Files
gio/text/shaper_test.go
T
Chris Waldon d71f170c29 text: truncate multi-paragraph text correctly
This commit fixes a subtle problem when trunating text widgets that contain
multiple newline-delimited paragraphs.

Paragraphs are the unit of text shaping, so we divide the text into paragraphs
and then iterate those paragraphs performing shaping and line wrapping. If we
have a maximum number of lines to fill, we stop iterating paragraphs when we
use all of the available lines. Usually, if we fill all of the lines the text
shaper will insert the truncator symbol. However, if we exactly fill all of the
lines with the end of a paragraph, the line wrapper is able to fill the line
quota without actually truncating any of the text in that paragraph. Thus it
doesn't insert a truncator even though subsequent paragraphs were truncated (it
has no way to know).

To fix this, I've taught the line wrapper about an explicit scenario in which
we always want to show the truncator symbol *if* we hit the line limit, even if
all of the text in the current paragraph fit. I've then plumbed support for
that through our text stack.

Signed-off-by: Chris Waldon <christopher.waldon.dev@gmail.com>
2023-03-30 12:04:08 -06:00

434 lines
14 KiB
Go

package text
import (
"fmt"
"strings"
"testing"
nsareg "eliasnaur.com/font/noto/sans/arabic/regular"
"gioui.org/font/opentype"
"gioui.org/io/system"
"golang.org/x/exp/slices"
"golang.org/x/image/font/gofont/goregular"
"golang.org/x/image/math/fixed"
)
// TestWrappingTruncation checks that the line wrapper's truncation features
// behave as expected.
func TestWrappingTruncation(t *testing.T) {
// Use a test string containing multiple newlines to ensure that they are shaped
// as separate paragraphs.
textInput := "Lorem ipsum dolor sit amet, consectetur adipiscing elit,\nsed do eiusmod tempor incididunt ut labore et\ndolore magna aliqua.\n"
ltrFace, _ := opentype.Parse(goregular.TTF)
collection := []FontFace{{Face: ltrFace}}
cache := NewShaper(collection)
cache.LayoutString(Parameters{
Alignment: Middle,
PxPerEm: fixed.I(10),
MinWidth: 200,
MaxWidth: 200,
Locale: english,
}, textInput)
untruncatedCount := len(cache.txt.lines)
for i := untruncatedCount + 1; i > 0; i-- {
t.Run(fmt.Sprintf("truncated to %d/%d lines", i, untruncatedCount), func(t *testing.T) {
cache.LayoutString(Parameters{
Alignment: Middle,
PxPerEm: fixed.I(10),
MaxLines: i,
MinWidth: 200,
MaxWidth: 200,
Locale: english,
}, textInput)
lineCount := 0
lastGlyphWasLineBreak := false
glyphs := []Glyph{}
untruncatedRunes := 0
truncatedRunes := 0
for g, ok := cache.NextGlyph(); ok; g, ok = cache.NextGlyph() {
glyphs = append(glyphs, g)
if g.Flags&FlagTruncator != 0 && g.Flags&FlagClusterBreak != 0 {
truncatedRunes += g.Runes
} else {
untruncatedRunes += g.Runes
}
if g.Flags&FlagLineBreak != 0 {
lineCount++
lastGlyphWasLineBreak = true
} else {
lastGlyphWasLineBreak = false
}
}
if lastGlyphWasLineBreak && truncatedRunes == 0 {
// There was no actual line of text following this break.
lineCount--
}
if i <= untruncatedCount {
if lineCount != i {
t.Errorf("expected %d lines, got %d", i, lineCount)
}
} else if i > untruncatedCount {
if lineCount != untruncatedCount {
t.Errorf("expected %d lines, got %d", untruncatedCount, lineCount)
}
}
if expected := len([]rune(textInput)); truncatedRunes+untruncatedRunes != expected {
t.Errorf("expected %d total runes, got %d (%d truncated)", expected, truncatedRunes+untruncatedRunes, truncatedRunes)
}
})
}
}
// TestWrappingForcedTruncation checks that the line wrapper's truncation features
// activate correctly on multi-paragraph text when later paragraphs are truncated.
func TestWrappingForcedTruncation(t *testing.T) {
// Use a test string containing multiple newlines to ensure that they are shaped
// as separate paragraphs.
