v0.4.1

Co-authored-by: Rod Vagg <rod@vagg.org>

.github/workflows/automerge.yml

@@ -0,0 +1,11 @@
+# File managed by web3-bot. DO NOT EDIT.
+# See https://github.com/protocol/.github/ for details.
+
+name: Automerge
+on: [ pull_request ]
+
+jobs:
+  automerge:
+    uses: protocol/.github/.github/workflows/automerge.yml@master
+    with:
+      job: 'automerge'

.github/workflows/go-check.yml

@@ -0,0 +1,67 @@
+# File managed by web3-bot. DO NOT EDIT.
+# See https://github.com/protocol/.github/ for details.
+
+on: [push, pull_request]
+name: Go Checks
+
+jobs:
+  unit:
+    runs-on: ubuntu-latest
+    name: All
+    steps:
+      - uses: actions/checkout@v3
+        with:
+          submodules: recursive
+      - id: config
+        uses: protocol/.github/.github/actions/read-config@master
+      - uses: actions/setup-go@v3
+        with:
+          go-version: 1.20.x
+      - name: Run repo-specific setup
+        uses: ./.github/actions/go-check-setup
+        if: hashFiles('./.github/actions/go-check-setup') != ''
+      - name: Install staticcheck
+        run: go install honnef.co/go/tools/cmd/staticcheck@4970552d932f48b71485287748246cf3237cebdf # 2023.1 (v0.4.0)
+      - name: Check that go.mod is tidy
+        uses: protocol/multiple-go-modules@v1.2
+        with:
+          run: |
+            go mod tidy
+            if [[ -n $(git ls-files --other --exclude-standard --directory -- go.sum) ]]; then
+              echo "go.sum was added by go mod tidy"
+              exit 1
+            fi
+            git diff --exit-code -- go.sum go.mod
+      - name: gofmt
+        if: success() || failure() # run this step even if the previous one failed
+        run: |
+          out=$(gofmt -s -l .)
+          if [[ -n "$out" ]]; then
+            echo $out | awk '{print "::error file=" $0 ",line=0,col=0::File is not gofmt-ed."}'
+            exit 1
+          fi
+      - name: go vet
+        if: success() || failure() # run this step even if the previous one failed
+        uses: protocol/multiple-go-modules@v1.2
+        with:
+          run: go vet ./...
+      - name: staticcheck
+        if: success() || failure() # run this step even if the previous one failed
+        uses: protocol/multiple-go-modules@v1.2
+        with:
+          run: |
+            set -o pipefail
+            staticcheck ./... | sed -e 's@\(.*\)\.go@./\1.go@g'
+      - name: go generate
+        uses: protocol/multiple-go-modules@v1.2
+        if: (success() || failure()) && fromJSON(steps.config.outputs.json).gogenerate == true
+        with:
+          run: |
+            git clean -fd # make sure there aren't untracked files / directories
+            go generate -x ./...
+            # check if go generate modified or added any files
+            if ! $(git add . && git diff-index HEAD --exit-code --quiet); then
+              echo "go generated caused changes to the repository:"
+              git status --short
+              exit 1
+            fi

.github/workflows/go-test.yml

@@ -0,0 +1,76 @@
+# File managed by web3-bot. DO NOT EDIT.
+# See https://github.com/protocol/.github/ for details.
+
+on: [push, pull_request]
+name: Go Test
+
+jobs:
+  unit:
+    strategy:
+      fail-fast: false
+      matrix:
+        os: [ "ubuntu", "windows", "macos" ]
+        go: ["1.19.x","1.20.x"]
+    env:
+      COVERAGES: ""
+    runs-on: ${{ fromJSON(vars[format('UCI_GO_TEST_RUNNER_{0}', matrix.os)] || format('"{0}-latest"', matrix.os)) }}
+    name: ${{ matrix.os }} (go ${{ matrix.go }})
+    steps:
+      - uses: actions/checkout@v3
+        with:
+          submodules: recursive
+      - id: config
+        uses: protocol/.github/.github/actions/read-config@master
+      - uses: actions/setup-go@v3
+        with:
+          go-version: ${{ matrix.go }}
+      - name: Go information
+        run: |
+          go version
+          go env
+      - name: Use msys2 on windows
+        if: matrix.os == 'windows'
+        shell: bash
+        # The executable for msys2 is also called bash.cmd
+        #   https://github.com/actions/virtual-environments/blob/main/images/win/Windows2019-Readme.md#shells
+        # If we prepend its location to the PATH
+        #   subsequent 'shell: bash' steps will use msys2 instead of gitbash
+        run: echo "C:/msys64/usr/bin" >> $GITHUB_PATH
+      - name: Run repo-specific setup
+        uses: ./.github/actions/go-test-setup
+        if: hashFiles('./.github/actions/go-test-setup') != ''
+      - name: Run tests
+        if: contains(fromJSON(steps.config.outputs.json).skipOSes, matrix.os) == false
+        uses: protocol/multiple-go-modules@v1.2
+        with:
+          # Use -coverpkg=./..., so that we include cross-package coverage.
+          # If package ./A imports ./B, and ./A's tests also cover ./B,
+          # this means ./B's coverage will be significantly higher than 0%.
+          run: go test -v -shuffle=on -coverprofile=module-coverage.txt -coverpkg=./... ./...
+      - name: Run tests (32 bit)
+        # can't run 32 bit tests on OSX.
+        if: matrix.os != 'macos' &&
+          fromJSON(steps.config.outputs.json).skip32bit != true &&
+          contains(fromJSON(steps.config.outputs.json).skipOSes, matrix.os) == false
+        uses: protocol/multiple-go-modules@v1.2
+        env:
+          GOARCH: 386
+        with:
+          run: |
+            export "PATH=$PATH_386:$PATH"
+            go test -v -shuffle=on ./...
+      - name: Run tests with race detector
+        # speed things up. Windows and OSX VMs are slow
+        if: matrix.os == 'ubuntu' &&
+          contains(fromJSON(steps.config.outputs.json).skipOSes, matrix.os) == false
+        uses: protocol/multiple-go-modules@v1.2
+        with:
+          run: go test -v -race ./...
+      - name: Collect coverage files
+        shell: bash
+        run: echo "COVERAGES=$(find . -type f -name 'module-coverage.txt' | tr -s '\n' ',' | sed 's/,$//')" >> $GITHUB_ENV
+      - name: Upload coverage to Codecov
+        uses: codecov/codecov-action@d9f34f8cd5cb3b3eb79b3e4b5dae3a16df499a70 # v3.1.1
+        with:
+          files: '${{ env.COVERAGES }}'
+          env_vars: OS=${{ matrix.os }}, GO=${{ matrix.go }}

.github/workflows/release-check.yml

@@ -0,0 +1,13 @@
+# File managed by web3-bot. DO NOT EDIT.
+# See https://github.com/protocol/.github/ for details.
+
+name: Release Checker
+on:
+  pull_request_target:
+    paths: [ 'version.json' ]
+
+jobs:
+  release-check:
+    uses: protocol/.github/.github/workflows/release-check.yml@master
+    with:
+      go-version: 1.20.x

.github/workflows/releaser.yml

@@ -0,0 +1,11 @@
+# File managed by web3-bot. DO NOT EDIT.
+# See https://github.com/protocol/.github/ for details.
+
+name: Releaser
+on:
+  push:
+    paths: [ 'version.json' ]
+
+jobs:
+  releaser:
+    uses: protocol/.github/.github/workflows/releaser.yml@master

.github/workflows/stale.yml

@@ -0,0 +1,26 @@
+name: Close and mark stale issue
+
+on:
+  schedule:
+  - cron: '0 0 * * *'
+
+jobs:
+  stale:
+
+    runs-on: ubuntu-latest
+    permissions:
+      issues: write
+      pull-requests: write
+
+    steps:
+    - uses: actions/stale@v3
+      with:
+        repo-token: ${{ secrets.GITHUB_TOKEN }}
+        stale-issue-message: 'Oops, seems like we needed more information for this issue, please comment with more details or this issue will be closed in 7 days.'
+        close-issue-message: 'This issue was closed because it is missing author input.'
+        stale-issue-label: 'kind/stale'
+        any-of-labels: 'need/author-input'
+        exempt-issue-labels: 'need/triage,need/community-input,need/maintainer-input,need/maintainers-input,need/analysis,status/blocked,status/in-progress,status/ready,status/deferred,status/inactive'
+        days-before-issue-stale: 6
+        days-before-issue-close: 7
+        enable-statistics: true

