diff options
Diffstat (limited to 'util/text')
| -rw-r--r-- | util/text/left_padded.go | 42 | ||||
| -rw-r--r-- | util/text/left_padded_test.go | 56 | ||||
| -rw-r--r-- | util/text/text.go | 330 | ||||
| -rw-r--r-- | util/text/text_test.go | 376 |
4 files changed, 0 insertions, 804 deletions
diff --git a/util/text/left_padded.go b/util/text/left_padded.go deleted file mode 100644 index eae65d34..00000000 --- a/util/text/left_padded.go +++ /dev/null @@ -1,42 +0,0 @@ -package text - -import ( - "bytes" - "fmt" - "github.com/mattn/go-runewidth" - "strings" -) - -// LeftPadMaxLine pads a string on the left by a specified amount and pads the -// string on the right to fill the maxLength -func LeftPadMaxLine(text string, length, leftPad int) string { - var rightPart string = text - - scrWidth := runewidth.StringWidth(text) - // truncate and ellipse if needed - if scrWidth+leftPad > length { - rightPart = runewidth.Truncate(text, length-leftPad, "…") - } else if scrWidth+leftPad < length { - rightPart = runewidth.FillRight(text, length-leftPad) - } - - return fmt.Sprintf("%s%s", - strings.Repeat(" ", leftPad), - rightPart, - ) -} - -// LeftPad left pad each line of the given text -func LeftPad(text string, leftPad int) string { - var result bytes.Buffer - - pad := strings.Repeat(" ", leftPad) - - for _, line := range strings.Split(text, "\n") { - result.WriteString(pad) - result.WriteString(line) - result.WriteString("\n") - } - - return result.String() -} diff --git a/util/text/left_padded_test.go b/util/text/left_padded_test.go deleted file mode 100644 index 0be79e32..00000000 --- a/util/text/left_padded_test.go +++ /dev/null @@ -1,56 +0,0 @@ -package text - -import "testing" - -func TestLeftPadMaxLine(t *testing.T) { - cases := []struct { - input, output string - maxValueLength int - leftPad int - }{ - { - "foo", - "foo ", - 4, - 0, - }, - { - "foofoofoo", - "foo…", - 4, - 0, - }, - { - "foo", - "foo ", - 10, - 0, - }, - { - "foo", - " f…", - 4, - 2, - }, - { - "foofoofoo", - " foo…", - 6, - 2, - }, - { - "foo", - " foo ", - 10, - 2, - }, - } - - for i, tc := range cases { - result := LeftPadMaxLine(tc.input, tc.maxValueLength, tc.leftPad) - if result != tc.output { - t.Fatalf("Case %d Input:\n\n`%s`\n\nExpected Output:\n\n`%s`\n\nActual Output:\n\n`%s`", - i, tc.input, tc.output, result) - } - } -} diff --git a/util/text/text.go b/util/text/text.go deleted file mode 100644 index 39584d5d..00000000 --- a/util/text/text.go +++ /dev/null @@ -1,330 +0,0 @@ -package text - -import ( - "github.com/mattn/go-runewidth" - "strings" - "unicode/utf8" -) - -// Force runewidth not to treat ambiguous runes as wide chars, so that things -// like unicode ellipsis/up/down/left/right glyphs can have correct runewidth -// and can be displayed correctly in terminals. -func init() { - runewidth.DefaultCondition.EastAsianWidth = false -} - -// Wrap a text for an exact line size -// Handle properly terminal color escape code -func Wrap(text string, lineWidth int) (string, int) { - return WrapLeftPadded(text, lineWidth, 0) -} - -// Wrap a text for an exact line size with a left padding -// Handle properly terminal color escape code -func WrapLeftPadded(text string, lineWidth int, leftPad int) (string, int) { - var lines []string - nbLine := 0 - pad := strings.Repeat(" ", leftPad) - - // tabs are formatted as 4 spaces - text = strings.Replace(text, "\t", " ", -1) - // NOTE: text is first segmented into lines so that softwrapLine can handle. - for _, line := range strings.Split(text, "\n") { - if line == "" || strings.TrimSpace(line) == "" { - lines = append(lines, "") - nbLine++ - } else { - wrapped := softwrapLine(line, lineWidth-leftPad) - firstLine := true - for _, seg := range strings.Split(wrapped, "\n") { - if firstLine { - lines = append(lines, pad+strings.TrimRight(seg, " ")) - firstLine = false - } else { - lines = append(lines, pad+strings.TrimSpace(seg)) - } - nbLine++ - } - } - } - return strings.Join(lines, "\n"), nbLine -} - -// Break a line into several lines so that each line consumes at most -// 'textWidth' cells. Lines break at groups of white spaces and multibyte -// chars. Nothing is removed from the original text so that it behaves like a -// softwrap. -// -// Required: The line shall not contain '\n' -// -// WRAPPING ALGORITHM: The line is broken into non-breakable chunks, then line -// breaks ("\n") are inserted between these groups so that the total length -// between breaks does not exceed the required width. Words that are longer than -// the textWidth are broken into pieces no longer than textWidth. -// -func softwrapLine(line string, textWidth int) string { - // NOTE: terminal escapes are stripped out of the line so the algorithm is - // simpler. Do not try to mix them in the wrapping algorithm, as it can get - // complicated quickly. - line1, termEscapes := extractTermEscapes(line) - - chunks := segmentLine(line1) - // Reverse the chunk array so we can use it as a stack. - for i, j := 0, len(chunks)-1; i < j; i, j = i+1, j-1 { - chunks[i], chunks[j] = chunks[j], chunks[i] - } - var line2 string = "" - var width int = 0 - for len(chunks) > 0 { - thisWord := chunks[len(chunks)-1] - wl := wordLen(thisWord) - if width+wl <= textWidth { - line2 += chunks[len(chunks)-1] - chunks = chunks[:len(chunks)-1] - width += wl - if width == textWidth && len(chunks) > 0 { - // NOTE: new line begins when current line is full and there are more - // chunks to come. - line2 += "\n" - width = 0 - } - } else if wl > textWidth { - // NOTE: By default, long words are splited to fill the remaining space. - // But if the long words is the first non-space word in the middle of the - // line, preceeding spaces shall not be counted in word spliting. - splitWidth := textWidth - width - if strings.HasSuffix(line2, "\n"+strings.Repeat(" ", width)) { - splitWidth += width - } - left, right := splitWord(chunks[len(chunks)-1], splitWidth) - chunks[len(chunks)-1] = right - line2 += left + "\n" - width = 0 - } else { - line2 += "\n" - width = 0 - } - } - - line3 := applyTermEscapes(line2, termEscapes) - return line3 -} - -// EscapeItem: Storage of terminal escapes in a line. 