mirror of
https://github.com/internetarchive/brozzler.git
synced 2025-02-23 16:19:49 -05:00

1. If thread is not currently accepting exceptions, queue it and raise if and when it does start accepting them. This fixes problem of thread_raise exceptions being ignored when raised just before the target thread starts accepting exceptions. 2. Avoid problems caused by raising multiple exceptions in the same thread in quick succession by ensuring that only one is actually raised for a given `with` block. This type of occurrence had been putting brozzler into a borked/frozen state.
97 lines
3.6 KiB
Python
97 lines
3.6 KiB
Python
#!/usr/bin/env python
|
|
'''
|
|
setup.py - brozzler setup script
|
|
|
|
Copyright (C) 2014-2017 Internet Archive
|
|
|
|
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 setuptools
|
|
import os
|
|
|
|
def find_package_data(package):
|
|
pkg_data = []
|
|
depth = len(package.split('.'))
|
|
path = os.path.join(*package.split('.'))
|
|
for dirpath, dirnames, filenames in os.walk(path):
|
|
if not os.path.exists(os.path.join(dirpath, '__init__.py')):
|
|
relpath = os.path.join(*dirpath.split(os.sep)[depth:])
|
|
pkg_data.extend(os.path.join(relpath, f) for f in filenames)
|
|
return pkg_data
|
|
|
|
setuptools.setup(
|
|
name='brozzler',
|
|
version='1.1b11.dev248',
|
|
description='Distributed web crawling with browsers',
|
|
url='https://github.com/internetarchive/brozzler',
|
|
author='Noah Levitt',
|
|
author_email='nlevitt@archive.org',
|
|
long_description=open('README.rst', mode='rb').read().decode('UTF-8'),
|
|
license='Apache License 2.0',
|
|
packages=['brozzler', 'brozzler.dashboard'],
|
|
package_data={
|
|
'brozzler': [
|
|
'js-templates/*.js*', 'behaviors.yaml', 'job_schema.yaml'],
|
|
'brozzler.dashboard': find_package_data('brozzler.dashboard'),
|
|
},
|
|
entry_points={
|
|
'console_scripts': [
|
|
'brozzle-page=brozzler.cli:brozzle_page',
|
|
'brozzler-new-job=brozzler.cli:brozzler_new_job',
|
|
'brozzler-new-site=brozzler.cli:brozzler_new_site',
|
|
'brozzler-worker=brozzler.cli:brozzler_worker',
|
|
'brozzler-ensure-tables=brozzler.cli:brozzler_ensure_tables',
|
|
'brozzler-list-captures=brozzler.cli:brozzler_list_captures',
|
|
'brozzler-list-jobs=brozzler.cli:brozzler_list_jobs',
|
|
'brozzler-list-sites=brozzler.cli:brozzler_list_sites',
|
|
'brozzler-list-pages=brozzler.cli:brozzler_list_pages',
|
|
'brozzler-stop-crawl=brozzler.cli:brozzler_stop_crawl',
|
|
'brozzler-dashboard=brozzler.dashboard:main',
|
|
'brozzler-easy=brozzler.easy:main',
|
|
'brozzler-wayback=brozzler.pywb:main',
|
|
],
|
|
},
|
|
install_requires=[
|
|
'PyYAML',
|
|
'youtube-dl',
|
|
'reppy==0.3.4',
|
|
'requests',
|
|
'websocket-client!=0.39.0',
|
|
'pillow==3.3.0',
|
|
'urlcanon>=0.1.dev16',
|
|
'doublethink>=0.2.0.dev71',
|
|
'rethinkdb>=2.3,<2.4',
|
|
'cerberus==1.0.1',
|
|
'jinja2',
|
|
],
|
|
extras_require={
|
|
'dashboard': ['flask>=0.11', 'gunicorn'],
|
|
'easy': [
|
|
'warcprox>=2.1b1.dev60',
|
|
'pywb',
|
|
'flask>=0.11',
|
|
'gunicorn'
|
|
],
|
|
},
|
|
zip_safe=False,
|
|
classifiers=[
|
|
'Development Status :: 4 - Beta',
|
|
'Environment :: Console',
|
|
'License :: OSI Approved :: Apache Software License',
|
|
'Programming Language :: Python :: 3.4',
|
|
'Programming Language :: Python :: 3.5',
|
|
'Topic :: Internet :: WWW/HTTP',
|
|
'Topic :: System :: Archiving',
|
|
])
|