.github/workflows/tagpush.yml

@@ -0,0 +1,12 @@
+# File managed by web3-bot. DO NOT EDIT.
+# See https://github.com/protocol/.github/ for details.
+
+name: Tag Push Checker
+on:
+  push:
+    tags:
+      - v*
+
+jobs:
+  releaser:
+    uses: protocol/.github/.github/workflows/tagpush.yml@master

.travis.yml

@@ -1,30 +0,0 @@
-os:
-  - linux
-
-language: go
-
-go:
-  - 1.11.x
-
-env:
-  global:
-    - GOTFLAGS="-race"
-  matrix:
-    - BUILD_DEPTYPE=gomod
-
-
-# disable travis install
-install:
-  - true
-
-script:
-  - bash <(curl -s https://raw.githubusercontent.com/ipfs/ci-helpers/master/travis-ci/run-standard-tests.sh)
-
-
-cache:
-  directories:
-    - $GOPATH/pkg/mod
-    - $HOME/.cache/go-build
-
-notifications:
-  email: false

README.md

@@ -59,10 +59,17 @@ fmt.Println("Got CID: ", c)
 #### Creating a CID from scratch
 
 ```go
+
+import (
+  cid "github.com/ipfs/go-cid"
+  mc "github.com/multiformats/go-multicodec"
+  mh "github.com/multiformats/go-multihash"
+)
+
 // Create a cid manually by specifying the 'prefix' parameters
 pref := cid.Prefix{
 	Version: 1,
-	Codec: cid.Raw,
+	Codec: uint64(mc.Raw),
 	MhType: mh.SHA2_256,
 	MhLength: -1, // default length
 }
@@ -88,7 +95,7 @@ if c1.Equals(c2) {
 ```go
 // To check if some data matches a given cid, 
 // Get your CIDs prefix, and use that to sum the data in question:
-other, err := c.Prefix().SumStream(myFile)
+other, err := c.Prefix().Sum(mydata)
 if err != nil {...}
 
 if !c.Equals(other) {

_rsrch/cidiface/cid.go

@@ -39,7 +39,8 @@ type Cid interface {
 // and the Multihash length. It does not contains
 // any actual content information.
 // NOTE: The use -1 in MhLength to mean default length is deprecated,
-//   use the V0Builder or V1Builder structures instead
+//
+//	use the V0Builder or V1Builder structures instead
 type Prefix struct {
 	Version  uint64
 	Codec    uint64

_rsrch/cidiface/cidBoxingBench_test.go

@@ -12,14 +12,14 @@ import (
 //
 // Sample results on linux amd64 go1.11beta:
 //
-//   BenchmarkCidMap_CidStr-8          100000             16317 ns/op
-//   BenchmarkCidMap_CidIface-8        100000             20516 ns/op
+//	BenchmarkCidMap_CidStr-8          100000             16317 ns/op
+//	BenchmarkCidMap_CidIface-8        100000             20516 ns/op
 //
 // With benchmem on:
 //
-//   BenchmarkCidMap_CidStr-8          100000             15579 ns/op           11223 B/op        207 allocs/op
-//   BenchmarkCidMap_CidIface-8        100000             19500 ns/op           12824 B/op        307 allocs/op
-//   BenchmarkCidMap_StrPlusHax-8      200000             10451 ns/op            7589 B/op        202 allocs/op
+//	BenchmarkCidMap_CidStr-8          100000             15579 ns/op           11223 B/op        207 allocs/op
+//	BenchmarkCidMap_CidIface-8        100000             19500 ns/op           12824 B/op        307 allocs/op
+//	BenchmarkCidMap_StrPlusHax-8      200000             10451 ns/op            7589 B/op        202 allocs/op
 //
 // We can see here that the impact of interface boxing is significant:
 // it increases the time taken to do the inserts to 133%, largely because
@@ -36,7 +36,6 @@ import (
 // re-arranges itself, it involves more or less an O(n) copy of the content
 // in addition to the alloc itself).  This isn't topical to the question of
 // whether or not interfaces are a good idea; just for contextualizing.
-//
 func BenchmarkCidMap_CidStr(b *testing.B) {
 	for i := 0; i < b.N; i++ {
 		mp := map[CidStr]int{}

_rsrch/cidiface/cidString.go

@@ -114,7 +114,7 @@ func NewCidStr(version uint64, codecType uint64, mhash mh.Multihash) CidStr {
 //
 // For CidV1, the data buffer is in the form:
 //
-//     <version><codec-type><multihash>
+//	<version><codec-type><multihash>
 //
 // CidV0 are also supported. In particular, data buffers starting
 // with length 34 bytes, which starts with bytes [18,32...] are considered

_rsrch/cidiface/cidStruct.go

@@ -110,7 +110,7 @@ func (c CidStruct) Prefix() Prefix {
 //
 // For CidV1, the data buffer is in the form:
 //
-//     <version><codec-type><multihash>
+//	<version><codec-type><multihash>
 //
 // CidV0 are also supported. In particular, data buffers starting
 // with length 34 bytes, which starts with bytes [18,32...] are considered

benchmark_test.go

@@ -0,0 +1,61 @@
+package cid_test
+
+import (
+	"math/rand"
+	"testing"
+
+	"github.com/ipfs/go-cid"
+	"github.com/multiformats/go-multihash"
+)
+
+// BenchmarkIdentityCheck benchmarks two ways of checking whether a CIDv1 has multihash.IDENTITY
+// code:
+//  1. Cid.Prefix(), and
+//  2. decoding the Cid.Hash().
+//
+// This benchmark illustrates that using Cid.Prefix is more efficient than multihash.Decode.
+// Users wishing to perform such a check should use Cid.Prefix.
+//
+// Consider that `Cid.Prefix` is already efficient enough and introducing a dedicated API for
+// performing this check will likely result in small gains.
+func BenchmarkIdentityCheck(b *testing.B) {
+	rng := rand.New(rand.NewSource(1413))
+
+	data := make([]byte, rng.Intn(100)+1024)
+	if _, err := rng.Read(data); err != nil {
+		b.Fatal(err)
+	}
+	mh, err := multihash.Sum(data, multihash.IDENTITY, -1)
+	if err != nil {
+		b.Fatal(err)
+	}
+	cv1 := cid.NewCidV1(cid.Raw, mh)
+
+	b.SetBytes(int64(cv1.ByteLen()))
+	b.ReportAllocs()
+	b.ResetTimer()
+
+	b.Run("Prefix", func(b *testing.B) {
+		b.RunParallel(func(pb *testing.PB) {
+			for pb.Next() {
+				if cv1.Prefix().MhType != multihash.IDENTITY {
+					b.Fatal("expected IDENTITY CID")
+				}
+			}
+		})
+	})
+
+	b.Run("MultihashDecode", func(b *testing.B) {
+		b.RunParallel(func(pb *testing.PB) {
+			for pb.Next() {
+				dmh, err := multihash.Decode(cv1.Hash())
+				if err != nil {
+					b.Fatal(err)
+				}
+				if dmh.Code != multihash.IDENTITY {
+					b.Fatal("expected IDENTITY CID")
+				}
+			}
+		})
+	})
+}

builder.go

@@ -1,8 +1,6 @@
 package cid
 
 import (
-	"io"
-
 	mh "github.com/multiformats/go-multihash"
 )
 
@@ -43,14 +41,6 @@ func (p V0Builder) Sum(data []byte) (Cid, error) {
 	return Cid{string(hash)}, nil
 }
 
-func (p V0Builder) SumStream(r io.Reader) (Cid, error) {
-	hash, err := mh.SumStream(r, mh.SHA2_256, -1)
-	if err != nil {
-		return Undef, err
-	}
-	return Cid{string(hash)}, nil
-}
-
 func (p V0Builder) GetCodec() uint64 {
 	return DagProtobuf
 }
@@ -74,18 +64,6 @@ func (p V1Builder) Sum(data []byte) (Cid, error) {
 	return NewCidV1(p.Codec, hash), nil
 }
 
-func (p V1Builder) SumStream(r io.Reader) (Cid, error) {
-	mhLen := p.MhLength
-	if mhLen <= 0 {
-		mhLen = -1
-	}
-	hash, err := mh.SumStream(r, p.MhType, mhLen)
-	if err != nil {
-		return Undef, err
-	}
-	return NewCidV1(p.Codec, hash), nil
-}
-
 func (p V1Builder) GetCodec() uint64 {
 	return p.Codec
 }

builder_test.go

@@ -1,7 +1,6 @@
 package cid
 
 import (
-	"bytes"
 	"testing"
 
 	mh "github.com/multiformats/go-multihash"
@@ -17,15 +16,6 @@ func TestV0Builder(t *testing.T) {
 		t.Fatal(err)
 	}
 
