#
# SPDX-License-Identifier: AGPL-3.0
+from __future__ import absolute_import
+from __future__ import division
+from future.utils import viewitems
+from future.utils import itervalues
+from builtins import dict
import logging
import re
import time
import errno
import time
-from fusefile import StringFile, ObjectFile, FuncToJSONFile, FuseArvadosFile
-from fresh import FreshBase, convertTime, use_counter, check_update
+from .fusefile import StringFile, ObjectFile, FuncToJSONFile, FuseArvadosFile
+from .fresh import FreshBase, convertTime, use_counter, check_update
import arvados.collection
from arvados.util import portable_data_hash_pattern, uuid_pattern, collection_uuid_pattern, group_uuid_pattern, user_uuid_pattern, link_uuid_pattern
# appear as underscores in the fuse mount.)
_disallowed_filename_characters = re.compile('[\x00/]')
-# '.' and '..' are not reachable if API server is newer than #6277
-def sanitize_filename(dirty):
- """Replace disallowed filename characters with harmless "_"."""
- if dirty is None:
- return None
- elif dirty == '':
- return '_'
- elif dirty == '.':
- return '_'
- elif dirty == '..':
- return '__'
- else:
- return _disallowed_filename_characters.sub('_', dirty)
-
class Directory(FreshBase):
"""Generic directory object, backed by a dict.
and the value referencing a File or Directory object.
"""
- def __init__(self, parent_inode, inodes):
+ def __init__(self, parent_inode, inodes, apiconfig):
"""parent_inode is the integer inode number"""
super(Directory, self).__init__()
raise Exception("parent_inode should be an int")
self.parent_inode = parent_inode
self.inodes = inodes
+ self.apiconfig = apiconfig
self._entries = {}
self._mtime = time.time()
- # Overriden by subclasses to implement logic to update the entries dict
- # when the directory is stale
+ def forward_slash_subst(self):
+ if not hasattr(self, '_fsns'):
+ self._fsns = None
+ config = self.apiconfig()
+ try:
+ self._fsns = config["Collections"]["ForwardSlashNameSubstitution"]
+ except KeyError:
+ # old API server with no FSNS config
+ self._fsns = '_'
+ else:
+ if self._fsns == '' or self._fsns == '/':
+ self._fsns = None
+ return self._fsns
+
+ def unsanitize_filename(self, incoming):
+ """Replace ForwardSlashNameSubstitution value with /"""
+ fsns = self.forward_slash_subst()
+ if isinstance(fsns, str):
+ return incoming.replace(fsns, '/')
+ else:
+ return incoming
+
+ def sanitize_filename(self, dirty):
+ """Replace disallowed filename characters according to
+ ForwardSlashNameSubstitution in self.api_config."""
+ # '.' and '..' are not reachable if API server is newer than #6277
+ if dirty is None:
+ return None
+ elif dirty == '':
+ return '_'
+ elif dirty == '.':
+ return '_'
+ elif dirty == '..':
+ return '__'
+ else:
+ fsns = self.forward_slash_subst()
+ if isinstance(fsns, str):
+ dirty = dirty.replace('/', fsns)
+ return _disallowed_filename_characters.sub('_', dirty)
+
+
+ # Overridden by subclasses to implement logic to update the
+ # entries dict when the directory is stale
@use_counter
def update(self):
pass
self._entries = {}
changed = False
for i in items:
- name = sanitize_filename(fn(i))
+ name = self.sanitize_filename(fn(i))
if name:
if name in oldentries and same(oldentries[name], i):
# move existing directory entry over
# delete any other directory entries that were not in found in 'items'
for i in oldentries:
_logger.debug("Forgetting about entry '%s' on inode %i", i, self.inode)
- self.inodes.invalidate_entry(self.inode, i.encode(self.inodes.encoding))
+ self.inodes.invalidate_entry(self, i)
self.inodes.del_entry(oldentries[i])
changed = True
if changed:
- self.inodes.invalidate_inode(self.inode)
+ self.inodes.invalidate_inode(self)
self._mtime = time.time()
self.fresh()
def in_use(self):
if super(Directory, self).in_use():
return True
- for v in self._entries.itervalues():
+ for v in itervalues(self._entries):
if v.in_use():
return True
return False
def has_ref(self, only_children):
if super(Directory, self).has_ref(only_children):
return True
- for v in self._entries.itervalues():
+ for v in itervalues(self._entries):
if v.has_ref(False):
return True
return False
self._entries = {}
for n in oldentries:
oldentries[n].clear()
- self.inodes.invalidate_entry(self.inode, n.encode(self.inodes.encoding))
self.inodes.del_entry(oldentries[n])
- self.inodes.invalidate_inode(self.inode)
self.invalidate()
def kernel_invalidate(self):
- for n, e in self._entries.iteritems():
- self.inodes.invalidate_entry(self.inode, n.encode(self.inodes.encoding))
- e.kernel_invalidate()
- self.inodes.invalidate_inode(self.inode)
+ # Invalidating the dentry on the parent implies invalidating all paths
+ # below it as well.
