def end_write(self, with_errors, isnewfile=True):
if not with_errors:
parent, filename = os.path.split(self.rel_path)
contentlength = os.stat(self.tmpfile_path).st_size
if not self.check_repo_owner_quota(isnewfile=isnewfile, contentlength=contentlength):
raise DAVError(HTTP_FORBIDDEN, "The quota of the repo owner is exceeded")
seafile_api.put_file(self.repo.id, self.tmpfile_path, parent, filename,
self.username, None)
if self.tmpfile_path:
try:
os.unlink(self.tmpfile_path)
finally:
self.tmpfile_path = None
def endWrite(self, withErrors, isnewfile=True):
if not withErrors:
parent, filename = os.path.split(self.rel_path)
contentlength = os.stat(self.tmpfile_path).st_size
if not self.check_repo_owner_quota(isnewfile=isnewfile, contentlength=contentlength):
raise DAVError(HTTP_FORBIDDEN, "The quota of the repo owner is exceeded")
seafile_api.put_file(self.repo.id, self.tmpfile_path, parent, filename,
self.username, None)
if self.tmpfile_path:
try:
os.unlink(self.tmpfile_path)
finally:
self.tmpfile_path = None
def create_file_with_content(self, file_name, parent_dir='/', content='abc',
username=''):
seafile_api.post_empty_file(self.repo.id, parent_dir, file_name, username)
# first dump the file content to a tmp file, then update the file
fd, tmp_file = mkstemp()
try:
bytesWritten = os.write(fd, 'junk content')
except:
bytesWritten = -1
finally:
os.close(fd)
assert bytesWritten > 0
seafile_api.put_file(self.repo.id, tmp_file, parent_dir, file_name,
'', None)
return parent_dir + file_name
def create_file_with_content(self, file_name, parent_dir='/', content='junk content',
username=''):
seafile_api.post_empty_file(self.repo.id, parent_dir, file_name, username)
# first dump the file content to a tmp file, then update the file
fd, tmp_file = mkstemp()
try:
bytesWritten = os.write(fd, content.encode('utf-8'))
except:
bytesWritten = -1
finally:
os.close(fd)
assert bytesWritten > 0
seafile_api.put_file(self.repo.id, tmp_file, parent_dir, file_name,
'', None)
return parent_dir + file_name
def test_file_operation():
t_repo_version = 1
t_repo_id1 = api.create_repo('test_file_operation1', '', USER, passwd=None)
create_the_file()
# test post_file
assert api.post_file(t_repo_id1, file_path, '/', file_name, USER) == 0
t_file_id = api.get_file_id_by_path(t_repo_id1, '/' + file_name)
t_file_size = len(file_content)
assert t_file_size == api.get_file_size(t_repo_id1, t_repo_version,
t_file_id)
# test post_dir
assert api.post_dir(t_repo_id1, '/', dir_name, USER) == 0
# test copy_file (synchronize)
t_copy_file_result1 = api.copy_file(t_repo_id1, '/', file_name, t_repo_id1,
'/', new_file_name, USER, 0, 1)
assert t_copy_file_result1
assert t_copy_file_result1.task_id is None
assert not t_copy_file_result1.background
t_file_id = api.get_file_id_by_path(t_repo_id1, '/' + new_file_name)
assert t_file_size == api.get_file_size(t_repo_id1, t_repo_version,
t_file_id)
# test copy_file (asynchronous)
t_repo_id2 = api.create_repo('test_file_operation2', '', USER, passwd=None)
usage = api.get_user_self_usage(USER)
api.set_user_quota(USER, usage + 1)
t_copy_file_result2 = api.copy_file(t_repo_id1, '/', file_name, t_repo_id2,
'/', file_name, USER, 1, 0)
assert t_copy_file_result2
assert t_copy_file_result2.background
while True:
time.sleep(0.1)
t_copy_task = api.get_copy_task(t_copy_file_result2.task_id)
assert t_copy_task.failed
assert t_copy_task.failed_reason == 'Quota is full'