textInput := "Lorem ipsum\ndolor sit\namet"
ltrFace, _ := opentype.Parse(goregular.TTF)
collection := []FontFace{{Face: ltrFace}}
cache := NewShaper(collection)
cache.LayoutString(Parameters{
Alignment: Middle,
PxPerEm: fixed.I(10),
MinWidth: 200,
MaxWidth: 200,
Locale: english,
}, textInput)
untruncatedCount := len(cache.txt.lines)
for i := untruncatedCount + 1; i > 0; i-- {
t.Run(fmt.Sprintf("truncated to %d/%d lines", i, untruncatedCount), func(t *testing.T) {
cache.LayoutString(Parameters{
Alignment: Middle,
PxPerEm: fixed.I(10),
MaxLines: i,
MinWidth: 200,
MaxWidth: 200,
Locale: english,
}, textInput)
lineCount := 0
glyphs := []Glyph{}
untruncatedRunes := 0
truncatedRunes := 0
for g, ok := cache.NextGlyph(); ok; g, ok = cache.NextGlyph() {
glyphs = append(glyphs, g)
if g.Flags&FlagTruncator != 0 && g.Flags&FlagClusterBreak != 0 {
truncatedRunes += g.Runes
} else {
untruncatedRunes += g.Runes
}
if g.Flags&FlagLineBreak != 0 {
lineCount++
}
}
expectedTruncated := false
expectedLines := 0
if i < untruncatedCount {
expectedLines = i
expectedTruncated = true
} else if i == untruncatedCount {
expectedLines = i
expectedTruncated = false
} else if i > untruncatedCount {
expectedLines = untruncatedCount
expectedTruncated = false
}
if lineCount != expectedLines {
t.Errorf("expected %d lines, got %d", expectedLines, lineCount)
}
if truncatedRunes > 0 != expectedTruncated {
t.Errorf("expected expectedTruncated=%v, truncatedRunes=%d", expectedTruncated, truncatedRunes)
}
if expected := len([]rune(textInput)); truncatedRunes+untruncatedRunes != expected {
t.Errorf("expected %d total runes, got %d (%d truncated)", expected, truncatedRunes+untruncatedRunes, truncatedRunes)
}
})
}
}
// TestShapingNewlineHandling checks that the shaper's newline splitting behaves
// consistently and does not create spurious lines of text.
func TestShapingNewlineHandling(t *testing.T) {
type testcase struct {
textInput string
expectedLines int
expectedGlyphs int
}
for _, tc := range []testcase{
{textInput: "a\n", expectedLines: 1, expectedGlyphs: 3},
{textInput: "a\nb", expectedLines: 2, expectedGlyphs: 3},
{textInput: "", expectedLines: 1, expectedGlyphs: 1},
} {
t.Run(fmt.Sprintf("%q", tc.textInput), func(t *testing.T) {
ltrFace, _ := opentype.Parse(goregular.TTF)
collection := []FontFace{{Face: ltrFace}}
cache := NewShaper(collection)
checkGlyphs := func() {
glyphs := []Glyph{}
for g, ok := cache.NextGlyph(); ok; g, ok = cache.NextGlyph() {
glyphs = append(glyphs, g)
}
if len(glyphs) != tc.expectedGlyphs {
t.Errorf("expected %d glyphs, got %d", tc.expectedGlyphs, len(glyphs))
}
findBreak := func(g Glyph) bool {
return g.Flags&FlagParagraphBreak != 0
}
found := 0
for idx := slices.IndexFunc(glyphs, findBreak); idx != -1; idx = slices.IndexFunc(glyphs, findBreak) {
found++
breakGlyph := glyphs[idx]
startGlyph := glyphs[idx+1]
glyphs = glyphs[idx+1:]
if flags := breakGlyph.Flags; flags&FlagParagraphBreak == 0 {
t.Errorf("expected newline glyph to have P flag, got %s", flags)
}
if flags := startGlyph.Flags; flags&FlagParagraphStart == 0 {
t.Errorf("expected newline glyph to have S flag, got %s", flags)
}
breakX, breakY := breakGlyph.X, breakGlyph.Y
startX, startY := startGlyph.X, startGlyph.Y
if breakX == startX {
t.Errorf("expected paragraph start glyph to have cursor x")
}
if breakY == startY {
t.Errorf("expected paragraph start glyph to have cursor y")
}
}
if count := strings.Count(tc.textInput, "\n"); found != count {
t.Errorf("expected %d paragraph breaks, found %d", count, found)
}
}
cache.LayoutString(Parameters{
Alignment: Middle,
PxPerEm: fixed.I(10),
MinWidth: 200,
MaxWidth: 200,
Locale: english,
}, tc.textInput)
if lineCount := len(cache.txt.lines); lineCount > tc.expectedLines {
t.Errorf("shaping string %q created %d lines", tc.textInput, lineCount)
}
checkGlyphs()
cache.Layout(Parameters{
Alignment: Middle,
PxPerEm: fixed.I(10),
MinWidth: 200,
MaxWidth: 200,
Locale: english,
}, strings.NewReader(tc.textInput))
if lineCount := len(cache.txt.lines); lineCount > tc.expectedLines {
t.Errorf("shaping reader %q created %d lines", tc.textInput, lineCount)
}
checkGlyphs()
})
}
}
// TestCacheEmptyString ensures that shaping the empty string returns a
// single synthetic glyph with ascent/descent info.
