def __iter__(self):
for item in self.previous:
keys = item.keys()
pathkey = self.pathkey(*keys)[0]
objpath = item[pathkey]
obj = self.context.unrestrictedTraverse(objpath)
# The contructor generated an empty object (which triggered our
# stats event handler.) Let's update the stats here manually.
created = IAnnotations(obj.REQUEST).get(
'org.bccvl.site.stats.created', False)
if created:
# our created subscriber has been invoked
getUtility(IStatsUtility).count_dataset(
source=obj.dataSource, portal_type=obj.portal_type)
# reset the flag ... we can do that, because the pipeline runs
# sequentially
IAnnotations(
obj.REQUEST)['org.bccvl.site.stats.created'] = False
# Attach a job tracker only for species dataset from multispecies
if item.get('_partof', {}):
jt = IJobTracker(obj)
jt.new_job('TODO: generate id',
'generate taskname: ala_import',
function=obj.dataSource,
type=obj.portal_type,
state='COMPLETED')
yield item
def update_metadata(self):
uuid = self.request.form.get('uuid', None)
try:
if uuid:
brain = uuidToCatalogBrain(uuid)
if brain is None:
raise Exception("Brain not found")
obj = brain.getObject()
else:
obj = self.context
# get username
member = ploneapi.user.get_current()
if member.getId():
user = {
'id': member.getUserName(),
'email': member.getProperty('email'),
'fullname': member.getProperty('fullname')
}
else:
raise Exception("Invalid user")
# build download url
# 1. get context (site) relative path
obj_url = obj.absolute_url()
if obj.portal_type == 'org.bccvl.content.dataset':
filename = obj.file.filename
obj_url = '{}/@@download/file/{}'.format(obj_url, filename)
elif obj.portal_type == 'org.bccvl.content.remotedataset':
filename = os.path.basename(obj.remoteUrl)
obj_url = '{}/@@download/{}'.format(obj_url, filename)
elif obj.portal_type == 'org.bccvl.content.multispeciesdataset':
filename = obj.file.filename
obj_url = '{}/@@download/file/{}'.format(obj_url, filename)
else:
raise Exception("Wrong content type")
from org.bccvl.tasks.celery import app
update_task = app.signature(
"org.bccvl.tasks.datamover.tasks.update_metadata",
kwargs={
'url': obj_url,
'filename': filename,
'contenttype': obj.format,
'context': {
'context': '/'.join(obj.getPhysicalPath()),
'user': user,
}
},
immutable=True)
from org.bccvl.tasks.plone import after_commit_task
after_commit_task(update_task)
# track background job state
jt = IJobTracker(obj)
job = jt.new_job('TODO: generate id',
'generate taskname: update_metadata',
function=obj.dataSource,
type=obj.portal_type)
jt.set_progress('PENDING', 'Metadata update pending')
return job.id
except Exception as e:
LOG.error('Caught exception %s', e)
raise NotFound(self, 'update_metadata', self.request)
def add(self, object):