-	reader := bytes.NewReader(data)
-	c1a, err := format.SumStream(reader)
-	if err != nil {
-		t.Fatal(err)
-	}
-	if !c1a.Equals(c1) {
-		t.Fatal("Sum and SumStream create different cids")
-	}
-
 	// Construct c2
 	hash, err := mh.Sum(data, mh.SHA2_256, -1)
 	if err != nil {
@@ -39,16 +29,6 @@ func TestV0Builder(t *testing.T) {
 	if c1.Prefix() != c2.Prefix() {
 		t.Fatal("prefixes mismatch")
 	}
-
-	reader.Seek(0, 0)
-	hash, err = mh.SumStream(reader, mh.SHA2_256, -1)
-	if err != nil {
-		t.Fatal(err)
-	}
-	c2 = NewCidV0(hash)
-	if !c1.Equals(c2) {
-		t.Fatal("Sum and SumStream create different cids")
-	}
 }
 
 func TestV1Builder(t *testing.T) {
@@ -61,15 +41,6 @@ func TestV1Builder(t *testing.T) {
 		t.Fatal(err)
 	}
 
-	reader := bytes.NewReader(data)
-	c1a, err := format.SumStream(reader)
-	if err != nil {
-		t.Fatal(err)
-	}
-	if !c1a.Equals(c1) {
-		t.Fatal("Sum and SumStream create different cids")
-	}
-
 	// Construct c2
 	hash, err := mh.Sum(data, mh.SHA2_256, -1)
 	if err != nil {

cid.go

@@ -10,7 +10,7 @@
 //
 // A CIDv1 has four parts:
 //
-//     <cidv1> ::= <multibase-prefix><cid-version><multicodec-packed-content-type><multihash-content-address>
+//	<cidv1> ::= <multibase-prefix><cid-version><multicodec-packed-content-type><multihash-content-address>
 //
 // As shown above, the CID implementation relies heavily on Multiformats,
 // particularly Multibase
@@ -22,6 +22,7 @@ package cid
 import (
 	"bytes"
 	"encoding"
+	"encoding/binary"
 	"encoding/json"
 	"errors"
 	"fmt"
@@ -36,29 +37,53 @@ import (
 // UnsupportedVersionString just holds an error message
 const UnsupportedVersionString = "<unsupported cid version>"
 
+// ErrInvalidCid is an error that indicates that a CID is invalid.
+type ErrInvalidCid struct {
+	Err error
+}
+
+func (e ErrInvalidCid) Error() string {
+	return fmt.Sprintf("invalid cid: %s", e.Err)
+}
+
+func (e ErrInvalidCid) Unwrap() error {
+	return e.Err
+}
+
+func (e ErrInvalidCid) Is(err error) bool {
+	switch err.(type) {
+	case ErrInvalidCid, *ErrInvalidCid:
+		return true
+	default:
+		return false
+	}
+}
+
 var (
 	// ErrCidTooShort means that the cid passed to decode was not long
 	// enough to be a valid Cid
-	ErrCidTooShort = errors.New("cid too short")
+	ErrCidTooShort = ErrInvalidCid{errors.New("cid too short")}
 
 	// ErrInvalidEncoding means that selected encoding is not supported
 	// by this Cid version
 	ErrInvalidEncoding = errors.New("invalid base encoding")
 )
 
-// These are multicodec-packed content types. The should match
-// the codes described in the authoritative document:
-// https://github.com/multiformats/multicodec/blob/master/table.csv
+// Consts below are DEPRECATED and left only for legacy reasons:
+// <https://github.com/ipfs/go-cid/pull/137>
+// Modern code should use consts from go-multicodec instead:
+// <https://github.com/multiformats/go-multicodec>
 const (
-	Raw = 0x55
+	// common ones
+	Raw         = 0x55
+	DagProtobuf = 0x70   // https://ipld.io/docs/codecs/known/dag-pb/
+	DagCBOR     = 0x71   // https://ipld.io/docs/codecs/known/dag-cbor/
+	DagJSON     = 0x0129 // https://ipld.io/docs/codecs/known/dag-json/
+	Libp2pKey   = 0x72   // https://github.com/libp2p/specs/blob/master/peer-ids/peer-ids.md#peer-ids
 
-	DagProtobuf = 0x70
-	DagCBOR     = 0x71
-	Libp2pKey   = 0x72
-
-	GitRaw = 0x78
-
-	DagJOSE               = 0x85
+	// other
+	GitRaw                = 0x78
+	DagJOSE               = 0x85 // https://ipld.io/specs/codecs/dag-jose/spec/
 	EthBlock              = 0x90
 	EthBlockList          = 0x91
 	EthTxTrie             = 0x92
@@ -80,64 +105,6 @@ const (
 	FilCommitmentSealed   = 0xf102
 )
 
-// Codecs maps the name of a codec to its type
-var Codecs = map[string]uint64{
-	"v0":                      DagProtobuf,
-	"raw":                     Raw,
-	"protobuf":                DagProtobuf,
-	"cbor":                    DagCBOR,
-	"libp2p-key":              Libp2pKey,
-	"git-raw":                 GitRaw,
-	"eth-block":               EthBlock,
-	"eth-block-list":          EthBlockList,
-	"eth-tx-trie":             EthTxTrie,
-	"eth-tx":                  EthTx,
-	"eth-tx-receipt-trie":     EthTxReceiptTrie,
-	"eth-tx-receipt":          EthTxReceipt,
-	"eth-state-trie":          EthStateTrie,
-	"eth-account-snapshot":    EthAccountSnapshot,
-	"eth-storage-trie":        EthStorageTrie,
-	"bitcoin-block":           BitcoinBlock,
-	"bitcoin-tx":              BitcoinTx,
-	"zcash-block":             ZcashBlock,
-	"zcash-tx":                ZcashTx,
-	"decred-block":            DecredBlock,
-	"decred-tx":               DecredTx,
-	"dash-block":              DashBlock,
-	"dash-tx":                 DashTx,
-	"fil-commitment-unsealed": FilCommitmentUnsealed,
-	"fil-commitment-sealed":   FilCommitmentSealed,
-	"dag-jose":                DagJOSE,
-}
-
-// CodecToStr maps the numeric codec to its name
-var CodecToStr = map[uint64]string{
-	Raw:                   "raw",
-	DagProtobuf:           "protobuf",
-	DagCBOR:               "cbor",
-	GitRaw:                "git-raw",
-	EthBlock:              "eth-block",
-	EthBlockList:          "eth-block-list",
-	EthTxTrie:             "eth-tx-trie",
-	EthTx:                 "eth-tx",
-	EthTxReceiptTrie:      "eth-tx-receipt-trie",
-	EthTxReceipt:          "eth-tx-receipt",
-	EthStateTrie:          "eth-state-trie",
-	EthAccountSnapshot:    "eth-account-snapshot",
-	EthStorageTrie:        "eth-storage-trie",
-	BitcoinBlock:          "bitcoin-block",
-	BitcoinTx:             "bitcoin-tx",
-	ZcashBlock:            "zcash-block",
-	ZcashTx:               "zcash-tx",
-	DecredBlock:           "decred-block",
-	DecredTx:              "decred-tx",
-	DashBlock:             "dash-block",
-	DashTx:                "dash-tx",
-	FilCommitmentUnsealed: "fil-commitment-unsealed",
-	FilCommitmentSealed:   "fil-commitment-sealed",
-	DagJOSE:               "dag-jose",
-}
-
 // tryNewCidV0 tries to convert a multihash into a CIDv0 CID and returns an
 // error on failure.
 func tryNewCidV0(mhash mh.Multihash) (Cid, error) {
@@ -145,10 +112,10 @@ func tryNewCidV0(mhash mh.Multihash) (Cid, error) {
 	// incorrectly detect it as CidV1 in the Version() method
 	dec, err := mh.Decode(mhash)
 	if err != nil {
-		return Undef, err
+		return Undef, ErrInvalidCid{err}
 	}
 	if dec.Code != mh.SHA2_256 || dec.Length != 32 {
-		return Undef, fmt.Errorf("invalid hash for cidv0 %d-%d", dec.Code, dec.Length)
+		return Undef, ErrInvalidCid{fmt.Errorf("invalid hash for cidv0 %d-%d", dec.Code, dec.Length)}
 	}
 	return Cid{string(mhash)}, nil
 }
@@ -173,22 +140,32 @@ func NewCidV0(mhash mh.Multihash) Cid {
 // Panics if the multihash is invalid.
 func NewCidV1(codecType uint64, mhash mh.Multihash) Cid {
 	hashlen := len(mhash)
-	// two 8 bytes (max) numbers plus hash
-	buf := make([]byte, 1+varint.UvarintSize(codecType)+hashlen)
-	n := varint.PutUvarint(buf, 1)
-	n += varint.PutUvarint(buf[n:], codecType)
-	cn := copy(buf[n:], mhash)
+
+	// Two 8 bytes (max) numbers plus hash.
+	// We use strings.Builder to only allocate once.
+	var b strings.Builder
+	b.Grow(1 + varint.UvarintSize(codecType) + hashlen)
+
+	b.WriteByte(1)
+
+	var buf [binary.MaxVarintLen64]byte
+	n := varint.PutUvarint(buf[:], codecType)
+	b.Write(buf[:n])
+
+	cn, _ := b.Write(mhash)
 	if cn != hashlen {
 		panic("copy hash length is inconsistent")
 	}
 
