def test_queue():
'Test creation, modification and download of URL queues.'
# test conversion and storage
inputdict = add_to_compressed_dict(['ftps://www.example.org/'])
assert inputdict == dict()
inputdict = add_to_compressed_dict(['https://www.example.org/'])
# CLI args
testargs = ['', '--list']
with patch.object(sys, 'argv', testargs):
args = parse_args(testargs)
assert url_processing_pipeline(args, inputdict) is None
# single/multiprocessing
testargs = ['', '-v']
with patch.object(sys, 'argv', testargs):
args = parse_args(testargs)
domain_dict = dict()
domain_dict['https://httpbin.org'] = deque([
'/status/301', '/status/304', '/status/200', '/status/300',
'/status/400', '/status/505'
])
args.archived = True
args.config_file = os.path.join(RESOURCES_DIR, 'newsettings.cfg')
config = use_config(filename=args.config_file)
results = download_queue_processing(domain_dict, args, None, config)
assert len(results[0]) == 6 and results[1] is None
# test backoff algorithm
testdict = dict()
backoffdict = dict()
testdict['http://test.org'] = deque(['/1'])
assert draw_backoff_url(testdict, backoffdict, 0,
set()) == ('http://test.org/1', dict(), dict(),
'http://test.org')
testdict['http://test.org'] = deque(['/1'])
backoffdict['http://test.org'] = datetime(2019, 5, 18, 15, 17, 8, 132263)
assert draw_backoff_url(testdict, backoffdict, 0,
set()) == ('http://test.org/1', dict(), dict(),
'http://test.org')
# code hangs, logical:
#testdict['http://test.org'] = deque(['/1'])
#backoffdict['http://test.org'] = datetime(2030, 5, 18, 15, 17, 8, 132263)
#assert cli_utils.draw_backoff_url(testdict, backoffdict, 0, 3) == ('http://test.org/1', dict(), dict(), 0)
# download buffer
domain_dict = {
'https://test.org': deque(['/1', '/2', '/3']),
'https://test2.org': deque(['/1', '/2', '/3']),
'https://test3.org': deque(['/1', '/2', '/3']),
'https://test4.org': deque(['/1', '/2', '/3']),
'https://test5.org': deque(['/1', '/2', '/3']),
'https://test6.org': deque(['/1', '/2', '/3'])
}
bufferlist, _, _, _ = load_download_buffer(domain_dict,
dict(),
0,
threads=1)
assert len(bufferlist) == 6
bufferlist, _, _, _ = load_download_buffer(domain_dict,
dict(),
0,
threads=2)
assert len(bufferlist) == 6
def test_climain():
'''test arguments and main CLI entrypoint'''
# exit status required: 0
# Windows platforms
if os.name == 'nt':
trafilatura_bin = os.path.join(sys.prefix, "Scripts", "trafilatura")
# other platforms
else:
trafilatura_bin = 'trafilatura'
# help display
assert subprocess.run([trafilatura_bin, '--help']).returncode == 0
# piped input
empty_input = b'<html><body></body></html>'
assert subprocess.run([trafilatura_bin], input=empty_input).returncode == 0
# input directory walking and processing
env = os.environ.copy()
if os.name == 'nt':
# Force encoding to utf-8 for Windows (seem to be a problem only in GitHub Actions)
env['PYTHONIOENCODING'] = 'utf-8'
assert subprocess.run([trafilatura_bin, '--inputdir', RESOURCES_DIR],
env=env).returncode == 0
# dump urls
inputdict = add_to_compressed_dict(['https://www.example.org'])
f = io.StringIO()
with redirect_stdout(f):
cli.dump_on_exit(inputdict)
assert f.getvalue() == 'todo: https://www.example.org/\n'
def test_input_filtering():
'''test internal functions to filter urls'''
testargs = ['']
with patch.object(sys, 'argv', testargs):
args = cli.parse_args(testargs)
# load dictionary
args.inputfile = os.path.join(RESOURCES_DIR, 'list-process.txt')
inputdict = cli.load_input_dict(args)
assert inputdict['https://httpbin.org'] == deque(['/status/200', '/status/404'])
args.inputfile = os.path.join(RESOURCES_DIR, 'list-process.txt')
args.blacklist = {'httpbin.org/status/404'}
inputdict = cli.load_input_dict(args)
assert inputdict['https://httpbin.org'] == deque(['/status/200'])
# deduplication and filtering
myinput = ['https://example.org/1', 'https://example.org/2', 'https://example.org/2', 'https://example.org/3', 'https://example.org/4', 'https://example.org/5', 'https://example.org/6']
myblacklist = {'example.org/1', 'example.org/3', 'example.org/5'}
inputdict = add_to_compressed_dict(myinput, myblacklist)
assert inputdict['https://example.org'] == deque(['/2', '/4', '/6'])
# URL in blacklist
args.inputfile = os.path.join(RESOURCES_DIR, 'list-process.txt')