# FIXME: this is a workaround, which is fine for small uploaded files.
# large uploads should go through another process anyway
# TODO: re implementing this method is the only way to know
# the full path of the object. We need the path to apply
# the transmogrifier chain.
# fti = getUtility(IDexterityFTI, name=self.portal_type)
container = aq_inner(self.context)
try:
# traverse to subfolder if possible
container = container.restrictedTraverse('/'.join(self.subpath))
except Exception as e:
LOG.warn('Could not traverse to %s/%s',
'/'.join(container.getPhysicalPath()),
'/'.join(self.subpath))
new_object = addContentToContainer(container, object)
# set data genre:
if self.datagenre:
IBCCVLMetadata(new_object)['genre'] = self.datagenre
if self.categories:
IBCCVLMetadata(new_object)['categories'] = self.categories
new_object.subject = []
if self.domain:
new_object.subject = [self.domain]
if self.timeperiod:
new_object.subject += self.timeperiod
# rdf commit should happens in transmogrifier step later on
# if fti.immediate_view:
# self.immediate_view = "%s/%s/%s" % (container.absolute_url(), new_object.id, fti.immediate_view,)
# else:
# self.immediate_view = "%s/%s" % (container.absolute_url(), new_object.id)
# start background import process (just a metadata update)
# run transmogrify md extraction here
context_path = '/'.join(new_object.getPhysicalPath())
member = api.user.get_current()
# species extract task
if IMultiSpeciesDataset.providedBy(new_object):
# kick off csv split import tasks
import_task = app.signature(
"org.bccvl.tasks.datamover.tasks.import_multi_species_csv",
kwargs={
'url':
'{}/@@download/file/{}'.format(new_object.absolute_url(),
new_object.file.filename),
'results_dir':
get_results_dir(new_object,
self.request,
childSpecies=True),
'import_context': {
'context': '/'.join(container.getPhysicalPath()),
'user': {
'id': member.getUserName(),
'email': member.getProperty('email'),
'fullname': member.getProperty('fullname')
}
},
'context': {
'context': context_path,
'genre': self.datagenre,
'dataSource': new_object.dataSource,
'user': {
'id': member.getUserName(),
'email': member.getProperty('email'),
'fullname': member.getProperty('fullname')
}
}
},
immutable=True)
after_commit_task(import_task)
# create job tracking object
jt = IJobTracker(new_object)
jt.new_job('TODO: generate id',
'generate taskname: import_multi_species_csv',
function=new_object.dataSource,
type=new_object.portal_type)
jt.set_progress('PENDING', u'Multi species import pending')
else:
if hasattr(self, '_upload'):
file = self._upload['file']
new_object.format = file.contentType
uid = IUUID(new_object)
swiftsettings = getUtility(IRegistry).forInterface(
ISwiftSettings)
import os.path
swift_url = '{storage_url}/{container}/{path}/{name}'.format(
storage_url=swiftsettings.storage_url,
container=swiftsettings.result_container,
path=uid,
name=os.path.basename(file.filename))
new_object.remoteUrl = swift_url
else:
file = new_object.file
new_object.format = file.contentType
dlinfo = IDownloadInfo(new_object)
# single species upload
update_task = app.signature(
"org.bccvl.tasks.datamover.tasks.update_metadata",
kwargs={
'url': dlinfo['url'],
'filename': dlinfo['filename'],
'contenttype': dlinfo['contenttype'],
'context': {
'context': context_path,
'user': {
'id': member.getUserName(),
'email': member.getProperty('email'),
'fullname': member.getProperty('fullname')
}
}
},
immutable=True)
# create upload task in case we upload to external store
if hasattr(self, '_upload'):
# FIXME: we can't use ssh here.... we don't know which container we are in... and
# sshing here is bad as well....
# There is an upload ... we have to make sure the uploaded data ends up in external storage
# 3. put temp file aside
tmpdir = tempfile.mkdtemp(prefix='bccvl_upload')
tmpfile = os.path.join(tmpdir, os.path.basename(file.filename))
blobf = file.open()
try:
# try rename
os.rename(blobf.name, tmpfile)
except OSError:
# try copy
shutil.copy(blobf.name, tmpfile)