-	return Cid{string(buf[:n+hashlen])}
+	return Cid{b.String()}
 }
 
-var _ encoding.BinaryMarshaler = Cid{}
-var _ encoding.BinaryUnmarshaler = (*Cid)(nil)
-var _ encoding.TextMarshaler = Cid{}
-var _ encoding.TextUnmarshaler = (*Cid)(nil)
+var (
+	_ encoding.BinaryMarshaler   = Cid{}
+	_ encoding.BinaryUnmarshaler = (*Cid)(nil)
+	_ encoding.TextMarshaler     = Cid{}
+	_ encoding.TextUnmarshaler   = (*Cid)(nil)
+)
 
 // Cid represents a self-describing content addressed
 // identifier. It is formed by a Version, a Codec (which indicates
@@ -222,14 +199,23 @@ func Parse(v interface{}) (Cid, error) {
 	case Cid:
 		return v2, nil
 	default:
-		return Undef, fmt.Errorf("can't parse %+v as Cid", v2)
+		return Undef, ErrInvalidCid{fmt.Errorf("can't parse %+v as Cid", v2)}
 	}
 }
 
+// MustParse calls Parse but will panic on error.
+func MustParse(v interface{}) Cid {
+	c, err := Parse(v)
+	if err != nil {
+		panic(err)
+	}
+	return c
+}
+
 // Decode parses a Cid-encoded string and returns a Cid object.
 // For CidV1, a Cid-encoded string is primarily a multibase string:
 //
-//     <multibase-type-code><base-encoded-string>
+//	<multibase-type-code><base-encoded-string>
 //
 // The base-encoded string represents a:
 //
@@ -246,7 +232,7 @@ func Decode(v string) (Cid, error) {
 	if len(v) == 46 && v[:2] == "Qm" {
 		hash, err := mh.FromB58String(v)
 		if err != nil {
-			return Undef, err
+			return Undef, ErrInvalidCid{err}
 		}
 
 		return tryNewCidV0(hash)
@@ -254,7 +240,7 @@ func Decode(v string) (Cid, error) {
 
 	_, data, err := mbase.Decode(v)
 	if err != nil {
-		return Undef, err
+		return Undef, ErrInvalidCid{err}
 	}
 
 	return Cast(data)
@@ -276,7 +262,7 @@ func ExtractEncoding(v string) (mbase.Encoding, error) {
 	// check encoding is valid
 	_, err := mbase.NewEncoder(encoding)
 	if err != nil {
-		return -1, err
+		return -1, ErrInvalidCid{err}
 	}
 
 	return encoding, nil
@@ -285,7 +271,7 @@ func ExtractEncoding(v string) (mbase.Encoding, error) {
 // Cast takes a Cid data slice, parses it and returns a Cid.
 // For CidV1, the data buffer is in the form:
 //
-//     <version><codec-type><multihash>
+//	<version><codec-type><multihash>
 //
 // CidV0 are also supported. In particular, data buffers starting
 // with length 34 bytes, which starts with bytes [18,32...] are considered
@@ -296,11 +282,11 @@ func ExtractEncoding(v string) (mbase.Encoding, error) {
 func Cast(data []byte) (Cid, error) {
 	nr, c, err := CidFromBytes(data)
 	if err != nil {
-		return Undef, err
+		return Undef, ErrInvalidCid{err}
 	}
 
 	if nr != len(data) {
-		return Undef, fmt.Errorf("trailing bytes in data buffer passed to cid Cast")
+		return Undef, ErrInvalidCid{fmt.Errorf("trailing bytes in data buffer passed to cid Cast")}
 	}
 
 	return c, nil
@@ -414,7 +400,13 @@ func (c Cid) Hash() mh.Multihash {
 // Bytes returns the byte representation of a Cid.
 // The output of bytes can be parsed back into a Cid
 // with Cast().
+//
+// If c.Defined() == false, it return a nil slice and may not
+// be parsable with Cast().
 func (c Cid) Bytes() []byte {
+	if !c.Defined() {
+		return nil
+	}
 	return []byte(c.str)
 }
 
@@ -464,7 +456,7 @@ func (c Cid) Equals(o Cid) bool {
 // UnmarshalJSON parses the JSON representation of a Cid.
 func (c *Cid) UnmarshalJSON(b []byte) error {
 	if len(b) < 2 {
-		return fmt.Errorf("invalid cid json blob")
+		return ErrInvalidCid{fmt.Errorf("invalid cid json blob")}
 	}
 	obj := struct {
 		CidTarget string `json:"/"`
@@ -472,7 +464,7 @@ func (c *Cid) UnmarshalJSON(b []byte) error {
 	objptr := &obj
 	err := json.Unmarshal(b, &objptr)
 	if err != nil {
-		return err
+		return ErrInvalidCid{err}
 	}
 	if objptr == nil {
 		*c = Cid{}
@@ -480,12 +472,12 @@ func (c *Cid) UnmarshalJSON(b []byte) error {
 	}
 
 	if obj.CidTarget == "" {
-		return fmt.Errorf("cid was incorrectly formatted")
+		return ErrInvalidCid{fmt.Errorf("cid was incorrectly formatted")}
 	}
 
 	out, err := Decode(obj.CidTarget)
 	if err != nil {
-		return err
+		return ErrInvalidCid{err}
 	}
 
 	*c = out
@@ -495,7 +487,7 @@ func (c *Cid) UnmarshalJSON(b []byte) error {
 
 // MarshalJSON procudes a JSON representation of a Cid, which looks as follows:
 //
-//    { "/": "<cid-string>" }
+//	{ "/": "<cid-string>" }
 //
 // Note that this formatting comes from the IPLD specification
 // (https://github.com/ipld/specs/tree/master/ipld)
@@ -552,7 +544,8 @@ func (c Cid) Prefix() Prefix {
 // and the Multihash length. It does not contains
 // any actual content information.
 // NOTE: The use -1 in MhLength to mean default length is deprecated,
-//   use the V0Builder or V1Builder structures instead
+//
+//	use the V0Builder or V1Builder structures instead
 type Prefix struct {
 	Version  uint64
 	Codec    uint64
@@ -564,19 +557,19 @@ type Prefix struct {
 // and return a newly constructed Cid with the resulting multihash.
 func (p Prefix) Sum(data []byte) (Cid, error) {
 	length := p.MhLength
-	if p.MhType == mh.ID {
+	if p.MhType == mh.IDENTITY {
 		length = -1
 	}
 
 	if p.Version == 0 && (p.MhType != mh.SHA2_256 ||
 		(p.MhLength != 32 && p.MhLength != -1)) {
 
-		return Undef, fmt.Errorf("invalid v0 prefix")
+		return Undef, ErrInvalidCid{fmt.Errorf("invalid v0 prefix")}
 	}
 
 	hash, err := mh.Sum(data, p.MhType, length)
 	if err != nil {
-		return Undef, err
+		return Undef, ErrInvalidCid{err}
 	}
 
 	switch p.Version {
@@ -585,43 +578,13 @@ func (p Prefix) Sum(data []byte) (Cid, error) {
 	case 1:
 		return NewCidV1(p.Codec, hash), nil
 	default:
-		return Undef, fmt.Errorf("invalid cid version")
-	}
-}
-
-// SumStream uses the information in a prefix, and data read from a io.Reader,
-// to perform a multihash.SumStream() and return a newly constructed Cid with
-// the resulting multihash.
-func (p Prefix) SumStream(r io.Reader) (Cid, error) {
-	length := p.MhLength
-	if p.MhType == mh.ID {
-		length = -1
-	}
-
-	if p.Version == 0 && (p.MhType != mh.SHA2_256 ||
-		(p.MhLength != 32 && p.MhLength != -1)) {
-
-		return Undef, fmt.Errorf("invalid v0 prefix")
-	}
-
-	hash, err := mh.SumStream(r, p.MhType, length)
-	if err != nil {
-		return Undef, err
-	}
-
-	switch p.Version {
-	case 0:
-		return NewCidV0(hash), nil
-	case 1:
-		return NewCidV1(p.Codec, hash), nil
-	default:
-		return Undef, fmt.Errorf("invalid cid version")
+		return Undef, ErrInvalidCid{fmt.Errorf("invalid cid version")}
 	}
 }
 
