Skip to content

Commit

Permalink
shadow-dom: Add test conversion script.
Browse files Browse the repository at this point in the history 
This script converts tests under the shadow-dom/ directory to separate HTML
files with appropriate metadata format. The output is produced within
the out/ directory by default to avoid conflicts with existing files, but
I intend that the files under the out/ directory will take over the existing
contents of the shadow-dom/ directory.

The script file will be removed after the conversion.
  • Loading branch information
@yutak
yutak committed Jun 3, 2013
1 parent ee27fe8 commit 1bcae1d
Showing 1 changed file with 303 additions and 0 deletions.
303 changes: 303 additions & 0 deletions shadow-dom/convert.py
View file
Original file line number Diff line number Diff line change
@@ -0,0 +1,303 @@
#!/usr/bin/python3

# Distributed under both the W3C Test Suite License [1] and the W3C
# 3-clause BSD License [2]. To contribute to a W3C Test Suite, see the
# policies and contribution forms [3].
#
# [1] http://www.w3.org/Consortium/Legal/2008/04-testsuite-license
# [2] http://www.w3.org/Consortium/Legal/2008/03-bsd-license
# [3] http://www.w3.org/2004/10/27-testcases

import os
import os.path
import re
import shutil
import sys


__author__ = 'Yuta Kitamura <[email protected]>'


# The name of output directory.
# NOTE: If a directory with the same name already exists, the content will be
# REMOVED.
OUTPUT_DIRECTORY = 'out'


# <tuple indicating section hierarchy> => <section title>
SECTION_TITLES = {
( 4,): 'shadow-trees',
( 4, 1): 'upper-boundary-encapsulation',
( 4, 2): 'lower-boundary-encapsulation',
( 4, 3): 'satisfying-matching-criteria',
( 4, 4): 'distributed-pseudo-element',
( 4, 5): 'hosting-multiple-shadow-trees',
( 4, 6): 'reprojection',
( 4, 7): 'composition',
( 4, 8): 'nested-shadow-trees',
( 4, 9): 'rendering-shadow-trees',
( 4, 10): 'custom-pseudo-elements',
( 5,): 'events',
( 5, 1): 'event-retargeting',
( 5, 2): 'retargeting-relatedtarget',
( 5, 3): 'retargeting-focus-events',
( 5, 4): 'events-that-are-always-stopped',
( 5, 5): 'event-dispatch',
( 5, 6): 'event-retargeting-example',
( 6,): 'styles',
( 6, 1): 'css-variables',
( 6, 2): 'text-decoration-property',
( 6, 3): 'at-host-at-rule',
( 7,): 'user-interaction',
( 7, 1): 'ranges-and-selections',
( 7, 2): 'focus-navigation',
( 7, 3): 'active-element',
( 7, 4): 'editing',
( 7, 5): 'assistive-technology',
( 8,): 'html-elements-in-shadow-trees',
( 8, 1): 'inert-html-elements',
( 8, 2): 'html-forms',
( 9,): 'html-elements-and-their-shadow-trees',
(10,): 'elements-and-dom-objects',
(10, 1): 'shadowroot-object',
(10, 1, 1): 'shadowroot-attributes',
(10, 1, 2): 'shadowroot-methods',
(10, 2): 'extensions-to-element-interface',
(10, 2, 1): 'attributes',
(10, 2, 2): 'methods',
(10, 3): 'csshostrule-interface',
(10, 3, 1): 'csshostrule-attributes',
(10, 3, 2): 'csshostrule-methods',
(10, 4): 'the-content-html-element',
(10, 5): 'the-shadow-html-element',
}


