Source file src/go/scanner/scanner_test.go

     1  // Copyright 2009 The Go Authors. All rights reserved.
     2  // Use of this source code is governed by a BSD-style
     3  // license that can be found in the LICENSE file.
     4  
     5  package scanner
     6  
     7  import (
     8  	"go/token"
     9  	"os"
    10  	"path/filepath"
    11  	"runtime"
    12  	"strings"
    13  	"testing"
    14  )
    15  
    16  var fset = token.NewFileSet()
    17  
    18  const /* class */ (
    19  	special = iota
    20  	literal
    21  	operator
    22  	keyword
    23  )
    24  
    25  func tokenclass(tok token.Token) int {
    26  	switch {
    27  	case tok.IsLiteral():
    28  		return literal
    29  	case tok.IsOperator():
    30  		return operator
    31  	case tok.IsKeyword():
    32  		return keyword
    33  	}
    34  	return special
    35  }
    36  
    37  type elt struct {
    38  	tok   token.Token
    39  	lit   string
    40  	class int
    41  }
    42  
    43  var tokens = []elt{
    44  	// Special tokens
    45  	{token.COMMENT, "/* a comment */", special},
    46  	{token.COMMENT, "// a comment \n", special},
    47  	{token.COMMENT, "/*\r*/", special},
    48  	{token.COMMENT, "/**\r/*/", special}, // issue 11151
    49  	{token.COMMENT, "/**\r\r/*/", special},
    50  	{token.COMMENT, "//\r\n", special},
    51  
    52  	// Identifiers and basic type literals
    53  	{token.IDENT, "foobar", literal},
    54  	{token.IDENT, "a۰۱۸", literal},
    55  	{token.IDENT, "foo६४", literal},
    56  	{token.IDENT, "bar９８７６", literal},
    57  	{token.IDENT, "ŝ", literal},    // was bug (issue 4000)
    58  	{token.IDENT, "ŝfoo", literal}, // was bug (issue 4000)
    59  	{token.INT, "0", literal},
    60  	{token.INT, "1", literal},
    61  	{token.INT, "123456789012345678890", literal},
    62  	{token.INT, "01234567", literal},
    63  	{token.INT, "0xcafebabe", literal},
    64  	{token.FLOAT, "0.", literal},
    65  	{token.FLOAT, ".0", literal},
    66  	{token.FLOAT, "3.14159265", literal},
    67  	{token.FLOAT, "1e0", literal},
    68  	{token.FLOAT, "1e+100", literal},
    69  	{token.FLOAT, "1e-100", literal},
    70  	{token.FLOAT, "2.71828e-1000", literal},
    71  	{token.IMAG, "0i", literal},
    72  	{token.IMAG, "1i", literal},
    73  	{token.IMAG, "012345678901234567889i", literal},
    74  	{token.IMAG, "123456789012345678890i", literal},
    75  	{token.IMAG, "0.i", literal},
    76  	{token.IMAG, ".0i", literal},
    77  	{token.IMAG, "3.14159265i", literal},
    78  	{token.IMAG, "1e0i", literal},
    79  	{token.IMAG, "1e+100i", literal},
    80  	{token.IMAG, "1e-100i", literal},
    81  	{token.IMAG, "2.71828e-1000i", literal},
    82  	{token.CHAR, "'a'", literal},
    83  	{token.CHAR, "'\\000'", literal},
    84  	{token.CHAR, "'\\xFF'", literal},
    85  	{token.CHAR, "'\\uff16'", literal},
    86  	{token.CHAR, "'\\U0000ff16'", literal},
    87  	{token.STRING, "`foobar`", literal},
    88  	{token.STRING, "`" + `foo
    89  	                        bar` +
    90  		"`",
    91  		literal,
    92  	},
    93  	{token.STRING, "`\r`", literal},
    94  	{token.STRING, "`foo\r\nbar`", literal},
    95  
    96  	// Operators and delimiters
    97  	{token.ADD, "+", operator},
    98  	{token.SUB, "-", operator},
    99  	{token.MUL, "*", operator},
   100  	{token.QUO, "/", operator},
   101  	{token.REM, "%", operator},
   102  
   103  	{token.AND, "&", operator},
   104  	{token.OR, "|", operator},
   105  	{token.XOR, "^", operator},
   106  	{token.SHL, "<<", operator},
   107  	{token.SHR, ">>", operator},
   108  	{token.AND_NOT, "&^", operator},
   109  
   110  	{token.ADD_ASSIGN, "+=", operator},
   111  	{token.SUB_ASSIGN, "-=", operator},
   112  	{token.MUL_ASSIGN, "*=", operator},
   113  	{token.QUO_ASSIGN, "/=", operator},
   114  	{token.REM_ASSIGN, "%=", operator},
   115  
   116  	{token.AND_ASSIGN, "&=", operator},
   117  	{token.OR_ASSIGN, "|=", operator},
   118  	{token.XOR_ASSIGN, "^=", operator},
   119  	{token.SHL_ASSIGN, "<<=", operator},
   120  	{token.SHR_ASSIGN, ">>=", operator},
   121  	{token.AND_NOT_ASSIGN, "&^=", operator},
   122  
   123  	{token.LAND, "&&", operator},
   124  	{token.LOR, "||", operator},
   125  	{token.ARROW, "<-", operator},
   126  	{token.INC, "++", operator},
   127  	{token.DEC, "--", operator},
   128  
   129  	{token.EQL, "==", operator},
   130  	{token.LSS, "<", operator},
   131  	{token.GTR, ">", operator},
   132  	{token.ASSIGN, "=", operator},
   133  	{token.NOT, "!", operator},
   134  
