blob: f4f30d060096a0c073162098b01546d6a035264e [file] [log] [blame]
#!/usr/bin/env python3
# Copyright (c) 2023 Project CHIP Authors
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
import glob
import os
import re
import subprocess
import click
DEFAULT_CHIP_ROOT = os.path.abspath(os.path.join(os.path.dirname(__file__), '..', '..'))
DEFAULT_OUTPUT_DIR = os.path.abspath(os.path.join(DEFAULT_CHIP_ROOT, 'data_model'))
def get_xml_path(filename, output_dir):
xml = os.path.basename(filename).replace('.adoc', '.xml')
return os.path.abspath(os.path.join(output_dir, xml))
@click.command()
@click.option(
'--scraper',
required=True,
type=str,
help='Path to the location of the scraper tool')
@click.option(
'--spec-root',
required=True,
type=str,
help='Path to the spec root')
@click.option(
'--output-dir',
default=DEFAULT_OUTPUT_DIR,
help='Path to output xml files')
@click.option(
'--dry-run',
default=False,
is_flag=True,
help='Flag for dry run')
def main(scraper, spec_root, output_dir, dry_run):
# Clusters need to be scraped first because the cluster directory is passed to the device type directory
scrape_clusters(scraper, spec_root, output_dir, dry_run)
scrape_device_types(scraper, spec_root, output_dir, dry_run)
if not dry_run:
dump_versions(scraper, spec_root, output_dir)
def scrape_clusters(scraper, spec_root, output_dir, dry_run):
src_dir = os.path.abspath(os.path.join(spec_root, 'src'))
sdm_clusters_dir = os.path.abspath(os.path.join(src_dir, 'service_device_management'))
app_clusters_dir = os.path.abspath(os.path.join(src_dir, 'app_clusters'))
dm_clusters_dir = os.path.abspath(os.path.join(src_dir, 'data_model'))
media_clusters_dir = os.path.abspath(os.path.join(app_clusters_dir, 'media'))
clusters_output_dir = os.path.abspath(os.path.join(output_dir, 'clusters'))
dm_clusters_list = ['ACL-Cluster.adoc', 'Binding-Cluster.adoc', 'bridge-clusters.adoc',
'Descriptor-Cluster.adoc', 'Group-Key-Management-Cluster.adoc', 'ICDManagement.adoc',
'Label-Cluster.adoc']
sdm_exclude_list = ['AdminAssistedCommissioningFlows.adoc', 'BulkDataExchange.adoc', 'CommissioningFlows.adoc',
'DeviceCommissioningFlows.adoc', 'DistributedComplianceLedger.adoc', 'OTAFileFormat.adoc']
app_exclude_list = ['appliances.adoc', 'closures.adoc', 'general.adoc',
'hvac.adoc', 'lighting.adoc', 'meas_and_sense.adoc', 'robots.adoc']
media_exclude_list = ['media.adoc', 'VideoPlayerArchitecture.adoc']
if not os.path.exists(clusters_output_dir):
os.makedirs(clusters_output_dir)
def scrape_cluster(filename: str) -> None:
xml_path = get_xml_path(filename, clusters_output_dir)
cmd = [scraper, 'cluster', '-i', filename, '-o', xml_path, '-nd']
if dry_run:
print(cmd)
else:
subprocess.run(cmd)
def scrape_all_clusters(dir: str, exclude_list: list[str] = []) -> None:
for filename in glob.glob(f'{dir}/*.adoc'):
if os.path.basename(filename) in exclude_list:
continue
scrape_cluster(filename)
scrape_all_clusters(sdm_clusters_dir, sdm_exclude_list)
scrape_all_clusters(app_clusters_dir, app_exclude_list)
scrape_all_clusters(media_clusters_dir, media_exclude_list)
for f in dm_clusters_list:
filename = f'{dm_clusters_dir}/{f}'
scrape_cluster(filename)
def scrape_device_types(scraper, spec_root, output_dir, dry_run):
device_type_dir = os.path.abspath(os.path.join(spec_root, 'src', 'device_types'))
device_types_output_dir = os.path.abspath(os.path.join(output_dir, 'device_types'))
clusters_output_dir = os.path.abspath(os.path.join(output_dir, 'clusters'))
if not os.path.exists(device_types_output_dir):
os.makedirs(device_types_output_dir)
def scrape_device_type(filename: str) -> None:
xml_path = get_xml_path(filename, device_types_output_dir)
cmd = [scraper, 'devicetype', '-c', clusters_output_dir, '-nd', '-i', filename, '-o', xml_path]
if dry_run:
print(cmd)
else:
print(' '.join(cmd))
subprocess.run(cmd)
exclude_list = [r"section_*"]
for filename in glob.glob(f'{device_type_dir}/*.adoc'):
for exclude in exclude_list:
if not re.match(exclude, os.path.basename(filename)):
scrape_device_type(filename)
def dump_versions(scraper, spec_root, output_dir):
sha_file = os.path.abspath(os.path.join(output_dir, 'spec_sha'))
out = subprocess.run(['git', 'rev-parse', 'HEAD'], capture_output=True, encoding="utf8", cwd=spec_root)
sha = out.stdout
with open(sha_file, 'wt', encoding='utf8') as output:
output.write(sha)
scraper_file = os.path.abspath(os.path.join(output_dir, 'scraper_version'))
out = subprocess.run([scraper, '--version'], capture_output=True, encoding="utf8")
version = out.stdout
with open(scraper_file, "wt", encoding='utf8') as output:
output.write(version)
if __name__ == '__main__':
main()