From 4619260dc1255222d1819c60360c0913f7f3d564 Mon Sep 17 00:00:00 2001
From: Giuliano Mega <giuliano.mega@gmail.com>
Date: Fri, 21 Jun 2024 08:09:59 -0300
Subject: [PATCH] Fix StoreStream so it doesn't return parity bytes  (#838)

* fix storestream so it doesn\'t return parity bits for protected/verifiable manifests

* use Cid.example instead of creating a mock manually
---
 codex/manifest/manifest.nim     |  7 -----
 codex/streams/storestream.nim   |  7 ++---
 tests/codex/teststorestream.nim | 48 ++++++++++++++++++++++++++++++---
 3 files changed, 48 insertions(+), 14 deletions(-)

diff --git a/codex/manifest/manifest.nim b/codex/manifest/manifest.nim
index 486a8fc3..b9d40945 100644
--- a/codex/manifest/manifest.nim
+++ b/codex/manifest/manifest.nim
@@ -135,13 +135,6 @@ func isManifest*(mc: MultiCodec): ?!bool =
 # Various sizes and verification
 ############################################################
 
-func bytes*(self: Manifest, pad = true): NBytes =
-  ## Compute how many bytes corresponding StoreStream(Manifest, pad) will return
-  if pad or self.protected:
-    self.blocksCount.NBytes * self.blockSize
-  else:
-    self.datasetSize
-
 func rounded*(self: Manifest): int =
   ## Number of data blocks in *protected* manifest including padding at the end
   roundUp(self.originalBlocksCount, self.ecK)
diff --git a/codex/streams/storestream.nim b/codex/streams/storestream.nim
index ce89171c..8a3b1a3c 100644
--- a/codex/streams/storestream.nim
+++ b/codex/streams/storestream.nim
@@ -38,7 +38,6 @@ type
   StoreStream* = ref object of SeekableStream
     store*: BlockStore          # Store where to lookup block contents
     manifest*: Manifest         # List of block CIDs
-    pad*: bool                  # Pad last block to manifest.blockSize?
 
 method initStream*(s: StoreStream) =
   if s.objName.len == 0:
@@ -57,13 +56,15 @@ proc new*(
   result = StoreStream(
     store: store,
     manifest: manifest,
-    pad: pad,
     offset: 0)
 
   result.initStream()
 
 method `size`*(self: StoreStream): int =
-  bytes(self.manifest, self.pad).int
+  ## The size of a StoreStream is the size of the original dataset, without
+  ## padding or parity blocks.
+  let m = self.manifest
+  (if m.protected: m.originalDatasetSize else: m.datasetSize).int
 
 proc `size=`*(self: StoreStream, size: int)
   {.error: "Setting the size is forbidden".} =
diff --git a/tests/codex/teststorestream.nim b/tests/codex/teststorestream.nim
index eaf92c1d..b717a8ec 100644
--- a/tests/codex/teststorestream.nim
+++ b/tests/codex/teststorestream.nim
@@ -1,12 +1,15 @@
 import pkg/chronos
 import pkg/questionable/results
 
-import pkg/codex/streams
-import pkg/codex/stores
-import pkg/codex/manifest
-import pkg/codex/blocktype as bt
+import pkg/codex/[
+  streams,
+  stores,
+  indexingstrategy,
+  manifest,
+  blocktype as bt]
 
 import ../asynctest
+import ./examples
 import ./helpers
 
 asyncchecksuite "StoreStream":
@@ -99,3 +102,40 @@ asyncchecksuite "StoreStream":
 
     await stream.readExactly(addr buf[0], 15)
     check sequentialBytes(buf,15,0)
+
+suite "StoreStream - Size Tests":
+
+  var stream: StoreStream
+
+  teardown:
+    await stream.close()
+
+  test "Should return dataset size as stream size":
+    let manifest = Manifest.new(
+      treeCid = Cid.example,
+      datasetSize = 80.NBytes,
+      blockSize = 10.NBytes
+    )
+
+    stream = StoreStream.new(CacheStore.new(), manifest)
+
+    check stream.size == 80
+
+  test "Should not count parity/padding bytes as part of stream size":
+    let protectedManifest = Manifest.new(
+      treeCid = Cid.example,
+      datasetSize = 120.NBytes, # size including parity bytes
+      blockSize = 10.NBytes,
+      version = CIDv1,
+      hcodec = Sha256HashCodec,
+      codec = BlockCodec,
+      ecK = 2,
+      ecM = 1,
+      originalTreeCid = Cid.example,
+      originalDatasetSize = 80.NBytes, # size without parity bytes
+      strategy = StrategyType.SteppedStrategy
+    )
+
+    stream = StoreStream.new(CacheStore.new(), protectedManifest)
+
+    check stream.size == 80