if t_copy_task.failed:
break
api.set_user_quota(USER, -1)
t_copy_file_result2 = api.copy_file(t_repo_id1, '/', file_name, t_repo_id2,
'/', file_name, USER, 1, 0)
assert t_copy_file_result2
assert t_copy_file_result2.task_id
assert t_copy_file_result2.background
while True:
time.sleep(0.1)
t_copy_task = api.get_copy_task(t_copy_file_result2.task_id)
if t_copy_task.successful:
break
t_file_id = api.get_file_id_by_path(t_repo_id2, '/' + file_name)
assert t_file_size == api.get_file_size(t_repo_id2, t_repo_version,
t_file_id)
# test move_file (synchronize)
t_move_file_info1 = api.get_dirent_by_path(t_repo_id1, '/' + new_file_name)
t_move_file_result1 = api.move_file(t_repo_id1, '/', new_file_name,
t_repo_id1, '/' + dir_name,
new_file_name, 1, USER, 0, 1)
assert t_move_file_result1
t_move_file_info2 = api.get_dirent_by_path(
t_repo_id1, '/' + dir_name + '/' + new_file_name)
assert t_move_file_info1.mtime == t_move_file_info2.mtime
t_file_id = api.get_file_id_by_path(t_repo_id1, '/' + new_file_name)
assert t_file_id is None
# test move_file (synchronize)
t_move_file_result1 = api.move_file(t_repo_id1, '/' + dir_name,
new_file_name, t_repo_id1, '/',
new_file_name_2, 1, USER, 0, 1)
assert t_move_file_result1
t_file_id = api.get_file_id_by_path(t_repo_id1,
'/' + dir_name + '/' + new_file_name)
assert t_file_id is None
# test move_file (asynchronous)
usage = api.get_user_self_usage(USER)
api.set_user_quota(USER, usage + 1)
t_move_file_result2 = api.move_file(t_repo_id1, '/', file_name, t_repo_id2,
'/', new_file_name, 1, USER, 1, 0)
assert t_move_file_result2
assert t_move_file_result2.task_id
assert t_move_file_result2.background
while True:
time.sleep(0.1)
t_move_task = api.get_copy_task(t_move_file_result2.task_id)
assert t_move_task.failed
assert t_move_task.failed_reason == 'Quota is full'
if t_move_task.failed:
break
api.set_user_quota(USER, -1)
t_move_file_result2 = api.move_file(t_repo_id1, '/', file_name, t_repo_id2,
'/', new_file_name, 1, USER, 1, 0)
assert t_move_file_result2
assert t_move_file_result2.task_id
assert t_move_file_result2.background
while True:
time.sleep(0.1)
t_move_task = api.get_copy_task(t_move_file_result2.task_id)
if t_move_task.successful:
break
t_file_id = api.get_file_id_by_path(t_repo_id2, '/' + new_file_name)
assert t_file_size == api.get_file_size(t_repo_id2, t_repo_version,
t_file_id)
# test post_empty_file
assert api.post_empty_file(t_repo_id1, '/' + dir_name, empty_file_name,
USER) == 0
t_file_id = api.get_file_id_by_path(t_repo_id1,
'/' + dir_name + '/' + empty_file_name)
assert api.get_file_size(t_repo_id1, t_repo_version, t_file_id) == 0
# test rename_file
assert api.rename_file(t_repo_id1, '/' + dir_name, empty_file_name,
new_empty_file_name, USER) == 0
#test put_file
t_new_file_id = api.put_file(t_repo_id1, file_path, '/' + dir_name,
new_empty_file_name, USER, None)
assert t_new_file_id
# test get_file_revisions
t_commit_list = api.get_file_revisions(t_repo_id2, None, '/' + file_name,
2)
assert t_commit_list
assert len(t_commit_list) == 2
assert t_commit_list[0].creator_name == USER
# test del_file
assert api.del_file(t_repo_id2, '/', file_name, USER) == 0
# test get_deleted
t_deleted_file_list = api.get_deleted(t_repo_id2, 1)
assert t_deleted_file_list
assert len(t_deleted_file_list) == 2
assert t_deleted_file_list[0].obj_name == file_name
assert t_deleted_file_list[0].basedir == '/'
