From e356f1d88f398a0fdb098fb9f2309cc102d80d8a Mon Sep 17 00:00:00 2001 From: Fazlul Shahriar Date: Fri, 23 Jul 2010 12:34:35 -0700 Subject: [PATCH] bytes: port IndexFunc and LastIndexFunc from strings package This CL basically applies the same changes as http://code.google.com/p/go/source/detail?r=5e0a29014e8e but for bytes package. R=r, rog CC=golang-dev https://golang.org/cl/1670052 --- src/pkg/bytes/bytes.go | 122 +++++++++++++++++++++++--------- src/pkg/bytes/bytes_test.go | 94 +++++++++++++++++++++--- src/pkg/strings/strings.go | 6 +- src/pkg/strings/strings_test.go | 38 ++-------- 4 files changed, 180 insertions(+), 80 deletions(-) diff --git a/src/pkg/bytes/bytes.go b/src/pkg/bytes/bytes.go index dc0a41553c2..c0ee44ff96c 100644 --- a/src/pkg/bytes/bytes.go +++ b/src/pkg/bytes/bytes.go @@ -127,7 +127,7 @@ func LastIndex(s, sep []byte) int { return -1 } -// IndexAny interprets s as a sequence of UTF-8 encoded Unicode code points. +// IndexAny interprets s as a sequence of UTF-8-encoded Unicode code points. // It returns the byte index of the first occurrence in s of any of the Unicode // code points in chars. It returns -1 if chars is empty or if there is no code // point in common. @@ -278,7 +278,7 @@ func HasSuffix(s, suffix []byte) bool { // Map returns a copy of the byte array s with all its characters modified // according to the mapping function. If mapping returns a negative value, the character is // dropped from the string with no replacement. The characters in s and the -// output are interpreted as UTF-8 encoded Unicode code points. +// output are interpreted as UTF-8-encoded Unicode code points. func Map(mapping func(rune int) int, s []byte) []byte { // In the worst case, the array can grow when mapped, making // things unpleasant. But it's so rare we barge in assuming it's @@ -378,52 +378,104 @@ func Title(s []byte) []byte { s) } -// TrimLeftFunc returns a subslice of s by slicing off all leading UTF-8 encoded +// TrimLeftFunc returns a subslice of s by slicing off all leading UTF-8-encoded // Unicode code points c that satisfy f(c). func TrimLeftFunc(s []byte, f func(r int) bool) []byte { - var start, wid int - for start = 0; start < len(s); start += wid { - wid = 1 - rune := int(s[start]) - if rune >= utf8.RuneSelf { - rune, wid = utf8.DecodeRune(s[start:]) - } - if !f(rune) { - break - } + i := indexFunc(s, f, false) + if i == -1 { + return nil } - return s[start:] + return s[i:] } // TrimRightFunc returns a subslice of s by slicing off all trailing UTF-8 // encoded Unicode code points c that satisfy f(c). func TrimRightFunc(s []byte, f func(r int) bool) []byte { - var end, wid int - for end = len(s); end > 0; end -= wid { - wid = 1 - rune := int(s[end-wid]) - if rune >= utf8.RuneSelf { - // Back up & look for beginning of rune. Mustn't pass start. - for wid = 2; end-wid >= 0 && !utf8.RuneStart(s[end-wid]); wid++ { - } - if end-wid < 0 { // invalid UTF-8 sequence; stop processing - break - } - rune, wid = utf8.DecodeRune(s[end-wid : end]) - } - if !f(rune) { - break - } + i := lastIndexFunc(s, f, false) + if i >= 0 && s[i] >= utf8.RuneSelf { + _, wid := utf8.DecodeRune(s[i:]) + i += wid + } else { + i++ } - return s[0:end] + return s[0:i] } // TrimFunc returns a subslice of s by slicing off all leading and trailing -// UTF-8 encoded Unicode code points c that satisfy f(c). +// UTF-8-encoded Unicode code points c that satisfy f(c). func TrimFunc(s []byte, f