avoid double alloc in NewCidV1

We allocate once via "make([]byte, len)",
and again when that buffer is converted to a string.

Thankfully, since Go 1.10 we have strings.Builder,
designed specifically for this use case.

In a downstream benchmark in go-car,
which needs to reconstruct many CID values,
we see small but nice gains:

    name           old time/op    new time/op    delta
    ReadBlocks-16    1.09ms ± 4%    1.06ms ± 5%   -3.33%  (p=0.007 n=11+11)

    name           old speed      new speed      delta
    ReadBlocks-16   478MB/s ± 4%   494MB/s ± 5%   +3.46%  (p=0.007 n=11+11)

    name           old alloc/op   new alloc/op   delta
    ReadBlocks-16    1.30MB ± 0%    1.25MB ± 0%   -3.86%  (p=0.000 n=12+12)

    name           old allocs/op  new allocs/op  delta
    ReadBlocks-16     9.50k ± 0%     8.45k ± 0%  -11.05%  (p=0.000 n=12+12)

.github/workflows/automerge.yml

@@ -33,7 +33,9 @@ jobs:
   automerge:
     needs: automerge-check
     runs-on: ubuntu-latest
-    if: ${{ needs.automerge-check.outputs.status == 'true' }}
+    # The check for the user is redundant here, as this job depends on the automerge-check job,
+    # but it prevents this job from spinning up, just to be skipped shortly after.
+    if: github.event.pull_request.user.login == 'web3-bot' && needs.automerge-check.outputs.status == 'true'
     steps:
     - name: Wait on tests
       uses: lewagon/wait-on-check-action@bafe56a6863672c681c3cf671f5e10b20abf2eaa # v0.2

.github/workflows/go-check.yml

@@ -8,17 +8,28 @@ jobs:
   unit:
     runs-on: ubuntu-latest
     name: All
+    env:
+      RUNGOGENERATE: false
     steps:
       - uses: actions/checkout@v2
         with:
           submodules: recursive
       - uses: actions/setup-go@v2
         with:
-          go-version: "1.16.x"
+          go-version: "1.17.x"
+      - name: Run repo-specific setup
+        uses: ./.github/actions/go-check-setup
+        if: hashFiles('./.github/actions/go-check-setup') != ''
+      - name: Read config
+        if: hashFiles('./.github/workflows/go-check-config.json') != ''
+        run: |
+          if jq -re .gogenerate ./.github/workflows/go-check-config.json; then
+            echo "RUNGOGENERATE=true" >> $GITHUB_ENV
+          fi
       - name: Install staticcheck
-        run: go install honnef.co/go/tools/cmd/staticcheck@434f5f3816b358fe468fa83dcba62d794e7fe04b # 2021.1 (v0.2.0)
+        run: go install honnef.co/go/tools/cmd/staticcheck@df71e5d0e0ed317ebf43e6e59cf919430fa4b8f2 # 2021.1.1 (v0.2.1)
       - name: Check that go.mod is tidy
-        uses: protocol/multiple-go-modules@v1.0
+        uses: protocol/multiple-go-modules@v1.2
         with:
           run: |
             go mod tidy
@@ -37,14 +48,27 @@ jobs:
           fi
       - name: go vet
         if: ${{ success() || failure() }} # run this step even if the previous one failed
-        uses: protocol/multiple-go-modules@v1.0
+        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@v1.0
+        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()) && env.RUNGOGENERATE == 'true'
+        with:
+          run: |
+            git clean -fd # make sure there aren't untracked files / directories
+            go generate ./...
+            # 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

@@ -10,7 +10,9 @@ jobs:
       fail-fast: false
       matrix:
         os: [ "ubuntu", "windows", "macos" ]
-        go: [ "1.15.x", "1.16.x" ]
+        go: [ "1.16.x", "1.17.x" ]
+    env:
+      COVERAGES: ""
     runs-on: ${{ matrix.os }}-latest
     name: ${{ matrix.os}} (go ${{ matrix.go }})
     steps:
@@ -24,24 +26,30 @@ jobs:
         run: |
           go version
           go env
+      - name: Run repo-specific setup
+        uses: ./.github/actions/go-test-setup
+        if: hashFiles('./.github/actions/go-test-setup') != ''
       - name: Run tests
-        uses: protocol/multiple-go-modules@v1.0
+        uses: protocol/multiple-go-modules@v1.2
         with:
-          run: go test -v -coverprofile coverage.txt ./...
+          run: go test -v -coverprofile module-coverage.txt ./...
       - name: Run tests (32 bit)
         if: ${{ matrix.os != 'macos' }} # can't run 32 bit tests on OSX.
-        uses: protocol/multiple-go-modules@v1.0
+        uses: protocol/multiple-go-modules@v1.2
         env:
           GOARCH: 386
         with:
           run: go test -v ./...
       - name: Run tests with race detector
         if: ${{ matrix.os == 'ubuntu' }} # speed things up. Windows and OSX VMs are slow
-        uses: protocol/multiple-go-modules@v1.0
+        uses: protocol/multiple-go-modules@v1.2
         with:
           run: go test -v -race ./...
+      - name: Collect coverage files
+        shell: bash
+        run: echo "COVERAGES=$(find . -type f -name 'module-coverage.txt' | tr -s '\n' ',' | sed 's/,$//')" >> $GITHUB_ENV
       - name: Upload coverage to Codecov
-        uses: codecov/codecov-action@a1ed4b322b4b38cb846afb5a0ebfa17086917d27 # v1.5.0
+        uses: codecov/codecov-action@51d810878be5422784e86451c0e7c14e5860ec47 # v2.0.2
         with:
-          file: coverage.txt
+          files: '${{ env.COVERAGES }}'
           env_vars: OS=${{ matrix.os }}, GO=${{ matrix.go }}