# test del a non-exist file. should return 0.
assert api.del_file(t_repo_id2, '/', file_name, USER) == 0
assert api.del_file(t_repo_id1, '/' + dir_name, new_empty_file_name,
USER) == 0
assert api.del_file(t_repo_id1, '/' + dir_name, new_file_name, USER) == 0
assert api.del_file(t_repo_id2, '/', new_file_name, USER) == 0
assert api.del_file(t_repo_id1, '/', new_file_name_2, USER) == 0
time.sleep(1)
api.remove_repo(t_repo_id1)
def endWrite(self, withErrors):
if not withErrors:
parent, filename = os.path.split(self.rel_path)
seafile_api.put_file(self.repo.id, self.tmpfile_path, parent, filename,
self.username, None)
os.unlink(self.tmpfile_path)
def test_cdc_completely(create_tmp_repo):
"""Test CDC generation, complete workflow
"""
repo = create_tmp_repo
# TEST DEFAULT LIBRARY
"""check default library files are avialable"""
kdl = get_keeper_default_library()
if kdl:
check_set = set([d.obj_name for d in kdl['dirents']])
else:
pytest.fail(msg="Default Library is empty, please install!")
"""copy library default files to lib"""
copy_keeper_default_library(repo.id)
sleep(2)
# get copied files
# commits = seafile_api.get_commit_list(repo.id, 0, 1)
# commit = commit_mgr.load_commit(repo.id, repo.version, commits[0].id)
# dir = fs_mgr.load_seafdir(repo.id, repo.version, commit.root_id)
dir = get_root_dir(repo)
# check copied files
for fn in check_set:
if not dir.lookup(fn):
pytest.fail("Cannot find %s!" % fn)
# TEST CDC GENERATION
"""add some txt file"""
f = tempfile.NamedTemporaryFile()
f.write('Text')
f.flush()
os.chmod(f.name, 0o666)
seafile_api.post_file(repo.id, f.name, "/", "some_file.txt", SERVER_EMAIL)
f.close()
"""update md file with correct fields"""
f = tempfile.NamedTemporaryFile()
f.write(MD_GOOD)
f.flush()
os.chmod(f.name, 0o666)
seafile_api.put_file(repo.id, f.name, "/", ARCHIVE_METADATA_TARGET,
SERVER_EMAIL, None)
f.close()
sleep(10)
"""check cdc pdf exists"""
dir = get_root_dir(repo)
cdc_pdfs = [
f.name for f in dir.get_files_list()
if f.name.startswith(CDC_PDF_PREFIX) and f.name.endswith('.pdf')
]
if not cdc_pdfs:
pytest.fail("Cannot find cdc pdf in repo %s!" % repo.name)
"""check cdc pdf remove"""
seafile_api.del_file(repo.id, "/", cdc_pdfs[0], SERVER_EMAIL)
sleep(10)
dir = get_root_dir(repo)
cdc_pdfs = [
f.name for f in dir.get_files_list()
if f.name.startswith(CDC_PDF_PREFIX) and f.name.endswith('.pdf')
]
assert not cdc_pdfs, "cdc pdf should not be recreated if it has been deleted"
"""update md file with corrected field"""
f = tempfile.NamedTemporaryFile()
f.write(MD_GOOD.replace("2010", "2017"))
f.flush()
os.chmod(f.name, 0o666)
seafile_api.put_file(repo.id, f.name, "/", ARCHIVE_METADATA_TARGET,
SERVER_EMAIL, None)
f.close()
sleep(10)
dir = get_root_dir(repo)
cdc_pdfs = [
f.name for f in dir.get_files_list()
if f.name.startswith(CDC_PDF_PREFIX) and f.name.endswith('.pdf')
]
assert cdc_pdfs, "cdc pdf should be recreated if metadata md has been changed"
def endWrite(self, withErrors):
if not withErrors:
parent, filename = os.path.split(self.rel_path)
seafile_api.put_file(self.repo.id, self.tmpfile_path, parent,
filename, self.username, None)
os.unlink(self.tmpfile_path)
def generate_certificate(repo, commit):