func TestCacheEmptyString(t *testing.T) {
ltrFace, _ := opentype.Parse(goregular.TTF)
collection := []FontFace{{Face: ltrFace}}
cache := NewShaper(collection)
cache.LayoutString(Parameters{
Alignment: Middle,
PxPerEm: fixed.I(10),
MinWidth: 200,
MaxWidth: 200,
Locale: english,
}, "")
glyphs := make([]Glyph, 0, 1)
for g, ok := cache.NextGlyph(); ok; g, ok = cache.NextGlyph() {
glyphs = append(glyphs, g)
}
if len(glyphs) != 1 {
t.Errorf("expected %d glyphs, got %d", 1, len(glyphs))
}
glyph := glyphs[0]
checkFlag(t, true, FlagClusterBreak, glyph, 0)
checkFlag(t, true, FlagRunBreak, glyph, 0)
checkFlag(t, true, FlagLineBreak, glyph, 0)
checkFlag(t, false, FlagParagraphBreak, glyph, 0)
if glyph.Ascent == 0 {
t.Errorf("expected non-zero ascent")
}
if glyph.Descent == 0 {
t.Errorf("expected non-zero descent")
}
if glyph.Y == 0 {
t.Errorf("expected non-zero y offset")
}
if glyph.X == 0 {
t.Errorf("expected non-zero x offset")
}
}
// TestCacheAlignment ensures that shaping with different alignments or dominant
// text directions results in different X offsets.
func TestCacheAlignment(t *testing.T) {
ltrFace, _ := opentype.Parse(goregular.TTF)
collection := []FontFace{{Face: ltrFace}}
cache := NewShaper(collection)
params := Parameters{
Alignment: Start,
PxPerEm: fixed.I(10),
MinWidth: 200,
MaxWidth: 200,
Locale: english,
}
cache.LayoutString(params, "A")
glyph, _ := cache.NextGlyph()
startX := glyph.X
params.Alignment = Middle
cache.LayoutString(params, "A")
glyph, _ = cache.NextGlyph()
middleX := glyph.X
params.Alignment = End
cache.LayoutString(params, "A")
glyph, _ = cache.NextGlyph()
endX := glyph.X
if startX == middleX || startX == endX || endX == middleX {
t.Errorf("[LTR] shaping with with different alignments should not produce the same X, start %d, middle %d, end %d", startX, middleX, endX)
}
params.Locale = arabic
params.Alignment = Start
cache.LayoutString(params, "A")
glyph, _ = cache.NextGlyph()
rtlStartX := glyph.X
params.Alignment = Middle
cache.LayoutString(params, "A")
glyph, _ = cache.NextGlyph()
rtlMiddleX := glyph.X
params.Alignment = End
cache.LayoutString(params, "A")
glyph, _ = cache.NextGlyph()
rtlEndX := glyph.X
if rtlStartX == rtlMiddleX || rtlStartX == rtlEndX || rtlEndX == rtlMiddleX {
t.Errorf("[RTL] shaping with with different alignments should not produce the same X, start %d, middle %d, end %d", rtlStartX, rtlMiddleX, rtlEndX)
}
if startX == rtlStartX || endX == rtlEndX {
t.Errorf("shaping with with different dominant text directions and the same alignment should not produce the same X unless it's middle-aligned")
}
}
func TestCacheGlyphConverstion(t *testing.T) {
ltrFace, _ := opentype.Parse(goregular.TTF)
rtlFace, _ := opentype.Parse(nsareg.TTF)
collection := []FontFace{{Face: ltrFace}, {Face: rtlFace}}
type testcase struct {
name string
text string
locale system.Locale
expected []Glyph
}
for _, tc := range []testcase{
{
name: "bidi ltr",
text: "The quick سماء שלום لا fox تمط שלום\nغير the\nlazy dog.",