# TODO: we push the uploaded file directly to swift here..
# this really should be a background process
# best solution: ...
# user uploads to some temporary upload service (file never ends up here)
# we have a remote url here, and tell the datamover to pull it from there
# and move it to final destination. (or something like this)
# other good way: ...
# let user upload directly to swift (what about large file uploads?)
# and take care of clean up if necessary
# 4. move file to swift
# TODO: do we have enough information to upload to swift?
# need a temp url?
swiftopts = app.conf.get('bccvl', {}).get('swift', {})
src_url = build_source('file://{}'.format(tmpfile))
dest_url = build_destination(
'swift+{}'.format(new_object.remoteUrl),
settings={
'swift': {
'os_auth_url':
swiftopts.get('os_auth_url'),
'os_username':
swiftopts.get('os_username'),
'os_password':
swiftopts.get('os_password'),
'os_project_name':
swiftopts.get('os_project_name'),
'os_storage_url':
swiftopts.get('os_storage_url'),
'os_user_domain_name':
swiftopts.get('os_user_domain_name'),
'os_project_domain_name':
swiftopts.get('os_project_domain_name'),
'auth_version':
swiftopts.get('auth_version')
}
})
try:
movelib.move(src_url, dest_url)
except Exception as e:
# do error handling here
raise
finally:
# clean up temp location
path = os.path.dirname(tmpfile)
shutil.rmtree(path)
# queue job submission
after_commit_task(update_task)
# create job tracking object
jt = IJobTracker(new_object)
jt.new_job('TODO: generate id',
'generate taskname: update_metadata',
function=new_object.dataSource,
type=new_object.portal_type)
jt.set_progress('PENDING', u'Metadata update pending')
# We have to reindex after updating the object
new_object.reindexObject()
def add(self, object):
# FIXME: this is a workaround, which is fine for small uploaded files.
# large uploads should go through another process anyway
# TODO: re implementing this method is the only way to know
# the full path of the object. We need the path to apply
# the transmogrifier chain.
# fti = getUtility(IDexterityFTI, name=self.portal_type)
container = aq_inner(self.context)
try:
# traverse to subfolder if possible
container = container.restrictedTraverse('/'.join(self.subpath))
except Exception as e:
LOG.warn('Could not traverse to %s/%s',
'/'.join(container.getPhysicalPath()), '/'.join(self.subpath))
new_object = addContentToContainer(container, object)
# set data genre:
if self.datagenre:
IBCCVLMetadata(new_object)['genre'] = self.datagenre
if self.categories:
IBCCVLMetadata(new_object)['categories'] = self.categories
# rdf commit should happens in transmogrifier step later on
# if fti.immediate_view:
# self.immediate_view = "%s/%s/%s" % (container.absolute_url(), new_object.id, fti.immediate_view,)
# else:
# self.immediate_view = "%s/%s" % (container.absolute_url(), new_object.id)
# start background import process (just a metadata update)
# run transmogrify md extraction here
context_path = '/'.join(new_object.getPhysicalPath())
member = api.user.get_current()
# species extract task
if IMultiSpeciesDataset.providedBy(new_object):
# kick off csv split import tasks
import_task = app.signature(
"org.bccvl.tasks.datamover.tasks.import_multi_species_csv",
kwargs={
'url': '{}/@@download/file/{}'.format(new_object.absolute_url(), new_object.file.filename),
'results_dir': get_results_dir(container, self.request),
'import_context': {
'context': '/'.join(container.getPhysicalPath()),
'user': {
'id': member.getUserName(),
'email': member.getProperty('email'),
'fullname': member.getProperty('fullname')
}
},
'context': {
'context': context_path,
'user': {
'id': member.getUserName(),
'email': member.getProperty('email'),
'fullname': member.getProperty('fullname')
}
}
},
immutable=True)
after_commit_task(import_task)
# create job tracking object
jt = IJobTracker(new_object)
job = jt.new_job('TODO: generate id',
'generate taskname: import_multi_species_csv')
job.type = new_object.portal_type
jt.set_progress('PENDING', u'Multi species import pending')
else:
if hasattr(self, '_upload'):
file = self._upload['file']
new_object.format = file.contentType
uid = IUUID(new_object)
swiftsettings = getUtility(
IRegistry).forInterface(ISwiftSettings)
import os.path
swift_url = '{storage_url}/{container}/{path}/{name}'.format(
storage_url=swiftsettings.storage_url,