my_urls = cli_utils.load_input_urls(args)
my_blacklist = cli_utils.load_blacklist(os.path.join(RESOURCES_DIR, 'list-discard.txt'))
inputdict = add_to_compressed_dict(my_urls, my_blacklist)
assert len(inputdict) == 0
# URL filter
args.inputfile = os.path.join(RESOURCES_DIR, 'list-process.txt')
my_urls = cli_utils.load_input_urls(args)
assert len(add_to_compressed_dict(my_urls, None, ['status'], None)) == 1
assert len(add_to_compressed_dict(my_urls, None, ['teststring'], None)) == 0
assert len(add_to_compressed_dict(my_urls, None, ['status', 'teststring'], None)) == 1
# malformed URLs
inputdict = add_to_compressed_dict(['123345', 'https://www.example.org/1'], {}, None, None)
assert len(inputdict) == 1
def test_cli_pipeline():
'''test command-line processing pipeline'''
# straight command-line input
#testargs = ['', '<html><body>Text</body></html>']
#with patch.object(sys, 'argv', testargs):
# args = cli.parse_args(testargs)
#f = io.StringIO()
#with redirect_stdout(f):
# cli.process_args(args)
#assert len(f.getvalue()) == 0
# test URL listing
# Force encoding to utf-8 for Windows in future processes spawned by multiprocessing.Pool
os.environ['PYTHONIOENCODING'] = "utf-8"
testargs = ['', '--list']
with patch.object(sys, 'argv', testargs):
args = cli.parse_args(testargs)
assert cli_utils.url_processing_pipeline(args, dict()) is None
# test inputlist + blacklist
testargs = ['', '-i', os.path.join(RESOURCES_DIR, 'list-process.txt')]
with patch.object(sys, 'argv', testargs):
args = cli.parse_args(testargs)
my_urls = cli_utils.load_input_urls(args)
assert my_urls is not None and len(my_urls) == 2
testargs = [
'', '-i',
os.path.join(RESOURCES_DIR, 'list-process.txt'), '--blacklist',
os.path.join(RESOURCES_DIR, 'list-discard.txt'), '--archived'
]
with patch.object(sys, 'argv', testargs):
args = cli.parse_args(testargs)
assert args.blacklist is not None
# test backoff between domain requests
inputdict = add_to_compressed_dict(my_urls, args.blacklist, None, None)
reftime = datetime.now()
cli_utils.url_processing_pipeline(args, inputdict)
delta = (datetime.now() - reftime).total_seconds()
assert delta > 2
# test blacklist and empty dict
args.blacklist = cli_utils.load_blacklist(args.blacklist)
assert len(args.blacklist) == 2
inputdict = add_to_compressed_dict(my_urls, args.blacklist, None, None)
cli_utils.url_processing_pipeline(args, inputdict)
# test backup
testargs = ['', '--backup-dir', '/tmp/']
with patch.object(sys, 'argv', testargs):
args = cli.parse_args(testargs)
cli_utils.archive_html('00Test', args)
# test date-based exclusion
testargs = ['', '-out', 'xml', '--with-metadata']
with patch.object(sys, 'argv', testargs):
args = cli.parse_args(testargs)
with open(os.path.join(RESOURCES_DIR, 'httpbin_sample.html'), 'r') as f:
teststring = f.read()
assert cli.examine(teststring, args) is None
testargs = ['', '-out', 'xml', '--only-with-metadata', '--precision']
with patch.object(sys, 'argv', testargs):
args = cli.parse_args(testargs)
with open(os.path.join(RESOURCES_DIR, 'httpbin_sample.html'), 'r') as f:
teststring = f.read()
assert cli.examine(teststring, args) is None
# test JSON output
testargs = ['', '-out', 'json', '--recall']
with patch.object(sys, 'argv', testargs):
args = cli.parse_args(testargs)
with open(os.path.join(RESOURCES_DIR, 'httpbin_sample.html'), 'r') as f:
teststring = f.read()
assert cli.examine(teststring, args) is not None
# dry-run file processing pipeline
testargs = ['', '--parallel', '1', '--inputdir', '/dev/null']
with patch.object(sys, 'argv', testargs):
args = cli.parse_args(testargs)
cli_utils.file_processing_pipeline(args)
# file processing pipeline on resources/
args.inputdir = RESOURCES_DIR
cli_utils.file_processing_pipeline(args)
# sitemaps
testargs = ['', '--sitemap', 'https://httpbin.org/', '--list']
with patch.object(sys, 'argv', testargs):
args = cli.parse_args(testargs)
f = io.StringIO()
with redirect_stdout(f):
cli.process_args(args)
assert len(f.getvalue()) == 0
# config file
testargs = [
'', '--inputdir', '/dev/null', '--config-file', 'newsettings.cfg'
]
with patch.object(sys, 'argv', testargs):
args = cli.parse_args(testargs)
with open(os.path.join(RESOURCES_DIR, 'httpbin_sample.html'), 'r') as f:
teststring = f.read()
