From 079a117469d8f4322ff549bc9c02ceaef8a3bb08 Mon Sep 17 00:00:00 2001 From: Rob Pike Date: Thu, 12 Aug 2010 17:16:37 +1000 Subject: [PATCH] regexp: delete the deprecated methods and tests. R=golang-dev CC=golang-dev https://golang.org/cl/1956044 --- src/pkg/regexp/all_test.go | 264 +++---------------------------------- src/pkg/regexp/regexp.go | 110 +--------------- 2 files changed, 20 insertions(+), 354 deletions(-) diff --git a/src/pkg/regexp/all_test.go b/src/pkg/regexp/all_test.go index 4bdd6c67e8a..13a3d3ce51f 100644 --- a/src/pkg/regexp/all_test.go +++ b/src/pkg/regexp/all_test.go @@ -51,68 +51,6 @@ var bad_re = []stringError{ stringError{`\x`, ErrBadBackslash}, } -type vec []int - -type tester struct { - re string - text string - match vec -} - -var matches = []tester{ - tester{`^abcdefg`, "abcdefg", vec{0, 7}}, - tester{`a+`, "baaab", vec{1, 4}}, - tester{"abcd..", "abcdef", vec{0, 6}}, - tester{``, "", vec{0, 0}}, - tester{`a`, "a", vec{0, 1}}, - tester{`x`, "y", vec{}}, - tester{`b`, "abc", vec{1, 2}}, - tester{`.`, "a", vec{0, 1}}, - tester{`.*`, "abcdef", vec{0, 6}}, - tester{`^`, "abcde", vec{0, 0}}, - tester{`$`, "abcde", vec{5, 5}}, - tester{`^abcd$`, "abcd", vec{0, 4}}, - tester{`^bcd'`, "abcdef", vec{}}, - tester{`^abcd$`, "abcde", vec{}}, - tester{`a+`, "baaab", vec{1, 4}}, - tester{`a*`, "baaab", vec{0, 0}}, - tester{`[a-z]+`, "abcd", vec{0, 4}}, - tester{`[^a-z]+`, "ab1234cd", vec{2, 6}}, - tester{`[a\-\]z]+`, "az]-bcz", vec{0, 4}}, - tester{`[^\n]+`, "abcd\n", vec{0, 4}}, - tester{`[日本語]+`, "日本語日本語", vec{0, 18}}, - tester{`日本語+`, "日本語", vec{0, 9}}, - tester{`日本語+`, "日本語語語語", vec{0, 18}}, - tester{`()`, "", vec{0, 0, 0, 0}}, - tester{`(a)`, "a", vec{0, 1, 0, 1}}, - tester{`(.)(.)`, "日a", vec{0, 4, 0, 3, 3, 4}}, - tester{`(.*)`, "", vec{0, 0, 0, 0}}, - tester{`(.*)`, "abcd", vec{0, 4, 0, 4}}, - tester{`(..)(..)`, "abcd", vec{0, 4, 0, 2, 2, 4}}, - tester{`(([^xyz]*)(d))`, "abcd", vec{0, 4, 0, 4, 0, 3, 3, 4}}, - tester{`((a|b|c)*(d))`, "abcd", vec{0, 4, 0, 4, 2, 3, 3, 4}}, - tester{`(((a|b|c)*)(d))`, "abcd", vec{0, 4, 0, 4, 0, 3, 2, 3, 3, 4}}, - tester{`a*(|(b))c*`, "aacc", vec{0, 4, 2, 2, -1, -1}}, - tester{`(.*).*`, "ab", vec{0, 2, 0, 2}}, - tester{`[.]`, ".", vec{0, 1}}, - tester{`/$`, "/abc/", vec{4, 5}}, - tester{`/$`, "/abc", vec{}}, - - // fixed bugs - tester{`ab$`, "cab", vec{1, 3}}, - tester{`axxb$`, "axxcb", vec{}}, - tester{`data`, "daXY data", vec{5, 9}}, - tester{`da(.)a$`, "daXY data", vec{5, 9, 7, 8}}, - - // can backslash-escape any punctuation - tester{`\!\"\#\$\%\&\'\(\)\*\+\,\-\.\/\:\;\<\=\>\?\@\[\\\]\^\_\{\|\}\~`, - `!"#$%&'()*+,-./:;<=>?@[\]^_{|}~`, vec{0, 31}}, - tester{`[\!\"\#\$\%\&\'\(\)\*\+\,\-\.\/\:\;\<\=\>\?\@\[\\\]\^\_\{\|\}\~]+`, - `!"#$%&'()*+,-./:;<=>?@[\]^_{|}~`, vec{0, 31}}, - tester{"\\`", "`", vec{0, 1}}, - tester{"[\\`]+", "`", vec{0, 1}}, -} - func compileTest(t *testing.T, expr string, error os.Error) *Regexp { re, err := Compile(expr) if err != error { @@ -121,66 +59,6 @@ func compileTest(t *testing.T, expr string, error os.Error) *Regexp { return re } -func printVec(t *testing.T, m []int) { - l := len(m) - if l == 0 { - t.Log("\t") - } else { - if m[len(m)-1] == -1 { - m = m[0 : len(m)-2] - } - t.Log("\t", m) - } -} - -func equal(m1, m2 []int) bool { - l := len(m1) - if l != len(m2) { - return false - } - for i := 0; i < l; i++ { - if m1[i] != m2[i] { - return false - } - } - return true -} - -func equalStrings(m1, m2 []string) bool { - l := len(m1) - if l != len(m2) { - return false - } - for i := 0; i < l; i++ { - if m1[i] != m2[i] { - return false - } - } - return true -} - -func executeTest(t *testing.T, expr string, str string, match []int) { - re := compileTest(t, expr, nil) - if re == nil { - return - } - m := re.ExecuteString(str) - if !equal(m, match) { - t.Errorf("ExecuteString failure on %#q matching %q:", expr, str) - printVec(t, m) - t.Log("should be:") - printVec(t, match) - } - // now try bytes - m = re.Execute([]byte(str)) - if !equal(m, match) { - t.Errorf("Execute failure on %#q matching %q:", expr, str) - printVec(t, m) - t.Log("should be:") - printVec(t, match) - } -} - func TestGoodCompile(t *testing.T) { for i := 0; i < len(good_re); i++ { compileTest(t, good_re[i], nil) @@ -193,57 +71,41 @@ func TestBadCompile(t *testing.T) { } } -func TestExecute(t *testing.T) { - for i := 0; i < len(matches); i++ { - test := &matches[i] - executeTest(t, test.re, test.text, test.match) - } -} - -func matchTest(t *testing.T, expr string, str string, match []int) { - re := compileTest(t, expr, nil) +func matchTest(t *testing.T, test *FindTest) { + re := compileTest(t, test.pat, nil) if re == nil { return } - m := re.MatchString(str) - if m != (len(match) > 0) { - t.Errorf("MatchString failure on %#q matching %q: %t should be %t", expr, str, m, len(match) > 0) + m := re.MatchString(test.text) + if m != (len(test.matches) > 0) { + t.Errorf("MatchString failure on %s: %t should be %t", test, m, len(test.matches) > 0) } // now try bytes - m = re.Match([]byte(str)) - if m != (len(match) > 0) { - t.Errorf("Match failure on %#q matching %q: %t should be %t", expr, str, m, len(match) > 0) + m = re.Match([]byte(test.text)) + if m != (len(test.matches) > 0) { + t.Errorf("Match failure on %s: %t should be %t", test, m, len(test.matches) > 0) } } func TestMatch(t *testing.T) { - for i := 0; i < len(matches); i++ { - test := &matches[i] - matchTest(t, test.re, test.text, test.match) + for _, test := range findTests { + matchTest(t, &test) } } -func TestMatchStrings(t *testing.T) { - for i := 0; i < len(matches); i++ { - test := &matches[i] - matchTest(t, test.re, test.text, test.match) - } -} - -func matchFunctionTest(t *testing.T, expr string, str string, match []int) { - m, err := MatchString(expr, str) +func matchFunctionTest(t *testing.T, test *FindTest) { + m, err := MatchString(test.pat, test.text) if err == nil { return } - if m != (len(match) > 0) { - t.Errorf("Match failure on %#q matching %q: %d should be %d", expr, str, m, len(match) > 0) + if m != (len(test.matches) > 0) { + t.Errorf("Match failure on %s: %t should be %t", test, m, len(test.matches) > 0) } } func TestMatchFunction(t *testing.T) { - for i := 0; i < len(matches); i++ { - test := &matches[i] - matchFunctionTest(t, test.re, test.text, test.match) + for _, test := range findTests { + matchFunctionTest(t, &test) } } @@ -407,100 +269,6 @@ func TestQuoteMeta(t *testing.T) { } } -type matchCase struct { - matchfunc string - input string - n int - regexp string - expected []string -} - -var matchCases = []matchCase{ - matchCase{"match", " aa b", 0, "[^ ]+", []string{"aa", "b"}}, - matchCase{"match", " aa b", 0, "[^ ]*", []string{"", "aa", "b"}}, - matchCase{"match", "a b c", 0, "[^ ]*", []string{"a", "b", "c"}}, - matchCase{"match", "a:a: a:", 0, "^.:", []string{"a:"}}, - matchCase{"match", "", 0, "[^ ]*", []string{""}}, - matchCase{"match", "", 0, "", []string{""}}, - matchCase{"match", "a", 0, "", []string{"", ""}}, - matchCase{"match", "ab", 0, "^", []string{""}}, - matchCase{"match", "ab", 0, "$", []string{""}}, - matchCase{"match", "ab", 0, "X*", []string{"", "", ""}}, - matchCase{"match", "aX", 0, "X*", []string{"", "X"}}, - matchCase{"match", "XabX", 