From 66d0ce841c9d89bbca3663e4c51b59120838d074 Mon Sep 17 00:00:00 2001
From: Stefan Haller <stefan@haller-berlin.de>
Date: Sat, 27 Apr 2024 17:16:06 +0200
Subject: [PATCH] Implement ScanLinesAndTruncateWhenLongerThanBuffer

---
 pkg/utils/lines.go      | 59 ++++++++++++++++++++++++++++++++++++-
 pkg/utils/lines_test.go | 64 +++++++++++++++++++++++++++++++++++++++++
 2 files changed, 122 insertions(+), 1 deletion(-)

diff --git a/pkg/utils/lines.go b/pkg/utils/lines.go
index 662ba2f9b..c70d02ffc 100644
--- a/pkg/utils/lines.go
+++ b/pkg/utils/lines.go
@@ -1,6 +1,9 @@
 package utils
 
-import "strings"
+import (
+	"bytes"
+	"strings"
+)
 
 // SplitLines takes a multiline string and splits it on newlines
 // currently we are also stripping \r's which may have adverse effects for
@@ -43,3 +46,57 @@ func EscapeSpecialChars(str string) string {
 		"\v", "\\v",
 	).Replace(str)
 }
+
+func dropCR(data []byte) []byte {
+	if len(data) > 0 && data[len(data)-1] == '\r' {
+		return data[0 : len(data)-1]
+	}
+	return data
+}
+
+// ScanLinesAndTruncateWhenLongerThanBuffer returns a split function that can be
+// used with bufio.Scanner.Split(). It is very similar to bufio.ScanLines,
+// except that it will truncate lines that are longer than the scanner's read
+// buffer (whereas bufio.ScanLines will return an error in that case, which is
+// often difficult to handle).
+//
+// If you are using your own buffer for the scanner, you must set maxBufferSize
+// to the same value as the max parameter that you passed to scanner.Buffer().
+// Otherwise, maxBufferSize must be set to bufio.MaxScanTokenSize.
+func ScanLinesAndTruncateWhenLongerThanBuffer(maxBufferSize int) func(data []byte, atEOF bool) (int, []byte, error) {
+	skipOverRemainderOfLongLine := false
+
+	return func(data []byte, atEOF bool) (int, []byte, error) {
+		if atEOF && len(data) == 0 {
+			// Done
+			return 0, nil, nil
+		}
+		if i := bytes.IndexByte(data, '\n'); i >= 0 {
+			if skipOverRemainderOfLongLine {
+				skipOverRemainderOfLongLine = false
+				return i + 1, nil, nil
+			}
+			return i + 1, dropCR(data[0:i]), nil
+		}
+		if atEOF {
+			if skipOverRemainderOfLongLine {
+				return len(data), nil, nil
+			}
+
+			return len(data), dropCR(data), nil
+		}
+
+		// Buffer is full, so we can't get more data
+		if len(data) >= maxBufferSize {
+			if skipOverRemainderOfLongLine {
+				return len(data), nil, nil
+			}
+
+			skipOverRemainderOfLongLine = true
+			return len(data), data, nil
+		}
+
+		// Request more data.
+		return 0, nil, nil
+	}
+}
diff --git a/pkg/utils/lines_test.go b/pkg/utils/lines_test.go
index e7171022b..2192a3780 100644
--- a/pkg/utils/lines_test.go
+++ b/pkg/utils/lines_test.go
@@ -1,6 +1,8 @@
 package utils
 
 import (
+	"bufio"
+	"strings"
 	"testing"
 
 	"github.com/stretchr/testify/assert"
@@ -100,3 +102,65 @@ func TestNormalizeLinefeeds(t *testing.T) {
 		assert.EqualValues(t, string(s.expected), NormalizeLinefeeds(string(s.byteArray)))
 	}
 }
+
+func TestScanLinesAndTruncateWhenLongerThanBuffer(t *testing.T) {
+	type scenario struct {
+		input         string
+		expectedLines []string
+	}
+
+	scenarios := []scenario{
+		{
+			"",
+			[]string{},
+		},
+		{
+			"\n",
+			[]string{""},
+		},
+		{
+			"abc",
+			[]string{"abc"},
+		},
+		{
+			"abc\ndef",
+			[]string{"abc", "def"},
+		},
+		{
+			"abc\n\ndef",
+			[]string{"abc", "", "def"},
+		},
+		{
+			"abc\r\ndef\r",
+			[]string{"abc", "def"},
+		},
+		{
+			"abcdef",
+			[]string{"abcde"},
+		},
+		{
+			"abcdef\n",
+			[]string{"abcde"},
+		},
+		{
+			"abcdef\nghijkl\nx",
+			[]string{"abcde", "ghijk", "x"},
+		},
+		{
+			"abc\ndefghijklmnopqrstuvw\nx",
+			[]string{"abc", "defgh", "x"},
+		},
+	}
+
+	for _, s := range scenarios {
+		scanner := bufio.NewScanner(strings.NewReader(s.input))
+		scanner.Buffer(make([]byte, 5), 5)
+		scanner.Split(ScanLinesAndTruncateWhenLongerThanBuffer(5))
+		result := []string{}
+		for scanner.Scan() {
+			result = append(result, scanner.Text())
+		}
+		assert.NoError(t, scanner.Err())
+		assert.EqualValues(t, s.expectedLines, result)
+	}
+}