""" Generate Cared Data Certificate according to markdown file """
#exit if repo encrypted
if repo.encrypted:
return False
# exit if repo is system template
if repo.rep_desc == TEMPLATE_DESC:
return False
dir = fs_mgr.load_seafdir(repo.id, repo.version, commit.root_id)
# certificate already exists in root
# file_names = [f.name for f in dir.get_files_list()]
# if any(file_name.startswith(CDC_PDF_PREFIX) and file_name.endswith('.pdf') for file_name in file_names):
# return False
# get latest version of the ARCHIVE_METADATA_TARGET
file = dir.lookup(ARCHIVE_METADATA_TARGET)
#exit if no metadata file exists
if not file:
return False
#check wether there is at least one creative dirent
if not has_at_least_one_creative_dirent(dir):
return False
logging.info('Repo has creative dirents')
try:
db = get_db(KEEPER_DB_NAME)
cur = db.cursor()
# if is_certified(db, cur, repo.id):
# return False
owner = seafile_api.get_repo_owner(repo.id)
logging.info("Certifying repo id: %s, name: %s, owner: %s ..." % (repo.id, repo.name, owner))
cdc_dict = parse_markdown(file.get_content())
if validate(cdc_dict):
cdc_id = register_cdc_in_db(db, cur, repo.id, owner)
logging.info("Generate CDC PDF...")
cdc_pdf = CDC_PDF_PREFIX + cdc_id + ".pdf"
# TODO: specify which url should be in CDC
# as tmp decision: SERVICE_URL
# repo_share_url = get_repo_share_url(repo.id, owner)
repo_share_url = SERVICE_URL
jars = ":".join(map(lambda e : MODULE_PATH + '/' + e, CDC_GENERATOR_JARS))
args = [ "java", "-cp", jars, CDC_GENERATOR_MAIN_CLASS,
"-i", "\"" + cdc_id + "\"",
"-t", "\"" + cdc_dict['Title'] + "\"",
"-aa", "\"" + cdc_dict['Author'] + "\"",
"-d", "\"" + cdc_dict['Description'] + "\"",
"-c", "\"" + owner + "\"",
"-u", "\"" + repo_share_url + "\"",
cdc_pdf ]
check_call(args)
tmp_path = os.path.abspath(cdc_pdf)
logging.info("PDF sucessfully generated")
logging.info("Add " + cdc_pdf + " to the repo...")
if UPDATE:
seafile_api.put_file(repo.id, tmp_path, "/", cdc_pdf, SERVER_EMAIL, None)
logging.info("Sucessfully updated")
else:
seafile_api.post_file(repo.id, tmp_path, "/", cdc_pdf, SERVER_EMAIL)
logging.info("Sucessfully added")
if not DEBUG:
send_email(owner, {'USER_NAME': get_user_name(owner), 'PROJECT_NAME':repo.name, 'PROJECT_URL':get_repo_pivate_url(repo.id) })
#TODO: Send seafile notification
except Exception as err:
logging.info(str(err))
finally:
# other final stuff
db.close()
if 'tmp_path' in vars() and os.path.exists(tmp_path):
os.remove(tmp_path)
return True
def generate_certificate(repo, commit):
""" Generate Cared Data Certificate according to markdown file """
event = None
# exit if repo encrypted
if repo.encrypted:
return False
# exit if repo is system template
if repo.rep_desc == TEMPLATE_DESC:
return False
# TODO: if cdc pdf is deleted: set cert status to False and exit
# see https://github.com/MPDL/KEEPER/issues/41
if re.match(EVENT_PATTERNS['CDC_PDF_DELETED'], commit.desc):
event = EVENT.pdf_delete
Catalog.objects.update_cert_status_by_repo_id(repo.id, False)
return False
if re.match(EVENT_PATTERNS['ARCHIVE_METADATA_TARGET_MODIFIED'],
commit.desc):
event = EVENT.md_modified
try:
cdc_id = get_cdc_id_by_repo(repo.id)
if cdc_id is not None:
if re.match(EVENT_PATTERNS['CDC_PDF_DELETED'], commit.desc):
