def test_date_comparison(self):
target_filepath = get_temporary_file(directory="/tmp")
file_last_modified = get_file_modified_date(target_filepath)
comparison_date = "2013-07-18T16:31:46-0400"
self.assertFalse(compare_datetime_to_iso8601_date(file_last_modified, comparison_date))
comparison_date = "2013-07-18T16:31:46Z"
self.assertFalse(compare_datetime_to_iso8601_date(file_last_modified, comparison_date))
def test_date_comparison(self):
target_filepath = get_temporary_file(directory="/tmp")
file_last_modified = get_file_modified_date(target_filepath)
comparison_date = "2013-07-18T16:31:46-0400"
self.assertFalse(
compare_datetime_to_iso8601_date(file_last_modified,
comparison_date))
comparison_date = "2013-07-18T16:31:46Z"
self.assertFalse(
compare_datetime_to_iso8601_date(file_last_modified,
comparison_date))
def perform_fulltext_harvest(self, record_list, parameters):
"""
For every record in given list APSRecord(record ID, DOI, date last
updated), yield a APSRecord with added FFT dictionary containing URL to
fulltext/metadata XML downloaded locally.
If a download is unsuccessful, an error message is given.
@return: tuple of (APSRecord, error_message)
"""
count = 0
request_end = None
request_start = None
for record in record_list:
task_sleep_now_if_required(can_stop_too=False)
# Unless this is the first request, lets sleep a bit
if request_end and request_start:
request_dt = request_end-request_start
write_message("Checking request time (%d)"
% (request_dt,), verbose=3)
if count and request_dt > 0 and request_dt < CFG_APSHARVEST_REQUEST_TIMEOUT:
write_message("Initiating sleep for %.1f seconds"
% (request_dt,), verbose=3)
time.sleep(request_dt)
count += 1
task_update_progress("Harvesting record (%d/%d)" % (count,
len(record_list)))
if not record.doi:
msg = "No DOI found for record %d" % (record.recid or "",)
write_message("Error: %s" % (msg,), stream=sys.stderr)
yield record, msg
continue
url = CFG_APSHARVEST_FULLTEXT_URL % {'doi': record.doi}
result_file = os.path.join(self.zip_folder,
"%s.zip" % (record.doi.replace('/', '_')))
try:
request_start = time.time()
if os.path.exists(result_file):
# File already downloaded recently, lets see if it is the same
file_last_modified = get_file_modified_date(result_file)
if record.last_modified and not compare_datetime_to_iso8601_date(file_last_modified, record.last_modified):
# File is not older than APS version, we should not download.
raise APSHarvesterFileExits
write_message("Trying to save to %s" % (result_file,), verbose=5)
result_file = download_url(url=url,
download_to_file=result_file,
content_type="zip",
accept="application/zip",
retry_count=5,
timeout=60.0)
write_message("Downloaded %s to %s" % (url, result_file), verbose=2)
except InvenioFileDownloadError, e:
msg = "URL could not be opened: %s" % (url,)
write_message("Error: %s" % (msg,),
stream=sys.stderr)
yield record, msg
continue
except APSHarvesterFileExits:
write_message("File exists at %s" % (result_file,), verbose=2)
def perform_fulltext_harvest(self, record_list, parameters):
"""
For every record in given list APSRecord(record ID, DOI, date last
updated), yield a APSRecord with added FFT dictionary containing URL to
fulltext/metadata XML downloaded locally.
If a download is unsuccessful, an error message is given.
@return: tuple of (APSRecord, error_message)
"""
count = 0
request_end = None
request_start = None
for record in record_list:
task_sleep_now_if_required(can_stop_too=False)
# Unless this is the first request, lets sleep a bit
if request_end and request_start:
request_dt = request_end-request_start
write_message("Checking request time (%d)"
% (request_dt,), verbose=3)
if count and request_dt > 0 and request_dt < CFG_APSHARVEST_REQUEST_TIMEOUT:
write_message("Initiating sleep for %.1f seconds"
% (request_dt,), verbose=3)
time.sleep(request_dt)
count += 1
task_update_progress("Harvesting record (%d/%d)" % (count,
len(record_list)))
if not record.doi:
msg = "No DOI found for record %d" % (record.recid or "",)
write_message("Error: %s" % (msg,), stream=sys.stderr)
yield record, msg
continue
url = CFG_APSHARVEST_FULLTEXT_URL % {'doi': record.doi}
result_file = os.path.join(self.out_folder,
"%s.zip" % (record.doi.replace('/', '_')))
try:
request_start = time.time()
if os.path.exists(result_file):
# File already downloaded recently, lets see if it is the same
file_last_modified = get_file_modified_date(result_file)
if not compare_datetime_to_iso8601_date(file_last_modified, record.last_modified):
# File is not older than APS version, we should not download.
raise APSHarvesterFileExits
write_message("Trying to save to %s" % (result_file,), verbose=5)
result_file = download_url(url=url,
download_to_file=result_file,
content_type="zip",
retry_count=5,
timeout=60.0)
write_message("Downloaded %s to %s" % (url, result_file), verbose=2)
except InvenioFileDownloadError, e:
msg = "URL could not be opened: %s" % (url,)
write_message("Error: %s" % (msg,),
stream=sys.stderr)
yield record, msg
continue
except APSHarvesterFileExits:
write_message("File exists at %s" % (result_file,), verbose=2)