   135  	{token.NEQ, "!=", operator},
   136  	{token.LEQ, "<=", operator},
   137  	{token.GEQ, ">=", operator},
   138  	{token.DEFINE, ":=", operator},
   139  	{token.ELLIPSIS, "...", operator},
   140  
   141  	{token.LPAREN, "(", operator},
   142  	{token.LBRACK, "[", operator},
   143  	{token.LBRACE, "{", operator},
   144  	{token.COMMA, ",", operator},
   145  	{token.PERIOD, ".", operator},
   146  
   147  	{token.RPAREN, ")", operator},
   148  	{token.RBRACK, "]", operator},
   149  	{token.RBRACE, "}", operator},
   150  	{token.SEMICOLON, ";", operator},
   151  	{token.COLON, ":", operator},
   152  	{token.TILDE, "~", operator},
   153  
   154  	// Keywords
   155  	{token.BREAK, "break", keyword},
   156  	{token.CASE, "case", keyword},
   157  	{token.CHAN, "chan", keyword},
   158  	{token.CONST, "const", keyword},
   159  	{token.CONTINUE, "continue", keyword},
   160  
   161  	{token.DEFAULT, "default", keyword},
   162  	{token.DEFER, "defer", keyword},
   163  	{token.ELSE, "else", keyword},
   164  	{token.FALLTHROUGH, "fallthrough", keyword},
   165  	{token.FOR, "for", keyword},
   166  
   167  	{token.FUNC, "func", keyword},
   168  	{token.GO, "go", keyword},
   169  	{token.GOTO, "goto", keyword},
   170  	{token.IF, "if", keyword},
   171  	{token.IMPORT, "import", keyword},
   172  
   173  	{token.INTERFACE, "interface", keyword},
   174  	{token.MAP, "map", keyword},
   175  	{token.PACKAGE, "package", keyword},
   176  	{token.RANGE, "range", keyword},
   177  	{token.RETURN, "return", keyword},
   178  
   179  	{token.SELECT, "select", keyword},
   180  	{token.STRUCT, "struct", keyword},
   181  	{token.SWITCH, "switch", keyword},
   182  	{token.TYPE, "type", keyword},
   183  	{token.VAR, "var", keyword},
   184  }
   185  
   186  const whitespace = "  \t  \n\n\n" // to separate tokens
   187  
   188  var source = func() []byte {
   189  	var src []byte
   190  	for _, t := range tokens {
   191  		src = append(src, t.lit...)
   192  		src = append(src, whitespace...)
   193  	}
   194  	return src
   195  }()
   196  
   197  func newlineCount(s string) int {
   198  	n := 0
   199  	for i := 0; i < len(s); i++ {
   200  		if s[i] == '\n' {
   201  			n++
   202  		}
   203  	}
   204  	return n
   205  }
   206  
   207  func checkPos(t *testing.T, lit string, p token.Pos, expected token.Position) {
   208  	pos := fset.Position(p)
   209  	// Check cleaned filenames so that we don't have to worry about
   210  	// different os.PathSeparator values.
   211  	if pos.Filename != expected.Filename && filepath.Clean(pos.Filename) != filepath.Clean(expected.Filename) {
   212  		t.Errorf("bad filename for %q: got %s, expected %s", lit, pos.Filename, expected.Filename)
   213  	}
   214  	if pos.Offset != expected.Offset {
   215  		t.Errorf("bad position for %q: got %d, expected %d", lit, pos.Offset, expected.Offset)
   216  	}
   217  	if pos.Line != expected.Line {
   218  		t.Errorf("bad line for %q: got %d, expected %d", lit, pos.Line, expected.Line)
   219  	}
   220  	if pos.Column != expected.Column {
   221  		t.Errorf("bad column for %q: got %d, expected %d", lit, pos.Column, expected.Column)
   222  	}
   223  }
   224  
   225  // Verify that calling Scan() provides the correct results.
   226  func TestScan(t *testing.T) {
   227  	whitespace_linecount := newlineCount(whitespace)
   228  
   229  	// error handler
   230  	eh := func(_ token.Position, msg string) {
   231  		t.Errorf("error handler called (msg = %s)", msg)
   232  	}
   233  
   234  	// verify scan
   235  	var s Scanner
   236  	s.Init(fset.AddFile("", fset.Base(), len(source)), source, eh, ScanComments|dontInsertSemis)
   237  
   238  	// set up expected position
   239  	epos := token.Position{
   240  		Filename: "",
   241  		Offset:   0,
   242  		Line:     1,
   243  		Column:   1,
   244  	}
   245  
   246  	index := 0
   247  	for {
   248  		pos, tok, lit := s.Scan()
   249  
   250  		// check position
   251  		if tok == token.EOF {
   252  			// correction for EOF
   253  			epos.Line = newlineCount(string(source))
   254  			epos.Column = 2
   255  		}
   256  		checkPos(t, lit, pos, epos)
   257  
   258  		// check token
   259  		e := elt{token.EOF, "", special}
   260  		if index < len(tokens) {
   261  			e = tokens[index]
   262  			index++
   263  		}
   264  		if tok != e.tok {
   265  			t.Errorf("bad token for %q: got %s, expected %s", lit, tok, e.tok)
   266  		}
   267  
   268  		// check token class
   269  		if tokenclass(tok) != e.class {
   270  			t.Errorf("bad class for %q: got %d, expected %d", lit, tokenclass(tok), e.class)
