1 # Copyright (C) The Arvados Authors. All rights reserved.
3 # SPDX-License-Identifier: Apache-2.0
5 from __future__ import absolute_import
6 from __future__ import division
7 from future import standard_library
8 from future.utils import listitems, listvalues
9 standard_library.install_aliases()
10 from builtins import range
11 from builtins import object
28 from .errors import KeepWriteError, AssertionError, ArgumentError
29 from .keep import KeepLocator
30 from ._normalize_stream import normalize_stream
31 from ._ranges import locators_and_ranges, replace_range, Range, LocatorAndRange
32 from .retry import retry_method
37 _logger = logging.getLogger('arvados.arvfile')
40 """split(path) -> streamname, filename
42 Separate the stream name and file name in a /-separated stream path and
43 return a tuple (stream_name, file_name). If no stream name is available,
48 stream_name, file_name = path.rsplit('/', 1)
49 except ValueError: # No / in string
50 stream_name, file_name = '.', path
51 return stream_name, file_name
54 class UnownedBlockError(Exception):
55 """Raised when there's an writable block without an owner on the BlockManager."""
59 class _FileLikeObjectBase(object):
60 def __init__(self, name, mode):
66 def _before_close(orig_func):
67 @functools.wraps(orig_func)
68 def before_close_wrapper(self, *args, **kwargs):
70 raise ValueError("I/O operation on closed stream file")
71 return orig_func(self, *args, **kwargs)
72 return before_close_wrapper
77 def __exit__(self, exc_type, exc_value, traceback):
88 class ArvadosFileReaderBase(_FileLikeObjectBase):
89 def __init__(self, name, mode, num_retries=None):
90 super(ArvadosFileReaderBase, self).__init__(name, mode)
92 self.num_retries = num_retries
93 self._readline_cache = (None, None)
97 data = self.readline()
102 def decompressed_name(self):
103 return re.sub('\.(bz2|gz)$', '', self.name)
105 @_FileLikeObjectBase._before_close
106 def seek(self, pos, whence=os.SEEK_SET):
107 if whence == os.SEEK_CUR:
109 elif whence == os.SEEK_END:
112 raise IOError(errno.EINVAL, "Tried to seek to negative file offset.")
128 @_FileLikeObjectBase._before_close
130 def readall(self, size=2**20, num_retries=None):
132 data = self.read(size, num_retries=num_retries)
137 @_FileLikeObjectBase._before_close
139 def readline(self, size=float('inf'), num_retries=None):
140 cache_pos, cache_data = self._readline_cache
141 if self.tell() == cache_pos:
143 self._filepos += len(cache_data)
146 data_size = len(data[-1])
147 while (data_size < size) and (b'\n' not in data[-1]):
148 next_read = self.read(2 ** 20, num_retries=num_retries)
151 data.append(next_read)
152 data_size += len(next_read)
153 data = b''.join(data)
155 nextline_index = data.index(b'\n') + 1
157 nextline_index = len(data)
158 nextline_index = min(nextline_index, size)
159 self._filepos -= len(data) - nextline_index
160 self._readline_cache = (self.tell(), data[nextline_index:])
161 return data[:nextline_index].decode()
163 @_FileLikeObjectBase._before_close
165 def decompress(self, decompress, size, num_retries=None):
166 for segment in self.readall(size, num_retries=num_retries):
167 data = decompress(segment)
171 @_FileLikeObjectBase._before_close
173 def readall_decompressed(self, size=2**20, num_retries=None):
175 if self.name.endswith('.bz2'):
176 dc = bz2.BZ2Decompressor()
177 return self.decompress(dc.decompress, size,
178 num_retries=num_retries)
179 elif self.name.endswith('.gz'):
180 dc = zlib.decompressobj(16+zlib.MAX_WBITS)
181 return self.decompress(lambda segment: dc.decompress(dc.unconsumed_tail + segment),
182 size, num_retries=num_retries)
184 return self.readall(size, num_retries=num_retries)
186 @_FileLikeObjectBase._before_close
188 def readlines(self, sizehint=float('inf'), num_retries=None):
191 for s in self.readall(num_retries=num_retries):
194 if data_size >= sizehint:
196 return b''.join(data).decode().splitlines(True)
199 raise IOError(errno.ENOSYS, "Not implemented")
201 def read(self, size, num_retries=None):
202 raise IOError(errno.ENOSYS, "Not implemented")
204 def readfrom(self, start, size, num_retries=None):
205 raise IOError(errno.ENOSYS, "Not implemented")
208 class StreamFileReader(ArvadosFileReaderBase):
209 class _NameAttribute(str):
210 # The Python file API provides a plain .name attribute.
211 # Older SDK provided a name() method.
212 # This class provides both, for maximum compatibility.