HTML_TEMPLATE = """<!DOCTYPE html>
<!--
Distributed under both the W3C Test Suite License [1] and the W3C
3-clause BSD License [2]. To contribute to a W3C Test Suite, see the
policies and contribution forms [3].
[1] http://www.w3.org/Consortium/Legal/2008/04-testsuite-license
[2] http://www.w3.org/Consortium/Legal/2008/03-bsd-license
[3] http://www.w3.org/2004/10/27-testcases
-->
<html>
<head>
<title>Shadow DOM Test: {title}</title>
{authors}<link rel="help" href="{help}">
<meta name="assert" content="{assert_}">
<script src="/resources/testharness.js"></script>
<script src="/resources/testharnessreport.js"></script>
<script src="{root_dir}/testcommon.js"></script>
<link rel="stylesheet" href="/resources/testharness.css">
</head>
<body>
<div id="log"></div>
<script>
{test_content}
</script>
</body>
</html>
"""


AUTHOR_TEMPLATE = '<link rel="author" title="{name}" href="mailto:{email}">\n'


def Main():
CreateOutputDirectory()

test_dir_pattern = re.compile(r'^\d\d_')
for path in sorted(os.listdir('.')):
if os.path.isdir(path) and test_dir_pattern.match(path):
RunTestDirectory(path)

files_to_copy = ['LICENSE', 'testcommon.js']
dirs_to_copy = ['resources']

for file in files_to_copy:
print('Copying file {}...'.format(file))
shutil.copy2(file, os.path.join(OUTPUT_DIRECTORY, file))

for dir in dirs_to_copy:
print('Copying directory {}...'.format(dir))
shutil.copytree(dir, os.path.join(OUTPUT_DIRECTORY, dir))

print('Done!')


def CreateOutputDirectory():
if os.path.exists(OUTPUT_DIRECTORY):
if not os.path.islink(OUTPUT_DIRECTORY) and os.path.isdir(OUTPUT_DIRECTORY):
shutil.rmtree(OUTPUT_DIRECTORY)
else:
print('Path "{}" exists and is not a directory.'.format(OUTPUT_DIRECTORY),
file=sys.stderrr)
sys.exit(1)

os.mkdir(OUTPUT_DIRECTORY)


def RunTestDirectory(src_dir):
section_path = ExtractSectionPath(src_dir)
depth = len(section_path)
dir_components = []
for i in range(depth):
component = '{:02d}-{}'.format(section_path[i],
SECTION_TITLES[section_path[:(i + 1)]])
dir_components.append(component)
dest_dir = os.path.join(OUTPUT_DIRECTORY, *dir_components)

os.makedirs(dest_dir, exist_ok=True)

# TODO(yutak): Should we take care of <test>_<subtest> structure?
# Might be better to look into after mass conversion.
serial = 1
for src_name in sorted(os.listdir(src_dir)):
if os.path.splitext(src_name)[1] == '.js':
# This test contains no tests. Let's skip it, and don't even bother
# assigning a test serial.
if src_name == 'A_06_00_05.js':
continue

dest_name = 'test-{:03d}.html'.format(serial)
RunTestFile(src_dir, src_name, dest_dir, dest_name, depth)
serial += 1


def ExtractSectionPath(path):
section_path = []
prefix_pattern = re.compile(r'^(\d\d)_')
while True:
match = prefix_pattern.match(path)
if not match:
break
section_num = int(match.group(1))
if section_num > 0:
section_path.append(section_num)
path = path[3:]
return tuple(section_path)


def RunTestFile(src_dir, src_name, dest_dir, dest_name, depth):
src_path = os.path.join(src_dir, src_name)
dest_path = os.path.join(dest_dir, dest_name)
root_dir = ('../' * depth)[:-1]

print('{} =>\n {}'.format(src_path, dest_path))

with open(src_path, 'rt') as src_fp, open(dest_path, 'wt') as dest_fp:
src = src_fp.read()
prologue_regex = re.compile(r'^var A_[0-9_]+ = ({$(?:.|\n)*?^});$',
flags=re.MULTILINE)
match = prologue_regex.search(src)
assert match

# HACK! Let's make the JavaScript mapping into something that Python can
# understand.
prologue_str = re.sub(r'^\s*([a-z]+):', r'"\1":', match.group(1),
flags=re.MULTILINE)

# A_04_06_01.js contains "//TODO" comment in the prologue block.
prologue_str = re.sub(r'^\s*//', r'#', prologue_str, flags=re.MULTILINE)

prologue = eval(prologue_str)