args.config_file = os.path.join(RESOURCES_DIR, args.config_file)
# config = use_config(filename=args.config_file)
assert cli.examine(teststring, args) is None
# CLI options
testargs = ['', '--links', '--images']
with patch.object(sys, 'argv', testargs):
args = cli.parse_args(testargs)
#with open(os.path.join(RESOURCES_DIR, 'http_sample.html'), 'r') as f:
# teststring = f.read()
#result = cli.examine(teststring, args)
#assert '[link](testlink.html)' in result # and 'test.jpg' in result
# Crawling
testargs = ['', '--crawl', 'https://httpbin.org/html']
with patch.object(sys, 'argv', testargs):
args = cli.parse_args(testargs)
f = io.StringIO()
with redirect_stdout(f):
cli_utils.cli_crawler(args)
assert len(f.getvalue()) == 0
# links permitted
testargs = [
'', '--crawl', 'https://httpbin.org/links/1/1', '--list', '--parallel',
'1'
]
with patch.object(sys, 'argv', testargs):
args = cli.parse_args(testargs)
f = io.StringIO()
with redirect_stdout(f):
cli_utils.cli_crawler(args)
assert f.getvalue() == 'https://httpbin.org/links/1/0\n'
# 0 links permitted
args.crawl = 'https://httpbin.org/links/4/4'
f = io.StringIO()
with redirect_stdout(f):
cli_utils.cli_crawler(args, n=0)
# print(f.getvalue())
assert len(f.getvalue().split('\n')) == 5
# Exploration (Sitemap + Crawl)
testargs = ['', '--explore', 'https://httpbin.org/html']
with patch.object(sys, 'argv', testargs):
args = cli.parse_args(testargs)
f = io.StringIO()
with redirect_stdout(f):
cli.process_args(args)
assert len(f.getvalue()) == 0
def test_queue():
'Test creation, modification and download of URL queues.'
# test conversion and storage
inputdict = add_to_compressed_dict(['ftps://www.example.org/', 'http://'])
assert inputdict == dict()
inputdict = add_to_compressed_dict(['https://www.example.org/'])
# CLI args
testargs = ['', '--list']
with patch.object(sys, 'argv', testargs):
args = parse_args(testargs)
assert url_processing_pipeline(args, inputdict) is None
# single/multiprocessing
testargs = ['', '-v']
with patch.object(sys, 'argv', testargs):
args = parse_args(testargs)
domain_dict = {
'https://httpbin.org': deque(
[
'/status/301',
'/status/304',
'/status/200',
'/status/300',
'/status/400',
'/status/505',
]
)
}
args.archived = True
args.config_file = os.path.join(RESOURCES_DIR, 'newsettings.cfg')
config = use_config(filename=args.config_file)
config['DEFAULT']['SLEEP_TIME'] = '0.2'
results = download_queue_processing(domain_dict, args, None, config)
## fixed: /301 missing, probably for a good reason...
assert len(results[0]) == 5 and results[1] is None
# test backoff algorithm
backoffdict = {}
testdict = {'http://test.org': deque(['/1'])}
assert draw_backoff_url(testdict, backoffdict, 0) == ('http://test.org/1', dict(), dict())
testdict['http://test.org'] = deque(['/1'])
backoffdict['http://test.org'] = datetime(2019, 5, 18, 15, 17, 8, 132263)
assert draw_backoff_url(testdict, backoffdict, 0) == ('http://test.org/1', dict(), dict())
# concurrent domains
testdict = {}
backoffdict = {}
testdict['http://test.org'] = deque(['/1'])
testdict['http://example.org'] = deque(['/1'])
# simulate recent request
backoffdict['http://test.org'] = datetime.now()
# must return the other domain
test = draw_backoff_url(testdict, backoffdict, 5)
assert test[0], test[1] == ('http://example.org/1', {'http://test.org': deque(['/1'])})
print(test)
assert test[2] != {}
# sleeps and returns the rest
assert draw_backoff_url(testdict, backoffdict, 1) == ('http://test.org/1', {}, {})
# code hangs, logical:
#testdict['http://test.org'] = deque(['/1'])
#backoffdict['http://test.org'] = datetime(2030, 5, 18, 15, 17, 8, 132263)
#assert draw_backoff_url(testdict, backoffdict, 0) == ('http://test.org/1', dict(), dict())
# download buffer
domain_dict = {'https://test.org': deque(['/1', '/2', '/3']), 'https://test2.org': deque(['/1', '/2', '/3']), 'https://test3.org': deque(['/1', '/2', '/3']), 'https://test4.org': deque(['/1', '/2', '/3']), 'https://test5.org': deque(['/1', '/2', '/3']), 'https://test6.org': deque(['/1', '/2', '/3'])}
bufferlist, _, _, _ = load_download_buffer(domain_dict, dict(), sleep_time=5, threads=1)
assert len(bufferlist) == 6
bufferlist, _, _, _ = load_download_buffer(domain_dict, dict(), sleep_time=5, threads=2)
assert len(bufferlist) == 6