216 def __init__(self, stream, segments, name):
217 super(StreamFileReader, self).__init__(self._NameAttribute(name), 'rb', num_retries=stream.num_retries)
218 self._stream = stream
219 self.segments = segments
221 def stream_name(self):
222 return self._stream.name()
225 n = self.segments[-1]
226 return n.range_start + n.range_size
228 @_FileLikeObjectBase._before_close
230 def read(self, size, num_retries=None):
231 """Read up to 'size' bytes from the stream, starting at the current file position"""
236 available_chunks = locators_and_ranges(self.segments, self._filepos, size)
238 lr = available_chunks[0]
239 data = self._stream.readfrom(lr.locator+lr.segment_offset,
241 num_retries=num_retries)
243 self._filepos += len(data)
246 @_FileLikeObjectBase._before_close
248 def readfrom(self, start, size, num_retries=None):
249 """Read up to 'size' bytes from the stream, starting at 'start'"""
254 for lr in locators_and_ranges(self.segments, start, size):
255 data.append(self._stream.readfrom(lr.locator+lr.segment_offset, lr.segment_size,
256 num_retries=num_retries))
257 return b''.join(data)
259 def as_manifest(self):
261 for r in self.segments:
262 segs.extend(self._stream.locators_and_ranges(r.locator, r.range_size))
263 return " ".join(normalize_stream(".", {self.name: segs})) + "\n"
266 def synchronized(orig_func):
267 @functools.wraps(orig_func)
268 def synchronized_wrapper(self, *args, **kwargs):
270 return orig_func(self, *args, **kwargs)
271 return synchronized_wrapper
274 class StateChangeError(Exception):
275 def __init__(self, message, state, nextstate):
276 super(StateChangeError, self).__init__(message)
278 self.nextstate = nextstate
280 class _BufferBlock(object):
281 """A stand-in for a Keep block that is in the process of being written.
283 Writers can append to it, get the size, and compute the Keep locator.
284 There are three valid states:
290 Block is in the process of being uploaded to Keep, append is an error.
293 The block has been written to Keep, its internal buffer has been
294 released, fetching the block will fetch it via keep client (since we
295 discarded the internal copy), and identifiers referring to the BufferBlock
296 can be replaced with the block locator.
306 def __init__(self, blockid, starting_capacity, owner):
309 the identifier for this block
312 the initial buffer capacity
315 ArvadosFile that owns this block
318 self.blockid = blockid
319 self.buffer_block = bytearray(starting_capacity)
320 self.buffer_view = memoryview(self.buffer_block)
321 self.write_pointer = 0
322 self._state = _BufferBlock.WRITABLE
325 self.lock = threading.Lock()
326 self.wait_for_commit = threading.Event()
330 def append(self, data):
331 """Append some data to the buffer.
333 Only valid if the block is in WRITABLE state. Implements an expanding
334 buffer, doubling capacity as needed to accomdate all the data.
337 if self._state == _BufferBlock.WRITABLE:
338 if not isinstance(data, bytes) and not isinstance(data, memoryview):
340 while (self.write_pointer+len(data)) > len(self.buffer_block):
341 new_buffer_block = bytearray(len(self.buffer_block) * 2)
342 new_buffer_block[0:self.write_pointer] = self.buffer_block[0:self.write_pointer]
343 self.buffer_block = new_buffer_block
344 self.buffer_view = memoryview(self.buffer_block)
345 self.buffer_view[self.write_pointer:self.write_pointer+len(data)] = data
346 self.write_pointer += len(data)
349 raise AssertionError("Buffer block is not writable")
351 STATE_TRANSITIONS = frozenset([
353 (PENDING, COMMITTED),
358 def set_state(self, nextstate, val=None):
359 if (self._state, nextstate) not in self.STATE_TRANSITIONS:
360 raise StateChangeError("Invalid state change from %s to %s" % (self._state, nextstate), self._state, nextstate)
361 self._state = nextstate
363 if self._state == _BufferBlock.PENDING:
364 self.wait_for_commit.clear()
366 if self._state == _BufferBlock.COMMITTED:
368 self.buffer_view = None
369 self.buffer_block = None
370 self.wait_for_commit.set()
372 if self._state == _BufferBlock.ERROR:
374 self.wait_for_commit.set()
381 """The amount of data written to the buffer."""
382 return self.write_pointer
386 """The Keep locator for this buffer's contents."""
387 if self._locator is None:
388 self._locator = "%s+%i" % (hashlib.md5(self.buffer_view[0:self.write_pointer]).hexdigest(), self.size())
392 def clone(self, new_blockid, owner):
393 if self._state == _BufferBlock.COMMITTED:
394 raise AssertionError("Cannot duplicate committed buffer block")
395 bufferblock = _BufferBlock(new_blockid, self.size(), owner)
396 bufferblock.append(self.buffer_view[0:self.size()])
401 self._state = _BufferBlock.DELETED
403 self.buffer_block = None
404 self.buffer_view = None
407 def repack_writes(self):
408 """Optimize buffer block by repacking segments in file sequence.