# if cdc pdf is deleted, add pdf again!
event = EVENT.pdf_delete
elif event != EVENT.md_modified:
# exit if already certified and MD has not been changed
return False
dir = fs_mgr.load_seafdir(repo.id, repo.version, commit.root_id)
# certificate already exists in root
# file_names = [f.name for f in dir.get_files_list()]
# if any(file_name.startswith(CDC_PDF_PREFIX) and file_name.endswith('.pdf') for file_name in file_names):
# return False
# get latest version of the ARCHIVE_METADATA_TARGET
file = dir.lookup(ARCHIVE_METADATA_TARGET)
#exit if no metadata file exists
if not file:
return False
# check whether there is at least one creative dirent
if not has_at_least_one_creative_dirent(dir):
return False
logger.info('Repo has creative dirents')
owner = seafile_api.get_repo_owner(repo.id)
logger.info("Certifying repo id: %s, name: %s, owner: %s ..." %
(repo.id, repo.name, owner))
content = file.get_content().decode('utf-8')
cdc_dict = parse_markdown(content)
status = 'metadata are not valid'
is_successful = False
if validate(cdc_dict):
status = 'metadata are valid'
if event == EVENT.pdf_delete:
# only modified update
cdc_id = register_cdc_in_db(repo.id, owner)[0]
else:
cdc_id, event = register_cdc_in_db(repo.id, owner)
cdc_id = str(cdc_id)
logger.info("Generate CDC PDF...")
cdc_pdf = CDC_PDF_PREFIX + cdc_id + ".pdf"
jars = ":".join(
[MODULE_PATH + '/' + e for e in CDC_GENERATOR_JARS])
tmp_path = tempfile.gettempdir() + "/" + cdc_pdf
args = [
"date;",
"java",
"-cp",
jars,
CDC_GENERATOR_MAIN_CLASS,
"-i",
quote_arg(cdc_id),
"-t",
quote_arg(cdc_dict['Title']),
"-aa",
quote_arg(cdc_dict['Author']),
"-d",
quote_arg(cdc_dict['Description']),
"-c",
quote_arg(owner),
"-u",
quote_arg(SERVICE_URL),
tmp_path,
"1>&2;",
]
try:
call_str = " ".join(args)
logger.info(call_str)
# subprocess.call(call_str, shell=True)
# p = subprocess.Popen(call_str, stdout=subprocess.PIPE, stderr=subprocess.PIPE)
p = subprocess.Popen(call_str,
stdout=subprocess.PIPE,
stderr=subprocess.PIPE,
shell=True)
stdout, stderr = p.communicate()
if stdout:
logger.info(stdout)
if stderr:
logger.error(stderr)
except Exception as err:
logger.error('Cannot call command')
logger.error(traceback.format_exc())
raise err
if os.path.isfile(tmp_path):
logger.info("PDF sucessfully generated, tmp_path=%s" %
tmp_path)
else:
logger.error(
"Cannot find generated CDC PDF, tmp_path=%s, exiting..." %
tmp_path)
if event == EVENT.db_create:
# TODO: test
CDC.objects.get(cdc_id=cdc_id).delete()
return False
logger.info("Add " + cdc_pdf + " to the repo...")