'item' is the actural -// escape command, and 'pos' is the index in the rune array where the 'item' -// shall be inserted back. For example, the escape item in "F\x1b33mox" is -// {"\x1b33m", 1}. -type escapeItem struct { - item string - pos int -} - -// Extract terminal escapes out of a line, returns a new line without terminal -// escapes and a slice of escape items. The terminal escapes can be inserted -// back into the new line at rune index 'item.pos' to recover the original line. -// -// Required: The line shall not contain "\n" -// -func extractTermEscapes(line string) (string, []escapeItem) { - var termEscapes []escapeItem - var line1 string - - pos := 0 - item := "" - occupiedRuneCount := 0 - inEscape := false - for i, r := range []rune(line) { - if r == '\x1b' { - pos = i - item = string(r) - inEscape = true - continue - } - if inEscape { - item += string(r) - if r == 'm' { - termEscapes = append(termEscapes, escapeItem{item, pos - occupiedRuneCount}) - occupiedRuneCount += utf8.RuneCountInString(item) - inEscape = false - } - continue - } - line1 += string(r) - } - - return line1, termEscapes -} - -// Apply the extracted terminal escapes to the edited line. The only edit -// allowed is to insert "\n" like that in softwrapLine. Callers shall ensure -// this since this function is not able to check it. -func applyTermEscapes(line string, escapes []escapeItem) string { - if len(escapes) == 0 { - return line - } - - var out string = "" - - currPos := 0 - currItem := 0 - for _, r := range line { - if currItem < len(escapes) && currPos == escapes[currItem].pos { - // NOTE: We avoid terminal escapes at the end of a line by move them one - // pass the end of line, so that algorithms who trim right spaces are - // happy. But algorithms who trim left spaces are still unhappy. - if r == '\n' { - out += "\n" + escapes[currItem].item - } else { - out += escapes[currItem].item + string(r) - currPos++ - } - currItem++ - } else { - if r != '\n' { - currPos++ - } - out += string(r) - } - } - - // Don't forget the trailing escape, if any. - if currItem == len(escapes)-1 && currPos == escapes[currItem].pos { - out += escapes[currItem].item - } - - return out -} - -// Segment a line into chunks, where each chunk consists of chars with the same -// type and is not breakable. -func segmentLine(s string) []string { - var chunks []string - - var word string - wordType := none - flushWord := func() { - chunks = append(chunks, word) - word = "" - wordType = none - } - - for _, r := range s { - // A WIDE_CHAR itself constitutes a chunk. - thisType := runeType(r) - if thisType == wideChar { - if wordType != none { - flushWord() - } - chunks = append(chunks, string(r)) - continue - } - // Other type of chunks starts with a char of that type, and ends with a - // char with different type or end of string. - if thisType != wordType { - if wordType != none { - flushWord() - } - word = string(r) - wordType = thisType - } else { - word += string(r) - } - } - if word != "" { - flushWord() - } - - return chunks -} - -// Rune categories -// -// These categories are so defined that each category forms a non-breakable -// chunk. It IS NOT the same as unicode code point categories. -// -const ( - none int = iota - wideChar - invisible - shortUnicode - space - visibleAscii -) - -// Determine the category of a rune. -func runeType(r rune) int { - rw := runewidth.RuneWidth(r) - if rw > 1 { - return wideChar - } else if rw == 0 { - return invisible - } else if r > 127 { - return shortUnicode - } else if r == ' ' { - return space - } else { - return visibleAscii - } -} - -// wordLen return the length of a word, while ignoring the terminal