EXIT_CODE)"
log "Full error output:"
if [[ -s "$MIGRATION_OUTPUT" ]]; then
cat "$MIGRATION_OUTPUT"
else
log "No output captured - migration may have failed before producing output"
fi
# Get additional debugging info
log "Debugging information:"
log "Current migration revision:"
flask db current 2>&1 || log " (could not determine)"
if [[ "$db_url" == sqlite://* ]]; then
local db_file="${db_url#sqlite://}"
log "Database file: $db_file"
if [[ -f "$db_file" ]]; then
log " File exists: yes"
log " File size: $(stat -c%s "$db_file" 2>/dev/null || echo 'unknown') bytes"
log " Tables in database:"
python -c "
import sqlite3
try:
conn = sqlite3.connect('$db_file')
cursor = conn.cursor()
cursor.execute(\"SELECT name FROM sqlite_master WHERE type='table'\")
tables = [row[0] for row in cursor.fetchall()]
conn.close()
for table in tables:
print(f' - {table}')
except Exception as e:
print(f' Error: {e}')
" 2>&1 || true
else
log " File exists: no"
fi
fi
rm -f "$MIGRATION_OUTPUT"
return 1
fi
rm -f "$MIGRATION_OUTPUT"
log "✓ Migrations applied"
# Wait a moment for tables to be fully committed
log "Waiting for tables to be fully committed..."
sleep 2
return 0
else
log "⚠ Migrations directory exists but is incomplete, removing it..."
rm -rf /app/migrations
fi
fi
# Check if we're in the right directory
if [[ ! -f "/app/app.py" ]]; then
log "⚠ Not in correct directory, changing to /app"
cd /app
fi
# Initialize Flask-Migrate
log "Initializing Flask-Migrate..."
if ! flask db init; then
log "✗ Flask-Migrate initialization failed"
log "Error details:"
flask db init 2>&1 | head -20
return 1
fi
log "✓ Flask-Migrate initialized"
# Create initial migration
log "Creating initial migration..."
if ! flask db migrate -m "Initial database schema"; then
log "✗ Initial migration creation failed"
log "Error details:"
flask db migrate -m "Initial database schema" 2>&1 | head -20
return 1
fi
log "✓ Initial migration created"
# Apply migration
log "Applying initial migration..."
MIGRATION_OUTPUT=$(mktemp)
set +e # Don't exit on error immediately
# Use 'heads' to handle multiple migration branches
PYTHONUNBUFFERED=1 flask db upgrade heads > "$MIGRATION_OUTPUT" 2>&1
MIGRATION_EXIT_CODE=$?
set -e # Re-enable exit on error
if [[ $MIGRATION_EXIT_CODE -ne 0 ]]; then
log "✗ Initial migration application failed (exit code: $MIGRATION_EXIT_CODE)"
log "Error details:"
if [[ -s "$MIGRATION_OUTPUT" ]]; then
cat "$MIGRATION_OUTPUT"
else
log "No output captured from migration command"
fi
rm -f "$MIGRATION_OUTPUT"
return 1
fi
# Show migration output even on success for debugging
if [[ -s "$MIGRATION_OUTPUT" ]]; then
log "Migration output:"
cat "$MIGRATION_OUTPUT" | head -50
fi
rm -f "$MIGRATION_OUTPUT"
log "✓ Initial migration applied"
return 0
}
# Function to check and apply pending migrations
execute_check_migrations() {
local db_url="$1"
log "Checking for pending migrations..."
# Check current migration status
local current_revision=$(flask db current 2>/dev/null | tr -d '\n' || echo "unknown")
log "Current migration revision: $current_revision"
# If Alembic reports multiple heads and the DB is already at a head, attempting to
# upgrade to "heads" can fail with:
# "Requested revision X overlaps with other requested revisions ..."
# In that situation, treat it as "no pending migrations for this database" and
# continue startup (log a warning so operators can clean up the history).
local heads_output
heads_output=$(flask db heads 2>/dev/null || echo "")
local head_count
head_count=$(echo "$heads_output" | grep -v '^$' | wc -l | tr -d ' ')
if [[ "$current_revision" == *"(head)"* ]] && [[ "$head_count" -gt 1 ]]; then
log "⚠ Multiple migration heads detected ($head_count), but database is already at a head."
