[docker/cli] Add `--no-trunc` option to `docker stack deploy` for full error messages

### ROOT CAUSE The `docker stack deploy` command truncates error messages by default, particularly when encountering scheduling constraint failures. This truncation occurs due to hardcoded output formatting that does not respect terminal width or the `--no-trunc` flag, making it difficult to diagnose deployment issues. The truncation is hardcoded and does not adapt to the user's environment or terminal settings. ### CODE FIX To resolve this issue, implement a `--no-trunc` flag for the `docker stack deploy` command, similar to the existing flag in `docker service ps`. This flag should disable truncation of error messages, providing full details for debugging. #### Changes: 1. **Add the `--no-trunc` flag** to the `Deploy` function in `docker/stack/stack.go`. 2. **Modify error handling** to respect the `--no-trunc` flag when printing errors. ```go // In docker/stack/stack.go, modify the Deploy function func (s *stack) Deploy(ctx context.Context, cli *cli.Client, noTrunc bool) error { // ... existing code ... // When an error occurs, print it with noTrunc if err := compose.ParseFile(s.ComposeFile); err != nil { clioutput.PrintError(os.Stderr, err, noTrunc) return err } // ... existing code ... } ``` 3. **Update the command-line flag** in `docker/stack/stack.go` to include `--no-trunc`. ```go // In docker/stack/stack.go, add the flag to the Deploy command func (s *stack) Deploy() *cobra.Command { cmd := &cobra.Command{ Use: "deploy [service name or Compose file]", Short: "Deploy a stack", Long: "Deploy a stack to the swarm", Run: s.Deploy, Flags: func() *pflag.FlagSet { fs := pflag.NewFlagSet("deploy", pflag.ExitOnError) fs.BoolVarP(&deployFlag, "no-trunc", "n", false, "Do not truncate output") return fs }(), } return cmd } ``` 4. **Ensure the `noTrunc` flag is passed** to the `Deploy` function when invoked. This fix ensures that error messages are fully displayed when `--no-trunc` is provided, improving the debugging experience for deployment issues.