+ parent = self.inodes[self.parent_inode]
+
+ # Find self on the parent in order to invalidate this path.
+ # Calling the public items() method might trigger a refresh,
+ # which we definitely don't want, so read the internal dict directly.
+ for k,v in viewitems(parent._entries):
+ if v is self:
+ self.inodes.invalidate_entry(parent, k)
+ break
def mtime(self):
return self._mtime
"""
- def __init__(self, parent_inode, inodes, collection):
- super(CollectionDirectoryBase, self).__init__(parent_inode, inodes)
+ def __init__(self, parent_inode, inodes, apiconfig, collection):
+ super(CollectionDirectoryBase, self).__init__(parent_inode, inodes, apiconfig)
+ self.apiconfig = apiconfig
self.collection = collection
def new_entry(self, name, item, mtime):
- name = sanitize_filename(name)
+ name = self.sanitize_filename(name)
if hasattr(item, "fuse_entry") and item.fuse_entry is not None:
if item.fuse_entry.dead is not True:
raise Exception("Can only reparent dead inode entry")
item.fuse_entry.dead = False
self._entries[name] = item.fuse_entry
elif isinstance(item, arvados.collection.RichCollectionBase):
- self._entries[name] = self.inodes.add_entry(CollectionDirectoryBase(self.inode, self.inodes, item))
+ self._entries[name] = self.inodes.add_entry(CollectionDirectoryBase(self.inode, self.inodes, self.apiconfig, item))
self._entries[name].populate(mtime)
else:
self._entries[name] = self.inodes.add_entry(FuseArvadosFile(self.inode, item, mtime))
def on_event(self, event, collection, name, item):
if collection == self.collection:
- name = sanitize_filename(name)
+ name = self.sanitize_filename(name)
_logger.debug("collection notify %s %s %s %s", event, collection, name, item)
with llfuse.lock:
if event == arvados.collection.ADD:
elif event == arvados.collection.DEL:
ent = self._entries[name]
del self._entries[name]
- self.inodes.invalidate_entry(self.inode, name.encode(self.inodes.encoding))
+ self.inodes.invalidate_entry(self, name)
self.inodes.del_entry(ent)
elif event == arvados.collection.MOD:
if hasattr(item, "fuse_entry") and item.fuse_entry is not None:
- self.inodes.invalidate_inode(item.fuse_entry.inode)
+ self.inodes.invalidate_inode(item.fuse_entry)
elif name in self._entries:
- self.inodes.invalidate_inode(self._entries[name].inode)
+ self.inodes.invalidate_inode(self._entries[name])
def populate(self, mtime):
self._mtime = mtime
self.collection.subscribe(self.on_event)
- for entry, item in self.collection.items():
+ for entry, item in viewitems(self.collection):
self.new_entry(entry, item, self.mtime())
def writable(self):
"""Represents the root of a directory tree representing a collection."""