locale: english,
},
{
name: "bidi rtl",
text: "الحب سماء brown привет fox تمط jumps\nпривет over\nغير الأحلام.",
locale: arabic,
},
} {
t.Run(tc.name, func(t *testing.T) {
cache := NewShaper(collection)
cache.LayoutString(Parameters{
PxPerEm: fixed.I(10),
MaxWidth: 200,
Locale: tc.locale,
}, tc.text)
doc := cache.txt
glyphs := make([]Glyph, 0, len(tc.expected))
for g, ok := cache.NextGlyph(); ok; g, ok = cache.NextGlyph() {
glyphs = append(glyphs, g)
}
glyphCursor := 0
for _, line := range doc.lines {
for runIdx, run := range line.runs {
lastRun := runIdx == len(line.runs)-1
start := 0
end := len(run.Glyphs) - 1
inc := 1
towardOrigin := false
if run.Direction.Progression() == system.TowardOrigin {
start = len(run.Glyphs) - 1
end = 0
inc = -1
towardOrigin = true
}
for glyphIdx := start; ; glyphIdx += inc {
endOfRun := glyphIdx == end
glyph := run.Glyphs[glyphIdx]
endOfCluster := glyphIdx == end || run.Glyphs[glyphIdx+inc].clusterIndex != glyph.clusterIndex
actual := glyphs[glyphCursor]
if actual.ID != glyph.id {
t.Errorf("glyphs[%d] expected id %d, got id %d", glyphCursor, glyph.id, actual.ID)
}
// Synthetic glyphs should only ever show up at the end of lines.
endOfLine := lastRun && endOfRun
synthetic := glyph.glyphCount == 0 && endOfLine
checkFlag(t, endOfLine, FlagLineBreak, actual, glyphCursor)
checkFlag(t, endOfRun, FlagRunBreak, actual, glyphCursor)
checkFlag(t, towardOrigin, FlagTowardOrigin, actual, glyphCursor)
checkFlag(t, synthetic, FlagParagraphBreak, actual, glyphCursor)
checkFlag(t, endOfCluster, FlagClusterBreak, actual, glyphCursor)
glyphCursor++
if glyphIdx == end {
break
}
}
}
}
printLinePositioning(t, doc.lines, glyphs)
})
}
}
func checkFlag(t *testing.T, shouldHave bool, flag Flags, actual Glyph, glyphCursor int) {
t.Helper()
if shouldHave && actual.Flags&flag == 0 {
t.Errorf("glyphs[%d] should have %s set", glyphCursor, flag)
} else if !shouldHave && actual.Flags&flag != 0 {
t.Errorf("glyphs[%d] should not have %s set", glyphCursor, flag)
}
}
func printLinePositioning(t *testing.T, lines []line, glyphs []Glyph) {
t.Helper()
glyphCursor := 0
for i, line := range lines {
t.Logf("line %d, dir %s, width %d, visual %v, runeCount: %d", i, line.direction, line.width, line.visualOrder, line.runeCount)
for k, run := range line.runs {
t.Logf("run: %d, dir %s, width %d, runes {count: %d, offset: %d}", k, run.Direction, run.Advance, run.Runes.Count, run.Runes.Offset)
start := 0
end := len(run.Glyphs) - 1
inc := 1
if run.Direction.Progression() == system.TowardOrigin {
start = len(run.Glyphs) - 1
end = 0
inc = -1
}
for g := start; ; g += inc {
glyph := run.Glyphs[g]
if glyphCursor < len(glyphs) {
t.Logf("glyph %2d, adv %3d, runes %2d, glyphs %d - glyphs[%2d] flags %s", g, glyph.xAdvance, glyph.runeCount, glyph.glyphCount, glyphCursor, glyphs[glyphCursor].Flags)
t.Logf("glyph %2d, adv %3d, runes %2d, glyphs %d - n/a", g, glyph.xAdvance, glyph.runeCount, glyph.glyphCount)
}
glyphCursor++
if g == end {
break
}
}
}
}
}