410 When the client makes random writes, they appear in the buffer block in
411 the sequence they were written rather than the sequence they appear in
412 the file. This makes for inefficient, fragmented manifests. Attempt
413 to optimize by repacking writes in file sequence.
416 if self._state != _BufferBlock.WRITABLE:
417 raise AssertionError("Cannot repack non-writable block")
419 segs = self.owner.segments()
421 # Collect the segments that reference the buffer block.
422 bufferblock_segs = [s for s in segs if s.locator == self.blockid]
424 # Collect total data referenced by segments (could be smaller than
425 # bufferblock size if a portion of the file was written and
427 write_total = sum([s.range_size for s in bufferblock_segs])
429 if write_total < self.size() or len(bufferblock_segs) > 1:
430 # If there's more than one segment referencing this block, it is
431 # due to out-of-order writes and will produce a fragmented
432 # manifest, so try to optimize by re-packing into a new buffer.
433 contents = self.buffer_view[0:self.write_pointer].tobytes()
434 new_bb = _BufferBlock(None, write_total, None)
435 for t in bufferblock_segs:
436 new_bb.append(contents[t.segment_offset:t.segment_offset+t.range_size])
437 t.segment_offset = new_bb.size() - t.range_size
439 self.buffer_block = new_bb.buffer_block
440 self.buffer_view = new_bb.buffer_view
441 self.write_pointer = new_bb.write_pointer
444 self.owner.set_segments(segs)
447 return "<BufferBlock %s>" % (self.blockid)
450 class NoopLock(object):
454 def __exit__(self, exc_type, exc_value, traceback):
457 def acquire(self, blocking=False):
464 def must_be_writable(orig_func):
465 @functools.wraps(orig_func)
466 def must_be_writable_wrapper(self, *args, **kwargs):
467 if not self.writable():
468 raise IOError(errno.EROFS, "Collection is read-only.")
469 return orig_func(self, *args, **kwargs)
470 return must_be_writable_wrapper
473 class _BlockManager(object):
474 """BlockManager handles buffer blocks.
476 Also handles background block uploads, and background block prefetch for a
477 Collection of ArvadosFiles.
481 DEFAULT_PUT_THREADS = 2
482 DEFAULT_GET_THREADS = 2
484 def __init__(self, keep, copies=None, put_threads=None, num_retries=None):
485 """keep: KeepClient object to use"""
487 self._bufferblocks = collections.OrderedDict()
488 self._put_queue = None
489 self._put_threads = None
490 self._prefetch_queue = None
491 self._prefetch_threads = None
492 self.lock = threading.Lock()
493 self.prefetch_enabled = True
495 self.num_put_threads = put_threads
497 self.num_put_threads = _BlockManager.DEFAULT_PUT_THREADS
498 self.num_get_threads = _BlockManager.DEFAULT_GET_THREADS
500 self._pending_write_size = 0
501 self.threads_lock = threading.Lock()
502 self.padding_block = None
503 self.num_retries = num_retries
506 def alloc_bufferblock(self, blockid=None, starting_capacity=2**14, owner=None):
507 """Allocate a new, empty bufferblock in WRITABLE state and return it.
510 optional block identifier, otherwise one will be automatically assigned
513 optional capacity, otherwise will use default capacity
516 ArvadosFile that owns this block
519 return self._alloc_bufferblock(blockid, starting_capacity, owner)
521 def _alloc_bufferblock(self, blockid=None, starting_capacity=2**14, owner=None):
523 blockid = str(uuid.uuid4())
524 bufferblock = _BufferBlock(blockid, starting_capacity=starting_capacity, owner=owner)
525 self._bufferblocks[bufferblock.blockid] = bufferblock
529 def dup_block(self, block, owner):
530 """Create a new bufferblock initialized with the content of an existing bufferblock.
533 the buffer block to copy.
536 ArvadosFile that owns the new block
539 new_blockid = str(uuid.uuid4())
540 bufferblock = block.clone(new_blockid, owner)
541 self._bufferblocks[bufferblock.blockid] = bufferblock
545 def is_bufferblock(self, locator):
546 return locator in self._bufferblocks
548 def _commit_bufferblock_worker(self):
549 """Background uploader thread."""
553 bufferblock = self._put_queue.get()
554 if bufferblock is None:
557 if self.copies is None:
558 loc = self._keep.put(bufferblock.buffer_view[0:bufferblock.write_pointer].tobytes(), num_retries=self.num_retries)
560 loc = self._keep.put(bufferblock.buffer_view[0:bufferblock.write_pointer].tobytes(), num_retries=self.num_retries, copies=self.copies)
561 bufferblock.set_state(_BufferBlock.COMMITTED, loc)
562 except Exception as e:
563 bufferblock.set_state(_BufferBlock.ERROR, e)
565 if self._put_queue is not None:
566 self._put_queue.task_done()
568 def start_put_threads(self):
569 with self.threads_lock:
570 if self._put_threads is None:
571 # Start uploader threads.