escape -// sequences -func wordLen(word string) int { - length := 0 - escape := false - - for _, char := range word { - if char == '\x1b' { - escape = true - } - if !escape { - length += runewidth.RuneWidth(rune(char)) - } - if char == 'm' { - escape = false - } - } - - return length -} - -// splitWord split a word at the given length, while ignoring the terminal escape sequences -func splitWord(word string, length int) (string, string) { - runes := []rune(word) - var result []rune - added := 0 - escape := false - - if length == 0 { - return "", word - } - - for _, r := range runes { - if r == '\x1b' { - escape = true - } - - width := runewidth.RuneWidth(r) - if width+added > length { - // wide character made the length overflow - break - } - - result = append(result, r) - - if !escape { - added += width - if added >= length { - break - } - } - - if r == 'm' { - escape = false - } - } - - leftover := runes[len(result):] - - return string(result), string(leftover) -} diff --git a/util/text/text_test.go b/util/text/text_test.go deleted file mode 100644 index 5be25409..00000000 --- a/util/text/text_test.go +++ /dev/null @@ -1,376 +0,0 @@ -package text - -import ( - "reflect" - "strings" - "testing" -) - -func TestWrap(t *testing.T) { - cases := []struct { - Input, Output string - Lim int - }{ - // A simple word passes through. - { - "foo", - "foo", - 4, - }, - // Word breaking - { - "foobarbaz", - "foob\narba\nz", - 4, - }, - // Lines are broken at whitespace. - { - "foo bar baz", - "foo\nbar\nbaz", - 4, - }, - // Word breaking - { - "foo bars bazzes", - "foo\nbars\nbazz\nes", - 4, - }, - // A word that would run beyond the width is wrapped. - { - "fo sop", - "fo\nsop", - 4, - }, - // A tab counts as 4 characters. - { - "foo\nb\t r\n baz", - "foo\nb\nr\n baz", - 4, - }, - // Trailing whitespace is removed after used for wrapping. - // Runs of whitespace on which a line is broken are removed. - { - "foo \nb ar ", - "foo\n\nb\nar\n", - 4, - }, - // An explicit line break at the end of the input is preserved. - { - "foo bar baz\n", - "foo\nbar\nbaz\n", - 4, - }, - // Explicit break are always preserved. - { - "\nfoo bar\n\n\nbaz\n", - "\nfoo\nbar\n\n\nbaz\n", - 4, - }, - // Ignore complete words with terminal color sequence - { - "foo \x1b[31mbar\x1b[0m baz", - "foo\n\x1b[31mbar\x1b[0m\nbaz", - 4, - }, - // Handle words with colors sequence inside the word - { - "foo b\x1b[31mbar\x1b[0mr baz", - "foo\nb\x1b[31mbar\n\x1b[0mr\nbaz", - 4, - }, - // Break words with colors sequence inside the word - { - "foo bb\x1b[31mbar\x1b[0mr baz", - "foo\nbb\x1b[31mba\nr\x1b[0mr\nbaz", - 4, - }, - // Complete example: - { - " This is a list: \n\n\t* foo\n\t* bar\n\n\n\t* baz \nBAM ", - " This\nis a\nlist:\n\n *\nfoo\n *\nbar\n\n\n *\nbaz\nBAM\n", - 6, - }, - // Handle chinese (wide characters) - { - "一只敏捷的狐狸跳过了一只懒狗。", - "一只敏捷的狐\n狸跳过了一只\n懒狗。", - 12, - }, - // Handle chinese with colors - { - "一只敏捷的\x1b[31m狐狸跳过\x1b[0m了一只懒狗。", - "一只敏捷的\x1b[31m狐\n狸跳过\x1b[0m了一只\n懒狗。", - 12, - }, - // Handle mixed wide and short characters - { - "敏捷 A quick 的狐狸 fox 跳过 jumps over a lazy 了一只懒狗 dog。", - "敏捷 A quick\n的狐狸 fox\n跳过 jumps\nover a lazy\n了一只懒狗\ndog。", - 12, - }, - // Handle mixed wide and short characters with color - { - "敏捷 A \x1b31mquick 的狐狸 fox 