def __init__(self, parent_inode, inodes, api, num_retries, collection_record=None, explicit_collection=None):
- super(CollectionDirectory, self).__init__(parent_inode, inodes, None)
+ super(CollectionDirectory, self).__init__(parent_inode, inodes, api.config, None)
self.api = api
self.num_retries = num_retries
self.collection_record_file = None
self.collection_record = None
self._poll = True
try:
- self._poll_time = (api._rootDesc.get('blobSignatureTtl', 60*60*2)/2)
+ self._poll_time = (api._rootDesc.get('blobSignatureTtl', 60*60*2) // 2)
except:
_logger.debug("Error getting blobSignatureTtl from discovery document: %s", sys.exc_info()[0])
self._poll_time = 60*60
keep_client=api_client.keep,
num_retries=num_retries)
super(TmpCollectionDirectory, self).__init__(
- parent_inode, inodes, collection)
+ parent_inode, inodes, api_client.config, collection)
self.collection_record_file = None
self.populate(self.mtime())
if self.collection_record_file:
with llfuse.lock:
self.collection_record_file.invalidate()
- self.inodes.invalidate_inode(self.collection_record_file.inode)
+ self.inodes.invalidate_inode(self.collection_record_file)
_logger.debug("%s invalidated collection record", self)
def collection_record(self):
README_TEXT = """
This directory provides access to Arvados collections as subdirectories listed
by uuid (in the form 'zzzzz-4zz18-1234567890abcde') or portable data hash (in
-the form '1234567890abcdef0123456789abcdef+123').
+the form '1234567890abcdef0123456789abcdef+123'), and Arvados projects by uuid
+(in the form 'zzzzz-j7d0g-1234567890abcde').
Note that this directory will appear empty until you attempt to access a
-specific collection subdirectory (such as trying to 'cd' into it), at which
-point the collection will actually be looked up on the server and the directory
-will appear if it exists.
+specific collection or project subdirectory (such as trying to 'cd' into it),
+at which point the collection or project will actually be looked up on the server
+and the directory will appear if it exists.
""".lstrip()
def __init__(self, parent_inode, inodes, api, num_retries, pdh_only=False):
- super(MagicDirectory, self).__init__(parent_inode, inodes)
+ super(MagicDirectory, self).__init__(parent_inode, inodes, api.config)
self.api = api
self.num_retries = num_retries
self.pdh_only = pdh_only
return False
try:
- e = self.inodes.add_entry(CollectionDirectory(
- self.inode, self.inodes, self.api, self.num_retries, k))
+ e = None
+
+ if group_uuid_pattern.match(k):
+ project = self.api.groups().list(
+ filters=[['group_class', '=', 'project'], ["uuid", "=", k]]).execute(num_retries=self.num_retries)
+ if project[u'items_available'] == 0:
+ return False
+ e = self.inodes.add_entry(ProjectDirectory(
+ self.inode, self.inodes, self.api, self.num_retries, project[u'items'][0]))
+ else:
+ import sys
+ e = self.inodes.add_entry(CollectionDirectory(
+ self.inode, self.inodes, self.api, self.num_retries, k))
if e.update():
if k not in self._entries:
self.inodes.del_entry(e)
return True
else:
- self.inodes.invalidate_entry(self.inode, k)
+ self.inodes.invalidate_entry(self, k)
self.inodes.del_entry(e)
return False
except Exception as ex:
- _logger.debug('arv-mount exception keep %s', ex)
- self.inodes.del_entry(e)
+ _logger.exception("arv-mount lookup '%s':", k)
+ if e is not None:
+ self.inodes.del_entry(e)
return False
def __getitem__(self, item):
"""A special directory that contains as subdirectories all tags visible to the user."""