573 # If we don't limit the Queue size, the upload queue can quickly
574 # grow to take up gigabytes of RAM if the writing process is
575 # generating data more quickly than it can be send to the Keep
578 # With two upload threads and a queue size of 2, this means up to 4
579 # blocks pending. If they are full 64 MiB blocks, that means up to
580 # 256 MiB of internal buffering, which is the same size as the
581 # default download block cache in KeepClient.
582 self._put_queue = queue.Queue(maxsize=2)
584 self._put_threads = []
585 for i in range(0, self.num_put_threads):
586 thread = threading.Thread(target=self._commit_bufferblock_worker)
587 self._put_threads.append(thread)
591 def _block_prefetch_worker(self):
592 """The background downloader thread."""
595 b = self._prefetch_queue.get()
600 _logger.exception("Exception doing block prefetch")
603 def start_get_threads(self):
604 if self._prefetch_threads is None:
605 self._prefetch_queue = queue.Queue()
606 self._prefetch_threads = []
607 for i in range(0, self.num_get_threads):
608 thread = threading.Thread(target=self._block_prefetch_worker)
609 self._prefetch_threads.append(thread)
615 def stop_threads(self):
616 """Shut down and wait for background upload and download threads to finish."""
618 if self._put_threads is not None:
619 for t in self._put_threads:
620 self._put_queue.put(None)
621 for t in self._put_threads:
623 self._put_threads = None
624 self._put_queue = None
626 if self._prefetch_threads is not None:
627 for t in self._prefetch_threads:
628 self._prefetch_queue.put(None)
629 for t in self._prefetch_threads:
631 self._prefetch_threads = None
632 self._prefetch_queue = None
637 def __exit__(self, exc_type, exc_value, traceback):
641 def repack_small_blocks(self, force=False, sync=False, closed_file_size=0):
642 """Packs small blocks together before uploading"""
644 self._pending_write_size += closed_file_size
646 # Check if there are enough small blocks for filling up one in full
647 if not (force or (self._pending_write_size >= config.KEEP_BLOCK_SIZE)):
650 # Search blocks ready for getting packed together before being
652 # A WRITABLE block always has an owner.
653 # A WRITABLE block with its owner.closed() implies that its
654 # size is <= KEEP_BLOCK_SIZE/2.
656 small_blocks = [b for b in listvalues(self._bufferblocks)
657 if b.state() == _BufferBlock.WRITABLE and b.owner.closed()]
658 except AttributeError:
659 # Writable blocks without owner shouldn't exist.
660 raise UnownedBlockError()
662 if len(small_blocks) <= 1:
663 # Not enough small blocks for repacking
666 for bb in small_blocks:
669 # Update the pending write size count with its true value, just in case
670 # some small file was opened, written and closed several times.
671 self._pending_write_size = sum([b.size() for b in small_blocks])
673 if self._pending_write_size < config.KEEP_BLOCK_SIZE and not force:
676 new_bb = self._alloc_bufferblock()
679 while len(small_blocks) > 0 and (new_bb.write_pointer + small_blocks[0].size()) <= config.KEEP_BLOCK_SIZE:
680 bb = small_blocks.pop(0)
681 new_bb.owner.append(bb.owner)
682 self._pending_write_size -= bb.size()
683 new_bb.append(bb.buffer_view[0:bb.write_pointer].tobytes())
684 files.append((bb, new_bb.write_pointer - bb.size()))
686 self.commit_bufferblock(new_bb, sync=sync)
688 for bb, new_bb_segment_offset in files:
689 newsegs = bb.owner.segments()
691 if s.locator == bb.blockid:
692 s.locator = new_bb.blockid
693 s.segment_offset = new_bb_segment_offset+s.segment_offset
694 bb.owner.set_segments(newsegs)
695 self._delete_bufferblock(bb.blockid)
697 def commit_bufferblock(self, block, sync):
698 """Initiate a background upload of a bufferblock.
701 The block object to upload
704 If `sync` is True, upload the block synchronously.
705 If `sync` is False, upload the block asynchronously. This will
706 return immediately unless the upload queue is at capacity, in
707 which case it will wait on an upload queue slot.
711 # Mark the block as PENDING so to disallow any more appends.