container=swiftsettings.result_container,
path=uid,
name=os.path.basename(file.filename))
new_object.remoteUrl = swift_url
else:
file = new_object.file
new_object.format = file.contentType
dlinfo = IDownloadInfo(new_object)
# single species upload
update_task = app.signature(
"org.bccvl.tasks.datamover.tasks.update_metadata",
kwargs={
'url': dlinfo['url'],
'filename': dlinfo['filename'],
'contenttype': dlinfo['contenttype'],
'context': {
'context': context_path,
'user': {
'id': member.getUserName(),
'email': member.getProperty('email'),
'fullname': member.getProperty('fullname')
}
}
},
immutable=True)
# create upload task in case we upload to external store
if hasattr(self, '_upload'):
# There is an upload ... we have to make sure the uploaded data ends up in external storage
# 3. put temp file aside
tmpdir = tempfile.mkdtemp(prefix='bccvl_upload')
tmpfile = os.path.join(tmpdir, os.path.basename(file.filename))
blobf = file.open()
try:
# try rename
os.rename(blobf.name, tmpfile)
except OSError:
# try copy
shutil.copy(blobf.name, tmpfile)
# 4. update task chain
src_url = 'scp://{uid}@{ip}:{port}{file}'.format(
uid=pwd.getpwuid(os.getuid()).pw_name,
ip=get_hostname(self.request),
port=os.environ.get('SSH_PORT', 22),
file=tmpfile)
dest_url = 'swift+{}'.format(new_object.remoteUrl)
move_task = app.signature(
'org.bccvl.tasks.datamover.tasks.move',
kwargs={
'move_args': [(src_url, dest_url)],
'context': {
'context': context_path,
'user': {
'id': member.getUserName(),
'email': member.getProperty('email'),
'fullname': member.getProperty('fullname')
}
}
},
immutable=True)
cleanup_task = app.signature(
'org.bccvl.tasks.plone.import_cleanup',
kwargs={
'path': os.path.dirname(tmpfile),
'context': {
'context': context_path,
'user': {
'id': member.getUserName(),
'email': member.getProperty('email'),
'fullname': member.getProperty('fullname')
}
}
},
immutable=True)
update_task = move_task | update_task | cleanup_task
# need some more workflow states here to support e.g. zip file upload (multiple rasters),
# give user a chance to better define metadata
# make sure update_metadata does not change user edited metadata
# -> layer, unit, projection, whatever
# FIXME: clean up tmp upload directory as well
# queue job submission
after_commit_task(update_task)
# create job tracking object
jt = IJobTracker(new_object)
job = jt.new_job('TODO: generate id',
'generate taskname: update_metadata')
job.type = new_object.portal_type
jt.set_progress('PENDING', u'Metadata update pending')
# We have to reindex after updating the object
new_object.reindexObject()
def update_metadata(self):
uuid = self.request.form.get("uuid", None)
try:
if uuid:
brain = uuidToCatalogBrain(uuid)
if brain is None:
raise Exception("Brain not found")
obj = brain.getObject()
else:
obj = self.context
# get username
member = ploneapi.user.get_current()
if member.getId():
user = {
"id": member.getUserName(),
"email": member.getProperty("email"),
"fullname": member.getProperty("fullname"),
}
else:
raise Exception("Invalid user")
# build download url
# 1. get context (site) relative path
obj_url = obj.absolute_url()
if obj.portal_type == "org.bccvl.content.dataset":
filename = obj.file.filename
obj_url = "{}/@@download/file/{}".format(obj_url, filename)
elif obj.portal_type == "org.bccvl.content.remotedataset":
filename = os.path.basename(obj.remoteUrl)
obj_url = "{}/@@download/{}".format(obj_url, filename)
elif obj.portal_type == "org.bccvl.content.multispeciesdataset":
filename = obj.file.filename
obj_url = "{}/@@download/file/{}".format(obj_url, filename)
else:
raise Exception("Wrong content type")
from org.bccvl.tasks.celery import app
update_task = app.signature(
"org.bccvl.tasks.datamover.tasks.update_metadata",
kwargs={
"url": obj_url,
"filename": filename,
"contenttype": obj.format,
"context": {"context": "/".join(obj.getPhysicalPath()), "user": user},
},
immutable=True,
)
from org.bccvl.tasks.plone import after_commit_task
after_commit_task(update_task)
# track background job state
jt = IJobTracker(obj)
job = jt.new_job("TODO: generate id", "generate taskname: update_metadata")
job.type = obj.portal_type
jt.set_progress("PENDING", "Metadata update pending")
return job.id
except Exception as e:
LOG.error("Caught exception %s", e)
raise NotFound(self, "update_metadata", self.request)
def add(self, object):