log "⚠ Skipping 'flask db upgrade heads' to avoid Alembic overlap error during startup."
log "Heads:"
echo "$heads_output" | sed 's/^/[HEAD] /'
return 0
fi
# Check if database actually has tables (not just alembic_version)
# This handles the case where alembic_version exists but migrations were never applied
local has_core_tables=false
if [[ "$db_url" == sqlite://* ]]; then
local db_file="${db_url#sqlite://}"
if [[ -f "$db_file" ]]; then
local core_table_count=$(python -c "
import sqlite3
try:
conn = sqlite3.connect('$db_file')
cursor = conn.cursor()
cursor.execute('SELECT name FROM sqlite_master WHERE type=\"table\" AND name IN (\"users\", \"projects\", \"time_entries\", \"settings\", \"clients\")')
core_tables = cursor.fetchall()
conn.close()
print(len(core_tables))
except:
print(0)
" 2>/dev/null || echo "0")
if [[ "$core_table_count" -ge 3 ]]; then
has_core_tables=true
fi
fi
elif [[ "$db_url" == postgresql* ]]; then
local core_table_count=$(python -c "
import psycopg2
try:
conn_str = '$db_url'.replace('+psycopg2://', '://')
conn = psycopg2.connect(conn_str)
cursor = conn.cursor()
cursor.execute(\"\"\"
SELECT COUNT(*)
FROM information_schema.tables
WHERE table_schema = 'public'
AND table_name IN ('users', 'projects', 'time_entries', 'settings', 'clients')
\"\"\")
count = cursor.fetchone()[0]
conn.close()
print(count)
except:
print(0)
" 2>/dev/null || echo "0")
if [[ "$core_table_count" -ge 3 ]]; then
has_core_tables=true
fi
fi
# If alembic_version exists but no core tables, treat as fresh and apply all migrations
if [[ "$current_revision" != "none" ]] && [[ "$has_core_tables" == "false" ]]; then
log "⚠ Database has alembic_version but no core tables detected"
log "This indicates migrations were never applied. Applying all migrations from beginning..."
# Drop alembic_version table to start fresh
if [[ "$db_url" == sqlite://* ]]; then
local db_file="${db_url#sqlite://}"
python -c "
import sqlite3
try:
conn = sqlite3.connect('$db_file')
cursor = conn.cursor()
cursor.execute('DROP TABLE IF EXISTS alembic_version')
conn.commit()
conn.close()
print('Dropped stale alembic_version table')
except Exception as e:
print(f'Error: {e}')
" 2>/dev/null || true
elif [[ "$db_url" == postgresql* ]]; then
python -c "
import psycopg2
try:
conn_str = '$db_url'.replace('+psycopg2://', '://')
conn = psycopg2.connect(conn_str)
cursor = conn.cursor()
cursor.execute('DROP TABLE IF EXISTS alembic_version')
conn.commit()
conn.close()
print('Dropped stale alembic_version table')
except Exception as e:
print(f'Error: {e}')
" 2>/dev/null || true
fi
log "Now applying all migrations..."
fi
# Check for pending migrations
MIGRATION_OUTPUT=$(mktemp)
set +e # Don't exit on error immediately
# Use 'heads' to handle multiple migration branches
PYTHONUNBUFFERED=1 flask db upgrade heads > "$MIGRATION_OUTPUT" 2>&1
MIGRATION_EXIT_CODE=$?
set -e # Re-enable exit on error
if [[ $MIGRATION_EXIT_CODE -ne 0 ]]; then
# If we're already at a head, tolerate Alembic overlap errors (multi-head history).
if [[ "$current_revision" == *"(head)"* ]] && [[ -s "$MIGRATION_OUTPUT" ]] && grep -q "overlaps with other requested revisions" "$MIGRATION_OUTPUT"; then
log "⚠ Migration upgrade reported overlapping heads, but DB is already at a head. Continuing startup."