712 block.set_state(_BufferBlock.PENDING)
713 except StateChangeError as e:
714 if e.state == _BufferBlock.PENDING:
716 block.wait_for_commit.wait()
719 if block.state() == _BufferBlock.COMMITTED:
721 elif block.state() == _BufferBlock.ERROR:
728 if self.copies is None:
729 loc = self._keep.put(block.buffer_view[0:block.write_pointer].tobytes(), num_retries=self.num_retries)
731 loc = self._keep.put(block.buffer_view[0:block.write_pointer].tobytes(), num_retries=self.num_retries, copies=self.copies)
732 block.set_state(_BufferBlock.COMMITTED, loc)
733 except Exception as e:
734 block.set_state(_BufferBlock.ERROR, e)
737 self.start_put_threads()
738 self._put_queue.put(block)
741 def get_bufferblock(self, locator):
742 return self._bufferblocks.get(locator)
745 def get_padding_block(self):
746 """Get a bufferblock 64 MB in size consisting of all zeros, used as padding
747 when using truncate() to extend the size of a file.
749 For reference (and possible future optimization), the md5sum of the
750 padding block is: 7f614da9329cd3aebf59b91aadc30bf0+67108864
754 if self.padding_block is None:
755 self.padding_block = self._alloc_bufferblock(starting_capacity=config.KEEP_BLOCK_SIZE)
756 self.padding_block.write_pointer = config.KEEP_BLOCK_SIZE
757 self.commit_bufferblock(self.padding_block, False)
758 return self.padding_block
761 def delete_bufferblock(self, locator):
762 self._delete_bufferblock(locator)
764 def _delete_bufferblock(self, locator):
765 bb = self._bufferblocks[locator]
767 del self._bufferblocks[locator]
769 def get_block_contents(self, locator, num_retries, cache_only=False):
772 First checks to see if the locator is a BufferBlock and return that, if
773 not, passes the request through to KeepClient.get().
777 if locator in self._bufferblocks:
778 bufferblock = self._bufferblocks[locator]
779 if bufferblock.state() != _BufferBlock.COMMITTED:
780 return bufferblock.buffer_view[0:bufferblock.write_pointer].tobytes()
782 locator = bufferblock._locator
784 return self._keep.get_from_cache(locator)
786 return self._keep.get(locator, num_retries=num_retries)
788 def commit_all(self):
789 """Commit all outstanding buffer blocks.
791 This is a synchronous call, and will not return until all buffer blocks
792 are uploaded. Raises KeepWriteError() if any blocks failed to upload.
795 self.repack_small_blocks(force=True, sync=True)
798 items = listitems(self._bufferblocks)
801 if v.state() != _BufferBlock.COMMITTED and v.owner:
802 # Ignore blocks with a list of owners, as if they're not in COMMITTED
803 # state, they're already being committed asynchronously.
804 if isinstance(v.owner, ArvadosFile):
805 v.owner.flush(sync=False)
808 if self._put_queue is not None:
809 self._put_queue.join()
813 if v.state() == _BufferBlock.ERROR:
814 err.append((v.locator(), v.error))
816 raise KeepWriteError("Error writing some blocks", err, label="block")
819 # flush again with sync=True to remove committed bufferblocks from
822 if isinstance(v.owner, ArvadosFile):
823 v.owner.flush(sync=True)
824 elif isinstance(v.owner, list) and len(v.owner) > 0:
825 # This bufferblock is referenced by many files as a result
826 # of repacking small blocks, so don't delete it when flushing
827 # its owners, just do it after flushing them all.
828 for owner in v.owner:
829 owner.flush(sync=True)
830 self.delete_bufferblock(k)
832 def block_prefetch(self, locator):
833 """Initiate a background download of a block.
835 This assumes that the underlying KeepClient implements a block cache,
836 so repeated requests for the same block will not result in repeated
837 downloads (unless the block is evicted from the cache.) This method
842 if not self.prefetch_enabled:
845 if self._keep.get_from_cache(locator) is not None:
849 if locator in self._bufferblocks:
852 self.start_get_threads()
853 self._prefetch_queue.put(locator)
856 class ArvadosFile(object):
857 """Represent a file in a Collection.
859 ArvadosFile manages the underlying representation of a file in Keep as a
860 sequence of segments spanning a set of blocks, and implements random
863 This object may be accessed from multiple threads.
867 __slots__ = ('parent', 'name', '_writers', '_committed',
868 '_segments', 'lock', '_current_bblock', 'fuse_entry')
870 def __init__(self, parent, name, stream=[], segments=[]):
872 ArvadosFile constructor.
875 a list of Range objects representing a block stream
878 a list of Range objects representing segments
882 self._writers = set()
883 self._committed = False
885 self.lock = parent.root_collection().lock
887 self._add_segment(stream, s.locator, s.range_size)
888 self._current_bblock = None
891 return self.parent.writable()
894 def permission_expired(self, as_of_dt=None):
895 """Returns True if any of the segment's locators is expired"""
896 for r in self._segments:
897 if KeepLocator(r.locator).permission_expired(as_of_dt):
902 def has_remote_blocks(self):
903 """Returns True if any of the segment's locators has a +R signature"""
905 for s in self._segments:
906 if '+R' in s.locator:
911 def _copy_remote_blocks(self, remote_blocks={}):
912 """Ask Keep to copy remote blocks and point to their local copies.