# FIXME: this is a workaround, which is fine for small uploaded files.
# large uploads should go through another process anyway
# TODO: re implementing this method is the only way to know
# the full path of the object. We need the path to apply
# the transmogrifier chain.
# fti = getUtility(IDexterityFTI, name=self.portal_type)
container = aq_inner(self.context)
try:
# traverse to subfolder if possible
container = container.restrictedTraverse('/'.join(self.subpath))
except Exception as e:
LOG.warn('Could not traverse to %s/%s',
'/'.join(container.getPhysicalPath()), '/'.join(self.subpath))
new_object = addContentToContainer(container, object)
# set data genre:
if self.datagenre:
IBCCVLMetadata(new_object)['genre'] = self.datagenre
if self.categories:
IBCCVLMetadata(new_object)['categories'] = self.categories
new_object.subject = []
if self.domain:
new_object.subject = [self.domain]
if self.timeperiod:
new_object.subject += self.timeperiod
# rdf commit should happens in transmogrifier step later on
# if fti.immediate_view:
# self.immediate_view = "%s/%s/%s" % (container.absolute_url(), new_object.id, fti.immediate_view,)
# else:
# self.immediate_view = "%s/%s" % (container.absolute_url(), new_object.id)
# start background import process (just a metadata update)
# run transmogrify md extraction here
context_path = '/'.join(new_object.getPhysicalPath())
member = api.user.get_current()
# species extract task
if IMultiSpeciesDataset.providedBy(new_object):
# kick off csv split import tasks
import_task = app.signature(
"org.bccvl.tasks.datamover.tasks.import_multi_species_csv",
kwargs={
'url': '{}/@@download/file/{}'.format(new_object.absolute_url(), new_object.file.filename),
'results_dir': get_results_dir(new_object, self.request, childSpecies=True),
'import_context': {
'context': '/'.join(container.getPhysicalPath()),
'user': {
'id': member.getUserName(),
'email': member.getProperty('email'),
'fullname': member.getProperty('fullname')
}
},
'context': {
'context': context_path,
'genre': self.datagenre,
'dataSource': new_object.dataSource,
'user': {
'id': member.getUserName(),
'email': member.getProperty('email'),
'fullname': member.getProperty('fullname')
}
}
},
immutable=True)
after_commit_task(import_task)
# create job tracking object
jt = IJobTracker(new_object)
jt.new_job('TODO: generate id',
'generate taskname: import_multi_species_csv',
function=new_object.dataSource,
type=new_object.portal_type)
jt.set_progress('PENDING', u'Multi species import pending')
else:
if hasattr(self, '_upload'):
file = self._upload['file']
new_object.format = file.contentType
uid = IUUID(new_object)
swiftsettings = getUtility(
IRegistry).forInterface(ISwiftSettings)
import os.path
swift_url = '{storage_url}/{container}/{path}/{name}'.format(
storage_url=swiftsettings.storage_url,
container=swiftsettings.result_container,
path=uid,
name=os.path.basename(file.filename))
new_object.remoteUrl = swift_url
else:
file = new_object.file
new_object.format = file.contentType
dlinfo = IDownloadInfo(new_object)
# single species upload
update_task = app.signature(
"org.bccvl.tasks.datamover.tasks.update_metadata",
kwargs={
'url': dlinfo['url'],
'filename': dlinfo['filename'],
'contenttype': dlinfo['contenttype'],
'context': {
'context': context_path,
'user': {
'id': member.getUserName(),
'email': member.getProperty('email'),
'fullname': member.getProperty('fullname')
}
}
},
immutable=True)
# create upload task in case we upload to external store
if hasattr(self, '_upload'):