0, "X*", []string{"X", "", "X"}}, - - matchCase{"matchit", "", 0, ".", []string{}}, - matchCase{"matchit", "abc", 2, ".", []string{"a", "b"}}, - matchCase{"matchit", "abc", 0, ".", []string{"a", "b", "c"}}, -} - -func printStringSlice(t *testing.T, s []string) { - t.Logf("%#v", s) -} - -func TestAllMatches(t *testing.T) { - ch := make(chan matchCase) - go func() { - for _, c := range matchCases { - ch <- c - stringCase := matchCase{ - "string" + c.matchfunc, - c.input, - c.n, - c.regexp, - c.expected, - } - ch <- stringCase - } - close(ch) - }() - - for c := range ch { - var result []string - re, _ := Compile(c.regexp) - - switch c.matchfunc { - case "matchit": - result = make([]string, len(c.input)+1) - i := 0 - b := []byte(c.input) - for match := range re.AllMatchesIter(b, c.n) { - result[i] = string(match) - i++ - } - result = result[0:i] - case "stringmatchit": - result = make([]string, len(c.input)+1) - i := 0 - for match := range re.AllMatchesStringIter(c.input, c.n) { - result[i] = match - i++ - } - result = result[0:i] - case "match": - result = make([]string, len(c.input)+1) - b := []byte(c.input) - i := 0 - for _, match := range re.AllMatches(b, c.n) { - result[i] = string(match) - i++ - } - result = result[0:i] - case "stringmatch": - result = re.AllMatchesString(c.input, c.n) - } - - if !equalStrings(result, c.expected) { - t.Errorf("testing '%s'.%s('%s', %d), expected: ", - c.regexp, c.matchfunc, c.input, c.n) - printStringSlice(t, c.expected) - t.Log("got: ") - printStringSlice(t, result) - t.Log("\n") - } - } -} - type numSubexpCase struct { input string expected int diff --git a/src/pkg/regexp/regexp.go b/src/pkg/regexp/regexp.go index 39a6b0cdeed..ea17cfe06f2 100644 --- a/src/pkg/regexp/regexp.go +++ b/src/pkg/regexp/regexp.go @@ -936,81 +936,15 @@ func (re *Regexp) doExecute(str string, bytestr []byte, pos int) []int { return final.match.m } - -// ExecuteString matches the Regexp against the string s. -// The return value is an array of integers, in pairs, identifying the positions of -// substrings matched by the expression. -// s[a[0]:a[1]] is the substring matched by the entire expression. -// s[a[2*i]:a[2*i+1]] for i > 0 is the substring matched by the ith parenthesized subexpression. -// A negative value means the subexpression did not match any element of the string. -// An empty array means "no match". -// Deprecated; use FindString. -func (re *Regexp) ExecuteString(s string) (a []int) { - return re.doExecute(s, nil, 0) -} - - -// Execute matches the Regexp against the byte slice b. -// The return value is an array of integers, in pairs, identifying the positions of -// subslices matched by the expression. -// b[a[0]:a[1]] is the subslice matched by the entire expression. -// b[a[2*i]:a[2*i+1]] for i > 0 is the subslice matched by the ith parenthesized subexpression. -// A negative value means the subexpression did not match any element of the slice. -// An empty array means "no match". -// Deprecated; use Find. -func (re *Regexp) Execute(b []byte) (a []int) { return re.doExecute("", b, 0) } - - // MatchString returns whether the Regexp matches the string s. // The return value is a boolean: true for match, false for no match. func (re *Regexp) MatchString(s string) bool { return len(re.doExecute(s, nil, 0)) > 0 } - // Match returns whether the Regexp matches the byte slice b. // The return value is a boolean: true for match, false for no match. func (re *Regexp) Match(b []byte) bool { return len(re.doExecute("", b, 0)) > 0 } -// MatchStrings matches the Regexp against the string