914 This is called from the parent Collection.
917 Shared cache of remote to local block mappings. This is used to avoid
918 doing extra work when blocks are shared by more than one file in
919 different subdirectories.
922 for s in self._segments:
923 if '+R' in s.locator:
925 loc = remote_blocks[s.locator]
927 loc = self.parent._my_keep().refresh_signature(s.locator)
928 remote_blocks[s.locator] = loc
930 self.parent.set_committed(False)
935 return copy.copy(self._segments)
938 def clone(self, new_parent, new_name):
939 """Make a copy of this file."""
940 cp = ArvadosFile(new_parent, new_name)
941 cp.replace_contents(self)
946 def replace_contents(self, other):
947 """Replace segments of this file with segments from another `ArvadosFile` object."""
951 for other_segment in other.segments():
952 new_loc = other_segment.locator
953 if other.parent._my_block_manager().is_bufferblock(other_segment.locator):
954 if other_segment.locator not in map_loc:
955 bufferblock = other.parent._my_block_manager().get_bufferblock(other_segment.locator)
956 if bufferblock.state() != _BufferBlock.WRITABLE:
957 map_loc[other_segment.locator] = bufferblock.locator()
959 map_loc[other_segment.locator] = self.parent._my_block_manager().dup_block(bufferblock, self).blockid
960 new_loc = map_loc[other_segment.locator]
962 self._segments.append(Range(new_loc, other_segment.range_start, other_segment.range_size, other_segment.segment_offset))
964 self.set_committed(False)
966 def __eq__(self, other):
969 if not isinstance(other, ArvadosFile):
972 othersegs = other.segments()
974 if len(self._segments) != len(othersegs):
976 for i in range(0, len(othersegs)):
977 seg1 = self._segments[i]
982 if self.parent._my_block_manager().is_bufferblock(loc1):
983 loc1 = self.parent._my_block_manager().get_bufferblock(loc1).locator()
985 if other.parent._my_block_manager().is_bufferblock(loc2):
986 loc2 = other.parent._my_block_manager().get_bufferblock(loc2).locator()
988 if (KeepLocator(loc1).stripped() != KeepLocator(loc2).stripped() or
989 seg1.range_start != seg2.range_start or
990 seg1.range_size != seg2.range_size or
991 seg1.segment_offset != seg2.segment_offset):
996 def __ne__(self, other):
997 return not self.__eq__(other)
1000 def set_segments(self, segs):
1001 self._segments = segs
1004 def set_committed(self, value=True):
1005 """Set committed flag.
1007 If value is True, set committed to be True.
1009 If value is False, set committed to be False for this and all parents.
1011 if value == self._committed:
1013 self._committed = value
1014 if self._committed is False and self.parent is not None:
1015 self.parent.set_committed(False)
1018 def committed(self):
1019 """Get whether this is committed or not."""
1020 return self._committed
1023 def add_writer(self, writer):
1024 """Add an ArvadosFileWriter reference to the list of writers"""
1025 if isinstance(writer, ArvadosFileWriter):
1026 self._writers.add(writer)
1029 def remove_writer(self, writer, flush):
1031 Called from ArvadosFileWriter.close(). Remove a writer reference from the list
1032 and do some block maintenance tasks.
1034 self._writers.remove(writer)
1036 if flush or self.size() > config.KEEP_BLOCK_SIZE // 2:
1037 # File writer closed, not small enough for repacking
1040 # All writers closed and size is adequate for repacking
1041 self.parent._my_block_manager().repack_small_blocks(closed_file_size=self.size())
1045 Get whether this is closed or not. When the writers list is empty, the file
1046 is supposed to be closed.
1048 return len(self._writers) == 0
1052 def truncate(self, size):
1053 """Shrink or expand the size of the file.
1055 If `size` is less than the size of the file, the file contents after
1056 `size` will be discarded. If `size` is greater than the current size
1057 of the file, it will be filled with zero bytes.
1060 if size < self.size():
1062 for r in self._segments:
1063 range_end = r.range_start+r.range_size
1064 if r.range_start >= size:
1065 # segment is past the trucate size, all done
1067 elif size < range_end:
1068 nr = Range(r.locator, r.range_start, size - r.range_start, 0)
1069 nr.segment_offset = r.segment_offset
1075 self._segments = new_segs
1076 self.set_committed(False)
1077 elif size > self.size():
1078 padding = self.parent._my_block_manager().get_padding_block()
1079 diff = size - self.size()
1080 while diff > config.KEEP_BLOCK_SIZE:
1081 self._segments.append(Range(padding.blockid, self.size(), config.KEEP_BLOCK_SIZE, 0))
1082 diff -= config.KEEP_BLOCK_SIZE
1084 self._segments.append(Range(padding.blockid, self.size(), diff, 0))
1085 self.set_committed(False)
1087 # size == self.size()
1090 def readfrom(self, offset, size, num_retries, exact=False):
1091 """Read up to `size` bytes from the file starting at `offset`.
