Polish the output for better UX?

This commit is contained in:
Alexandre Aubin 2021-02-03 18:30:26 +01:00
parent adec1fdf02
commit 0c13d92b6d

View file

@ -180,8 +180,8 @@ class TestReport:
def __init__(self, message):
self.message = message
def display(self):
_print(self.style % self.message)
def display(self, prefix=""):
_print(prefix + self.style % self.message)
class Warning(TestReport):
style = c.WARNING + " ! %s " + c.END
@ -190,7 +190,7 @@ class Error(TestReport):
style = c.FAIL + "%s" + c.END
class Info(TestReport):
style = " %s" + c.END
style = " - %s" + c.END
class Success(TestReport):
style = c.OKGREEN + "%s" + c.END
@ -198,20 +198,6 @@ class Success(TestReport):
class Critical(TestReport):
style = c.FAIL + " ✘✘✘ %s" + c.END
def header(app):
_print("""
[{header}{bold}YunoHost App Package Linter{end}]
App packaging documentation - https://yunohost.org/#/packaging_apps
App package example - https://github.com/YunoHost/example_ynh
Official helpers - https://yunohost.org/#/packaging_apps_helpers_en
Experimental helpers - https://github.com/YunoHost-Apps/Experimental_helpers
If you believe this linter returns false negative (warnings / errors which shouldn't happen),
please report them on https://github.com/YunoHost/package_linter/issues
Analyzing package {header}{app}{end}"""
.format(header=c.HEADER, bold=c.BOLD, end=c.END, app=app))
output = "plain"
@ -222,16 +208,10 @@ def _print(*args, **kwargs):
print(*args, **kwargs)
def print_header(str):
_print("\n [" + c.BOLD + c.HEADER + str.title() + c.END + "]\n")
def report_warning_not_reliable(str):
_print(c.MAYBE_FAIL + "?", str, c.END)
def print_happy(str):
_print(c.OKGREEN + "", str, "")
@ -282,22 +262,62 @@ def test(**kwargs):
class TestSuite():
def run_tests(self):
reports = []
for test, options in tests[self.__class__.__name__]:
if "only" in options and self.name not in options["only"]:
continue
if "ignore" in options and self.name in options["ignore"]:
continue
self.run_single_test(test)
reports += list(test(self))
def report_type(report):
return report.__class__.__name__.lower()
if output == "plain":
if any(report_type(r) in ["warning", "error", "critical"] for r in reports):
prefix = c.WARNING + '! '
elif any(report_type(r) in ["info"] for r in reports):
prefix = ''
else:
prefix = c.OKGREEN + ''
_print(" " + c.BOLD + prefix + c.OKBLUE + self.test_suite_name + c.END)
if len(reports):
_print("")
for report in reports:
report.display(prefix=" ")
if len(reports):
_print("")
else:
for report in reports:
test_name = test.__qualname__
tests_reports[report_type(report)].append((test_name, report))
def run_single_test(self, test):
reports = list(test(self))
for report in reports:
if output == "plain":
if output == "plain":
for report in reports:
report.display()
report_type = report.__class__.__name__.lower()
test_name = test.__qualname__
tests_reports[report_type].append((test_name, report))
else:
def report_type(report):
return report.__class__.__name__.lower()
for report in reports:
test_name = test.__qualname__
tests_reports[report_type(report)].append((test_name, report))
# ############################################################################
# Actual high-level checks
@ -310,7 +330,7 @@ class App(TestSuite):
def __init__(self, path):
print_header("LOADING APP")
_print(" Analyzing app %s ..." % path)
self.path = path
self.manifest_ = Manifest(self.path)
self.manifest = self.manifest_.manifest
@ -318,28 +338,27 @@ class App(TestSuite):
self.configurations = Configurations(self)
self.app_catalog = AppCatalog(self.manifest["id"])
self.test_suite_name = "General stuff, misc helper usage"
_print()
def analyze(self):
print_header("MANIFEST")
self.manifest_.run_tests()
for script in [self.scripts[s] for s in scriptnames if self.scripts[s].exists]:
print_header(script.name.upper() + " SCRIPT")
script.run_tests()
print_header("GENERAL STUFF, MISC HELPER USAGE")
self.run_tests()
print_header("CONFIGURATIONS")
self.configurations.run_tests()
print_header("APP CATALOG")
self.app_catalog.run_tests()
self.report()
def report(self):
_print(" =======")
# These are meant to be the last stuff running, they are based on
# previously computed errors/warning/successes
self.run_single_test(App.qualify_for_level_7)
@ -569,6 +588,7 @@ class Configurations(TestSuite):
def __init__(self, app):
self.app = app
self.test_suite_name = "Configuration files"
############################
# _____ __ #
@ -885,6 +905,7 @@ class Manifest(TestSuite):
def __init__(self, path):
self.path = path
self.test_suite_name = "manifest.json"
manifest_path = os.path.join(path, 'manifest.json')
assert os.path.exists(manifest_path), "manifest.json is missing"
@ -1088,6 +1109,7 @@ class AppCatalog(TestSuite):
def __init__(self, app_id):
self.app_id = app_id
self.test_suite_name = "Catalog infos"
self._fetch_app_repo()
@ -1249,6 +1271,7 @@ class Script(TestSuite):
if not self.exists:
return
self.lines = list(self.read_file())
self.test_suite_name = "scripts/" + self.name
def read_file(self):
with open(self.path) as f:
@ -1262,6 +1285,7 @@ class Script(TestSuite):
lines = '\n'.join(lines).replace("\\\n", "").split("\n")
some_parsing_failed = False
some_parsing_failed_closing_quotation = False
for line in lines:
@ -1269,9 +1293,15 @@ class Script(TestSuite):
line = shlex.split(line, True)
yield line
except Exception as e:
ignore_pattern = ["/etc/cron", "admin_panel=", "echo \"", "__PRE_TAG", "__URL_TAG", "maintenance.$app.conf", "mail_message=", "maintenance.$app.html", "> mail_to_send"]
if str(e) == "No closing quotation" and any(pattern in line for pattern in ignore_pattern):
continue
if not some_parsing_failed:
_print("Some lines could not be parsed in script %s. (That's probably not really critical)" % self.name)
some_parsing_failed = True
report_warning_not_reliable("%s : %s" % (e, line))
def occurences(self, command):
@ -1565,7 +1595,19 @@ def main():
global output
output = "json" if "--json" in sys.argv else "plain"
header(app_path)
_print("""
[{blue}{bold}YunoHost App Package Linter{end}]
App packaging documentation - https://yunohost.org/#/packaging_apps
App package example - https://github.com/YunoHost/example_ynh
Official helpers - https://yunohost.org/#/packaging_apps_helpers_en
Experimental helpers - https://github.com/YunoHost-Apps/Experimental_helpers
If you believe this linter returns false negative (warnings / errors which shouldn't happen),
please report them on https://github.com/YunoHost/package_linter/issues
""".format(blue=c.OKBLUE, bold=c.BOLD, end=c.END))
app = App(app_path)
app.analyze()