nim-codex/codex/manifest/manifest.nim

253 lines
6.6 KiB
Nim
Raw Normal View History

2022-05-19 19:56:03 +00:00
## Nim-Codex
## Copyright (c) 2022 Status Research & Development GmbH
## Licensed under either of
## * Apache License, version 2.0, ([LICENSE-APACHE](LICENSE-APACHE))
## * MIT license ([LICENSE-MIT](LICENSE-MIT))
## at your option.
## This file may not be copied, modified, or distributed except according to
## those terms.
Download files without padding (#218) The initial goal of this patch was to allow to download of a file via REST API in exactly the same size as it was uploaded, which required adding fields Chunker.offset and Manifest.originalBytes to keep that size. On top of that, we added more integrity checks to operations on Manifest, and reorganized TestNode.nim to test the actual interaction between node.store and node.retrieve operations. Note that the wire format of Manifest was changed, so we need to recreate all BlockStores. * Download without padding * Fixed chunker tests * Chunker: get rid of RabinChunker * Verify offset in the chunker tests * Use manifest.originalBytesPadded in StoreStream.size * StoreStream: replace emptyBlock with zeroMem * Manifest.bytes: compute how many bytes corresponding StoreStream(Manifest, pad) will return * Manifest: verify originalBytes and originalLen on new/encode/decode Also set originalBytes in each Manifest creation/update scenario * Manifest: comments, split code into sections * Reordered parameters to deal with int64 size in 32-bit builds * TestNode.nim: combine Store and Retrieve tests 1. Instead of copy-pasting code from node.nim, new test calls node.store() and node.retrieve() in order to check that they can correctly store and then retrieve data 2. New test compares only file contents, manifest contents considered an implementation detail 3. New test chunks at odd chunkSize=BlockSize/1.618 in order to ensure that data retrieved correctly even when buffer sizes mismatch * TestNode.nim: code refactoring * Manifest.add: one more test * Manifest.verify: return Result instead of raising Defect * Node.store: added blockSize parameter
2022-08-24 12:15:59 +00:00
# This module defines all operations on Manifest
import pkg/upraises
push: {.upraises: [].}
import pkg/libp2p/protobuf/minprotobuf
import pkg/libp2p
import pkg/questionable
import pkg/questionable/results
import pkg/chronicles
import ../errors
Download files without padding (#218) The initial goal of this patch was to allow to download of a file via REST API in exactly the same size as it was uploaded, which required adding fields Chunker.offset and Manifest.originalBytes to keep that size. On top of that, we added more integrity checks to operations on Manifest, and reorganized TestNode.nim to test the actual interaction between node.store and node.retrieve operations. Note that the wire format of Manifest was changed, so we need to recreate all BlockStores. * Download without padding * Fixed chunker tests * Chunker: get rid of RabinChunker * Verify offset in the chunker tests * Use manifest.originalBytesPadded in StoreStream.size * StoreStream: replace emptyBlock with zeroMem * Manifest.bytes: compute how many bytes corresponding StoreStream(Manifest, pad) will return * Manifest: verify originalBytes and originalLen on new/encode/decode Also set originalBytes in each Manifest creation/update scenario * Manifest: comments, split code into sections * Reordered parameters to deal with int64 size in 32-bit builds * TestNode.nim: combine Store and Retrieve tests 1. Instead of copy-pasting code from node.nim, new test calls node.store() and node.retrieve() in order to check that they can correctly store and then retrieve data 2. New test compares only file contents, manifest contents considered an implementation detail 3. New test chunks at odd chunkSize=BlockSize/1.618 in order to ensure that data retrieved correctly even when buffer sizes mismatch * TestNode.nim: code refactoring * Manifest.add: one more test * Manifest.verify: return Result instead of raising Defect * Node.store: added blockSize parameter
2022-08-24 12:15:59 +00:00
import ../utils
import ../utils/json
import ../units
import ../blocktype
import ./types
export types
type
Manifest* = ref object of RootObj
treeCid {.serialize.}: Cid # Root of the merkle tree
datasetSize {.serialize.}: NBytes # Total size of all blocks
blockSize {.serialize.}: NBytes # Size of each contained block (might not be needed if blocks are len-prefixed)
version: CidVersion # Cid version
hcodec: MultiCodec # Multihash codec
codec: MultiCodec # Data set codec
case protected {.serialize.}: bool # Protected datasets have erasure coded info
of true:
ecK: int # Number of blocks to encode
ecM: int # Number of resulting parity blocks
originalManifest: Manifest # The original Manifest being erasure coded
else:
discard
############################################################
# Accessors
############################################################
proc blockSize*(self: Manifest): NBytes =
self.blockSize
proc datasetSize*(self: Manifest): NBytes =
self.datasetSize
proc version*(self: Manifest): CidVersion =
self.version
proc hcodec*(self: Manifest): MultiCodec =
self.hcodec
proc codec*(self: Manifest): MultiCodec =
self.codec
proc protected*(self: Manifest): bool =
self.protected
proc ecK*(self: Manifest): int =
self.ecK
proc ecM*(self: Manifest): int =
self.ecM
proc originalManifest*(self: Manifest): Manifest =
self.originalManifest
proc originalTreeCid*(self: Manifest): Cid =
self.originalManifest.treeCid
proc originalBlocksCount*(self: Manifest): int =
divUp(self.originalManifest.datasetSize.int, self.blockSize.int)
proc originalDatasetSize*(self: Manifest): NBytes =
self.originalDatasetSize
proc treeCid*(self: Manifest): Cid =
self.treeCid
proc blocksCount*(self: Manifest): int =
divUp(self.datasetSize.int, self.blockSize.int)
############################################################
# Operations on block list
############################################################
func isManifest*(cid: Cid): ?!bool =
let res = ?cid.contentType().mapFailure(CodexError)
($(res) in ManifestContainers).success
func isManifest*(mc: MultiCodec): ?!bool =
($mc in ManifestContainers).success
Download files without padding (#218) The initial goal of this patch was to allow to download of a file via REST API in exactly the same size as it was uploaded, which required adding fields Chunker.offset and Manifest.originalBytes to keep that size. On top of that, we added more integrity checks to operations on Manifest, and reorganized TestNode.nim to test the actual interaction between node.store and node.retrieve operations. Note that the wire format of Manifest was changed, so we need to recreate all BlockStores. * Download without padding * Fixed chunker tests * Chunker: get rid of RabinChunker * Verify offset in the chunker tests * Use manifest.originalBytesPadded in StoreStream.size * StoreStream: replace emptyBlock with zeroMem * Manifest.bytes: compute how many bytes corresponding StoreStream(Manifest, pad) will return * Manifest: verify originalBytes and originalLen on new/encode/decode Also set originalBytes in each Manifest creation/update scenario * Manifest: comments, split code into sections * Reordered parameters to deal with int64 size in 32-bit builds * TestNode.nim: combine Store and Retrieve tests 1. Instead of copy-pasting code from node.nim, new test calls node.store() and node.retrieve() in order to check that they can correctly store and then retrieve data 2. New test compares only file contents, manifest contents considered an implementation detail 3. New test chunks at odd chunkSize=BlockSize/1.618 in order to ensure that data retrieved correctly even when buffer sizes mismatch * TestNode.nim: code refactoring * Manifest.add: one more test * Manifest.verify: return Result instead of raising Defect * Node.store: added blockSize parameter
2022-08-24 12:15:59 +00:00
############################################################
# Various sizes and verification
############################################################
func bytes*(self: Manifest, pad = true): NBytes =
Download files without padding (#218) The initial goal of this patch was to allow to download of a file via REST API in exactly the same size as it was uploaded, which required adding fields Chunker.offset and Manifest.originalBytes to keep that size. On top of that, we added more integrity checks to operations on Manifest, and reorganized TestNode.nim to test the actual interaction between node.store and node.retrieve operations. Note that the wire format of Manifest was changed, so we need to recreate all BlockStores. * Download without padding * Fixed chunker tests * Chunker: get rid of RabinChunker * Verify offset in the chunker tests * Use manifest.originalBytesPadded in StoreStream.size * StoreStream: replace emptyBlock with zeroMem * Manifest.bytes: compute how many bytes corresponding StoreStream(Manifest, pad) will return * Manifest: verify originalBytes and originalLen on new/encode/decode Also set originalBytes in each Manifest creation/update scenario * Manifest: comments, split code into sections * Reordered parameters to deal with int64 size in 32-bit builds * TestNode.nim: combine Store and Retrieve tests 1. Instead of copy-pasting code from node.nim, new test calls node.store() and node.retrieve() in order to check that they can correctly store and then retrieve data 2. New test compares only file contents, manifest contents considered an implementation detail 3. New test chunks at odd chunkSize=BlockSize/1.618 in order to ensure that data retrieved correctly even when buffer sizes mismatch * TestNode.nim: code refactoring * Manifest.add: one more test * Manifest.verify: return Result instead of raising Defect * Node.store: added blockSize parameter
