Add custom split logic for scanner

srt.go

@@ -1,7 +1,6 @@
 package astisub
 
 import (
-	"bufio"
 	"fmt"
 	"io"
 	"strconv"
@@ -36,7 +35,7 @@ func parseDurationSRT(i string) (d time.Duration, err error) {
 func ReadFromSRT(i io.Reader) (o *Subtitles, err error) {
 	// Init
 	o = NewSubtitles()
-	var scanner = bufio.NewScanner(i)
+	var scanner = NewScanner(i)
 
 	// Scan
 	var line string

ssa.go

@@ -1,7 +1,6 @@
 package astisub
 
 import (
-	"bufio"
 	"fmt"
 	"io"
 	"log"
@@ -134,7 +133,7 @@ func ReadFromSSA(i io.Reader) (o *Subtitles, err error) {
 func ReadFromSSAWithOptions(i io.Reader, opts SSAOptions) (o *Subtitles, err error) {
 	// Init
 	o = NewSubtitles()
-	var scanner = bufio.NewScanner(i)
+	var scanner = NewScanner(i)
 	var si = &ssaScriptInfo{}
 	var ss = []*ssaStyle{}
 	var es = []*ssaEvent{}

subtitles.go

@@ -1,8 +1,11 @@
 package astisub
 
 import (
+	"bufio"
+	"bytes"
 	"errors"
 	"fmt"
+	"io"
 	"math"
 	"os"
 	"path/filepath"
@@ -927,3 +930,32 @@ func escapeHTML(i string) string {
 func unescapeHTML(i string) string {
 	return htmlUnescaper.Replace(i)
 }
+
+func NewScanner(i io.Reader) *bufio.Scanner {
+	var scanner = bufio.NewScanner(i)
+	scanner.Split(splitLines)
+	return scanner
+}
+
+func splitLines(data []byte, atEOF bool) (advance int, token []byte, err error) {
+	if atEOF && len(data) == 0 {
+		return 0, nil, nil
+	}
+	if i := bytes.IndexAny(data, "\r\n"); i >= 0 {
+		if data[i] == '\n' {
+			// We have a line terminated by single newline.
+			return i + 1, data[0:i], nil
+		}
+		advance = i + 1
+		if len(data) > i+1 && data[i+1] == '\n' {
+			advance += 1
+		}
+		return advance, data[0:i], nil
+	}
+	// If we're at EOF, we have a final, non-terminated line. Return it.
+	if atEOF {
+		return len(data), data, nil
+	}
+	// Request more data.
+	return 0, nil, nil
+}

subtitles_test.go

@@ -340,3 +340,26 @@ func TestHTMLEntity(t *testing.T) {
 		assert.Equal(t, string(c), w.String())
 	}
 }
+
+func TestNewScanner(t *testing.T) {
+	exts := []string{"vtt", "srt", "ssa"}
+	for _, ext := range exts {
+		s, err := astisub.OpenFile("./testdata/example-in-scan-line." + ext)
+		assert.NoError(t, err)
+		assert.Len(t, s.Items, 3)
+		assert.Equal(t, time.Duration(0), s.Items[0].StartAt)
+		assert.Equal(t, 3*time.Second+766*time.Millisecond, s.Items[0].EndAt)
+		assert.Equal(t, "Did one of the last stories strike you as", s.Items[0].Lines[0].String())
+		assert.Equal(t, "more interesting than the other?", s.Items[0].Lines[1].String())
+
+		assert.Equal(t, 3*time.Second+767*time.Millisecond, s.Items[1].StartAt)
+		assert.Equal(t, 10*time.Second+732*time.Millisecond, s.Items[1].EndAt)
+		assert.Equal(t, "That's true. You don’t often find 632", s.Items[1].Lines[0].String())
+		assert.Equal(t, "pieces of gum stuck on a sidewalk", s.Items[1].Lines[1].String())
+
+		assert.Equal(t, 10*time.Second+733*time.Millisecond, s.Items[2].StartAt)
+		assert.Equal(t, 14*time.Second+66*time.Millisecond, s.Items[2].EndAt)
+		assert.Equal(t, "at a busy bus stop or anywhere", s.Items[2].Lines[0].String())
+		assert.Equal(t, "else for that matter.", s.Items[2].Lines[1].String())
+	}
+}

testdata/example-in-scan-line.srt

@@ -0,0 +1 @@
+100:00:00.000 --> 00:00:03.766Did one of the last stories strike you asmore interesting than the other?200:00:03.767 --> 00:00:10.732That's true. You don’t often find 632pieces of gum stuck on a sidewalk300:00:10.733 --> 00:00:14.066at a busy bus stop or anywhereelse for that matter.

testdata/example-in-scan-line.ssa

@@ -0,0 +1 @@
+[Script Info]; Comment 1; Comment 2Collisions: NormalOriginal Script: asticodePlayDepth: 0PlayResY: 600ScriptType: v4.00Script Updated By: version 2.8.01Timer: 100Title: SSA test[V4 Styles]Format: Name, Alignment, AlphaLevel, BackColour, Bold, BorderStyle, Encoding, Fontname, Fontsize, Italic, MarginL, MarginR, MarginV, Outline, OutlineColour, PrimaryColour, SecondaryColour, ShadowStyle: 1,7,0.100,&H80000008,1,7,0,f1,4.000,0,1,4,7,1.000,&H0000ffff,&H0000ffff,&H0000ffff,4.000Style: 2,8,0.200,&H000f0f0f,1,8,1,f2,5.000,0,2,5,8,2.000,&H0000ffff,&H00efefef,&H0000ffff,5.000Style: 3,9,0.300,&H00000008,0,9,2,f3,6.000,0,3,6,9,3.000,&H00000008,&H00b4fcfc,&H00b4fcfc,6.000[Events]Format: Marked, Start, End, Style, Name, MarginL, MarginR, MarginV, Effect, TextDialogue: Marked=0,00:00:00.00,00:00:03.766,1,Cher,1234,2345,3456,test,{\pos(400,570)}Did one of the last stories strike you as\nmore interesting than the other?Dialogue: Marked=1,00:00:03.767,00:00:10.732,2,autre,0,0,0,,That's true. You don’t often find 632\npieces of gum stuck on a sidewalkDialogue: Marked=1,00:00:10.733,00:00:14.066,3,autre,0,0,0,,at a busy bus stop or anywhere\nelse for that matter.

testdata/example-in-scan-line.vtt

@@ -0,0 +1 @@
+WEBVTT100:00:00.000 --> 00:00:03.766Did one of the last stories strike you asmore interesting than the other?200:00:03.767 --> 00:00:10.732That's true. You don’t often find 632pieces of gum stuck on a sidewalk300:00:10.733 --> 00:00:14.066at a busy bus stop or anywhereelse for that matter.

webvtt.go

@@ -1,7 +1,6 @@
 package astisub
 
 import (
-	"bufio"
 	"errors"
 	"fmt"
 	"io"
@@ -118,7 +117,8 @@ func parseWebVTTTimestampMap(line string) (timestampMap *WebVTTTimestampMap, err
 func ReadFromWebVTT(i io.Reader) (o *Subtitles, err error) {
 	// Init
 	o = NewSubtitles()
-	var scanner = bufio.NewScanner(i)
+	var scanner = NewScanner(i)
+
 	var line string
 	var lineNum int