def __init__(self, parent_inode, inodes, api, num_retries, poll_time=60):
- super(TagsDirectory, self).__init__(parent_inode, inodes)
+ super(TagsDirectory, self).__init__(parent_inode, inodes, api.config)
self.api = api
self.num_retries = num_retries
self._poll = True
def __init__(self, parent_inode, inodes, api, num_retries, tag,
poll=False, poll_time=60):
- super(TagDirectory, self).__init__(parent_inode, inodes)
+ super(TagDirectory, self).__init__(parent_inode, inodes, api.config)
self.api = api
self.num_retries = num_retries
self.tag = tag
def __init__(self, parent_inode, inodes, api, num_retries, project_object,
poll=False, poll_time=60):
- super(ProjectDirectory, self).__init__(parent_inode, inodes)
+ super(ProjectDirectory, self).__init__(parent_inode, inodes, api.config)
self.api = api
self.num_retries = num_retries
self.project_object = project_object
if 'name' in i:
if i['name'] is None or len(i['name']) == 0:
return None
- elif collection_uuid_pattern.match(i['uuid']) or group_uuid_pattern.match(i['uuid']):
+ elif "uuid" in i and (collection_uuid_pattern.match(i['uuid']) or group_uuid_pattern.match(i['uuid'])):
# collection or subproject
return i['name']
elif link_uuid_pattern.match(i['uuid']) and i['head_kind'] == 'arvados#collection':
self.inodes.add_entry(self.project_object_file)
if not self._full_listing:
- return
+ return True
def samefn(a, i):
if isinstance(a, CollectionDirectory) or isinstance(a, ProjectDirectory):
contents = arvados.util.list_all(self.api.groups().list,
self.num_retries,
filters=[["owner_uuid", "=", self.project_uuid],
- ["group_class", "=", "project"]],
- limit=1000, order="modified_at desc")
+ ["group_class", "=", "project"]])
contents.extend(arvados.util.list_all(self.api.collections().list,
self.num_retries,
- filters=[["owner_uuid", "=", self.project_uuid]],
- limit=1000, order="modified_at desc"))
+ filters=[["owner_uuid", "=", self.project_uuid]]))
# end with llfuse.lock_released, re-acquire lock
self.namefn,
samefn,
self.createDirectory)
+ return True
finally:
- self._full_listing = False
self._updating_lock.release()
+ def _add_entry(self, i, name):
+ ent = self.createDirectory(i)
+ self._entries[name] = self.inodes.add_entry(ent)
+ return self._entries[name]
+
@use_counter
@check_update
- def __getitem__(self, item):
- if item == '.arvados#project':
+ def __getitem__(self, k):
+ if k == '.arvados#project':
return self.project_object_file
- else:
- return super(ProjectDirectory, self).__getitem__(item)
+ elif self._full_listing or super(ProjectDirectory, self).__contains__(k):
+ return super(ProjectDirectory, self).__getitem__(k)
+ with llfuse.lock_released:
+ k2 = self.unsanitize_filename(k)
+ if k2 == k:
+ namefilter = ["name", "=", k]
+ else:
+ namefilter = ["name", "in", [k, k2]]
+ contents = self.api.groups().list(filters=[["owner_uuid", "=", self.project_uuid],
+ ["group_class", "=", "project"],
+ namefilter],
+ limit=2).execute(num_retries=self.num_retries)["items"]
+ if not contents:
+ contents = self.api.collections().list(filters=[["owner_uuid", "=", self.project_uuid],
+ namefilter],
+ limit=2).execute(num_retries=self.num_retries)["items"]
+ if contents:
+ if len(contents) > 1 and contents[1]['name'] == k:
+ # If "foo/bar" and "foo[SUBST]bar" both exist, use
+ # "foo[SUBST]bar".