log "⚠ (You likely have multiple heads in the migration history; consider upgrading to a build with a linearized migration chain.)"
rm -f "$MIGRATION_OUTPUT"
return 0
fi
log "✗ Migration check failed (exit code: $MIGRATION_EXIT_CODE)"
log "Error details:"
if [[ -s "$MIGRATION_OUTPUT" ]]; then
cat "$MIGRATION_OUTPUT"
else
log "No output captured from migration command"
fi
rm -f "$MIGRATION_OUTPUT"
return 1
fi
# Show migration output even on success for debugging
if [[ -s "$MIGRATION_OUTPUT" ]]; then
log "Migration output:"
cat "$MIGRATION_OUTPUT" | head -50
fi
rm -f "$MIGRATION_OUTPUT"
log "✓ Migrations checked and applied"
return 0
}
# Function to execute comprehensive migration
execute_comprehensive_migration() {
local db_url="$1"
log "Executing comprehensive migration..."
# Try to run the enhanced startup script
if [[ -f "/app/docker/startup_with_migration.py" ]]; then
log "Running enhanced startup script..."
if python /app/docker/startup_with_migration.py; then
log "✓ Enhanced startup script completed successfully"
return 0
else
log "⚠ Enhanced startup script failed, falling back to manual migration"
fi
fi
# Fallback to manual migration
log "Executing manual migration fallback..."
# Check what tables exist in the database
log "Analyzing existing database structure..."
if ! python -c "
import psycopg2
try:
conn_str = '$db_url'.replace('+psycopg2://', '://')
conn = psycopg2.connect(conn_str)
cursor = conn.cursor()
cursor.execute(\"\"\"
SELECT table_name
FROM information_schema.tables
WHERE table_schema = 'public'
ORDER BY table_name
\"\"\")
existing_tables = [row[0] for row in cursor.fetchall()]
cursor.execute(\"\"\"
SELECT table_name
FROM information_schema.tables
WHERE table_name = 'alembic_version'
AND table_schema = 'public'
\"\"\)
alembic_table = cursor.fetchone()
conn.close()
print(f'Existing tables: {existing_tables}')
print(f'Alembic version table exists: {bool(alembic_table)}')
if len(existing_tables) == 0:
print('Database is empty - no baseline migration needed')
exit(0)
elif alembic_table:
print('Database already has alembic_version table - no baseline migration needed')
exit(0)
else:
print('Database has existing tables but no alembic_version - baseline migration needed')
exit(1)
except Exception as e:
print(f'Error analyzing database: {e}')
exit(1)
"; then
log "Database analysis completed"
fi
# Initialize Flask-Migrate if not already done
if [[ ! -f "/app/migrations/env.py" ]]; then
log "Initializing Flask-Migrate..."
if ! flask db init; then
log "✗ Flask-Migrate initialization failed"
log "Error details:"
flask db init 2>&1 | head -20
return 1
fi
log "✓ Flask-Migrate initialized"
fi
# Check if we need to create a baseline migration
log "Checking if baseline migration is needed..."
if python -c "
import psycopg2
try:
conn_str = '$db_url'.replace('+psycopg2://', '://')
conn = psycopg2.connect(conn_str)
cursor = conn.cursor()
cursor.execute(\"\"\"
SELECT table_name
FROM information_schema.tables
WHERE table_name = 'alembic_version'
AND table_schema = 'public'
\"\"\")
alembic_table = cursor.fetchone()
conn.close()
if alembic_table:
print('Alembic version table already exists - skipping baseline migration')
exit(0)
else:
print('Alembic version table missing - baseline migration needed')
exit(1)
except Exception as e:
print(f'Error checking alembic version: {e}')
exit(1)
"; then
log "✓ No baseline migration needed - database already stamped"
return 0
fi
# If we get here, we need a baseline migration
# But first, let's check if there are any conflicting tables
log "Checking for potential table conflicts..."