   271  		}
   272  
   273  		// check literal
   274  		elit := ""
   275  		switch e.tok {
   276  		case token.COMMENT:
   277  			// no CRs in comments
   278  			elit = string(stripCR([]byte(e.lit), e.lit[1] == '*'))
   279  			//-style comment literal doesn't contain newline
   280  			if elit[1] == '/' {
   281  				elit = elit[0 : len(elit)-1]
   282  			}
   283  		case token.IDENT:
   284  			elit = e.lit
   285  		case token.SEMICOLON:
   286  			elit = ";"
   287  		default:
   288  			if e.tok.IsLiteral() {
   289  				// no CRs in raw string literals
   290  				elit = e.lit
   291  				if elit[0] == '`' {
   292  					elit = string(stripCR([]byte(elit), false))
   293  				}
   294  			} else if e.tok.IsKeyword() {
   295  				elit = e.lit
   296  			}
   297  		}
   298  		if lit != elit {
   299  			t.Errorf("bad literal for %q: got %q, expected %q", lit, lit, elit)
   300  		}
   301  
   302  		if tok == token.EOF {
   303  			break
   304  		}
   305  
   306  		// update position
   307  		epos.Offset += len(e.lit) + len(whitespace)
   308  		epos.Line += newlineCount(e.lit) + whitespace_linecount
   309  
   310  	}
   311  
   312  	if s.ErrorCount != 0 {
   313  		t.Errorf("found %d errors", s.ErrorCount)
   314  	}
   315  }
   316  
   317  func TestStripCR(t *testing.T) {
   318  	for _, test := range []struct{ have, want string }{
   319  		{"//\n", "//\n"},
   320  		{"//\r\n", "//\n"},
   321  		{"//\r\r\r\n", "//\n"},
   322  		{"//\r*\r/\r\n", "//*/\n"},
   323  		{"/**/", "/**/"},
   324  		{"/*\r/*/", "/*/*/"},
   325  		{"/*\r*/", "/**/"},
   326  		{"/**\r/*/", "/**\r/*/"},
   327  		{"/*\r/\r*\r/*/", "/*/*\r/*/"},
   328  		{"/*\r\r\r\r*/", "/**/"},
   329  	} {
   330  		got := string(stripCR([]byte(test.have), len(test.have) >= 2 && test.have[1] == '*'))
   331  		if got != test.want {
   332  			t.Errorf("stripCR(%q) = %q; want %q", test.have, got, test.want)
   333  		}
   334  	}
   335  }
   336  
   337  func checkSemi(t *testing.T, line string, mode Mode) {
   338  	var S Scanner
   339  	file := fset.AddFile("TestSemis", fset.Base(), len(line))
   340  	S.Init(file, []byte(line), nil, mode)
   341  	pos, tok, lit := S.Scan()
   342  	for tok != token.EOF {
   343  		if tok == token.ILLEGAL {
   344  			// the illegal token literal indicates what
   345  			// kind of semicolon literal to expect
   346  			semiLit := "\n"
   347  			if lit[0] == '#' {
   348  				semiLit = ";"
   349  			}
   350  			// next token must be a semicolon
   351  			semiPos := file.Position(pos)
   352  			semiPos.Offset++
   353  			semiPos.Column++
   354  			pos, tok, lit = S.Scan()
   355  			if tok == token.SEMICOLON {
   356  				if lit != semiLit {
   357  					t.Errorf(`bad literal for %q: got %q, expected %q`, line, lit, semiLit)
   358  				}
   359  				checkPos(t, line, pos, semiPos)
   360  			} else {
   361  				t.Errorf("bad token for %q: got %s, expected ;", line, tok)
   362  			}
   363  		} else if tok == token.SEMICOLON {
   364  			t.Errorf("bad token for %q: got ;, expected no ;", line)
   365  		}
   366  		pos, tok, lit = S.Scan()
   367  	}
   368  }
   369  
   370  var lines = []string{
   371  	// # indicates a semicolon present in the source
   372  	// $ indicates an automatically inserted semicolon
   373  	"",
   374  	"\ufeff#;", // first BOM is ignored
   375  	"#;",
   376  	"foo$\n",
   377  	"123$\n",
   378  	"1.2$\n",
   379  	"'x'$\n",
   380  	`"x"` + "$\n",
   381  	"`x`$\n",
   382  
   383  	"+\n",
   384  	"-\n",
   385  	"*\n",
   386  	"/\n",
   387  	"%\n",
   388  
   389  	"&\n",
   390  	"|\n",
   391  	"^\n",
   392  	"<<\n",
   393  	">>\n",
   394  	"&^\n",
   395  
   396  	"+=\n",
   397  	"-=\n",
   398  	"*=\n",
   399  	"/=\n",
   400  	"%=\n",
   401  
   402  	"&=\n",
   403  	"|=\n",
   404  	"^=\n",
   405  	"<<=\n",
   406  	">>=\n",
   407  	"&^=\n",
   408  
   409  	"&&\n",
   410  	"||\n",
   411  	"<-\n",
   412  	"++$\n",
   413  	"--$\n",
   414  
   415  	"==\n",
   416  	"<\n",
   417  	">\n",
   418  	"=\n",
   419  	"!\n",
   420  
   421  	"!=\n",
   422  	"<=\n",
   423  	">=\n",
   424  	":=\n",
   425  	"...\n",
   426  
   427  	"(\n",
   428  	"[\n",
   429  	"{\n",
   430  	",\n",
   431  	".\n",
   432  
   433  	")$\n",
   434  	"]$\n",
   435  	"}$\n",
   436  	"#;\n",
   437  	":\n",
   438  
   439  	"break$\n",
   440  	"case\n",
   441  	"chan\n",
   442  	"const\n",
   443  	"continue$\n",
   444  
   445  	"default\n",
   446  	"defer\n",
   447  	"else\n",
   448  	"fallthrough$\n",