+ contents = [contents[1]]
+ name = self.sanitize_filename(self.namefn(contents[0]))
+ if name != k:
+ raise KeyError(k)
+ return self._add_entry(contents[0], name)
+
+ # Didn't find item
+ raise KeyError(k)
def __contains__(self, k):
if k == '.arvados#project':
return True
- else:
- if super(ProjectDirectory, self).__contains__(k):
- return True
- elif not self._full_listing:
- with llfuse.lock_released:
- contents = self.api.groups().list(filters=[["owner_uuid", "=", self.project_uuid],
- ["group_class", "=", "project"],
- ["name", "=", k]],
- limit=1).execute(num_retries=self.num_retries)["items"]
- if not contents:
- contents = self.api.collections().list(filters=[["owner_uuid", "=", self.project_uuid],
- ["name", "=", k]],
- limit=1).execute(num_retries=self.num_retries)["items"]
- if contents:
- i = contents[0]
- name = sanitize_filename(self.namefn(i))
- ent = self.createDirectory(i)
- self._entries[name] = self.inodes.add_entry(ent)
- return True
+ try:
+ self[k]
+ return True
+ except KeyError:
+ pass
return False
@use_counter
with llfuse.lock_released:
if not self._current_user:
self._current_user = self.api.users().current().execute(num_retries=self.num_retries)
- return self._current_user["uuid"] in self.project_object["writable_by"]
+ return self._current_user["uuid"] in self.project_object.get("writable_by", [])
def persisted(self):
return True
# Acually move the entry from source directory to this directory.
del src._entries[name_old]
self._entries[name_new] = ent
- self.inodes.invalidate_entry(src.inode, name_old.encode(self.inodes.encoding))
+ self.inodes.invalidate_entry(src, name_old)
+
+ @use_counter
+ def child_event(self, ev):
+ properties = ev.get("properties") or {}
+ old_attrs = properties.get("old_attributes") or {}
+ new_attrs = properties.get("new_attributes") or {}
+ old_attrs["uuid"] = ev["object_uuid"]
+ new_attrs["uuid"] = ev["object_uuid"]
+ old_name = self.sanitize_filename(self.namefn(old_attrs))
+ new_name = self.sanitize_filename(self.namefn(new_attrs))
+
+ # create events will have a new name, but not an old name
+ # delete events will have an old name, but not a new name
+ # update events will have an old and new name, and they may be same or different
+ # if they are the same, an unrelated field changed and there is nothing to do.
+
+ if old_attrs.get("owner_uuid") != self.project_uuid:
+ # Was moved from somewhere else, so don't try to remove entry.
+ old_name = None
+ if ev.get("object_owner_uuid") != self.project_uuid:
+ # Was moved to somewhere else, so don't try to add entry
+ new_name = None
+
+ if old_attrs.get("is_trashed"):
+ # Was previously deleted
+ old_name = None
+ if new_attrs.get("is_trashed"):
+ # Has been deleted
+ new_name = None
+
+ if new_name != old_name:
+ ent = None
+ if old_name in self._entries:
+ ent = self._entries[old_name]
+ del self._entries[old_name]
+ self.inodes.invalidate_entry(self, old_name)
+
+ if new_name:
+ if ent is not None:
+ self._entries[new_name] = ent
+ else:
+ self._add_entry(new_attrs, new_name)
+ elif ent is not None:
+ self.inodes.del_entry(ent)
class SharedDirectory(Directory):
def __init__(self, parent_inode, inodes, api, num_retries, exclude,
poll=False, poll_time=60):
- super(SharedDirectory, self).__init__(parent_inode, inodes)
+ super(SharedDirectory, self).__init__(parent_inode, inodes, api.config)
self.api = api
self.num_retries = num_retries
self.current_user = api.users().current().execute(num_retries=num_retries)
self._poll = True
self._poll_time = poll_time
+ self._updating_lock = threading.Lock()
@use_counter
def update(self):
- with llfuse.lock_released:
