answerfile: make several usability improvements

- previously the generated answerfile didn't contain the
  expected choices for the boolean component (the only one
  allowed component for actor dialogs so far);
- leapp answer command now prints usage message in case of
  inappropriate invocation;
- answerfile path shown at the end of leapp preupgrade run
  even without --debug.

leapp/cli/upgrade/__init__.py

@@ -9,7 +9,7 @@
 from datetime import datetime
 
 from leapp.config import get_config
-from leapp.exceptions import CommandError, LeappError
+from leapp.exceptions import CommandError, LeappError, UsageError
 from leapp.logger import configure_logger
 from leapp.messaging.answerstore import AnswerStore
 from leapp.repository.scan import find_and_scan_repositories
@@ -188,6 +188,7 @@ def upgrade(args):
     context = str(uuid.uuid4())
     cfg = get_config()
     configuration = prepare_configuration(args)
+    answerfile_path = cfg.get('report', 'answerfile')
 
     if os.getuid():
         raise CommandError('This command has to be run under the root user.')
@@ -224,16 +225,12 @@ def upgrade(args):
     process_whitelist_experimental(repositories, workflow, configuration, logger)
     warn_if_unsupported(configuration)
     with beautify_actor_exception():
-        answerfile_path = cfg.get('report', 'answerfile')
         logger.info("Using answerfile at %s", answerfile_path)
         workflow.load_answerfile(answerfile_path)
         workflow.run(context=context, skip_phases_until=skip_phases_until, skip_dialogs=True)
 
     report_errors(workflow.errors)
     generate_report_files(context)
-
-    cfg = get_config()
-
     report_files = get_cfg_files('report', cfg)
     log_files = get_cfg_files('logs', cfg)
     report_info(report_files, log_files, fail=workflow.failure)
@@ -281,15 +278,15 @@ def preupgrade(args):
     report_errors(workflow.errors)
     report_files = get_cfg_files('report', cfg)
     log_files = get_cfg_files('logs', cfg)
-    report_info(report_files, log_files, fail=workflow.failure)
+    report_info(report_files, log_files, answerfile_path, fail=workflow.failure)
     if workflow.failure:
         sys.exit(1)
 
 
-@command('answer', help='Manage answerfile')
-@command_opt('answerfile', help='Path to an answerfile to update')
+@command('answer', help='Manage answerfile: register user choices for specific dialog sections')
+@command_opt('answerfile', help='Path to the answerfile to update')
 @command_opt('section', action='append', metavar='dialog_sections',
-             help='Record answer for specific section in answerfile')
+             help='Register answer for a specific section in the answerfile')
 @command_opt('add', is_flag=True,
              help='If set sections will be created even if missing in original answerfile')
 def answer(args):
@@ -298,12 +295,12 @@ def answer(args):
     if args.section:
         args.section = list(itertools.chain(*[i.split(',') for i in args.section]))
     else:
-        raise CommandError('At least one dialog section must be specified, ex. --section dialog.option=mychoice')
+        raise UsageError('At least one dialog section must be specified, ex. --section dialog.option=mychoice')
     try:
         sections = [tuple((dialog_option.split('.', 2) + [value]))
                     for dialog_option, value in [s.split('=', 2) for s in args.section]]
     except ValueError:
-        raise CommandError("A bad formatted section has been passed. Expected format is dialog.option=mychoice")
+        raise UsageError("A bad formatted section has been passed. Expected format is dialog.option=mychoice")
     answerfile_path = args.answerfile or cfg.get('report', 'answerfile')
     answerstore = AnswerStore()
     answerstore.load(answerfile_path)

leapp/dialogs/components.py

@@ -84,6 +84,7 @@ class BooleanComponent(Component):
     """
     BooleanComponent is used for boolean inputs such as Yes/No questions.
     """
+    choices = ('True', 'False')
     values = ('Yes', 'No')
     value_type = bool
 

leapp/messaging/answerstore.py

@@ -145,9 +145,7 @@ def generate(self, dialogs, answer_file_path):
                     choices = ''
                     answer_entry = ''
                     if hasattr(component, 'choices'):
-                        choices += '# Available choices:\n'
-                        for choice in component.choices:
-                            choices += '# - {}\n'.format(choice)
+                        choices += '# Available choices: {}\n'.format('/'.join(component.choices))
                         if component.value_type is tuple:
                             default = ';'.join(component.default)
                             answer = ';'.join(answer) if answer else answer

leapp/utils/output.py

@@ -46,7 +46,7 @@ def report_errors(errors):
         sys.stdout.write(pretty_block("END OF ERRORS", color=Color.red))
 
 
-def report_info(report_paths, log_paths, fail=False):
+def report_info(report_paths, log_paths, answerfile=None, fail=False):
     report_paths = [report_paths] if not isinstance(report_paths, list) else report_paths
     log_paths = [log_paths] if not isinstance(report_paths, list) else log_paths
 
@@ -63,6 +63,9 @@ def report_info(report_paths, log_paths, fail=False):
         sys.stdout.write(pretty_block("END OF REPORT", color=Color.bold if fail else Color.green))
         sys.stdout.write("\n")
 
+    if answerfile:
+        sys.stdout.write("Answerfile has been generated at {}\n".format(answerfile))
+
 
 def report_unsupported(devel_vars, experimental):
     sys.stdout.write(pretty_block("UNSUPPORTED UPGRADE", color=Color.yellow))