   449  	"for\n",
   450  
   451  	"func\n",
   452  	"go\n",
   453  	"goto\n",
   454  	"if\n",
   455  	"import\n",
   456  
   457  	"interface\n",
   458  	"map\n",
   459  	"package\n",
   460  	"range\n",
   461  	"return$\n",
   462  
   463  	"select\n",
   464  	"struct\n",
   465  	"switch\n",
   466  	"type\n",
   467  	"var\n",
   468  
   469  	"foo$//comment\n",
   470  	"foo$//comment",
   471  	"foo$/*comment*/\n",
   472  	"foo$/*\n*/",
   473  	"foo$/*comment*/    \n",
   474  	"foo$/*\n*/    ",
   475  
   476  	"foo    $// comment\n",
   477  	"foo    $// comment",
   478  	"foo    $/*comment*/\n",
   479  	"foo    $/*\n*/",
   480  	"foo    $/*  */ /* \n */ bar$/**/\n",
   481  	"foo    $/*0*/ /*1*/ /*2*/\n",
   482  
   483  	"foo    $/*comment*/    \n",
   484  	"foo    $/*0*/ /*1*/ /*2*/    \n",
   485  	"foo	$/**/ /*-------------*/       /*----\n*/bar       $/*  \n*/baa$\n",
   486  	"foo    $/* an EOF terminates a line */",
   487  	"foo    $/* an EOF terminates a line */ /*",
   488  	"foo    $/* an EOF terminates a line */ //",
   489  
   490  	"package main$\n\nfunc main() {\n\tif {\n\t\treturn /* */ }$\n}$\n",
   491  	"package main$",
   492  }
   493  
   494  func TestSemis(t *testing.T) {
   495  	for _, line := range lines {
   496  		checkSemi(t, line, 0)
   497  		checkSemi(t, line, ScanComments)
   498  
   499  		// if the input ended in newlines, the input must tokenize the
   500  		// same with or without those newlines
   501  		for i := len(line) - 1; i >= 0 && line[i] == '\n'; i-- {
   502  			checkSemi(t, line[0:i], 0)
   503  			checkSemi(t, line[0:i], ScanComments)
   504  		}
   505  	}
   506  }
   507  
   508  type segment struct {
   509  	srcline      string // a line of source text
   510  	filename     string // filename for current token; error message for invalid line directives
   511  	line, column int    // line and column for current token; error position for invalid line directives
   512  }
   513  
   514  var segments = []segment{
   515  	// exactly one token per line since the test consumes one token per segment
   516  	{"  line1", "TestLineDirectives", 1, 3},
   517  	{"\nline2", "TestLineDirectives", 2, 1},
   518  	{"\nline3  //line File1.go:100", "TestLineDirectives", 3, 1}, // bad line comment, ignored
   519  	{"\nline4", "TestLineDirectives", 4, 1},
   520  	{"\n//line File1.go:100\n  line100", "File1.go", 100, 0},
   521  	{"\n//line  \t :42\n  line1", " \t ", 42, 0},
   522  	{"\n//line File2.go:200\n  line200", "File2.go", 200, 0},
   523  	{"\n//line foo\t:42\n  line42", "foo\t", 42, 0},
   524  	{"\n //line foo:42\n  line43", "foo\t", 44, 0}, // bad line comment, ignored (use existing, prior filename)
   525  	{"\n//line foo 42\n  line44", "foo\t", 46, 0},  // bad line comment, ignored (use existing, prior filename)
   526  	{"\n//line /bar:42\n  line45", "/bar", 42, 0},
   527  	{"\n//line ./foo:42\n  line46", "foo", 42, 0},
   528  	{"\n//line a/b/c/File1.go:100\n  line100", "a/b/c/File1.go", 100, 0},
   529  	{"\n//line c:\\bar:42\n  line200", "c:\\bar", 42, 0},
   530  	{"\n//line c:\\dir\\File1.go:100\n  line201", "c:\\dir\\File1.go", 100, 0},
   531  
   532  	// tests for new line directive syntax
   533  	{"\n//line :100\na1", "", 100, 0}, // missing filename means empty filename
   534  	{"\n//line bar:100\nb1", "bar", 100, 0},
   535  	{"\n//line :100:10\nc1", "bar", 100, 10}, // missing filename means current filename
   536  	{"\n//line foo:100:10\nd1", "foo", 100, 10},
   537  
   538  	{"\n/*line :100*/a2", "", 100, 0}, // missing filename means empty filename
   539  	{"\n/*line bar:100*/b2", "bar", 100, 0},
   540  	{"\n/*line :100:10*/c2", "bar", 100, 10}, // missing filename means current filename
   541  	{"\n/*line foo:100:10*/d2", "foo", 100, 10},
   542  	{"\n/*line foo:100:10*/    e2", "foo", 100, 14}, // line-directive relative column
   543  	{"\n/*line foo:100:10*/\n\nf2", "foo", 102, 1},  // absolute column since on new line
   544  }
   545  
   546  var dirsegments = []segment{
   547  	// exactly one token per line since the test consumes one token per segment
   548  	{"  line1", "TestLineDir/TestLineDirectives", 1, 3},
   549  	{"\n//line File1.go:100\n  line100", "TestLineDir/File1.go", 100, 0},
   550  }
   551  
   552  var dirUnixSegments = []segment{
   553  	{"\n//line /bar:42\n  line42", "/bar", 42, 0},
   554  }
   555  
   556  var dirWindowsSegments = []segment{
   557  	{"\n//line c:\\bar:42\n  line42", "c:\\bar", 42, 0},
   558  }
   559  
   560  // Verify that line directives are interpreted correctly.
