def __init__(self, obj=None, url=None):
"""Create a new qfurl for object @obj"""
if obj:
typname = "%s.%s" % (type(obj).__module__, type(obj).__name__)
try:
qfid = obj.qf_id
except AttributeError:
raise QfurlError("%s has no qfurl" % obj)
self.url = _urlunparse((QFURL_SCHEME, "", qfid, "", "", typname))
else:
self.url = url
def __init__(self, obj=None, url=None):
"""Create a new qfurl for object @obj"""
if obj:
typname = "%s.%s" % (type(obj).__module__, type(obj).__name__)
try:
qfid = obj.qf_id
except AttributeError:
raise QfurlError("%s has no qfurl" % obj)
self.url = _urlunparse((QFURL_SCHEME, "", qfid, "", "", typname))
else:
self.url = url
def redacted_urlunparse(url, redact_with="***"):
"""``urlunparse()`` but redact password."""
netloc = []
if url.username:
netloc.append(url.username)
if url.password:
netloc.append(":")
netloc.append(redact_with)
if url.hostname:
if netloc:
netloc.append("@")
netloc.append(url.hostname)
url = url._replace(netloc="".join(netloc))
return _urlunparse(url)
def escape_channel_url(channel):
if channel.startswith("file:"):
if "%" in channel: # it's escaped already
return channel
if on_win:
channel = channel.replace("\\", "/")
parts = _urlparse(channel)
if parts.scheme:
components = parts.path.split("/")
if on_win:
if len(parts.netloc) == 2 and parts.netloc[1] == ":":
# with absolute paths (e.g. C:/something), C:, D:, etc might get parsed as netloc
path = "/".join([parts.netloc] +
[quote(p) for p in components])
parts = parts._replace(netloc="")
else:
path = "/".join(components[:2] +
[quote(p) for p in components[2:]])
else:
path = "/".join([quote(p) for p in components])
parts = parts._replace(path=path)
return _urlunparse(parts)
return channel
def urlunparse(parts):
result = _urlunparse(tuple([p.decode("charmap") for p in parts]))
return result.encode("charmap")
def urlunparse(parts):
result = _urlunparse(tuple([p.decode("charmap") for p in parts]))
return result.encode("charmap")
def download_file(url, path=None, *more_components):
"""Download file (if necessary) and yield filename if necessary."""
if path:
path = fs.path.join(path, *more_components)
full_path = fs.path.join(url.path, path)
else:
full_path = url.path
path = fs.path.basename(url.path)
url = url._replace(path=fs.path.dirname(url.path))
if url.scheme == "file":
if os.path.exists(full_path):
yield full_path
else:
yield None
else:
basename = fs.path.basename(full_path)
if url.scheme in PYFS_SCHEMES:
src_fs = make_fs(url)
# Download file if it exists.
if not src_fs.exists(path):
yield None
else:
with TempFS() as tmpfs:
logger.info("Downloading file %s from %s" % (path, redacted_urlunparse(url)))
with open(tmpfs.getospath(basename), "wb") as tmpf:
src_fs.download(path, tmpf)
logger.info("Download complete.")
yield tmpfs.getospath(basename)
logger.info("Releasing %s" % tmpfs)
elif url.scheme == "s3":
logger.info("Connecting via S3...")
anon = url.username is None and url.password is None
s3 = s3fs.S3FileSystem(anon=anon, key=url.username, secret=url.password)
with TempFS() as tmpfs:
logger.info("Downloading file %s from %s" % (path, redacted_urlunparse(url)))
with s3.open("%s/%s" % (url.hostname, path), "rb") as inputf:
with open(tmpfs.getospath(basename), "wb") as outputf:
shutil.copyfileobj(inputf, outputf, settings.MAX_UPLOAD_DATA_SIZE)
logger.info("Download complete.")
yield tmpfs.getospath(basename)
logger.info("Releasing %s" % tmpfs)
elif url.scheme.startswith("http"):
logger.info("Downloading via HTTP(S)...")
with TempFS() as tmpfs:
logger.info("Downloading file %s from %s" % (path, redacted_urlunparse(url)))
r = requests.get(_urlunparse(url._replace(path=full_path)), allow_redirects=True)
r.raise_for_status()
with open(tmpfs.getospath(basename), "wb") as outputf:
outputf.write(r.content)
logger.info("Download complete.")
yield tmpfs.getospath(basename)
logger.info("Releasing %s" % tmpfs)
elif url.scheme.startswith("irods"):
with create_irods_session(url) as irods_session:
logger.info("Downloading file...")
with TempFS() as tmpfs:
logger.info("Downloading file %s from %s" % (path, redacted_urlunparse(url)))
path_tmp = tmpfs.getospath(basename)
collection = irods_session.collections.get(fs.path.dirname(full_path))
name = fs.path.basename(full_path)
for data_object in collection.data_objects:
if data_object.name == name:
with data_object.open() as inputf:
with open(tmpfs.getospath(basename), "wb") as outputf:
shutil.copyfileobj(
inputf, outputf, settings.MAX_UPLOAD_DATA_SIZE
)
break
else:
raise ScelVisException(
"Could not find %s in %s" % (full_path, redacted_urlunparse(url))
)
logger.info("Download complete.")
yield path_tmp
logger.info("Releasing %s" % tmpfs)
else:
raise ScelVisException("Invalid URL scheme: %s" % url.scheme)