# Some tests have broken "author" field (missing email). Fix it up here.
src = re.sub(r'author:\'Aleksei Yu\. Semenov\'',
'author: \'Aleksei Yu. Semenov <[email protected]>\'',
src)

# Extract the authors.
author_regex = re.compile(r'author\s*:\s*[\'"](.+?)\s*<(.+)>[\'"]')
authors = set(author_regex.findall(src))
authors_html = ''.join([AUTHOR_TEMPLATE.format(name=name, email=email)
for name, email in authors])

# Some tests have duplicate test names, and that is rendered as an error
# with the latest test harness. Fix those by appending a suffix.
# TODO(yutak): More readable name is desirable, but I'll probably leave that
# as a post-mass-conversion task.
known_dup_tests = ['A_04_00_01_T12',
'A_10_01_01_03_02_T02',
'A_10_01_01_04_02_T01',
'A_10_04_02_T01']
for test_name in known_dup_tests:
count = src.count(test_name)
if count == 0:
continue
# Make sure the name occurs exactly twice.
assert count == 2

src = src.replace(test_name, test_name + '_02')
src = src.replace(test_name + '_02', test_name + '_01', 1)

# Some tests are known to be slow. Give these tests more time to run.
known_slow_tests = ['A_08_01_01_T01', 'A_08_02_03_T01']

# Remove unnecessary part of the code.
license_comment_regex = re.compile(
r'/\*\s*Distributed under both the W3C[^*]+\*/',
flags=re.MULTILINE)

src = license_comment_regex.sub('', src)
src = prologue_regex.sub('', src)

test_name_and_props_regex = re.compile(
r'\'([^\']+)\'\s*,\s*PROPS\([^\)]+\)',
flags=re.MULTILINE)
def DropProps(match):
if match.group(1) in known_slow_tests:
return "'{}', {{ timeout: 5000 }}".format(match.group(1))
return "'{}'".format(match.group(1))
src = test_name_and_props_regex.sub(DropProps, src)

# Some tests puts utility functions in an object whose name is same as
# the test name.
tests_wanting_object = ['A_04_01_01.js', 'A_04_01_09.js']
if src_name in tests_wanting_object:
src = 'var {} = new Object();\n\n{}'.format(src_name[:-3], src.lstrip())

# Update the relative URIs.
src = re.sub(r'resources/', root_dir + '/resources/', src)

# Some tests don't have "link" metadata!
if src_name.startswith('A_05_02_'):
prologue['link'] = ('http://www.w3.org/TR/2013/WD-shadow-dom-201305214'
'#retargeting-related-target')
elif src_name.startswith('A_05_03_'):
prologue['link'] = ('http://www.w3.org/TR/2013/WD-shadow-dom-20130514/'
'#retargeting-focus-events')
elif src_name.startswith('A_05_04_'):
prologue['link'] = ('http://www.w3.org/TR/2013/WD-shadow-dom-20130514/'
'#events-that-are-always-stopped')
elif src_name.startswith('A_05_05_'):
prologue['link'] = ('http://www.w3.org/TR/2013/WD-shadow-dom-20130514/'
'#event-dispatch')
else:
# Otherwise, adjust the reference so it points to a snapshot of the
# specification, not the latest one.
help_regex = re.compile(r'^[^#]+#(.+)$')
match = help_regex.match(prologue['link'])
assert match
prologue['link'] = (
'http://www.w3.org/TR/2013/WD-shadow-dom-20130514/#{}'.format(
match.group(1)))

# Remove extra whitespaces.
line_end_whitespaces_regex = re.compile(r'[ \t]+$', flags=re.MULTILINE)
src = line_end_whitespaces_regex.sub('', src)
src = src.strip()

dest_fp.write(HTML_TEMPLATE.format(
root_dir=root_dir,
title=src_name[:-3], # TODO(yutak): Needs better title.
authors=authors_html,
help=prologue['link'],
assert_=prologue['assert'],
test_content=src))

if __name__ == '__main__':
Main()

0 comments on commit 1bcae1d

Please sign in to comment.