if ! python -c "
import psycopg2
try:
conn_str = '$db_url'.replace('+psycopg2://', '://')
conn = psycopg2.connect(conn_str)
cursor = conn.cursor()
# Check for tables that might conflict with our migration
cursor.execute(\"\"\"
SELECT table_name
FROM information_schema.tables
WHERE table_schema = 'public'
AND table_name IN ('clients', 'users', 'projects', 'tasks', 'time_entries', 'settings', 'invoices', 'invoice_items')
ORDER BY table_name
\"\"\")
conflicting_tables = [row[0] for row in cursor.fetchall()]
conn.close()
if conflicting_tables:
print(f'Found potentially conflicting tables: {conflicting_tables}')
print('These tables might conflict with the migration. Consider backing up data first.')
exit(1)
else:
print('No conflicting tables found - safe to proceed with baseline migration')
exit(0)
except Exception as e:
print(f'Error checking for conflicts: {e}')
exit(1)
"; then
log "⚠ Potential table conflicts detected - proceeding with caution"
fi
# Create baseline migration
log "Creating baseline migration from existing database..."
if ! flask db migrate -m "Baseline from existing database"; then
log "✗ Baseline migration creation failed"
log "Error details:"
flask db migrate -m "Baseline from existing database" 2>&1 | head -20
# Try to understand why the migration failed
log "Attempting to understand migration failure..."
if ! python -c "
import psycopg2
try:
conn_str = '$db_url'.replace('+psycopg2://', '://')
conn = psycopg2.connect(conn_str)
cursor = conn.cursor()
cursor.execute(\"\"\"
SELECT table_name, column_name, data_type, is_nullable
FROM information_schema.columns
WHERE table_schema = 'public'
ORDER BY table_name, ordinal_position
\"\"\")
columns = cursor.fetchall()
conn.close()
print('Database schema:')
current_table = None
for table, column, data_type, nullable in columns:
if table != current_table:
print(f'\\nTable: {table}')
current_table = table
print(f' {column}: {data_type} (nullable: {nullable})')
except Exception as e:
print(f'Error getting schema: {e}')
"; then
log "Could not analyze database schema"
fi
log "✗ Baseline migration creation failed - trying alternative approach..."
# Try to stamp the database with the existing migration version
log "Attempting to stamp database with existing migration version..."
if python -c "
import psycopg2
try:
conn_str = '$db_url'.replace('+psycopg2://', '://')
conn = psycopg2.connect(conn_str)
cursor = conn.cursor()
# Check if we have any migration files
import os
migration_dir = '/app/migrations/versions'
if os.path.exists(migration_dir):
migration_files = [f for f in os.listdir(migration_dir) if f.endswith('.py')]
if migration_files:
# Get the first migration file (should be 001_initial_schema.py)
first_migration = sorted(migration_files)[0]
migration_id = first_migration.split('_')[0]
print(f'Found migration file: {first_migration} with ID: {migration_id}')
# Try to stamp the database with this migration
import subprocess
result = subprocess.run(['flask', 'db', 'stamp', migration_id],
capture_output=True, text=True)
if result.returncode == 0:
print('Successfully stamped database with existing migration')
exit(0)
else:
print(f'Failed to stamp database: {result.stderr}')
exit(1)
else:
print('No migration files found')
exit(1)
else:
print('Migration directory not found')
exit(1)
except Exception as e:
print(f'Error in alternative approach: {e}')
exit(1)
"; then
log "✓ Database stamped with existing migration version"
return 0
else
log "✗ Alternative approach also failed"
return 1
fi
fi
log "✓ Baseline migration created"
# Stamp database as current
log "Stamping database with current migration version..."
if ! flask db stamp head; then
log "✗ Database stamping failed"
log "Error details:"
flask db stamp head 2>&1 | head -20
return 1
fi
log "✓ Database stamped as current"
return 0
}
# Function to ensure data directory permissions
ensure_data_directory() {
log "Ensuring /data directory exists and has proper permissions..."