cid.go

@@ -22,6 +22,7 @@ package cid
 import (
 	"bytes"
 	"encoding"
+	"encoding/binary"
 	"encoding/json"
 	"errors"
 	"fmt"
@@ -173,16 +174,24 @@ 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)
+	// Two 8 bytes (max) numbers plus hash.
-	n := varint.PutUvarint(buf, 1)
+	// We use strings.Builder to only allocate once.
-	n += varint.PutUvarint(buf[n:], codecType)
+	var b strings.Builder
-	cn := copy(buf[n:], mhash)
+	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 (
@@ -680,3 +689,139 @@ func CidFromBytes(data []byte) (int, Cid, error) {
 
 	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.
+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 {
+		return len(br.dst), Undef, 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, err
+		}
+
+		br.dst = br.dst[:34]
+		h, err := mh.Cast(br.dst)
+		if err != nil {
+			return len(br.dst), Undef, err
+		}
+
+		return len(br.dst), Cid{string(h)}, nil
+	}
+
+	if vers != 1 {
+		return len(br.dst), Undef, 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, 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, err
+	}
+
+	// Multihash digest length.
+	mhl, err := varint.ReadUvarint(br)
+	if err != nil {
+		return len(br.dst), Undef, 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, 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, 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, err
+	}
+
+	return len(br.dst), Cid{string(br.dst)}, nil
+}

cid_fuzz.go

@@ -1,3 +1,4 @@
+//go:build gofuzz
 // +build gofuzz
 
 package cid

cid_test.go

@@ -4,10 +4,12 @@ import (
 	"bytes"
 	"encoding/json"
 	"fmt"
+	"io"
 	"math/rand"
 	"reflect"
 	"strings"
 	"testing"
+	"testing/iotest"
 
 	mbase "github.com/multiformats/go-multibase"
 	mh "github.com/multiformats/go-multihash"
@@ -665,6 +667,7 @@ func TestReadCidsFromBuffer(t *testing.T) {
 		"k2cwueckqkibutvhkr4p2ln2pjcaxaakpd9db0e7j7ax1lxhhxy3ekpv",
 		"Qmf5Qzp6nGBku7CEn2UQx4mgN8TW69YUok36DrGa6NN893",
 		"zb2rhZi1JR4eNc2jBGaRYJKYM8JEB4ovenym8L1CmFsRAytkz",
+		"bafkqarjpmzuwyzltorxxezjpkvcfgqkfjfbfcvslivje2vchkzdu6rckjjcfgtkolaze6mssjqzeyn2ekrcfatkjku2vowseky3fswkfkm2deqkrju3e2",
 	}
 
 	var cids []Cid
@@ -692,51 +695,98 @@ 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) {
+func TestBadCidInput(t *testing.T) {
-	l, c, err := CidFromBytes([]byte{mh.SHA2_256, 32, 0x00})
+	for _, name := range []string{
-	if err == nil {
+		"FromBytes",
-		t.Fatal("expected not-enough-bytes for V0 CidFromBytes")
+		"FromReader",
-	}
+	} {
-	if l != 0 {
+		t.Run(name, func(t *testing.T) {
-		t.Fatal("expected length=0 from bad CidFromBytes")
+			usingReader := name == "FromReader"
-	}
-	if c != Undef {
-		t.Fatal("expected Undef CID from bad CidFromBytes")
-	}
 
-	c, err = Decode("bafkreie5qrjvaw64n4tjm6hbnm7fnqvcssfed4whsjqxzslbd3jwhsk3mm")
+			fromBytes := CidFromBytes
-	if err != nil {
+			if usingReader {
-		t.Fatal(err)
+				fromBytes = func(data []byte) (int, Cid, error) {
-	}
+					return CidFromReader(bytes.NewReader(data))
-	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())
+			l, c, err := fromBytes([]byte{mh.SHA2_256, 32, 0x00})
-	byts[2] = 0x80 // bad multihash varint
+			if err == nil {
-	byts[3] = 0x00
+				t.Fatal("expected not-enough-bytes for V0 CID")
-	l, c, err = CidFromBytes(byts)
+			}
-	if err == nil {
+			if !usingReader && l != 0 {
-		t.Fatal("expected not-enough-bytes for V1 CidFromBytes")
+				t.Fatal("expected length==0 from bad CID")
-	}
+			} else if usingReader && l == 0 {
-	if l != 0 {
+				t.Fatal("expected length!=0 from bad CID")
-		t.Fatal("expected length=0 from bad CidFromBytes")
+			}
-	}
+			if c != Undef {
-	if c != Undef {
+				t.Fatal("expected Undef CID from bad CID")
-		t.Fatal("expected Undef CID from bad CidFromBytes")
+			}
+
+			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")
+			}
+		})
 	}
 }
 

go.mod

@@ -7,4 +7,4 @@ require (
 	golang.org/x/crypto v0.0.0-20210506145944-38f3c27a63bf // indirect
 )
 
-go 1.15
+go 1.16