 // Bytes returns a byte representation of a Prefix. It looks like:
 //
-//     <version><codec><mh-type><mh-length>
+//	<version><codec><mh-type><mh-length>
 func (p Prefix) Bytes() []byte {
 	size := varint.UvarintSize(p.Version)
 	size += varint.UvarintSize(p.Codec)
@@ -645,22 +608,22 @@ func PrefixFromBytes(buf []byte) (Prefix, error) {
 	r := bytes.NewReader(buf)
 	vers, err := varint.ReadUvarint(r)
 	if err != nil {
-		return Prefix{}, err
+		return Prefix{}, ErrInvalidCid{err}
 	}
 
 	codec, err := varint.ReadUvarint(r)
 	if err != nil {
-		return Prefix{}, err
+		return Prefix{}, ErrInvalidCid{err}
 	}
 
 	mhtype, err := varint.ReadUvarint(r)
 	if err != nil {
-		return Prefix{}, err
+		return Prefix{}, ErrInvalidCid{err}
 	}
 
 	mhlen, err := varint.ReadUvarint(r)
 	if err != nil {
-		return Prefix{}, err
+		return Prefix{}, ErrInvalidCid{err}
 	}
 
 	return Prefix{
@@ -674,12 +637,12 @@ func PrefixFromBytes(buf []byte) (Prefix, error) {
 func CidFromBytes(data []byte) (int, Cid, error) {
 	if len(data) > 2 && data[0] == mh.SHA2_256 && data[1] == 32 {
 		if len(data) < 34 {
-			return 0, Undef, fmt.Errorf("not enough bytes for cid v0")
+			return 0, Undef, ErrInvalidCid{fmt.Errorf("not enough bytes for cid v0")}
 		}
 
 		h, err := mh.Cast(data[:34])
 		if err != nil {
-			return 0, Undef, err
+			return 0, Undef, ErrInvalidCid{err}
 		}
 
 		return 34, Cid{string(h)}, nil
@@ -687,24 +650,168 @@ func CidFromBytes(data []byte) (int, Cid, error) {
 
 	vers, n, err := varint.FromUvarint(data)
 	if err != nil {
-		return 0, Undef, err
+		return 0, Undef, ErrInvalidCid{err}
 	}
 
 	if vers != 1 {
-		return 0, Undef, fmt.Errorf("expected 1 as the cid version number, got: %d", vers)
+		return 0, Undef, ErrInvalidCid{fmt.Errorf("expected 1 as the cid version number, got: %d", vers)}
 	}
 
 	_, cn, err := varint.FromUvarint(data[n:])
 	if err != nil {
-		return 0, Undef, err
+		return 0, Undef, ErrInvalidCid{err}
 	}
 
 	mhnr, _, err := mh.MHFromBytes(data[n+cn:])
 	if err != nil {
-		return 0, Undef, err
+		return 0, Undef, ErrInvalidCid{err}
 	}
 
 	l := n + cn + mhnr
 
 	return l, Cid{string(data[0:l])}, nil
 }
+
+func toBufByteReader(r io.Reader, dst []byte) *bufByteReader {
+	// If the reader already implements ByteReader, use it directly.
+	// Otherwise, use a fallback that does 1-byte Reads.
+	if br, ok := r.(io.ByteReader); ok {
+		return &bufByteReader{direct: br, dst: dst}
+	}
+	return &bufByteReader{fallback: r, dst: dst}
+}
+
+type bufByteReader struct {
+	direct   io.ByteReader
+	fallback io.Reader
+
+	dst []byte
+}
+
+func (r *bufByteReader) ReadByte() (byte, error) {
+	// The underlying reader has ReadByte; use it.
+	if br := r.direct; br != nil {
+		b, err := br.ReadByte()
+		if err != nil {
+			return 0, err
+		}
+		r.dst = append(r.dst, b)
+		return b, nil
+	}
+
+	// Fall back to a one-byte Read.
+	// TODO: consider reading straight into dst,
+	// once we have benchmarks and if they prove that to be faster.
+	var p [1]byte
+	if _, err := io.ReadFull(r.fallback, p[:]); err != nil {
+		return 0, err
+	}
+	r.dst = append(r.dst, p[0])
+	return p[0], nil
+}
+
+// CidFromReader reads a precise number of bytes for a CID from a given reader.
+// It returns the number of bytes read, the CID, and any error encountered.
+// The number of bytes read is accurate even if a non-nil error is returned.
+//
+// It's recommended to supply a reader that buffers and implements io.ByteReader,
+// as CidFromReader has to do many single-byte reads to decode varints.
+// If the argument only implements io.Reader, single-byte Read calls are used instead.
+//
+// If the Reader is found to yield zero bytes, an io.EOF error is returned directly, in all
+// other error cases, an ErrInvalidCid, wrapping the original error, is returned.
+func CidFromReader(r io.Reader) (int, Cid, error) {
+	// 64 bytes is enough for any CIDv0,
+	// and it's enough for most CIDv1s in practice.
+	// If the digest is too long, we'll allocate more.
+	br := toBufByteReader(r, make([]byte, 0, 64))
+
+	// We read the first varint, to tell if this is a CIDv0 or a CIDv1.
+	// The varint package wants a io.ByteReader, so we must wrap our io.Reader.
+	vers, err := varint.ReadUvarint(br)
+	if err != nil {
+		if err == io.EOF {
+			// First-byte read in ReadUvarint errors with io.EOF, so reader has no data.
+			// Subsequent reads with an EOF will return io.ErrUnexpectedEOF and be wrapped here.
+			return 0, Undef, err
+		}
+		return len(br.dst), Undef, ErrInvalidCid{err}
+	}
+
+	// If we have a CIDv0, read the rest of the bytes and cast the buffer.
+	if vers == mh.SHA2_256 {
+		if n, err := io.ReadFull(r, br.dst[1:34]); err != nil {
+			return len(br.dst) + n, Undef, ErrInvalidCid{err}
+		}
+
+		br.dst = br.dst[:34]
+		h, err := mh.Cast(br.dst)
+		if err != nil {
+			return len(br.dst), Undef, ErrInvalidCid{err}
+		}
+
+		return len(br.dst), Cid{string(h)}, nil
+	}
+
+	if vers != 1 {
+		return len(br.dst), Undef, ErrInvalidCid{fmt.Errorf("expected 1 as the cid version number, got: %d", vers)}
+	}
+
+	// CID block encoding multicodec.
+	_, err = varint.ReadUvarint(br)
+	if err != nil {
+		return len(br.dst), Undef, ErrInvalidCid{err}
+	}
+
+	// We could replace most of the code below with go-multihash's ReadMultihash.
+	// Note that it would save code, but prevent reusing buffers.
+	// Plus, we already have a ByteReader now.
+	mhStart := len(br.dst)
+
+	// Multihash hash function code.
+	_, err = varint.ReadUvarint(br)
+	if err != nil {
+		return len(br.dst), Undef, ErrInvalidCid{err}
+	}
+
+	// Multihash digest length.
+	mhl, err := varint.ReadUvarint(br)
+	if err != nil {
+		return len(br.dst), Undef, ErrInvalidCid{err}
+	}
+
+	// Refuse to make large allocations to prevent OOMs due to bugs.
+	const maxDigestAlloc = 32 << 20 // 32MiB
+	if mhl > maxDigestAlloc {
+		return len(br.dst), Undef, ErrInvalidCid{fmt.Errorf("refusing to allocate %d bytes for a digest", mhl)}
+	}
+
+	// Fine to convert mhl to int, given maxDigestAlloc.
+	prefixLength := len(br.dst)
+	cidLength := prefixLength + int(mhl)
+	if cidLength > cap(br.dst) {
+		// If the multihash digest doesn't fit in our initial 64 bytes,
+		// efficiently extend the slice via append+make.
+		br.dst = append(br.dst, make([]byte, cidLength-len(br.dst))...)
+	} else {
+		// The multihash digest fits inside our buffer,
+		// so just extend its capacity.
+		br.dst = br.dst[:cidLength]
+	}
+
+	if n, err := io.ReadFull(r, br.dst[prefixLength:cidLength]); err != nil {
+		// We can't use len(br.dst) here,
+		// as we've only read n bytes past prefixLength.
+		return prefixLength + n, Undef, ErrInvalidCid{err}
+	}
+
+	// This simply ensures the multihash is valid.
+	// TODO: consider removing this bit, as it's probably redundant;
+	// for now, it helps ensure consistency with CidFromBytes.
+	_, _, err = mh.MHFromBytes(br.dst[mhStart:])
+	if err != nil {
+		return len(br.dst), Undef, ErrInvalidCid{err}
+	}
+
+	return len(br.dst), Cid{string(br.dst)}, nil
+}

cid_fuzz.go

@@ -1,10 +1,9 @@
-// +build gofuzz
+//go:build gofuzz
 
 package cid
 
 func Fuzz(data []byte) int {
 	cid, err := Cast(data)
-
 	if err != nil {
 		return 0
 	}

cid_test.go

@@ -2,48 +2,21 @@ package cid
 
 import (
 	"bytes"
+	crand "crypto/rand"
 	"encoding/json"
+	"errors"
 	"fmt"
+	"io"
 	"math/rand"
 	"reflect"
 	"strings"
 	"testing"
+	"testing/iotest"
 
 	mbase "github.com/multiformats/go-multibase"
 	mh "github.com/multiformats/go-multihash"
 )
 