# Create /data directory if it doesn't exist
if [ ! -d "/data" ]; then
log "Creating /data directory..."
mkdir -p /data
fi
# Try to set permissions (best effort - may fail if we don't have permission)
# This is useful when the volume is mounted with different ownership
if [ -w "/data" ]; then
log "Setting permissions on /data directory..."
chmod 755 /data 2>/dev/null || true
# Get current user info
CURRENT_UID=$(id -u 2>/dev/null || echo "1000")
CURRENT_GID=$(id -g 2>/dev/null || echo "1000")
# Try to change ownership if we have permission (requires root or matching ownership)
if [ "$CURRENT_UID" = "0" ] || [ -O "/data" ]; then
log "Setting ownership of /data to current user (UID: $CURRENT_UID, GID: $CURRENT_GID)..."
chown "$CURRENT_UID:$CURRENT_GID" /data 2>/dev/null || true
else
log "Cannot change ownership of /data (not root and not owner), but directory is writable"
fi
# Ensure subdirectories exist
mkdir -p /data/uploads 2>/dev/null || true
chmod 755 /data/uploads 2>/dev/null || true
log "✓ /data directory setup complete"
else
log "⚠ /data directory is not writable - this may cause issues"
log "Current user: $(whoami) (UID: $(id -u 2>/dev/null || echo 'unknown'))"
log "Directory permissions: $(ls -ld /data 2>/dev/null || echo 'cannot read')"
fi
}
# Function to verify database integrity
verify_database_integrity() {
local db_url="$1"
log "Verifying database integrity..."
# Try up to 3 times with delays
local max_attempts=3
local attempt=1
while [[ $attempt -le $max_attempts ]]; do
log "Database integrity check attempt $attempt/$max_attempts..."
# Test basic database operations
if [[ "$db_url" == postgresql* ]]; then
log "Checking PostgreSQL database integrity..."
if python -c "
import psycopg2
try:
# Parse connection string to remove +psycopg2 if present
conn_str = '$db_url'.replace('+psycopg2://', '://')
conn = psycopg2.connect(conn_str)
cursor = conn.cursor()
# Check for key tables that should exist after migration
cursor.execute(\"\"\"
SELECT table_name
FROM information_schema.tables
WHERE table_name IN ('clients', 'users', 'projects', 'tasks', 'time_entries', 'settings', 'invoices', 'invoice_items')
AND table_schema = 'public'
ORDER BY table_name
\"\"\")
key_tables = [row[0] for row in cursor.fetchall()]
# Also check if alembic_version table exists
cursor.execute(\"\"\"
SELECT table_name
FROM information_schema.tables
WHERE table_name = 'alembic_version'
AND table_schema = 'public'
\"\"\")
alembic_table = cursor.fetchone()
conn.close()
print(f'Found tables: {key_tables}')
print(f'Alembic version table: {alembic_table[0] if alembic_table else \"missing\"}')
# Require at least the core tables and alembic_version
if len(key_tables) >= 5 and alembic_table:
exit(0)
else:
print(f'Expected at least 5 core tables, found {len(key_tables)}')
print(f'Expected alembic_version table: {bool(alembic_table)}')
exit(1)
except Exception as e:
print(f'Error during integrity check: {e}')
exit(1)
"; then
log "✓ Database integrity verified (PostgreSQL via Python)"
return 0
else
log "✗ Database integrity check failed (PostgreSQL)"
log "Error details:"
python -c "
import psycopg2
try:
conn_str = '$db_url'.replace('+psycopg2://', '://')
conn = psycopg2.connect(conn_str)
cursor = conn.cursor()
cursor.execute(\"\"\"
SELECT table_name
FROM information_schema.tables
WHERE table_schema = 'public'
ORDER BY table_name
\"\"\")
all_tables = [row[0] for row in cursor.fetchall()]
cursor.execute(\"\"\"
SELECT table_name
FROM information_schema.tables
WHERE table_name = 'alembic_version'
AND table_schema = 'public'
\"\"\)
alembic_table = cursor.fetchone()
conn.close()
print(f'All tables in database: {all_tables}')
print(f'Alembic version table exists: {bool(alembic_table)}')
except Exception as e:
print(f'Error getting table list: {e}')
" 2>&1 | head -20
if [[ $attempt -lt $max_attempts ]]; then
log "Waiting 3 seconds before retry..."