if event == EVENT.db_update:
# cdc pdf already exists, then put
if dir.lookup(cdc_pdf):
seafile_api.put_file(repo.id, tmp_path, "/", cdc_pdf,
SERVER_EMAIL, None)
logger.info("Sucessfully updated")
status = 'updated'
CDC_MSG.append("Certificate has been updated")
# post otherwise
else:
seafile_api.post_file(repo.id, tmp_path, "/", cdc_pdf,
SERVER_EMAIL)
logger.info("Sucessfully recreated")
status = 'recreated'
CDC_MSG.append("Certificate has been recreated")
logging.info(
"CDC has been successfully updated for repo %s, id: %s" %
(repo.id, cdc_id))
else:
seafile_api.post_file(repo.id, tmp_path, "/", cdc_pdf,
SERVER_EMAIL)
logger.info("Sucessfully created")
status = 'created'
CDC_MSG.append("Certificate has been sucessfully created")
# if not DEBUG:
send_email(
owner, {
'SERVICE_URL': SERVICE_URL,
'USER_NAME': get_user_name(owner),
'PROJECT_NAME': repo.name,
'PROJECT_TITLE': cdc_dict['Title'],
'PROJECT_URL': get_repo_pivate_url(repo.id),
'AUTHOR_LIST': get_authors_for_email(
cdc_dict['Author']),
'CDC_PDF_URL': get_file_pivate_url(repo.id, cdc_pdf),
'CDC_ID': cdc_id
})
logging.info(
"CDC has been successfully created for repo %s, id: %s" %
(repo.id, cdc_id))
is_successful = True
#send user notification
logger.info("Commit desc: " + commit.desc)
logger.info("event: {}".format(event))
if event in (EVENT.md_modified, EVENT.db_create, EVENT.db_update):
header = "Certificate processed for library" if is_successful else "Cannot process certificate for library"
UserNotification.objects._add_user_notification(
owner, MSG_TYPE_KEEPER_CDC_MSG,
json.dumps({
'header': header,
'status': status,
'message': ('; '.join(CDC_MSG)),
'msg_from': SERVER_EMAIL,
'lib': repo.id,
'lib_name': repo.name
}))
except Exception as err:
logger.error(traceback.format_exc())
logging.error(
"CDC generation for repo %s has been failed, check %s for details. \nTraceback:"
% (repo.id, CDC_LOG))
logging.error(traceback.format_exc())
finally:
if 'tmp_path' in vars() and os.path.exists(tmp_path):
os.remove(tmp_path)
return True
def save_archive_metadata(repo_id, md_dict):
"""
Convert archive metadata dictionary into markdown format
and put it into repo_id library
"""
p = "## %s\n%s\n\n"
md_markdown = ARCHIVE_METADATA_TARGET_HEADER
md_markdown += p % ('Title', md_dict.get('title', ''))
# authors
al = md_dict.get('authors', '')
a_res = []
if al and type(al) is list:
a_tmp = ""
for a in al:
# Name
n_tmp = []
ln = a.get('lastName', '').strip()
ln and n_tmp.append(ln)
fn = a.get('firstName', '').strip()
fn and n_tmp.append(fn)
a_tmp = ", ".join(n_tmp)
# Affiliations
affs = a.get('affs')
affs_str = ''
if affs:
affs = [aff for aff in affs if aff and aff.strip()]
if affs:
affs_str = " | ".join(affs)
a_res.append("; ".join(list(filter(None, [a_tmp, affs_str]))))
md_markdown += p % ('Author', '\n'.join(a_res))
md_markdown += p % ('Publisher', md_dict.get('publisher', ''))
md_markdown += p % ('Description', md_dict.get('description', ''))
md_markdown += p % ('Year', md_dict.get('year', ''))
# institute
dl = md_dict.get('directors', '')
d_tmp = []
if dl and type(dl) is list:
for d in dl:
n_tmp = []
ln = d.get('lastName', '').strip()
ln and n_tmp.append(ln)
fn = d.get('firstName', '').strip()
fn and n_tmp.append(fn)
n_tmp and d_tmp.append(", ".join(n_tmp))
md_markdown += p % (
'Institute',
"; ".join([
md_dict.get('institute', ''),
md_dict.get('department', ''),
# TODO: mutiply directors revision!!!!
" | ".join(d_tmp)
]))
md_markdown += p % ('Resource Type', md_dict.get('resourceType', ''))
md_markdown += p % ('License', md_dict.get('license', ''))
md_markdown = md_markdown[:-2]
try:
with tempfile.NamedTemporaryFile() as fp:
fp.write(md_markdown.encode())
fp.flush()
if seafile_api.get_dirent_by_path(repo_id,
ARCHIVE_METADATA_TARGET):
# file exists, update
seafile_api.put_file(repo_id, fp.name, "/",
ARCHIVE_METADATA_TARGET, SERVER_EMAIL,
None)
else:
# create file
seafile_api.post_file(repo_id, fp.name, "/",
ARCHIVE_METADATA_TARGET, SERVER_EMAIL)
except Exception as e:
logger.error('Cannot save file %s: %r', ARCHIVE_METADATA_TARGET, e)