-// Copying the "silly test" idea from
-// https://github.com/multiformats/go-multihash/blob/7aa9f26a231c6f34f4e9fad52bf580fd36627285/multihash_test.go#L13
-// Makes it so changing the table accidentally has to happen twice.
-var tCodecs = map[uint64]string{
-	Raw:                   "raw",
-	DagProtobuf:           "protobuf",
-	DagCBOR:               "cbor",
-	Libp2pKey:             "libp2p-key",
-	GitRaw:                "git-raw",
-	EthBlock:              "eth-block",
-	EthBlockList:          "eth-block-list",
-	EthTxTrie:             "eth-tx-trie",
-	EthTx:                 "eth-tx",
-	EthTxReceiptTrie:      "eth-tx-receipt-trie",
-	EthTxReceipt:          "eth-tx-receipt",
-	EthStateTrie:          "eth-state-trie",
-	EthAccountSnapshot:    "eth-account-snapshot",
-	EthStorageTrie:        "eth-storage-trie",
-	BitcoinBlock:          "bitcoin-block",
-	BitcoinTx:             "bitcoin-tx",
-	ZcashBlock:            "zcash-block",
-	ZcashTx:               "zcash-tx",
-	DecredBlock:           "decred-block",
-	DecredTx:              "decred-tx",
-	DashBlock:             "dash-block",
-	DashTx:                "dash-tx",
-	FilCommitmentUnsealed: "fil-commitment-unsealed",
-	FilCommitmentSealed:   "fil-commitment-sealed",
-	DagJOSE:               "dag-jose",
-}
-
 func assertEqual(t *testing.T, a, b Cid) {
 	if a.Type() != b.Type() {
 		t.Fatal("mismatch on type")
@@ -58,31 +31,11 @@ func assertEqual(t *testing.T, a, b Cid) {
 	}
 }
 
-func TestTable(t *testing.T) {
-	if len(tCodecs) != len(Codecs)-1 {
-		t.Errorf("Item count mismatch in the Table of Codec. Should be %d, got %d", len(tCodecs)+1, len(Codecs))
-	}
-
-	for k, v := range tCodecs {
-		if Codecs[v] != k {
-			t.Errorf("Table mismatch: 0x%x %s", k, v)
-		}
-	}
-}
-
-// The table returns cid.DagProtobuf for "v0"
-// so we test it apart
-func TestTableForV0(t *testing.T) {
-	if Codecs["v0"] != DagProtobuf {
-		t.Error("Table mismatch: Codecs[\"v0\"] should resolve to DagProtobuf (0x70)")
-	}
-}
-
 func TestPrefixSum(t *testing.T) {
 	// Test creating CIDs both manually and with Prefix.
 	// Tests: https://github.com/ipfs/go-cid/issues/83
 	for _, hashfun := range []uint64{
-		mh.ID, mh.SHA3, mh.SHA2_256,
+		mh.IDENTITY, mh.SHA3, mh.SHA2_256,
 	} {
 		h1, err := mh.Sum([]byte("TEST"), hashfun, -1)
 		if err != nil {
@@ -210,6 +163,9 @@ func TestBasesMarshaling(t *testing.T) {
 	if err == nil {
 		t.Fatal("expected too-short error from ExtractEncoding")
 	}
+	if !errors.Is(err, ErrInvalidCid{}) {
+		t.Fatal("expected error to be ErrInvalidCid")
+	}
 	if ee != -1 {
 		t.Fatal("expected -1 from too-short ExtractEncoding")
 	}
@@ -275,6 +231,9 @@ func TestEmptyString(t *testing.T) {
 	if err == nil {
 		t.Fatal("shouldnt be able to parse an empty cid")
 	}
+	if !errors.Is(err, ErrInvalidCid{}) {
+		t.Fatal("error must be ErrInvalidCid")
+	}
 }
 
 func TestV0Handling(t *testing.T) {
@@ -330,6 +289,9 @@ func TestV0ErrorCases(t *testing.T) {
 	if err == nil {
 		t.Fatal("should have failed to decode that ref")
 	}
+	if !errors.Is(err, ErrInvalidCid{}) {
+		t.Fatal("error must be ErrInvalidCid")
+	}
 }
 
 func TestNewPrefixV1(t *testing.T) {
@@ -388,7 +350,6 @@ func TestNewPrefixV0(t *testing.T) {
 	if c1.Prefix() != c2.Prefix() {
 		t.Fatal("prefixes mismatch")
 	}
-
 }
 
 func TestInvalidV0Prefix(t *testing.T) {
@@ -421,6 +382,9 @@ func TestInvalidV0Prefix(t *testing.T) {
 		if err == nil {
 			t.Fatalf("should error (index %d)", i)
 		}
+		if !errors.Is(err, ErrInvalidCid{}) {
+			t.Fatal("expected error to be ErrInvalidCid")
+		}
 	}
 }
 
@@ -430,6 +394,9 @@ func TestBadPrefix(t *testing.T) {
 	if err == nil {
 		t.Fatalf("expected error on v3 prefix Sum")
 	}
+	if !errors.Is(err, ErrInvalidCid{}) {
+		t.Fatal("expected error to be ErrInvalidCid")
+	}
 }
 
 func TestPrefixRoundtrip(t *testing.T) {
@@ -466,18 +433,30 @@ func TestBadPrefixFromBytes(t *testing.T) {
 	if err == nil {
 		t.Fatal("expected error for bad byte 0")
 	}
+	if !errors.Is(err, ErrInvalidCid{}) {
+		t.Fatal("expected error to be ErrInvalidCid")
+	}
 	_, err = PrefixFromBytes([]byte{0x01, 0x80})
 	if err == nil {
 		t.Fatal("expected error for bad byte 1")
 	}
+	if !errors.Is(err, ErrInvalidCid{}) {
+		t.Fatal("expected error to be ErrInvalidCid")
+	}
 	_, err = PrefixFromBytes([]byte{0x01, 0x01, 0x80})
 	if err == nil {
 		t.Fatal("expected error for bad byte 2")
 	}
+	if !errors.Is(err, ErrInvalidCid{}) {
+		t.Fatal("expected error to be ErrInvalidCid")
+	}
 	_, err = PrefixFromBytes([]byte{0x01, 0x01, 0x01, 0x80})
 	if err == nil {
 		t.Fatal("expected error for bad byte 3")
 	}
+	if !errors.Is(err, ErrInvalidCid{}) {
+		t.Fatal("expected error to be ErrInvalidCid")
+	}
 }
 
 func Test16BytesVarint(t *testing.T) {
@@ -491,7 +470,7 @@ func TestFuzzCid(t *testing.T) {
 	buf := make([]byte, 128)
 	for i := 0; i < 200; i++ {
 		s := rand.Intn(128)
-		rand.Read(buf[:s])
+		crand.Read(buf[:s])
 		_, _ = Cast(buf[:s])
 	}
 }
@@ -504,6 +483,9 @@ func TestParse(t *testing.T) {
 	if !strings.Contains(err.Error(), "can't parse 123 as Cid") {
 		t.Fatalf("expected int error, got %s", err.Error())
 	}
+	if !errors.Is(err, ErrInvalidCid{}) {
+		t.Fatalf("expected ErrInvalidCid, got %s", err.Error())
+	}
 
 	theHash := "QmdfTbBqBPQ7VNxZEYEj14VmRuZBkqFbiwReogJgS1zR1n"
 	h, err := mh.FromB58String(theHash)
@@ -512,13 +494,13 @@ func TestParse(t *testing.T) {
 	}
 
 	assertions := [][]interface{}{
-		[]interface{}{NewCidV0(h), theHash},
-		[]interface{}{NewCidV0(h).Bytes(), theHash},
-		[]interface{}{h, theHash},
-		[]interface{}{theHash, theHash},
-		[]interface{}{"/ipfs/" + theHash, theHash},
-		[]interface{}{"https://ipfs.io/ipfs/" + theHash, theHash},
-		[]interface{}{"http://localhost:8080/ipfs/" + theHash, theHash},
+		{NewCidV0(h), theHash},
+		{NewCidV0(h).Bytes(), theHash},
+		{h, theHash},
+		{theHash, theHash},
+		{"/ipfs/" + theHash, theHash},
+		{"https://ipfs.io/ipfs/" + theHash, theHash},
+		{"http://localhost:8080/ipfs/" + theHash, theHash},
 	}
 
 	assert := func(arg interface{}, expected string) error {
@@ -621,17 +603,29 @@ func TestJsonRoundTrip(t *testing.T) {
 		t.Fatal("cids not equal for Cid")
 	}
 