   561  func TestLineDirectives(t *testing.T) {
   562  	testSegments(t, segments, "TestLineDirectives")
   563  	testSegments(t, dirsegments, "TestLineDir/TestLineDirectives")
   564  	if runtime.GOOS == "windows" {
   565  		testSegments(t, dirWindowsSegments, "TestLineDir/TestLineDirectives")
   566  	} else {
   567  		testSegments(t, dirUnixSegments, "TestLineDir/TestLineDirectives")
   568  	}
   569  }
   570  
   571  func testSegments(t *testing.T, segments []segment, filename string) {
   572  	var src string
   573  	for _, e := range segments {
   574  		src += e.srcline
   575  	}
   576  
   577  	// verify scan
   578  	var S Scanner
   579  	file := fset.AddFile(filename, fset.Base(), len(src))
   580  	S.Init(file, []byte(src), func(pos token.Position, msg string) { t.Error(Error{pos, msg}) }, dontInsertSemis)
   581  	for _, s := range segments {
   582  		p, _, lit := S.Scan()
   583  		pos := file.Position(p)
   584  		checkPos(t, lit, p, token.Position{
   585  			Filename: s.filename,
   586  			Offset:   pos.Offset,
   587  			Line:     s.line,
   588  			Column:   s.column,
   589  		})
   590  	}
   591  
   592  	if S.ErrorCount != 0 {
   593  		t.Errorf("got %d errors", S.ErrorCount)
   594  	}
   595  }
   596  
   597  // The filename is used for the error message in these test cases.
   598  // The first line directive is valid and used to control the expected error line.
   599  var invalidSegments = []segment{
   600  	{"\n//line :1:1\n//line foo:42 extra text\ndummy", "invalid line number: 42 extra text", 1, 12},
   601  	{"\n//line :2:1\n//line foobar:\ndummy", "invalid line number: ", 2, 15},
   602  	{"\n//line :5:1\n//line :0\ndummy", "invalid line number: 0", 5, 9},
   603  	{"\n//line :10:1\n//line :1:0\ndummy", "invalid column number: 0", 10, 11},
   604  	{"\n//line :1:1\n//line :foo:0\ndummy", "invalid line number: 0", 1, 13}, // foo is considered part of the filename
   605  }
   606  
   607  // Verify that invalid line directives get the correct error message.
   608  func TestInvalidLineDirectives(t *testing.T) {
   609  	// make source
   610  	var src string
   611  	for _, e := range invalidSegments {
   612  		src += e.srcline
   613  	}
   614  
   615  	// verify scan
   616  	var S Scanner
   617  	var s segment // current segment
   618  	file := fset.AddFile(filepath.Join("dir", "TestInvalidLineDirectives"), fset.Base(), len(src))
   619  	S.Init(file, []byte(src), func(pos token.Position, msg string) {
   620  		if msg != s.filename {
   621  			t.Errorf("got error %q; want %q", msg, s.filename)
   622  		}
   623  		if pos.Line != s.line || pos.Column != s.column {
   624  			t.Errorf("got position %d:%d; want %d:%d", pos.Line, pos.Column, s.line, s.column)
   625  		}
   626  	}, dontInsertSemis)
   627  	for _, s = range invalidSegments {
   628  		S.Scan()
   629  	}
   630  
   631  	if S.ErrorCount != len(invalidSegments) {
   632  		t.Errorf("got %d errors; want %d", S.ErrorCount, len(invalidSegments))
   633  	}
   634  }
   635  
   636  // Verify that initializing the same scanner more than once works correctly.
   637  func TestInit(t *testing.T) {
   638  	var s Scanner
   639  
   640  	// 1st init
   641  	src1 := "if true { }"
   642  	f1 := fset.AddFile("src1", fset.Base(), len(src1))
   643  	s.Init(f1, []byte(src1), nil, dontInsertSemis)
   644  	if f1.Size() != len(src1) {
   645  		t.Errorf("bad file size: got %d, expected %d", f1.Size(), len(src1))
   646  	}
   647  	s.Scan()              // if
   648  	s.Scan()              // true
   649  	_, tok, _ := s.Scan() // {
   650  	if tok != token.LBRACE {
   651  		t.Errorf("bad token: got %s, expected %s", tok, token.LBRACE)
   652  	}
   653  
   654  	// 2nd init
   655  	src2 := "go true { ]"
   656  	f2 := fset.AddFile("src2", fset.Base(), len(src2))
   657  	s.Init(f2, []byte(src2), nil, dontInsertSemis)
   658  	if f2.Size() != len(src2) {
   659  		t.Errorf("bad file size: got %d, expected %d", f2.Size(), len(src2))
   660  	}
   661  	_, tok, _ = s.Scan() // go
   662  	if tok != token.GO {
   663  		t.Errorf("bad token: got %s, expected %s", tok, token.GO)
   664  	}
   665  
   666  	if s.ErrorCount != 0 {
   667  		t.Errorf("found %d errors", s.ErrorCount)
   668  	}
   669  }
   670  
   671  func TestStdErrorHander(t *testing.T) {
   672  	const src = "@\n" + // illegal character, cause an error
   673  		"@ @\n" + // two errors on the same line
   674  		"//line File2:20\n" +
   675  		"@\n" + // different file, but same line
   676  		"//line File2:1\n" +
   677  		"@ @\n" + // same file, decreasing line number
   678  		"//line File1:1\n" +
   679  		"@ @ @" // original file, line 1 again
   680  
   681  	var list ErrorList
   682  	eh := func(pos token.Position, msg string) { list.Add(pos, msg) }
   683  
   684  	var s Scanner
   685  	s.Init(fset.AddFile("File1", fset.Base(), len(src)), []byte(src), eh, dontInsertSemis)