2022-08-24 12:15:59 +00:00
## Compute how many bytes corresponding StoreStream(Manifest, pad) will return
if pad or self.protected:
self.blocksCount.NBytes * self.blockSize
Download files without padding (#218) The initial goal of this patch was to allow to download of a file via REST API in exactly the same size as it was uploaded, which required adding fields Chunker.offset and Manifest.originalBytes to keep that size. On top of that, we added more integrity checks to operations on Manifest, and reorganized TestNode.nim to test the actual interaction between node.store and node.retrieve operations. Note that the wire format of Manifest was changed, so we need to recreate all BlockStores. * Download without padding * Fixed chunker tests * Chunker: get rid of RabinChunker * Verify offset in the chunker tests * Use manifest.originalBytesPadded in StoreStream.size * StoreStream: replace emptyBlock with zeroMem * Manifest.bytes: compute how many bytes corresponding StoreStream(Manifest, pad) will return * Manifest: verify originalBytes and originalLen on new/encode/decode Also set originalBytes in each Manifest creation/update scenario * Manifest: comments, split code into sections * Reordered parameters to deal with int64 size in 32-bit builds * TestNode.nim: combine Store and Retrieve tests 1. Instead of copy-pasting code from node.nim, new test calls node.store() and node.retrieve() in order to check that they can correctly store and then retrieve data 2. New test compares only file contents, manifest contents considered an implementation detail 3. New test chunks at odd chunkSize=BlockSize/1.618 in order to ensure that data retrieved correctly even when buffer sizes mismatch * TestNode.nim: code refactoring * Manifest.add: one more test * Manifest.verify: return Result instead of raising Defect * Node.store: added blockSize parameter
2022-08-24 12:15:59 +00:00
else:
self.datasetSize
Download files without padding (#218) The initial goal of this patch was to allow to download of a file via REST API in exactly the same size as it was uploaded, which required adding fields Chunker.offset and Manifest.originalBytes to keep that size. On top of that, we added more integrity checks to operations on Manifest, and reorganized TestNode.nim to test the actual interaction between node.store and node.retrieve operations. Note that the wire format of Manifest was changed, so we need to recreate all BlockStores. * Download without padding * Fixed chunker tests * Chunker: get rid of RabinChunker * Verify offset in the chunker tests * Use manifest.originalBytesPadded in StoreStream.size * StoreStream: replace emptyBlock with zeroMem * Manifest.bytes: compute how many bytes corresponding StoreStream(Manifest, pad) will return * Manifest: verify originalBytes and originalLen on new/encode/decode Also set originalBytes in each Manifest creation/update scenario * Manifest: comments, split code into sections * Reordered parameters to deal with int64 size in 32-bit builds * TestNode.nim: combine Store and Retrieve tests 1. Instead of copy-pasting code from node.nim, new test calls node.store() and node.retrieve() in order to check that they can correctly store and then retrieve data 2. New test compares only file contents, manifest contents considered an implementation detail 3. New test chunks at odd chunkSize=BlockSize/1.618 in order to ensure that data retrieved correctly even when buffer sizes mismatch * TestNode.nim: code refactoring * Manifest.add: one more test * Manifest.verify: return Result instead of raising Defect * Node.store: added blockSize parameter
2022-08-24 12:15:59 +00:00
func rounded*(self: Manifest): int =
## Number of data blocks in *protected* manifest including padding at the end
roundUp(self.originalBlocksCount, self.ecK)