- all_projects = arvados.util.list_all(
- self.api.groups().list, self.num_retries,
- filters=[['group_class','=','project']])
- objects = {}
- for ob in all_projects:
- objects[ob['uuid']] = ob
-
- roots = []
- root_owners = {}
- for ob in all_projects:
- if ob['owner_uuid'] != self.current_user['uuid'] and ob['owner_uuid'] not in objects:
- roots.append(ob)
- root_owners[ob['owner_uuid']] = True
-
- lusers = arvados.util.list_all(
- self.api.users().list, self.num_retries,
- filters=[['uuid','in', list(root_owners)]])
- lgroups = arvados.util.list_all(
- self.api.groups().list, self.num_retries,
- filters=[['uuid','in', list(root_owners)]])
-
- users = {}
- groups = {}
-
- for l in lusers:
- objects[l["uuid"]] = l
- for l in lgroups:
- objects[l["uuid"]] = l
-
- contents = {}
- for r in root_owners:
- if r in objects:
- obr = objects[r]
- if obr.get("name"):
- contents[obr["name"]] = obr
- #elif obr.get("username"):
- # contents[obr["username"]] = obr
- elif "first_name" in obr:
- contents[u"{} {}".format(obr["first_name"], obr["last_name"])] = obr
-
-
- for r in roots:
- if r['owner_uuid'] not in objects:
- contents[r['name']] = r
-
- # end with llfuse.lock_released, re-acquire lock
-
try:
- self.merge(contents.items(),
+ with llfuse.lock_released:
+ self._updating_lock.acquire()
+ if not self.stale():
+ return
+
+ contents = {}
+ roots = []
+ root_owners = set()
+ objects = {}
+
+ methods = self.api._rootDesc.get('resources')["groups"]['methods']
+ if 'httpMethod' in methods.get('shared', {}):
+ page = []
+ while True:
+ resp = self.api.groups().shared(filters=[['group_class', '=', 'project']]+page,
+ order="uuid",
+ limit=10000,
+ count="none",
+ include="owner_uuid").execute()
+ if not resp["items"]:
+ break
+ page = [["uuid", ">", resp["items"][len(resp["items"])-1]["uuid"]]]
+ for r in resp["items"]:
+ objects[r["uuid"]] = r
+ roots.append(r["uuid"])
+ for r in resp["included"]:
+ objects[r["uuid"]] = r
+ root_owners.add(r["uuid"])
+ else:
+ all_projects = arvados.util.list_all(
+ self.api.groups().list, self.num_retries,
+ filters=[['group_class','=','project']],
+ select=["uuid", "owner_uuid"])
+ for ob in all_projects:
+ objects[ob['uuid']] = ob
+
+ current_uuid = self.current_user['uuid']
+ for ob in all_projects:
+ if ob['owner_uuid'] != current_uuid and ob['owner_uuid'] not in objects:
+ roots.append(ob['uuid'])
+ root_owners.add(ob['owner_uuid'])
+
+ lusers = arvados.util.list_all(
+ self.api.users().list, self.num_retries,
+ filters=[['uuid','in', list(root_owners)]])
+ lgroups = arvados.util.list_all(
+ self.api.groups().list, self.num_retries,
+ filters=[['uuid','in', list(root_owners)+roots]])
+
+ for l in lusers:
+ objects[l["uuid"]] = l
+ for l in lgroups:
+ objects[l["uuid"]] = l
+
+ for r in root_owners:
+ if r in objects:
+ obr = objects[r]
+ if obr.get("name"):
+ contents[obr["name"]] = obr
+ #elif obr.get("username"):
+ # contents[obr["username"]] = obr
+ elif "first_name" in obr:
+ contents[u"{} {}".format(obr["first_name"], obr["last_name"])] = obr
+
+ for r in roots:
+ if r in objects:
+ obr = objects[r]
+ if obr['owner_uuid'] not in objects:
+ contents[obr["name"]] = obr
+
+ # end with llfuse.lock_released, re-acquire lock
+
+ self.merge(viewitems(contents),
lambda i: i[0],
lambda a, i: a.uuid() == i[1]['uuid'],
lambda i: ProjectDirectory(self.inode, self.inodes, self.api, self.num_retries, i[1], poll=self._poll, poll_time=self._poll_time))
except Exception:
- _logger.exception()
+ _logger.exception("arv-mount shared dir error")
+ finally:
+ self._updating_lock.release()
def want_event_subscribe(self):
return True
projects / arvados.git / blobdiff