This PR fixes multiple issues with the provided docker-compose.yml setup that were preventing a clean deployment of Worklenz.

Problems identified

Problem 1

Inline init commands in Compose were triggering Docker variable substitution warnings ($i, $f), causing broken init scripts and migration failures.

WARN[0000] The "i" variable is not set. Defaulting to a blank string. 
WARN[0000] The "f" variable is not set. Defaulting to a blank string. 
WARN[0000] The "f" variable is not set. Defaulting to a blank string.

So what happens here is that docker compose tries to substitute all $VARIABLE strings. So inline scripts such as:

command: >
    bash -c '
      find /docker-entrypoint-initdb.d -type f -name "*.sh" -exec sh -c '"'"'
        for f; do
          dos2unix "$f" 2>/dev/null || true
          chmod +x "$f"
        done
      '"'"' sh {} +
      exec docker-entrypoint.sh postgres
    '

Will have their $f replaced with empty strings that leads to:

chmod: cannot access '': No such file or directory

Problem 2

Postgres healthcheck was running too early (before initdb, dos2unix install, and migrations finished), causing the container to be marked unhealthy and dependent services to fail.

worklenz_db | CREATE DATABASE 
worklenz_db | 
worklenz_db | 
worklenz_db | /usr/local/bin/docker-entrypoint.sh: sourcing /docker-entrypoint-initdb.d/00_init.sh 
worklenz_db | Starting database initialization... 
dependency failed to start: container worklenz_db is unhealthy

Fixes applied

1. Moved inline init commands into standalone scripts:

• docker/db-init-wrapper.sh – runs dos2unix + chmod safely without Compose substitution conflicts.
• docker/createbuckets-entrypoint.sh – encapsulates MinIO bucket creation logic.

2. Updated Postgres db service to use healthcheck start_period of 30s so the healthcheck waits for initialization before running.

Should fix #325