sleep 3
attempt=$((attempt + 1))
continue
else
return 1
fi
fi
elif [[ "$db_url" == sqlite://* ]]; then
local db_file="${db_url#sqlite://}"
if [[ ! -f "$db_file" ]]; then
log "✗ SQLite database file not found"
return 1
fi
log "Checking SQLite database integrity..."
if python -c "
import sqlite3
try:
conn = sqlite3.connect('$db_file')
cursor = conn.cursor()
cursor.execute('SELECT name FROM sqlite_master WHERE type=\"table\" AND name IN (\"clients\", \"users\", \"projects\", \"tasks\", \"time_entries\", \"settings\", \"invoices\", \"invoice_items\")')
key_tables = [row[0] for row in cursor.fetchall()]
cursor.execute('SELECT name FROM sqlite_master WHERE type=\"table\" AND name=\"alembic_version\"')
alembic_table = cursor.fetchone()
conn.close()
print(f'Found tables: {key_tables}')
print(f'Alembic version table: {alembic_table[0] if alembic_table else \"missing\"}')
if len(key_tables) >= 5 and alembic_table:
exit(0)
else:
print(f'Expected at least 5 core tables, found {len(key_tables)}')
print(f'Expected alembic_version table: {bool(alembic_table)}')
exit(1)
except Exception as e:
print(f'Error during integrity check: {e}')
exit(1)
"; then
log "✓ Database integrity verified (SQLite)"
return 0
else
log "✗ Database integrity check failed (SQLite)"
if [[ $attempt -lt $max_attempts ]]; then
log "Waiting 3 seconds before retry..."
sleep 3
attempt=$((attempt + 1))
continue
else
return 1
fi
fi
else
log "✗ Unsupported database type: $db_url"
return 1
fi
# If we get here, the check failed
if [[ $attempt -lt $max_attempts ]]; then
log "Waiting 3 seconds before retry..."
sleep 3
attempt=$((attempt + 1))
else
return 1
fi
done
return 1
}
# Main execution
main() {
log "=== TimeTracker Docker Entrypoint with Migration Detection ==="
# Set environment variables
export FLASK_APP=${FLASK_APP:-/app/app.py}
# Get database URL from environment
local db_url="${DATABASE_URL}"
if [[ -z "$db_url" ]]; then
log "✗ DATABASE_URL environment variable not set"
log "Available environment variables:"
env | grep -E "(DATABASE|FLASK|PYTHON)" | sort
exit 1
fi
log "Database URL: $db_url"
# Ensure data directory has proper permissions
ensure_data_directory
# Wait for database to be available
if ! wait_for_database "$db_url"; then
log "✗ Failed to connect to database"
log "Trying to run simple test script for debugging..."
if [[ -f "/app/docker/simple_test.sh" ]]; then
/app/docker/simple_test.sh
fi
exit 1
fi
# Detect database state
local db_state=$(detect_database_state "$db_url")
log "Raw db_state output: '${db_state}'"
log "Detected database state: '$db_state'"
# Choose migration strategy
local strategy=$(choose_migration_strategy "$db_state")
log "Raw strategy output: '${strategy}'"
log "Selected migration strategy: '$strategy'"
# Log the strategy selection details
case "$db_state" in
"fresh")
log "Fresh database detected - using standard initialization"
;;
"migrated")
log "Database already migrated - checking for pending migrations"
;;
"legacy")
log "Legacy database detected - using comprehensive migration"
;;
*)
log "Unknown database state: '$db_state' - using comprehensive migration as fallback"
;;
esac
# Execute migration strategy
if ! execute_migration_strategy "$strategy" "$db_url"; then
log "✗ Migration strategy execution failed"
exit 1
fi
# Verify database integrity
if ! verify_database_integrity "$db_url"; then
log "✗ Database integrity verification failed"
exit 1
fi
log "=== Startup and Migration Complete ==="
log "Database is ready for use"
# Show final migration status
local final_status=$(flask db current 2>/dev/null | tr -d '\n' || echo "unknown")
log "Final migration status: $final_status"
# Start the application
# Best-effort compile translations before starting
compile_translations || true
log "Starting TimeTracker application..."
exec "$@"
}
# Run main function with all arguments
main "$@"