-	if err = actual2.UnmarshalJSON([]byte("1")); err == nil {
+	err = actual2.UnmarshalJSON([]byte("1"))
+	if err == nil {
 		t.Fatal("expected error for too-short JSON")
 	}
-
-	if err = actual2.UnmarshalJSON([]byte(`{"nope":"nope"}`)); err == nil {
-		t.Fatal("expected error for bad CID JSON")
+	if !errors.Is(err, ErrInvalidCid{}) {
+		t.Fatal("expected error to be ErrInvalidCid")
 	}
 
-	if err = actual2.UnmarshalJSON([]byte(`bad "" json!`)); err == nil {
+	err = actual2.UnmarshalJSON([]byte(`{"nope":"nope"}`))
+	if err == nil {
+		t.Fatal("expected error for bad CID JSON")
+	}
+	if !errors.Is(err, ErrInvalidCid{}) {
+		t.Fatal("expected error to be ErrInvalidCid")
+	}
+
+	err = actual2.UnmarshalJSON([]byte(`bad "" json!`))
+	if err == nil {
 		t.Fatal("expected error for bad JSON")
 	}
+	if !errors.Is(err, ErrInvalidCid{}) {
+		t.Fatal("expected error to be ErrInvalidCid")
+	}
 
 	var actual3 Cid
 	enc, err = actual3.MarshalJSON()
@@ -666,6 +660,7 @@ func TestReadCidsFromBuffer(t *testing.T) {
 		"k2cwueckqkibutvhkr4p2ln2pjcaxaakpd9db0e7j7ax1lxhhxy3ekpv",
 		"Qmf5Qzp6nGBku7CEn2UQx4mgN8TW69YUok36DrGa6NN893",
 		"zb2rhZi1JR4eNc2jBGaRYJKYM8JEB4ovenym8L1CmFsRAytkz",
+		"bafkqarjpmzuwyzltorxxezjpkvcfgqkfjfbfcvslivje2vchkzdu6rckjjcfgtkolaze6mssjqzeyn2ekrcfatkjku2vowseky3fswkfkm2deqkrju3e2",
 	}
 
 	var cids []Cid
@@ -693,51 +688,122 @@ func TestReadCidsFromBuffer(t *testing.T) {
 	if cur != len(buf) {
 		t.Fatal("had trailing bytes")
 	}
+
+	// The same, but now with CidFromReader.
+	// In multiple forms, to catch more io interface bugs.
+	for _, r := range []io.Reader{
+		// implements io.ByteReader
+		bytes.NewReader(buf),
+
+		// tiny reads, no io.ByteReader
+		iotest.OneByteReader(bytes.NewReader(buf)),
+	} {
+		cur = 0
+		for _, expc := range cids {
+			n, c, err := CidFromReader(r)
+			if err != nil {
+				t.Fatal(err)
+			}
+			if c != expc {
+				t.Fatal("cids mismatched")
+			}
+			cur += n
+		}
+		if cur != len(buf) {
+			t.Fatal("had trailing bytes")
+		}
+	}
 }
 
-func TestBadCidFromBytes(t *testing.T) {
-	l, c, err := CidFromBytes([]byte{mh.SHA2_256, 32, 0x00})
-	if err == nil {
-		t.Fatal("expected not-enough-bytes for V0 CidFromBytes")
+func TestBadCidInput(t *testing.T) {
+	for _, name := range []string{
+		"FromBytes",
+		"FromReader",
+	} {
+		t.Run(name, func(t *testing.T) {
+			usingReader := name == "FromReader"
+
+			fromBytes := CidFromBytes
+			if usingReader {
+				fromBytes = func(data []byte) (int, Cid, error) {
+					return CidFromReader(bytes.NewReader(data))
+				}
+			}
+
+			l, c, err := fromBytes([]byte{mh.SHA2_256, 32, 0x00})
+			if err == nil {
+				t.Fatal("expected not-enough-bytes for V0 CID")
+			}
+			if !usingReader && l != 0 {
+				t.Fatal("expected length==0 from bad CID")
+			} else if usingReader && l == 0 {
+				t.Fatal("expected length!=0 from bad CID")
+			}
+			if c != Undef {
+				t.Fatal("expected Undef CID from bad CID")
+			}
+
+			c, err = Decode("bafkreie5qrjvaw64n4tjm6hbnm7fnqvcssfed4whsjqxzslbd3jwhsk3mm")
+			if err != nil {
+				t.Fatal(err)
+			}
+			byts := make([]byte, c.ByteLen())
+			copy(byts, c.Bytes())
+			byts[1] = 0x80 // bad codec varint
+			byts[2] = 0x00
+			l, c, err = fromBytes(byts)
+			if err == nil {
+				t.Fatal("expected not-enough-bytes for V1 CID")
+			}
+			if !usingReader && l != 0 {
+				t.Fatal("expected length==0 from bad CID")
+			} else if usingReader && l == 0 {
+				t.Fatal("expected length!=0 from bad CID")
+			}
+			if c != Undef {
+				t.Fatal("expected Undef CID from bad CID")
+			}
+
+			copy(byts, c.Bytes())
+			byts[2] = 0x80 // bad multihash varint
+			byts[3] = 0x00
+			l, c, err = fromBytes(byts)
+			if err == nil {
+				t.Fatal("expected not-enough-bytes for V1 CID")
+			}
+			if !usingReader && l != 0 {
+				t.Fatal("expected length==0 from bad CID")
+			} else if usingReader && l == 0 {
+				t.Fatal("expected length!=0 from bad CID")
+			}
+			if c != Undef {
+				t.Fatal("expected Undef CID from bad CidFromBytes")
+			}
+		})
 	}
-	if l != 0 {
-		t.Fatal("expected length=0 from bad CidFromBytes")
+}
+
+func TestFromReaderNoData(t *testing.T) {
+	// Reading no data from io.Reader should return io.EOF, not ErrInvalidCid.
+	n, cid, err := CidFromReader(bytes.NewReader(nil))
+	if err != io.EOF {
+		t.Fatal("Expected io.EOF error")
 	}
-	if c != Undef {
-		t.Fatal("expected Undef CID from bad CidFromBytes")
+	if cid != Undef {
+		t.Fatal("Expected Undef CID")
+	}
+	if n != 0 {
+		t.Fatal("Expected 0 data")
 	}
 
-	c, err = Decode("bafkreie5qrjvaw64n4tjm6hbnm7fnqvcssfed4whsjqxzslbd3jwhsk3mm")
-	if err != nil {
-		t.Fatal(err)
+	// Read byte indicatiing more data to and check error is ErrInvalidCid.
+	_, _, err = CidFromReader(bytes.NewReader([]byte{0x80}))
+	if !errors.Is(err, ErrInvalidCid{}) {
+		t.Fatal("Expected ErrInvalidCid error")
 	}
-	byts := make([]byte, c.ByteLen())
-	copy(byts, c.Bytes())
-	byts[1] = 0x80 // bad codec varint
-	byts[2] = 0x00
-	l, c, err = CidFromBytes(byts)
-	if err == nil {
-		t.Fatal("expected not-enough-bytes for V1 CidFromBytes")
-	}
-	if l != 0 {
-		t.Fatal("expected length=0 from bad CidFromBytes")
-	}
-	if c != Undef {
-		t.Fatal("expected Undef CID from bad CidFromBytes")
-	}
-
-	copy(byts, c.Bytes())
-	byts[2] = 0x80 // bad multihash varint
-	byts[3] = 0x00
-	l, c, err = CidFromBytes(byts)
-	if err == nil {
-		t.Fatal("expected not-enough-bytes for V1 CidFromBytes")
-	}
-	if l != 0 {
-		t.Fatal("expected length=0 from bad CidFromBytes")
-	}
-	if c != Undef {
-		t.Fatal("expected Undef CID from bad CidFromBytes")
+	// Check for expected wrapped error.
+	if !errors.Is(err, io.ErrUnexpectedEOF) {
+		t.Fatal("Expected error", io.ErrUnexpectedEOF)
 	}
 }
 