s. -// The return value is an array of strings matched by the expression. -// a[0] is the substring matched by the entire expression. -// a[i] for i > 0 is the substring matched by the ith parenthesized subexpression. -// An empty array means ``no match''. -// Deprecated; use FindStringSubmatch. -func (re *Regexp) MatchStrings(s string) (a []string) { - r := re.doExecute(s, nil, 0) - if r == nil { - return nil - } - a = make([]string, len(r)/2) - for i := 0; i < len(r); i += 2 { - if r[i] != -1 { // -1 means no match for this subexpression - a[i/2] = s[r[i]:r[i+1]] - } - } - return -} - -// MatchSlices matches the Regexp against the byte slice b. -// The return value is an array of subslices matched by the expression. -// a[0] is the subslice matched by the entire expression. -// a[i] for i > 0 is the subslice matched by the ith parenthesized subexpression. -// An empty array means ``no match''. -// Deprecated; use FindSubmatch. -func (re *Regexp) MatchSlices(b []byte) (a [][]byte) { - r := re.doExecute("", b, 0) - if r == nil { - return nil - } - a = make([][]byte, len(r)/2) - for i := 0; i < len(r); i += 2 { - if r[i] != -1 { // -1 means no match for this subexpression - a[i/2] = b[r[i]:r[i+1]] - } - } - return -} - // MatchString checks whether a textual regular expression // matches a string. More complicated queries need // to use Compile and the full Regexp interface. @@ -1202,46 +1136,7 @@ func (re *Regexp) allMatches(s string, b []byte, n int, deliver func([]int)) { } } -// AllMatches slices the byte slice b into substrings that are successive -// matches of the Regexp within b. If n > 0, the function returns at most n -// matches. Text that does not match the expression will be skipped. Empty -// matches abutting a preceding match are ignored. The function returns a slice -// containing the matching substrings. -// Deprecated; use FindAll. -func (re *Regexp) AllMatches(b []byte, n int) [][]byte { - if n <= 0 { - n = len(b) + 1 - } - // TODO: here and below, grow the result slice on demand - // to avoid allocating a huge slice for a small result and - // wasting memory. - result := make([][]byte, n) - i := 0 - re.allMatches("", b, n, func(match []int) { - result[i] = b[match[0]:match[1]] - i++ - }) - return result[0:i] -} - -// AllMatchesString slices the string s into substrings that are successive -// matches of the Regexp within s. If n > 0, the function returns at most n -// matches. Text that does not match the expression will be skipped. Empty -// matches abutting a preceding match are ignored. The function returns a slice -// containing the matching substrings. -// Deprecated; use FindAllString. -func (re *Regexp) AllMatchesString(s string, n int) []string { - if n <= 0 { - n = len(s) + 1 - } - result := make([]string, n) - i := 0 - re.allMatches(s, nil, n, func(match []int) { - result[i] = s[match[0]:match[1]] - i++ - }) - return result[0:i] -} +// TODO: AllMatchesIter and AllMatchesStringIter should change to return submatches as well. // AllMatchesIter slices the byte slice b into substrings that are successive // matches of the Regexp within b. If n > 0, the function returns at most n @@ -1380,6 +1275,9 @@ func (re *Regexp) FindStringSubmatchIndex(s string) []int { return re.doExecute(s, nil, 0) } +// TODO: The routines that return slices of unpredictable length should +// grow them on demand rather than just allocating them maximum-length. + // FindAll is the 'All' version of Find; it returns a slice of all successive // matches of the expression, as defined by the 'All' description in the // package comment.