1094 If False (default), return less data than requested if the read
1095 crosses a block boundary and the next block isn't cached. If True,
1096 only return less data than requested when hitting EOF.
1100 if size == 0 or offset >= self.size():
1102 readsegs = locators_and_ranges(self._segments, offset, size)
1103 prefetch = locators_and_ranges(self._segments, offset + size, config.KEEP_BLOCK_SIZE, limit=32)
1108 block = self.parent._my_block_manager().get_block_contents(lr.locator, num_retries=num_retries, cache_only=(bool(data) and not exact))
1110 blockview = memoryview(block)
1111 data.append(blockview[lr.segment_offset:lr.segment_offset+lr.segment_size].tobytes())
1112 locs.add(lr.locator)
1117 if lr.locator not in locs:
1118 self.parent._my_block_manager().block_prefetch(lr.locator)
1119 locs.add(lr.locator)
1121 return b''.join(data)
1125 def writeto(self, offset, data, num_retries):
1126 """Write `data` to the file starting at `offset`.
1128 This will update existing bytes and/or extend the size of the file as
1132 if not isinstance(data, bytes) and not isinstance(data, memoryview):
1133 data = data.encode()
1137 if offset > self.size():
1138 self.truncate(offset)
1140 if len(data) > config.KEEP_BLOCK_SIZE:
1141 # Chunk it up into smaller writes
1143 dataview = memoryview(data)
1144 while n < len(data):
1145 self.writeto(offset+n, dataview[n:n + config.KEEP_BLOCK_SIZE].tobytes(), num_retries)
1146 n += config.KEEP_BLOCK_SIZE
1149 self.set_committed(False)
1151 if self._current_bblock is None or self._current_bblock.state() != _BufferBlock.WRITABLE:
1152 self._current_bblock = self.parent._my_block_manager().alloc_bufferblock(owner=self)
1154 if (self._current_bblock.size() + len(data)) > config.KEEP_BLOCK_SIZE:
1155 self._current_bblock.repack_writes()
1156 if (self._current_bblock.size() + len(data)) > config.KEEP_BLOCK_SIZE:
1157 self.parent._my_block_manager().commit_bufferblock(self._current_bblock, sync=False)
1158 self._current_bblock = self.parent._my_block_manager().alloc_bufferblock(owner=self)
1160 self._current_bblock.append(data)
1162 replace_range(self._segments, offset, len(data), self._current_bblock.blockid, self._current_bblock.write_pointer - len(data))
1164 self.parent.notify(WRITE, self.parent, self.name, (self, self))
1169 def flush(self, sync=True, num_retries=0):
1170 """Flush the current bufferblock to Keep.
1173 If True, commit block synchronously, wait until buffer block has been written.
1174 If False, commit block asynchronously, return immediately after putting block into
1177 if self.committed():
1180 if self._current_bblock and self._current_bblock.state() != _BufferBlock.COMMITTED:
1181 if self._current_bblock.state() == _BufferBlock.WRITABLE:
1182 self._current_bblock.repack_writes()
1183 if self._current_bblock.state() != _BufferBlock.DELETED:
1184 self.parent._my_block_manager().commit_bufferblock(self._current_bblock, sync=sync)
1188 for s in self._segments:
1189 bb = self.parent._my_block_manager().get_bufferblock(s.locator)
1191 if bb.state() != _BufferBlock.COMMITTED:
1192 self.parent._my_block_manager().commit_bufferblock(bb, sync=True)
1193 to_delete.add(s.locator)
1194 s.locator = bb.locator()
1196 # Don't delete the bufferblock if it's owned by many files. It'll be
1197 # deleted after all of its owners are flush()ed.
1198 if self.parent._my_block_manager().get_bufferblock(s).owner is self:
1199 self.parent._my_block_manager().delete_bufferblock(s)
1201 self.parent.notify(MOD, self.parent, self.name, (self, self))
1205 def add_segment(self, blocks, pos, size):
1206 """Add a segment to the end of the file.
1208 `pos` and `offset` reference a section of the stream described by
1209 `blocks` (a list of Range objects)
1212 self._add_segment(blocks, pos, size)
1214 def _add_segment(self, blocks, pos, size):
1215 """Internal implementation of add_segment."""
1216 self.set_committed(False)
1217 for lr in locators_and_ranges(blocks, pos, size):
1218 last = self._segments[-1] if self._segments else Range(0, 0, 0, 0)
1219 r = Range(lr.locator, last.range_start+last.range_size, lr.segment_size, lr.segment_offset)
1220 self._segments.append(r)
1224 """Get the file size."""