@@ -750,6 +816,9 @@ func TestBadParse(t *testing.T) {
 	if err == nil {
 		t.Fatal("expected to fail to parse an invalid CIDv1 CID")
 	}
+	if !errors.Is(err, ErrInvalidCid{}) {
+		t.Fatal("error must be ErrInvalidCid")
+	}
 }
 
 func TestLoggable(t *testing.T) {
@@ -764,3 +833,80 @@ func TestLoggable(t *testing.T) {
 		t.Fatalf("did not get expected loggable form (got %v)", actual)
 	}
 }
+
+func TestErrInvalidCidIs(t *testing.T) {
+	for i, test := range []struct {
+		err    error
+		target error
+	}{
+		{&ErrInvalidCid{}, ErrInvalidCid{}},
+		{ErrInvalidCid{}, &ErrInvalidCid{}},
+		{ErrInvalidCid{}, ErrInvalidCid{}},
+		{&ErrInvalidCid{}, &ErrInvalidCid{}},
+	} {
+		if !errors.Is(test.err, test.target) {
+			t.Fatalf("expected error to be ErrInvalidCid, case %d", i)
+		}
+	}
+}
+
+func TestErrInvalidCid(t *testing.T) {
+	run := func(err error) {
+		if err == nil {
+			t.Fatal("expected error")
+		}
+
+		if !strings.HasPrefix(err.Error(), "invalid cid: ") {
+			t.Fatal(`expected error message to contain "invalid cid: "`)
+		}
+
+		is := errors.Is(err, ErrInvalidCid{})
+		if !is {
+			t.Fatal("expected error to be ErrInvalidCid")
+		}
+
+		if !errors.Is(err, &ErrInvalidCid{}) {
+			t.Fatal("expected error to be &ErrInvalidCid")
+		}
+	}
+
+	_, err := Decode("")
+	run(err)
+
+	_, err = Decode("not-a-cid")
+	run(err)
+
+	_, err = Decode("bafyInvalid")
+	run(err)
+
+	_, err = Decode("QmdfTbBqBPQ7VNxZEYEj14VmRuZBkqFbiwReogJgS1zIII")
+	run(err)
+
+	_, err = Cast([]byte("invalid"))
+	run(err)
+
+	_, err = Parse("not-a-cid")
+	run(err)
+
+	_, err = Parse("bafyInvalid")
+	run(err)
+
+	_, err = Parse("QmdfTbBqBPQ7VNxZEYEj14VmRuZBkqFbiwReogJgS1zIII")
+	run(err)
+
+	_, err = Parse(123)
+	run(err)
+
+	_, _, err = CidFromBytes([]byte("invalid"))
+	run(err)
+
+	_, err = Prefix{}.Sum([]byte("data"))
+	run(err)
+
+	_, err = PrefixFromBytes([]byte{0x80})
+	run(err)
+
+	_, err = ExtractEncoding("invalid ")
+	run(err)
+
+}

go.mod

@@ -2,8 +2,19 @@ module github.com/ipfs/go-cid
 
 require (
 	github.com/multiformats/go-multibase v0.0.3
-	github.com/multiformats/go-multihash v0.0.16-0.20210406191223-75ae3688857d
+	github.com/multiformats/go-multihash v0.0.15
 	github.com/multiformats/go-varint v0.0.6
 )
 
-go 1.15
+require (
+	github.com/klauspost/cpuid/v2 v2.0.4 // indirect
+	github.com/minio/blake2b-simd v0.0.0-20160723061019-3f5f724cb5b1 // indirect
+	github.com/minio/sha256-simd v1.0.0 // indirect
+	github.com/mr-tron/base58 v1.2.0 // indirect
+	github.com/multiformats/go-base32 v0.0.3 // indirect
+	github.com/multiformats/go-base36 v0.1.0 // indirect
+	golang.org/x/crypto v0.1.0 // indirect
+	golang.org/x/sys v0.1.0 // indirect
+)
+
+go 1.19

go.sum

@@ -13,17 +13,19 @@ github.com/multiformats/go-base36 v0.1.0 h1:JR6TyF7JjGd3m6FbLU2cOxhC0Li8z8dLNGQ8
 github.com/multiformats/go-base36 v0.1.0/go.mod h1:kFGE83c6s80PklsHO9sRn2NCoffoRdUUOENyW/Vv6sM=
 github.com/multiformats/go-multibase v0.0.3 h1:l/B6bJDQjvQ5G52jw4QGSYeOTZoAwIO77RblWplfIqk=
 github.com/multiformats/go-multibase v0.0.3/go.mod h1:5+1R4eQrT3PkYZ24C3W2Ue2tPwIdYQD509ZjSb5y9Oc=
-github.com/multiformats/go-multihash v0.0.16-0.20210406191223-75ae3688857d h1:xUl/JcLVtafS0n5X8guAExxDS/mJUGoL31p/FlNtC8o=
-github.com/multiformats/go-multihash v0.0.16-0.20210406191223-75ae3688857d/go.mod h1:D6aZrWNLFTV/ynMpKsNtB40mJzmCl4jb1alC0OvHiHg=
+github.com/multiformats/go-multihash v0.0.15 h1:hWOPdrNqDjwHDx82vsYGSDZNyktOJJ2dzZJzFkOV1jM=
+github.com/multiformats/go-multihash v0.0.15/go.mod h1:D6aZrWNLFTV/ynMpKsNtB40mJzmCl4jb1alC0OvHiHg=
 github.com/multiformats/go-varint v0.0.6 h1:gk85QWKxh3TazbLxED/NlDVv8+q+ReFJk7Y2W/KhfNY=
 github.com/multiformats/go-varint v0.0.6/go.mod h1:3Ls8CIEsrijN6+B7PbrXRPxHRPuXSrVKRY101jdMZYE=
 golang.org/x/crypto v0.0.0-20190308221718-c2843e01d9a2/go.mod h1:djNgcEr1/C05ACkg1iLfiJU5Ep61QUkGW8qpdssI0+w=
-golang.org/x/crypto v0.0.0-20210220033148-5ea612d1eb83 h1:/ZScEX8SfEmUGRHs0gxpqteO5nfNW6axyZbBdw9A12g=
 golang.org/x/crypto v0.0.0-20210220033148-5ea612d1eb83/go.mod h1:jdWPYTVW3xRLrWPugEBEK3UY2ZEsg3UU495nc5E+M+I=
+golang.org/x/crypto v0.1.0 h1:MDRAIl0xIo9Io2xV565hzXHw3zVseKrJKodhohM5CjU=
+golang.org/x/crypto v0.1.0/go.mod h1:RecgLatLF4+eUMCP1PoPZQb+cVrJcOPbHkTkbkB9sbw=
 golang.org/x/net v0.0.0-20190404232315-eb5bcb51f2a3/go.mod h1:t9HGtf8HONx5eT2rtn7q6eTqICYqUVnKs3thJo3Qplg=
 golang.org/x/sys v0.0.0-20190215142949-d0b11bdaac8a/go.mod h1:STP8DvDyc/dI5b8T5hshtkjS+E42TnysNCUPdjciGhY=
 golang.org/x/sys v0.0.0-20191026070338-33540a1f6037/go.mod h1:h1NjWce9XRLGQEsW7wpKNCjG9DtNlClVuFLEZdDNbEs=
-golang.org/x/sys v0.0.0-20210309074719-68d13333faf2 h1:46ULzRKLh1CwgRq2dC5SlBzEqqNCi8rreOZnNrbqcIY=
 golang.org/x/sys v0.0.0-20210309074719-68d13333faf2/go.mod h1:h1NjWce9XRLGQEsW7wpKNCjG9DtNlClVuFLEZdDNbEs=
+golang.org/x/sys v0.1.0 h1:kunALQeHf1/185U1i0GOB/fy1IPRDDpuoOOqRReG57U=
+golang.org/x/sys v0.1.0/go.mod h1:oPkhp1MJrh7nUepCBck5+mAzfO9JrbApNNgaTdGDITg=
 golang.org/x/term v0.0.0-20201117132131-f5c789dd3221/go.mod h1:Nr5EML6q2oocZ2LXRh80K7BxOlk5/8JxuGnuhpl+muw=
 golang.org/x/text v0.3.0/go.mod h1:NqM8EUOU14njkJ3fqMW+pc6Ldnwhi/IjpwHt7yyuwOQ=

version.json

@@ -0,0 +1,3 @@
+{
+  "version": "v0.4.1"
+}