func(r int) bool) []byte { return TrimRightFunc(TrimLeftFunc(s, f), f) } +// IndexFunc interprets s as a sequence of UTF-8-encoded Unicode code points. +// It returns the byte index in s of the first Unicode +// code point satisfying f(c), or -1 if none do. +func IndexFunc(s []byte, f func(r int) bool) int { + return indexFunc(s, f, true) +} + +// LastIndexFunc interprets s as a sequence of UTF-8-encoded Unicode code points. +// It returns the byte index in s of the last Unicode +// code point satisfying f(c), or -1 if none do. +func LastIndexFunc(s []byte, f func(r int) bool) int { + return lastIndexFunc(s, f, true) +} + +// indexFunc is the same as IndexFunc except that if +// truth==false, the sense of the predicate function is +// inverted. +func indexFunc(s []byte, f func(r int) bool, truth bool) int { + start := 0 + for start < len(s) { + wid := 1 + rune := int(s[start]) + if rune >= utf8.RuneSelf { + rune, wid = utf8.DecodeRune(s[start:]) + } + if f(rune) == truth { + return start + } + start += wid + } + return -1 +} + +// lastIndexFunc is the same as LastIndexFunc except that if +// truth==false, the sense of the predicate function is +// inverted. +func lastIndexFunc(s []byte, f func(r int) bool, truth bool) int { + end := len(s) + for end > 0 { + start := end - 1 + rune := int(s[start]) + if rune >= utf8.RuneSelf { + // Back up & look for beginning of rune. Mustn't pass start. + for start--; start >= 0; start-- { + if utf8.RuneStart(s[start]) { + break + } + } + if start < 0 { + return -1 + } + var wid int + rune, wid = utf8.DecodeRune(s[start:end]) + + // If we've decoded fewer bytes than we expected, + // we've got some invalid UTF-8, so make sure we return + // the last possible index in s. + if start+wid < end && f(utf8.RuneError) == truth { + return end - 1 + } + } + if f(rune) == truth { + return start + } + end = start + } + return -1 +} + func makeCutsetFunc(cutset string) func(rune int) bool { return func(rune int) bool { for _, c := range cutset { @@ -436,19 +488,19 @@ func makeCutsetFunc(cutset string) func(rune int) bool { } // Trim returns a subslice of s by slicing off all leading and -// trailing UTF-8 encoded Unicode code points contained in cutset. +// trailing UTF-8-encoded Unicode code points contained in cutset. func Trim(s []byte, cutset string) []byte { return TrimFunc(s, makeCutsetFunc(cutset)) } // TrimLeft returns a subslice of s by slicing off all leading -// UTF-8 encoded Unicode code points contained in cutset. +// UTF-8-encoded Unicode code points contained in cutset. func TrimLeft(s []byte, cutset string) []byte { return TrimLeftFunc(s, makeCutsetFunc(cutset)) } // TrimRight returns a subslice of s by slicing off all trailing -// UTF-8 encoded Unicode code points that are contained in cutset. +// UTF-8-encoded Unicode code points that are contained in cutset. func TrimRight(s []byte, cutset string) []byte { return TrimRightFunc(s, makeCutsetFunc(cutset)) } diff --git a/src/pkg/bytes/bytes_test.go b/src/pkg/bytes/bytes_test.go index de5edd120fc..5d3e9f66a1e 100644 --- a/src/pkg/bytes/bytes_test.go +++ b/src/pkg/bytes/bytes_test.go @@ -8,6 +8,7 @@ import ( . "bytes" "testing" "unicode" + "utf8" ) func eq(a, b []string) bool { @@ -367,8 +368,14 @@ var trimSpaceTests = []StringTest{ StringTest{" \t\r\n x\t\t\r\r\n\n ", "x"}, StringTest{" \u2000\t\r\n x\t\t\r\r\ny\n \u3000", "x\t\t\r\r\ny"}, StringTest{"1 \t\r\n2", "1 \t\r\n2"}, - StringTest{" x\x80", "x\x80"}, // invalid UTF-8 on end - StringTest{" x\xc0", "x\xc0"}, // invalid UTF-8 on end + StringTest{" x\x80", "x\x80"}, + StringTest{" x\xc0", "x\xc0"}, + StringTest{"x \xc0\xc0 ", "x \xc0\xc0"}, + StringTest{"x \xc0", "x \xc0"}, + StringTest{"x \xc0 ", "x \xc0"}, + StringTest{"x \xc0\xc0 ", "x \xc0\xc0"}, + StringTest{"x ☺\xc0\xc0 ", "x ☺\xc0\xc0"}, + StringTest{"x ☺ ", "x ☺"}, } // Bytes returns a new slice containing the bytes in s. @@ -607,6 +614,7 @@ var trimTests = []TrimTest{ TrimTest{TrimRight, "abba", "", "abba"}, TrimTest{TrimRight, "", "123", ""}, TrimTest{TrimRight, "", "", ""}, + TrimTest{TrimRight, "☺\xc0", "☺", "☺\xc0"}, } func TestTrim(t *testing.T) { @@ -629,22 +637,90 @@ func TestTrim(t *testing.T) { } } +type predicate struct { + f func(r int) bool + name string +} + +var isSpace = predicate{unicode.IsSpace, "IsSpace"} +var isDigit = predicate{unicode.IsDigit, "IsDigit"} +var isUpper = predicate{unicode.IsUpper, "IsUpper"} +var isValidRune = predicate{ + func(r int) bool { + return r != utf8.RuneError + }, + "IsValidRune", +} + type TrimFuncTest struct { - f func(r int) bool - name, in, out string + f predicate + in, out string +} + +func not(p predicate) predicate { + return predicate{ + func(r int) bool { + return !p.f(r) + }, + "not " + p.name, + } } var trimFuncTests = []TrimFuncTest{ - TrimFuncTest{unicode.IsSpace, "IsSpace", space + " hello " + space, "hello"}, - TrimFuncTest{unicode.IsDigit, "IsDigit", "\u0e50\u0e5212hello34\u0e50\u0e51", "hello"}, - TrimFuncTest{unicode.IsUpper, "IsUpper", "\u2C6F\u2C6F\u2C6F\u2C6FABCDhelloEF\u2C6F\u2C6FGH\u2C6F\u2C6F", "hello"}, + TrimFuncTest{isSpace, space + " hello " + space, "hello"}, + TrimFuncTest{isDigit, "\u0e50\u0e5212hello34\u0e50\u0e51", "hello"}, + TrimFuncTest{isUpper, "\u2C6F\u2C6F\u2C6F\u2C6FABCDhelloEF\u2C6F\u2C6FGH\u2C6F\u2C6F", "hello"}, + TrimFuncTest{not(isSpace), "hello" + space + "hello", space}, + TrimFuncTest{not(isDigit), "hello\u0e50\u0e521234\u0e50\u0e51helo", "\u0e50\u0e521234\u0e50\u0e51"}, + TrimFuncTest{isValidRune, "ab\xc0a\xc0cd", "\xc0a\xc0"}, + TrimFuncTest{not(isValidRune), "\xc0a\xc0", "a"}, } func TestTrimFunc(t *testing.T) { for _, tc := range trimFuncTests { - actual := string(TrimFunc([]byte(tc.in), tc.f)) + actual := string(TrimFunc([]byte(tc.in), tc.f.f)) if actual != tc.out { - t.Errorf("TrimFunc(%q, %q) = %q; want %q", tc.in, tc.name, actual, tc.out) + t.Errorf("TrimFunc(%q, %q) = %q; want %q", tc.in, tc.f.name, actual, tc.out) + } + } +} + +type IndexFuncTest struct { + in string + f predicate + first, last int +} + +var indexFuncTests = []IndexFuncTest{ + IndexFuncTest{"", isValidRune, -1, -1}, + IndexFuncTest{"abc", isDigit, -1, -1}, + IndexFuncTest{"0123", isDigit, 0, 3}, + IndexFuncTest{"a1b", isDigit, 1, 1}, + IndexFuncTest{space, isSpace, 0, len(space) - 3}, // last rune in space is 3 bytes + IndexFuncTest{"\u0e50\u0e5212hello34\u0e50\u0e51", isDigit, 0, 18}, + IndexFuncTest{"\u2C6F\u2C6F\u2C6F\u2C6FABCDhelloEF\u2C6F\u2C6FGH\u2C6F\u2C6F", isUpper, 0, 34}, + IndexFuncTest{"12\u0e50\u0e52hello34\u0e50\u0e51", not(isDigit), 8, 12}, + + // tests of invalid UTF-8 + IndexFuncTest{"\x801", isDigit, 1, 1}, + IndexFuncTest{"\x80abc", isDigit, -1, -1}, + IndexFuncTest{"\xc0a\xc0", isValidRune, 1, 1}, + IndexFuncTest{"\xc0a\xc0", not(isValidRune), 0, 2}, + IndexFuncTest{"\xc0☺\xc0", not(isValidRune), 0, 4}, + IndexFuncTest{"\xc0☺\xc0\xc0", not(isValidRune), 