1226 n = self._segments[-1]
1227 return n.range_start + n.range_size
1232 def manifest_text(self, stream_name=".", portable_locators=False,
1233 normalize=False, only_committed=False):
1236 for segment in self._segments:
1237 loc = segment.locator
1238 if self.parent._my_block_manager().is_bufferblock(loc):
1241 loc = self.parent._my_block_manager().get_bufferblock(loc).locator()
1242 if portable_locators:
1243 loc = KeepLocator(loc).stripped()
1244 filestream.append(LocatorAndRange(loc, KeepLocator(loc).size,
1245 segment.segment_offset, segment.range_size))
1246 buf += ' '.join(normalize_stream(stream_name, {self.name: filestream}))
1252 def _reparent(self, newparent, newname):
1253 self.set_committed(False)
1254 self.flush(sync=True)
1255 self.parent.remove(self.name)
1256 self.parent = newparent
1258 self.lock = self.parent.root_collection().lock
1261 class ArvadosFileReader(ArvadosFileReaderBase):
1262 """Wraps ArvadosFile in a file-like object supporting reading only.
1264 Be aware that this class is NOT thread safe as there is no locking around
1265 updating file pointer.
1269 def __init__(self, arvadosfile, mode="r", num_retries=None):
1270 super(ArvadosFileReader, self).__init__(arvadosfile.name, mode=mode, num_retries=num_retries)
1271 self.arvadosfile = arvadosfile
1274 return self.arvadosfile.size()
1276 def stream_name(self):
1277 return self.arvadosfile.parent.stream_name()
1279 def readinto(self, b):
1280 data = self.read(len(b))
1281 b[:len(data)] = data
1284 @_FileLikeObjectBase._before_close
1286 def read(self, size=None, num_retries=None):
1287 """Read up to `size` bytes from the file and return the result.
1289 Starts at the current file position. If `size` is None, read the
1290 entire remainder of the file.
1294 rd = self.arvadosfile.readfrom(self._filepos, config.KEEP_BLOCK_SIZE, num_retries)
1297 self._filepos += len(rd)
1298 rd = self.arvadosfile.readfrom(self._filepos, config.KEEP_BLOCK_SIZE, num_retries)
1299 return b''.join(data)
1301 data = self.arvadosfile.readfrom(self._filepos, size, num_retries, exact=True)
1302 self._filepos += len(data)
1305 @_FileLikeObjectBase._before_close
1307 def readfrom(self, offset, size, num_retries=None):
1308 """Read up to `size` bytes from the stream, starting at the specified file offset.
1310 This method does not change the file position.
1312 return self.arvadosfile.readfrom(offset, size, num_retries)
1318 class ArvadosFileWriter(ArvadosFileReader):
1319 """Wraps ArvadosFile in a file-like object supporting both reading and writing.
1321 Be aware that this class is NOT thread safe as there is no locking around
1322 updating file pointer.
1326 def __init__(self, arvadosfile, mode, num_retries=None):
1327 super(ArvadosFileWriter, self).__init__(arvadosfile, mode=mode, num_retries=num_retries)
1328 self.arvadosfile.add_writer(self)
1333 @_FileLikeObjectBase._before_close
1335 def write(self, data, num_retries=None):
1336 if self.mode[0] == "a":
1337 self._filepos = self.size()
1338 self.arvadosfile.writeto(self._filepos, data, num_retries)
1339 self._filepos += len(data)
1342 @_FileLikeObjectBase._before_close
1344 def writelines(self, seq, num_retries=None):
1346 self.write(s, num_retries=num_retries)
1348 @_FileLikeObjectBase._before_close
1349 def truncate(self, size=None):
1351 size = self._filepos
1352 self.arvadosfile.truncate(size)
1354 @_FileLikeObjectBase._before_close
1356 self.arvadosfile.flush()
1358 def close(self, flush=True):
1360 self.arvadosfile.remove_writer(self, flush)
1361 super(ArvadosFileWriter, self).close()
1364 class WrappableFile(object):
1365 """An interface to an Arvados file that's compatible with io wrappers.
1368 def __init__(self, f):
1373 return self.f.close()
1375 return self.f.flush()
1376 def read(self, *args, **kwargs):
1377 return self.f.read(*args, **kwargs)
1379 return self.f.readable()
1380 def readinto(self, *args, **kwargs):
1381 return self.f.readinto(*args, **kwargs)
1382 def seek(self, *args, **kwargs):
1383 return self.f.seek(*args, **kwargs)
1385 return self.f.seekable()
1387 return self.f.tell()
1389 return self.f.writable()
1390 def write(self, *args, **kwargs):
1391 return self.f.write(*args, **kwargs)