跳\x1b0m过 jumps over a lazy 了一只懒狗 dog。", - "敏捷 A \x1b31mquick\n的狐狸 fox\n跳\x1b0m过 jumps\nover a lazy\n了一只懒狗\ndog。", - 12, - }, - } - - for i, tc := range cases { - actual, lines := Wrap(tc.Input, tc.Lim) - if actual != tc.Output { - t.Fatalf("Case %d Input:\n\n`%s`\n\nExpected Output:\n\n`%s`\n\nActual Output:\n\n`%s`", - i, tc.Input, tc.Output, actual) - } - - expected := len(strings.Split(tc.Output, "\n")) - if expected != lines { - t.Fatalf("Case %d Nb lines mismatch\nExpected:%d\nActual:%d", - i, expected, lines) - } - } -} - -func TestWrapLeftPadded(t *testing.T) { - cases := []struct { - input, output string - lim, pad int - }{ - { - "The Lorem ipsum text is typically composed of pseudo-Latin words. It is commonly used as placeholder text to examine or demonstrate the visual effects of various graphic design.", - ` The Lorem ipsum text is typically composed of - pseudo-Latin words. It is commonly used as placeholder - text to examine or demonstrate the visual effects of - various graphic design.`, - 59, 4, - }, - // Handle Chinese - { - "婞一枳郲逴靲屮蜧曀殳，掫乇峔掮傎溒兀緉冘仜。郼牪艽螗媷錵朸一詅掜豗怙刉笀丌，楀棶乇矹迡搦囷圣亍昄漚粁仈祂。覂一洳袶揙楱亍滻瘯毌，掗屮柅軡菵腩乜榵毌夯。勼哻怌婇怤灟葠雺奷朾恦扰衪岨坋誁乇芚誙腞。冇笉妺悆浂鱦賌廌灱灱觓坋佫呬耴跣兀枔蓔輈。嵅咍犴膰痭瘰机一靬涽捊矷尒玶乇，煚塈丌岰陊鉖怞戉兀甿跾觓夬侄。棩岧汌橩僁螗玎一逭舴圂衪扐衲兀，嵲媕亍衩衿溽昃夯丌侄蒰扂丱呤。毰侘妅錣廇螉仴一暀淖蚗佶庂咺丌，輀鈁乇彽洢溦洰氶乇构碨洐巿阹。", - ` 婞一枳郲逴靲屮蜧曀殳，掫乇峔掮傎溒兀緉冘仜。郼牪艽螗媷 - 錵朸一詅掜豗怙刉笀丌，楀棶乇矹迡搦囷圣亍昄漚粁仈祂。覂 - 一洳袶揙楱亍滻瘯毌，掗屮柅軡菵腩乜榵毌夯。勼哻怌婇怤灟 - 葠雺奷朾恦扰衪岨坋誁乇芚誙腞。冇笉妺悆浂鱦賌廌灱灱觓坋 - 佫呬耴跣兀枔蓔輈。嵅咍犴膰痭瘰机一靬涽捊矷尒玶乇，煚塈 - 丌岰陊鉖怞戉兀甿跾觓夬侄。棩岧汌橩僁螗玎一逭舴圂衪扐衲 - 兀，嵲媕亍衩衿溽昃夯丌侄蒰扂丱呤。毰侘妅錣廇螉仴一暀淖 - 蚗佶庂咺丌，輀鈁乇彽洢溦洰氶乇构碨洐巿阹。`, - 59, 4, - }, - // Handle long unbreakable words in a full stentence - { - "OT: there are alternatives to maintainer-/user-set priority, e.g. \"[user pain](http://www.lostgarden.com/2008/05/improving-bug-triage-with-user-pain.html)\".", - ` OT: there are alternatives to maintainer-/user-set - priority, e.g. "[user pain](http://www.lostgarden.com/ - 2008/05/improving-bug-triage-with-user-pain.html)".`, - 58, 4, - }, - } - - for i, tc := range cases { - actual, lines := WrapLeftPadded(tc.input, tc.lim, tc.pad) - if actual != tc.output { - t.Fatalf("Case %d Input:\n\n`%s`\n\nExpected Output:\n`\n%s`\n\nActual Output:\n`\n%s\n%s`", - i, tc.input, tc.output, - "|"+strings.Repeat("-", tc.lim-2)+"|", - actual) - } - - expected := len(strings.Split(tc.output, "\n")) - if expected != lines { - t.Fatalf("Case %d Nb lines mismatch\nExpected:%d\nActual:%d", - i, expected, lines) - } - } -} - -func TestWordLen(t *testing.T) { - cases := []struct { - Input string - Length int - }{ - // A simple word - { - "foo", - 3, - }, - // A simple word with colors - { - "\x1b[31mbar\x1b[0m", - 3, - }, - // Handle prefix and suffix properly - { - "foo\x1b[31mfoobarHoy\x1b[0mbaaar", - 17, - }, - // Handle chinese - { - "快檢什麼望對", - 12, - }, - // Handle chinese with colors - { - "快\x1b[31m檢什麼\x1b[0m望對", - 12, - }, - } - - for i, tc := range cases { - l := wordLen(tc.Input) - if l != tc.Length { - t.Fatalf("Case %d Input:\n\n`%s`\n\nExpected Output:\n\n`%d`\n\nActual Output:\n\n`%d`", - i, tc.Input, tc.Length, l) - } - } -} - -func TestSplitWord(t *testing.T) { - cases := []struct { - Input string - Length int - Result, Leftover string - }{ - // A simple word passes through. - { - "foo", - 4, - "foo", "", - }, - // Cut at the right place - { - "foobarHoy", - 