0, 5}, + IndexFuncTest{"ab\xc0a\xc0cd", not(isValidRune), 2, 4}, + IndexFuncTest{"a\xe0\x80cd", not(isValidRune), 1, 2}, +} + +func TestIndexFunc(t *testing.T) { + for _, tc := range indexFuncTests { + first := IndexFunc([]byte(tc.in), tc.f.f) + if first != tc.first { + t.Errorf("IndexFunc(%q, %s) = %d; want %d", tc.in, tc.f.name, first, tc.first) + } + last := LastIndexFunc([]byte(tc.in), tc.f.f) + if last != tc.last { + t.Errorf("LastIndexFunc(%q, %s) = %d; want %d", tc.in, tc.f.name, last, tc.last) } } } diff --git a/src/pkg/strings/strings.go b/src/pkg/strings/strings.go index 594f03527e1..925566c744e 100644 --- a/src/pkg/strings/strings.go +++ b/src/pkg/strings/strings.go @@ -422,8 +422,7 @@ func LastIndexFunc(s string, f func(r int) bool) int { // indexFunc is the same as IndexFunc except that if // truth==false, the sense of the predicate function is -// inverted. We could use IndexFunc directly, but this -// way saves a closure allocation. +// inverted. func indexFunc(s string, f func(r int) bool, truth bool) int { start := 0 for start < len(s) { @@ -442,8 +441,7 @@ func indexFunc(s string, f func(r int) bool, truth bool) int { // lastIndexFunc is the same as LastIndexFunc except that if // truth==false, the sense of the predicate function is -// inverted. We could use IndexFunc directly, but this -// way saves a closure allocation. +// inverted. func lastIndexFunc(s string, f func(r int) bool, truth bool) int { end := len(s) for end > 0 { diff --git a/src/pkg/strings/strings_test.go b/src/pkg/strings/strings_test.go index 8b0c98113a6..9e8feceaa2a 100644 --- a/src/pkg/strings/strings_test.go +++ b/src/pkg/strings/strings_test.go @@ -420,25 +420,6 @@ var trimTests = []TrimTest{ TrimTest{TrimRight, "☺\xc0", "☺", "☺\xc0"}, } -// naiveTrimRight implements a version of TrimRight -// by scanning forwards from the start of s. -func naiveTrimRight(s string, cutset string) string { - i := -1 - for j, r := range s { - if IndexRune(cutset, r) == -1 { - i = j - } - } - if i >= 0 && s[i] >= utf8.RuneSelf { - _, wid := utf8.DecodeRuneInString(s[i:]) - i += wid - } else { - i++ - } - return s[0:i] -} - - func TestTrim(t *testing.T) { for _, tc := range trimTests { actual := tc.f(tc.in, tc.cutset) @@ -456,16 +437,14 @@ func TestTrim(t *testing.T) { if actual != tc.out { t.Errorf("%s(%q, %q) = %q; want %q", name, tc.in, tc.cutset, actual, tc.out) } - // test equivalence of TrimRight to naive version - if tc.f == TrimRight { - naive := naiveTrimRight(tc.in, tc.cutset) - if naive != actual { - t.Errorf("TrimRight(%q, %q) = %q, want %q", tc.in, tc.cutset, actual, naive) - } - } } } +type predicate struct { + f func(r int) bool + name string +} + var isSpace = predicate{unicode.IsSpace, "IsSpace"} var isDigit = predicate{unicode.IsDigit, "IsDigit"} var isUpper = predicate{unicode.IsUpper, "IsUpper"} @@ -476,11 +455,6 @@ var isValidRune = predicate{ "IsValidRune", } -type predicate struct { - f func(r int) bool - name string -} - type TrimFuncTest struct { f predicate in, out string @@ -530,7 +504,7 @@ var indexFuncTests = []IndexFuncTest{ IndexFuncTest{"\u2C6F\u2C6F\u2C6F\u2C6FABCDhelloEF\u2C6F\u2C6FGH\u2C6F\u2C6F", isUpper, 0, 34}, IndexFuncTest{"12\u0e50\u0e52hello34\u0e50\u0e51", not(isDigit), 8, 12}, - // broken unicode tests + // tests of invalid UTF-8 IndexFuncTest{"\x801", isDigit, 1, 1}, IndexFuncTest{"\x80abc", isDigit, -1, -1}, IndexFuncTest{"\xc0a\xc0", isValidRune, 1, 1},