   686  	for {
   687  		if _, tok, _ := s.Scan(); tok == token.EOF {
   688  			break
   689  		}
   690  	}
   691  
   692  	if len(list) != s.ErrorCount {
   693  		t.Errorf("found %d errors, expected %d", len(list), s.ErrorCount)
   694  	}
   695  
   696  	if len(list) != 9 {
   697  		t.Errorf("found %d raw errors, expected 9", len(list))
   698  		PrintError(os.Stderr, list)
   699  	}
   700  
   701  	list.Sort()
   702  	if len(list) != 9 {
   703  		t.Errorf("found %d sorted errors, expected 9", len(list))
   704  		PrintError(os.Stderr, list)
   705  	}
   706  
   707  	list.RemoveMultiples()
   708  	if len(list) != 4 {
   709  		t.Errorf("found %d one-per-line errors, expected 4", len(list))
   710  		PrintError(os.Stderr, list)
   711  	}
   712  }
   713  
   714  type errorCollector struct {
   715  	cnt int            // number of errors encountered
   716  	msg string         // last error message encountered
   717  	pos token.Position // last error position encountered
   718  }
   719  
   720  func checkError(t *testing.T, src string, tok token.Token, pos int, lit, err string) {
   721  	var s Scanner
   722  	var h errorCollector
   723  	eh := func(pos token.Position, msg string) {
   724  		h.cnt++
   725  		h.msg = msg
   726  		h.pos = pos
   727  	}
   728  	s.Init(fset.AddFile("", fset.Base(), len(src)), []byte(src), eh, ScanComments|dontInsertSemis)
   729  	_, tok0, lit0 := s.Scan()
   730  	if tok0 != tok {
   731  		t.Errorf("%q: got %s, expected %s", src, tok0, tok)
   732  	}
   733  	if tok0 != token.ILLEGAL && lit0 != lit {
   734  		t.Errorf("%q: got literal %q, expected %q", src, lit0, lit)
   735  	}
   736  	cnt := 0
   737  	if err != "" {
   738  		cnt = 1
   739  	}
   740  	if h.cnt != cnt {
   741  		t.Errorf("%q: got cnt %d, expected %d", src, h.cnt, cnt)
   742  	}
   743  	if h.msg != err {
   744  		t.Errorf("%q: got msg %q, expected %q", src, h.msg, err)
   745  	}
   746  	if h.pos.Offset != pos {
   747  		t.Errorf("%q: got offset %d, expected %d", src, h.pos.Offset, pos)
   748  	}
   749  }
   750  
   751  var errors = []struct {
   752  	src string
   753  	tok token.Token
   754  	pos int
   755  	lit string
   756  	err string
   757  }{
   758  	{"\a", token.ILLEGAL, 0, "", "illegal character U+0007"},
   759  	{`#`, token.ILLEGAL, 0, "", "illegal character U+0023 '#'"},
   760  	{`…`, token.ILLEGAL, 0, "", "illegal character U+2026 '…'"},
   761  	{"..", token.PERIOD, 0, "", ""}, // two periods, not invalid token (issue #28112)
   762  	{`' '`, token.CHAR, 0, `' '`, ""},
   763  	{`''`, token.CHAR, 0, `''`, "illegal rune literal"},
   764  	{`'12'`, token.CHAR, 0, `'12'`, "illegal rune literal"},
   765  	{`'123'`, token.CHAR, 0, `'123'`, "illegal rune literal"},
   766  	{`'\0'`, token.CHAR, 3, `'\0'`, "illegal character U+0027 ''' in escape sequence"},
   767  	{`'\07'`, token.CHAR, 4, `'\07'`, "illegal character U+0027 ''' in escape sequence"},
   768  	{`'\8'`, token.CHAR, 2, `'\8'`, "unknown escape sequence"},
   769  	{`'\08'`, token.CHAR, 3, `'\08'`, "illegal character U+0038 '8' in escape sequence"},
   770  	{`'\x'`, token.CHAR, 3, `'\x'`, "illegal character U+0027 ''' in escape sequence"},
   771  	{`'\x0'`, token.CHAR, 4, `'\x0'`, "illegal character U+0027 ''' in escape sequence"},
   772  	{`'\x0g'`, token.CHAR, 4, `'\x0g'`, "illegal character U+0067 'g' in escape sequence"},
   773  	{`'\u'`, token.CHAR, 3, `'\u'`, "illegal character U+0027 ''' in escape sequence"},
   774  	{`'\u0'`, token.CHAR, 4, `'\u0'`, "illegal character U+0027 ''' in escape sequence"},
   775  	{`'\u00'`, token.CHAR, 5, `'\u00'`, "illegal character U+0027 ''' in escape sequence"},
   776  	{`'\u000'`, token.CHAR, 6, `'\u000'`, "illegal character U+0027 ''' in escape sequence"},
   777  	{`'\u000`, token.CHAR, 6, `'\u000`, "escape sequence not terminated"},
   778  	{`'\u0000'`, token.CHAR, 0, `'\u0000'`, ""},
   779  	{`'\U'`, token.CHAR, 3, `'\U'`, "illegal character U+0027 ''' in escape sequence"},
   780  	{`'\U0'`, token.CHAR, 4, `'\U0'`, "illegal character U+0027 ''' in escape sequence"},
   781  	{`'\U00'`, token.CHAR, 5, `'\U00'`, "illegal character U+0027 ''' in escape sequence"},
   782  	{`'\U000'`, token.CHAR, 6, `'\U000'`, "illegal character U+0027 ''' in escape sequence"},
   783  	{`'\U0000'`, token.CHAR, 7, `'\U0000'`, "illegal character U+0027 ''' in escape sequence"},
   784  	{`'\U00000'`, token.CHAR, 8, `'\U00000'`, "illegal character U+0027 ''' in escape sequence"},
   785  	{`'\U000000'`, token.CHAR, 9, `'\U000000'`, "illegal character U+0027 ''' in escape sequence"},
   786  	{`'\U0000000'`, token.CHAR, 10, `'\U0000000'`, "illegal character U+0027 ''' in escape sequence"},
   787  	{`'\U0000000`, token.CHAR, 10, `'\U0000000`, "escape sequence not terminated"},
   788  	{`'\U00000000'`, token.CHAR, 0, `'\U00000000'`, ""},
