Read expected buffer size (#17409)

* Read expected buffer size.

* Changed name.

6 files changed, 33 insertions, 11 deletions

diff --git a/modules/charset/charset.go b/modules/charset/charset.go
index 47906e2638..ae5cf5aa1a 100644
--- a/modules/charset/charset.go
+++ b/modules/charset/charset.go
@@ -13,6 +13,7 @@ import (
 
 	"code.gitea.io/gitea/modules/log"
 	"code.gitea.io/gitea/modules/setting"
+	"code.gitea.io/gitea/modules/util"
 
 	"github.com/gogs/chardet"
 	"golang.org/x/net/html/charset"
@@ -25,9 +26,9 @@ var UTF8BOM = []byte{'\xef', '\xbb', '\xbf'}
 // ToUTF8WithFallbackReader detects the encoding of content and coverts to UTF-8 reader if possible
 func ToUTF8WithFallbackReader(rd io.Reader) io.Reader {
 	var buf = make([]byte, 2048)
-	n, err := rd.Read(buf)
+	n, err := util.ReadAtMost(rd, buf)
 	if err != nil {
-		return rd
+		return io.MultiReader(bytes.NewReader(RemoveBOMIfPresent(buf[:n])), rd)
 	}
 
 	charsetLabel, err := DetectEncoding(buf[:n])
diff --git a/modules/csv/csv.go b/modules/csv/csv.go
index 30698830a4..2e2611aa81 100644
--- a/modules/csv/csv.go
+++ b/modules/csv/csv.go
@@ -29,11 +29,8 @@ func CreateReader(input io.Reader, delimiter rune) *stdcsv.Reader {
 // CreateReaderAndGuessDelimiter tries to guess the field delimiter from the content and creates a csv.Reader.
 func CreateReaderAndGuessDelimiter(rd io.Reader) (*stdcsv.Reader, error) {
 	var data = make([]byte, 1e4)
-	size, err := rd.Read(data)
+	size, err := util.ReadAtMost(rd, data)
 	if err != nil {
-		if err == io.EOF {
-			return CreateReader(bytes.NewReader([]byte{}), rune(',')), nil
-		}
 		return nil, err
 	}
 
diff --git a/modules/git/blob.go b/modules/git/blob.go
index c7919f3c41..9567affd03 100644
--- a/modules/git/blob.go
+++ b/modules/git/blob.go
@@ -11,6 +11,7 @@ import (
 	"io"
 
 	"code.gitea.io/gitea/modules/typesniffer"
+	"code.gitea.io/gitea/modules/util"
 )
 
 // This file contains common functions between the gogit and !gogit variants for git Blobs
@@ -28,7 +29,7 @@ func (b *Blob) GetBlobContent() (string, error) {
 	}
 	defer dataRc.Close()
 	buf := make([]byte, 1024)
-	n, _ := dataRc.Read(buf)
+	n, _ := util.ReadAtMost(dataRc, buf)
 	buf = buf[:n]
 	return string(buf), nil
 }
diff --git a/modules/repofiles/update.go b/modules/repofiles/update.go
index dc2893cb1c..d25accff18 100644
--- a/modules/repofiles/update.go
+++ b/modules/repofiles/update.go
@@ -19,6 +19,7 @@ import (
 	repo_module "code.gitea.io/gitea/modules/repository"
 	"code.gitea.io/gitea/modules/setting"
 	"code.gitea.io/gitea/modules/structs"
+	"code.gitea.io/gitea/modules/util"
 
 	stdcharset "golang.org/x/net/html/charset"
 	"golang.org/x/text/transform"
@@ -61,7 +62,7 @@ func detectEncodingAndBOM(entry *git.TreeEntry, repo *models.Repository) (string
 	}
 	defer reader.Close()
 	buf := make([]byte, 1024)
-	n, err := reader.Read(buf)
+	n, err := util.ReadAtMost(reader, buf)
 	if err != nil {
 		// return default
 		return "UTF-8", false
@@ -84,7 +85,7 @@ func detectEncodingAndBOM(entry *git.TreeEntry, repo *models.Repository) (string
 				}
 				defer dataRc.Close()
 				buf = make([]byte, 1024)
-				n, err = dataRc.Read(buf)
+				n, err = util.ReadAtMost(dataRc, buf)
 				if err != nil {
 					// return default
 					return "UTF-8", false
diff --git a/modules/typesniffer/typesniffer.go b/modules/typesniffer/typesniffer.go
index d257b8179b..9e29b3557c 100644
--- a/modules/typesniffer/typesniffer.go
+++ b/modules/typesniffer/typesniffer.go
@@ -10,6 +10,8 @@ import (
 	"net/http"
 	"regexp"
 	"strings"
+
+	"code.gitea.io/gitea/modules/util"
 )
 
 // Use at most this many bytes to determine Content Type.
@@ -86,8 +88,8 @@ func DetectContentType(data []byte) SniffedType {
 // DetectContentTypeFromReader guesses the content type contained in the reader.
 func DetectContentTypeFromReader(r io.Reader) (SniffedType, error) {
 	buf := make([]byte, sniffLen)
-	n, err := r.Read(buf)
-	if err != nil && err != io.EOF {
+	n, err := util.ReadAtMost(r, buf)
+	if err != nil {
 		return SniffedType{}, fmt.Errorf("DetectContentTypeFromReader io error: %w", err)
 	}
 	buf = buf[:n]
diff --git a/modules/util/io.go b/modules/util/io.go
new file mode 100644
index 0000000000..b467c0ac8a
--- /dev/null
+++ b/modules/util/io.go
@@ -0,0 +1,20 @@
+// Copyright 2021 The Gitea Authors. All rights reserved.
+// Use of this source code is governed by a MIT-style
+// license that can be found in the LICENSE file.
+
+package util
+
+import (
+	"io"
+)
+
+// ReadAtMost reads at most len(buf) bytes from r into buf.
+// It returns the number of bytes copied. n is only less then len(buf) if r provides fewer bytes.
+// If EOF occurs while reading, err will be nil.
+func ReadAtMost(r io.Reader, buf []byte) (n int, err error) {
+	n, err = io.ReadFull(r, buf)
+	if err == io.EOF || err == io.ErrUnexpectedEOF {
+		err = nil
+	}
+	return
+}