Download files without padding (#218) The initial goal of this patch was to allow to download of a file via REST API in exactly the same size as it was uploaded, which required adding fields Chunker.offset and Manifest.originalBytes to keep that size. On top of that, we added more integrity checks to operations on Manifest, and reorganized TestNode.nim to test the actual interaction between node.store and node.retrieve operations. Note that the wire format of Manifest was changed, so we need to recreate all BlockStores. * Download without padding * Fixed chunker tests * Chunker: get rid of RabinChunker * Verify offset in the chunker tests * Use manifest.originalBytesPadded in StoreStream.size * StoreStream: replace emptyBlock with zeroMem * Manifest.bytes: compute how many bytes corresponding StoreStream(Manifest, pad) will return * Manifest: verify originalBytes and originalLen on new/encode/decode Also set originalBytes in each Manifest creation/update scenario * Manifest: comments, split code into sections * Reordered parameters to deal with int64 size in 32-bit builds * TestNode.nim: combine Store and Retrieve tests 1. Instead of copy-pasting code from node.nim, new test calls node.store() and node.retrieve() in order to check that they can correctly store and then retrieve data 2. New test compares only file contents, manifest contents considered an implementation detail 3. New test chunks at odd chunkSize=BlockSize/1.618 in order to ensure that data retrieved correctly even when buffer sizes mismatch * TestNode.nim: code refactoring * Manifest.add: one more test * Manifest.verify: return Result instead of raising Defect * Node.store: added blockSize parameter
2022-08-24 12:15:59 +00:00
func steps*(self: Manifest): int =
## Number of EC groups in *protected* manifest
divUp(self.originalBlocksCount, self.ecK)
Download files without padding (#218) The initial goal of this patch was to allow to download of a file via REST API in exactly the same size as it was uploaded, which required adding fields Chunker.offset and Manifest.originalBytes to keep that size. On top of that, we added more integrity checks to operations on Manifest, and reorganized TestNode.nim to test the actual interaction between node.store and node.retrieve operations. Note that the wire format of Manifest was changed, so we need to recreate all BlockStores. * Download without padding * Fixed chunker tests * Chunker: get rid of RabinChunker * Verify offset in the chunker tests * Use manifest.originalBytesPadded in StoreStream.size * StoreStream: replace emptyBlock with zeroMem * Manifest.bytes: compute how many bytes corresponding StoreStream(Manifest, pad) will return * Manifest: verify originalBytes and originalLen on new/encode/decode Also set originalBytes in each Manifest creation/update scenario * Manifest: comments, split code into sections * Reordered parameters to deal with int64 size in 32-bit builds * TestNode.nim: combine Store and Retrieve tests 1. Instead of copy-pasting code from node.nim, new test calls node.store() and node.retrieve() in order to check that they can correctly store and then retrieve data 2. New test compares only file contents, manifest contents considered an implementation detail 3. New test chunks at odd chunkSize=BlockSize/1.618 in order to ensure that data retrieved correctly even when buffer sizes mismatch * TestNode.nim: code refactoring * Manifest.add: one more test * Manifest.verify: return Result instead of raising Defect * Node.store: added blockSize parameter
2022-08-24 12:15:59 +00:00
func verify*(self: Manifest): ?!void =
## Check manifest correctness
##
if self.protected and (self.blocksCount != self.steps * (self.ecK + self.ecM)):
return failure newException(CodexError, "Broken manifest: wrong originalBlocksCount")
Download files without padding (#218) The initial goal of this patch was to allow to download of a file via REST API in exactly the same size as it was uploaded, which required adding fields Chunker.offset and Manifest.originalBytes to keep that size. On top of that, we added more integrity checks to operations on Manifest, and reorganized TestNode.nim to test the actual interaction between node.store and node.retrieve operations. Note that the wire format of Manifest was changed, so we need to recreate all BlockStores. * Download without padding * Fixed chunker tests * Chunker: get rid of RabinChunker * Verify offset in the chunker tests * Use manifest.originalBytesPadded in StoreStream.size * StoreStream: replace emptyBlock with zeroMem * Manifest.bytes: compute how many bytes corresponding StoreStream(Manifest, pad) will return * Manifest: verify originalBytes and originalLen on new/encode/decode Also set originalBytes in each Manifest creation/update scenario * Manifest: comments, split code into sections * Reordered parameters to deal with int64 size in 32-bit builds * TestNode.nim: combine Store and Retrieve tests 1. Instead of copy-pasting code from node.nim, new test calls node.store() and node.retrieve() in order to check that they can correctly store and then retrieve data 2. New test compares only file contents, manifest contents considered an implementation detail 3. New test chunks at odd chunkSize=BlockSize/1.618 in order to ensure that data retrieved correctly even when buffer sizes mismatch * TestNode.nim: code refactoring * Manifest.add: one more test * Manifest.verify: return Result instead of raising Defect * Node.store: added blockSize parameter