4, - "foob", "arHoy", - }, - // A simple word passes through with colors - { - "\x1b[31mbar\x1b[0m", - 4, - "\x1b[31mbar\x1b[0m", "", - }, - // Cut at the right place with colors - { - "\x1b[31mfoobarHoy\x1b[0m", - 4, - "\x1b[31mfoob", "arHoy\x1b[0m", - }, - // Handle prefix and suffix properly - { - "foo\x1b[31mfoobarHoy\x1b[0mbaaar", - 4, - "foo\x1b[31mf", "oobarHoy\x1b[0mbaaar", - }, - // Cut properly with length = 0 - { - "foo", - 0, - "", "foo", - }, - // Handle chinese - { - "快檢什麼望對", - 4, - "快檢", "什麼望對", - }, - { - "快檢什麼望對", - 5, - "快檢", "什麼望對", - }, - // Handle chinese with colors - { - "快\x1b[31m檢什麼\x1b[0m望對", - 4, - "快\x1b[31m檢", "什麼\x1b[0m望對", - }, - } - - for i, tc := range cases { - result, leftover := splitWord(tc.Input, tc.Length) - if result != tc.Result || leftover != tc.Leftover { - t.Fatalf("Case %d Input:\n\n`%s`\n\nExpected Output:\n\n`%s` - `%s`\n\nActual Output:\n\n`%s` - `%s`", - i, tc.Input, tc.Result, tc.Leftover, result, leftover) - } - } -} - -func TestExtractApplyTermEscapes(t *testing.T) { - cases := []struct { - Input string - Output string - TermEscapes []escapeItem - }{ - // A plain ascii line with escapes. - { - "This \x1b[31mis an\x1b[0m example.", - "This is an example.", - []escapeItem{{"\x1b[31m", 5}, {"\x1b[0m", 10}}, - }, - // Escape at the end - { - "This \x1b[31mis an example.\x1b[0m", - "This is an example.", - []escapeItem{{"\x1b[31m", 5}, {"\x1b[0m", 19}}, - }, - // A plain wide line with escapes. - { - "一只敏捷\x1b[31m的狐狸\x1b[0m跳过了一只懒狗。", - "一只敏捷的狐狸跳过了一只懒狗。", - []escapeItem{{"\x1b[31m", 4}, {"\x1b[0m", 7}}, - }, - // A normal-wide mixed line with escapes. - { - "一只 A Quick 敏捷\x1b[31m的狐 Fox 狸\x1b[0m跳过了Dog一只懒狗。", - "一只 A Quick 敏捷的狐 Fox 狸跳过了Dog一只懒狗。", - []escapeItem{{"\x1b[31m", 13}, {"\x1b[0m", 21}}, - }, - } - - for i, tc := range cases { - line2, escapes := extractTermEscapes(tc.Input) - if line2 != tc.Output || !reflect.DeepEqual(escapes, tc.TermEscapes) { - t.Fatalf("Case %d Input:\n\n`%s`\n\nExpected Output:\n\nLine: `%s`\nEscapes: `%+v`\n\nActual Output:\n\nLine: `%s`\nEscapes: `%+v`\n\n", - i, tc.Input, tc.Output, tc.TermEscapes, line2, escapes) - } - line3 := applyTermEscapes(line2, escapes) - if line3 != tc.Input { - t.Fatalf("Case %d Input:\n\n`%s`\n\nExpected Result:\n\n`%s`\n\nActual Result:\n\n`%s`\n\n", - i, tc.Input, tc.Input, line3) - } - } -} - -func TestSegmentLines(t *testing.T) { - cases := []struct { - Input string - Output []string - }{ - // A plain ascii line with escapes. - { - "This is an example.", - []string{"This", " ", "is", " ", "an", " ", "example."}, - }, - // A plain wide line with escapes. - { - "一只敏捷的狐狸跳过了一只懒狗。", - []string{"一", "只", "敏", "捷", "的", "狐", "狸", "跳", "过", - "了", "一", "只", "懒", "狗", "。"}, - }, - // A complex stentence. - { - "This is a 'complex' example, where 一只 and English 混合了。", - []string{"This", " ", "is", " ", "a", " ", "'complex'", " ", "example,", - " ", "where", " ", "一", "只", " ", "and", " ", "English", " ", "混", - "合", "了", "。"}, - }, - } - - for i, tc := range cases { - chunks := segmentLine(tc.Input) - if !reflect.DeepEqual(chunks, tc.Output) { - t.Fatalf("Case %d Input:\n\n`%s`\n\nExpected Output:\n\n`[%s]`\n\nActual Output:\n\n`[%s]`\n\n", - i, tc.Input, strings.Join(tc.Output, ", "), strings.Join(chunks, ", ")) - } - } -} |