   789  	{`'\Uffffffff'`, token.CHAR, 2, `'\Uffffffff'`, "escape sequence is invalid Unicode code point"},
   790  	{`'`, token.CHAR, 0, `'`, "rune literal not terminated"},
   791  	{`'\`, token.CHAR, 2, `'\`, "escape sequence not terminated"},
   792  	{"'\n", token.CHAR, 0, "'", "rune literal not terminated"},
   793  	{"'\n   ", token.CHAR, 0, "'", "rune literal not terminated"},
   794  	{`""`, token.STRING, 0, `""`, ""},
   795  	{`"abc`, token.STRING, 0, `"abc`, "string literal not terminated"},
   796  	{"\"abc\n", token.STRING, 0, `"abc`, "string literal not terminated"},
   797  	{"\"abc\n   ", token.STRING, 0, `"abc`, "string literal not terminated"},
   798  	{"``", token.STRING, 0, "``", ""},
   799  	{"`", token.STRING, 0, "`", "raw string literal not terminated"},
   800  	{"/**/", token.COMMENT, 0, "/**/", ""},
   801  	{"/*", token.COMMENT, 0, "/*", "comment not terminated"},
   802  	{"077", token.INT, 0, "077", ""},
   803  	{"078.", token.FLOAT, 0, "078.", ""},
   804  	{"07801234567.", token.FLOAT, 0, "07801234567.", ""},
   805  	{"078e0", token.FLOAT, 0, "078e0", ""},
   806  	{"0E", token.FLOAT, 2, "0E", "exponent has no digits"}, // issue 17621
   807  	{"078", token.INT, 2, "078", "invalid digit '8' in octal literal"},
   808  	{"07090000008", token.INT, 3, "07090000008", "invalid digit '9' in octal literal"},
   809  	{"0x", token.INT, 2, "0x", "hexadecimal literal has no digits"},
   810  	{"\"abc\x00def\"", token.STRING, 4, "\"abc\x00def\"", "illegal character NUL"},
   811  	{"\"abc\x80def\"", token.STRING, 4, "\"abc\x80def\"", "illegal UTF-8 encoding"},
   812  	{"\ufeff\ufeff", token.ILLEGAL, 3, "\ufeff\ufeff", "illegal byte order mark"},                        // only first BOM is ignored
   813  	{"//\ufeff", token.COMMENT, 2, "//\ufeff", "illegal byte order mark"},                                // only first BOM is ignored
   814  	{"'\ufeff" + `'`, token.CHAR, 1, "'\ufeff" + `'`, "illegal byte order mark"},                         // only first BOM is ignored
   815  	{`"` + "abc\ufeffdef" + `"`, token.STRING, 4, `"` + "abc\ufeffdef" + `"`, "illegal byte order mark"}, // only first BOM is ignored
   816  	{"abc\x00def", token.IDENT, 3, "abc", "illegal character NUL"},
   817  	{"abc\x00", token.IDENT, 3, "abc", "illegal character NUL"},
   818  }
   819  
   820  func TestScanErrors(t *testing.T) {
   821  	for _, e := range errors {
   822  		checkError(t, e.src, e.tok, e.pos, e.lit, e.err)
   823  	}
   824  }
   825  
   826  // Verify that no comments show up as literal values when skipping comments.
   827  func TestIssue10213(t *testing.T) {
   828  	const src = `
   829  		var (
   830  			A = 1 // foo
   831  		)
   832  
   833  		var (
   834  			B = 2
   835  			// foo
   836  		)
   837  
   838  		var C = 3 // foo
   839  
   840  		var D = 4
   841  		// foo
   842  
   843  		func anycode() {
   844  		// foo
   845  		}
   846  	`
   847  	var s Scanner
   848  	s.Init(fset.AddFile("", fset.Base(), len(src)), []byte(src), nil, 0)
   849  	for {
   850  		pos, tok, lit := s.Scan()
   851  		class := tokenclass(tok)
   852  		if lit != "" && class != keyword && class != literal && tok != token.SEMICOLON {
   853  			t.Errorf("%s: tok = %s, lit = %q", fset.Position(pos), tok, lit)
   854  		}
   855  		if tok <= token.EOF {
   856  			break
   857  		}
   858  	}
   859  }
   860  
   861  func TestIssue28112(t *testing.T) {
   862  	const src = "... .. 0.. .." // make sure to have stand-alone ".." immediately before EOF to test EOF behavior
   863  	tokens := []token.Token{token.ELLIPSIS, token.PERIOD, token.PERIOD, token.FLOAT, token.PERIOD, token.PERIOD, token.PERIOD, token.EOF}
   864  	var s Scanner
   865  	s.Init(fset.AddFile("", fset.Base(), len(src)), []byte(src), nil, 0)
   866  	for _, want := range tokens {
   867  		pos, got, lit := s.Scan()
   868  		if got != want {
   869  			t.Errorf("%s: got %s, want %s", fset.Position(pos), got, want)
   870  		}
   871  		// literals expect to have a (non-empty) literal string and we don't care about other tokens for this test
   872  		if tokenclass(got) == literal && lit == "" {
   873  			t.Errorf("%s: for %s got empty literal string", fset.Position(pos), got)
   874  		}
   875  	}
   876  }
   877  
   878  func BenchmarkScan(b *testing.B) {
   879  	b.StopTimer()
   880  	fset := token.NewFileSet()
   881  	file := fset.AddFile("", fset.Base(), len(source))
   882  	var s Scanner
   883  	b.StartTimer()
   884  	for i := 0; i < b.N; i++ {
   885  		s.Init(file, source, nil, ScanComments)
   886  		for {
   887  			_, tok, _ := s.Scan()
   888  			if tok == token.EOF {
   889  				break
   890  			}
   891  		}
   892  	}
   893  }
   894  
   895  func BenchmarkScanFiles(b *testing.B) {
   896  	// Scan a few arbitrary large files, and one small one, to provide some
   897  	// variety in benchmarks.