# FIXME: we can't use ssh here.... we don't know which container we are in... and
# sshing here is bad as well....
# There is an upload ... we have to make sure the uploaded data ends up in external storage
# 3. put temp file aside
tmpdir = tempfile.mkdtemp(prefix='bccvl_upload')
tmpfile = os.path.join(tmpdir, os.path.basename(file.filename))
blobf = file.open()
try:
# try rename
os.rename(blobf.name, tmpfile)
except OSError:
# try copy
shutil.copy(blobf.name, tmpfile)
# TODO: we push the uploaded file directly to swift here..
# this really should be a background process
# best solution: ...
# user uploads to some temporary upload service (file never ends up here)
# we have a remote url here, and tell the datamover to pull it from there
# and move it to final destination. (or something like this)
# other good way: ...
# let user upload directly to swift (what about large file uploads?)
# and take care of clean up if necessary
# 4. move file to swift
# TODO: do we have enough information to upload to swift?
# need a temp url?
swiftopts = app.conf.get('bccvl', {}).get('swift', {})
src_url = build_source('file://{}'.format(tmpfile))
dest_url = build_destination('swift+{}'.format(new_object.remoteUrl),
settings={'swift': {
'os_auth_url': swiftopts.get('os_auth_url'),
'os_username': swiftopts.get('os_username'),
'os_password': swiftopts.get('os_password'),
'os_tenant_name': swiftopts.get('os_tenant_name'),
'os_storage_url': swiftopts.get('os_storage_url')
}}
)
try:
movelib.move(src_url, dest_url)
except Exception as e:
# do error handling here
raise
finally:
# clean up temp location
path = os.path.dirname(tmpfile)
shutil.rmtree(path)
# queue job submission
after_commit_task(update_task)
# create job tracking object
jt = IJobTracker(new_object)
jt.new_job('TODO: generate id',
'generate taskname: update_metadata',
function=new_object.dataSource,
type=new_object.portal_type)
jt.set_progress('PENDING', u'Metadata update pending')
# We have to reindex after updating the object
new_object.reindexObject()
def update_metadata(self):
uuid = self.request.form.get('uuid', None)
try:
if uuid:
brain = uuidToCatalogBrain(uuid)
if brain is None:
raise Exception("Brain not found")
obj = brain.getObject()
else:
obj = self.context
# get username
member = ploneapi.user.get_current()
if member.getId():
user = {
'id': member.getUserName(),
'email': member.getProperty('email'),
'fullname': member.getProperty('fullname')
}
else:
raise Exception("Invalid user")
# build download url
# 1. get context (site) relative path
obj_url = obj.absolute_url()
if obj.portal_type == 'org.bccvl.content.dataset':
filename = obj.file.filename
obj_url = '{}/@@download/file/{}'.format(obj_url, filename)
elif obj.portal_type == 'org.bccvl.content.remotedataset':
filename = os.path.basename(obj.remoteUrl)
obj_url = '{}/@@download/{}'.format(obj_url, filename)
elif obj.portal_type == 'org.bccvl.content.multispeciesdataset':
filename = obj.file.filename
obj_url = '{}/@@download/file/{}'.format(obj_url, filename)
else:
raise Exception("Wrong content type")
from org.bccvl.tasks.celery import app
update_task = app.signature(
"org.bccvl.tasks.datamover.tasks.update_metadata",
kwargs={
'url': obj_url,
'filename': filename,
'contenttype': obj.format,
'context': {
'context': '/'.join(obj.getPhysicalPath()),
'user': user,
}
},
immutable=True)
from org.bccvl.tasks.plone import after_commit_task
after_commit_task(update_task)
# track background job state
jt = IJobTracker(obj)
job = jt.new_job('TODO: generate id',
'generate taskname: update_metadata',
function=obj.dataSource,
type=obj.portal_type)
jt.set_progress('PENDING', 'Metadata update pending')
return job.id
except Exception as e:
LOG.error('Caught exception %s', e)
raise NotFound(self, 'update_metadata', self.request)