2022-08-24 12:15:59 +00:00
return success()
proc cid*(self: Manifest): ?!Cid {.deprecated: "use treeCid instead".} =
self.treeCid.success
proc `==`*(a, b: Manifest): bool =
(a.treeCid == b.treeCid) and
(a.datasetSize == b.datasetSize) and
(a.blockSize == b.blockSize) and
(a.version == b.version) and
(a.hcodec == b.hcodec) and
(a.codec == b.codec) and
(a.protected == b.protected) and
(if a.protected:
(a.ecK == b.ecK) and
(a.ecM == b.ecM) and
(a.originalManifest == b.originalManifest)
else:
true)
proc `$`*(self: Manifest): string =
"treeCid: " & $self.treeCid &
", datasetSize: " & $self.datasetSize &
", blockSize: " & $self.blockSize &
", version: " & $self.version &
", hcodec: " & $self.hcodec &
", codec: " & $self.codec &
", protected: " & $self.protected &
(if self.protected:
", ecK: " & $self.ecK &
", ecM: " & $self.ecM &
", originalManifest: " & $self.originalManifest
else:
"")
Download files without padding (#218) The initial goal of this patch was to allow to download of a file via REST API in exactly the same size as it was uploaded, which required adding fields Chunker.offset and Manifest.originalBytes to keep that size. On top of that, we added more integrity checks to operations on Manifest, and reorganized TestNode.nim to test the actual interaction between node.store and node.retrieve operations. Note that the wire format of Manifest was changed, so we need to recreate all BlockStores. * Download without padding * Fixed chunker tests * Chunker: get rid of RabinChunker * Verify offset in the chunker tests * Use manifest.originalBytesPadded in StoreStream.size * StoreStream: replace emptyBlock with zeroMem * Manifest.bytes: compute how many bytes corresponding StoreStream(Manifest, pad) will return * Manifest: verify originalBytes and originalLen on new/encode/decode Also set originalBytes in each Manifest creation/update scenario * Manifest: comments, split code into sections * Reordered parameters to deal with int64 size in 32-bit builds * TestNode.nim: combine Store and Retrieve tests 1. Instead of copy-pasting code from node.nim, new test calls node.store() and node.retrieve() in order to check that they can correctly store and then retrieve data 2. New test compares only file contents, manifest contents considered an implementation detail 3. New test chunks at odd chunkSize=BlockSize/1.618 in order to ensure that data retrieved correctly even when buffer sizes mismatch * TestNode.nim: code refactoring * Manifest.add: one more test * Manifest.verify: return Result instead of raising Defect * Node.store: added blockSize parameter
2022-08-24 12:15:59 +00:00
############################################################
# Constructors
############################################################
proc new*(
T: type Manifest,
treeCid: Cid,
blockSize: NBytes,
datasetSize: NBytes,
version: CidVersion = CIDv1,
hcodec = multiCodec("sha2-256"),
codec = multiCodec("raw"),
protected = false,
): Manifest =
T(
treeCid: treeCid,
blockSize: blockSize,
datasetSize: datasetSize,
version: version,
codec: codec,
hcodec: hcodec,
protected: protected)
proc new*(
T: type Manifest,
manifest: Manifest,
treeCid: Cid,
datasetSize: NBytes,
ecK, ecM: int
): Manifest =
## Create an erasure protected dataset from an
## unprotected one
##
Manifest(
treeCid: treeCid,
datasetSize: datasetSize,
version: manifest.version,
codec: manifest.codec,
hcodec: manifest.hcodec,
blockSize: manifest.blockSize,
protected: true,
ecK: ecK, ecM: ecM,
originalManifest: manifest)
proc new*(
T: type Manifest,
manifest: Manifest
): Manifest =
## Create an unprotected dataset from an
## erasure protected one
##
manifest.originalManifest
proc new*(
T: type Manifest,
data: openArray[byte],
decoder = ManifestContainers[$DagPBCodec]
): ?!Manifest =
## Create a manifest instance from given data
##
Manifest.decode(data, decoder)
proc new*(
T: type Manifest,
treeCid: Cid,
datasetSize: NBytes,
blockSize: NBytes,
version: CidVersion,
hcodec: MultiCodec,
codec: MultiCodec,
ecK: int,
ecM: int,
originalManifest: Manifest
): Manifest =
Manifest(
treeCid: treeCid,
datasetSize: datasetSize,
blockSize: blockSize,
version: version,
hcodec: hcodec,
codec: codec,
protected: true,
ecK: ecK,
ecM: ecM,
originalManifest: originalManifest
)