   898  	for _, p := range []string{
   899  		"go/types/expr.go",
   900  		"go/parser/parser.go",
   901  		"net/http/server.go",
   902  		"go/scanner/errors.go",
   903  	} {
   904  		b.Run(p, func(b *testing.B) {
   905  			b.StopTimer()
   906  			filename := filepath.Join("..", "..", filepath.FromSlash(p))
   907  			src, err := os.ReadFile(filename)
   908  			if err != nil {
   909  				b.Fatal(err)
   910  			}
   911  			fset := token.NewFileSet()
   912  			file := fset.AddFile(filename, fset.Base(), len(src))
   913  			b.SetBytes(int64(len(src)))
   914  			var s Scanner
   915  			b.StartTimer()
   916  			for i := 0; i < b.N; i++ {
   917  				s.Init(file, src, nil, ScanComments)
   918  				for {
   919  					_, tok, _ := s.Scan()
   920  					if tok == token.EOF {
   921  						break
   922  					}
   923  				}
   924  			}
   925  		})
   926  	}
   927  }
   928  
   929  func TestNumbers(t *testing.T) {
   930  	for _, test := range []struct {
   931  		tok              token.Token
   932  		src, tokens, err string
   933  	}{
   934  		// binaries
   935  		{token.INT, "0b0", "0b0", ""},
   936  		{token.INT, "0b1010", "0b1010", ""},
   937  		{token.INT, "0B1110", "0B1110", ""},
   938  
   939  		{token.INT, "0b", "0b", "binary literal has no digits"},
   940  		{token.INT, "0b0190", "0b0190", "invalid digit '9' in binary literal"},
   941  		{token.INT, "0b01a0", "0b01 a0", ""}, // only accept 0-9
   942  
   943  		{token.FLOAT, "0b.", "0b.", "invalid radix point in binary literal"},
   944  		{token.FLOAT, "0b.1", "0b.1", "invalid radix point in binary literal"},
   945  		{token.FLOAT, "0b1.0", "0b1.0", "invalid radix point in binary literal"},
   946  		{token.FLOAT, "0b1e10", "0b1e10", "'e' exponent requires decimal mantissa"},
   947  		{token.FLOAT, "0b1P-1", "0b1P-1", "'P' exponent requires hexadecimal mantissa"},
   948  
   949  		{token.IMAG, "0b10i", "0b10i", ""},
   950  		{token.IMAG, "0b10.0i", "0b10.0i", "invalid radix point in binary literal"},
   951  
   952  		// octals
   953  		{token.INT, "0o0", "0o0", ""},
   954  		{token.INT, "0o1234", "0o1234", ""},
   955  		{token.INT, "0O1234", "0O1234", ""},
   956  
   957  		{token.INT, "0o", "0o", "octal literal has no digits"},
   958  		{token.INT, "0o8123", "0o8123", "invalid digit '8' in octal literal"},
   959  		{token.INT, "0o1293", "0o1293", "invalid digit '9' in octal literal"},
   960  		{token.INT, "0o12a3", "0o12 a3", ""}, // only accept 0-9
   961  
   962  		{token.FLOAT, "0o.", "0o.", "invalid radix point in octal literal"},
   963  		{token.FLOAT, "0o.2", "0o.2", "invalid radix point in octal literal"},
   964  		{token.FLOAT, "0o1.2", "0o1.2", "invalid radix point in octal literal"},
   965  		{token.FLOAT, "0o1E+2", "0o1E+2", "'E' exponent requires decimal mantissa"},
   966  		{token.FLOAT, "0o1p10", "0o1p10", "'p' exponent requires hexadecimal mantissa"},
   967  
   968  		{token.IMAG, "0o10i", "0o10i", ""},
   969  		{token.IMAG, "0o10e0i", "0o10e0i", "'e' exponent requires decimal mantissa"},
   970  
   971  		// 0-octals
   972  		{token.INT, "0", "0", ""},
   973  		{token.INT, "0123", "0123", ""},
   974  
   975  		{token.INT, "08123", "08123", "invalid digit '8' in octal literal"},
   976  		{token.INT, "01293", "01293", "invalid digit '9' in octal literal"},
   977  		{token.INT, "0F.", "0 F .", ""}, // only accept 0-9
   978  		{token.INT, "0123F.", "0123 F .", ""},
   979  		{token.INT, "0123456x", "0123456 x", ""},
   980  
   981  		// decimals
   982  		{token.INT, "1", "1", ""},
   983  		{token.INT, "1234", "1234", ""},
   984  
   985  		{token.INT, "1f", "1 f", ""}, // only accept 0-9
   986  
   987  		{token.IMAG, "0i", "0i", ""},
   988  		{token.IMAG, "0678i", "0678i", ""},
   989  
   990  		// decimal floats
   991  		{token.FLOAT, "0.", "0.", ""},
   992  		{token.FLOAT, "123.", "123.", ""},
   993  		{token.FLOAT, "0123.", "0123.", ""},
   994  
   995  		{token.FLOAT, ".0", ".0", ""},
   996  		{token.FLOAT, ".123", ".123", ""},
   997  		{token.FLOAT, ".0123", ".0123", ""},
   998  
   999  		{token.FLOAT, "0.0", "0.0", ""},
  1000  		{token.FLOAT, "123.123", "123.123", ""},
  1001  		{token.FLOAT, "0123.0123", "0123.0123", ""},
  1002  
  1003  		{token.FLOAT, "0e0", "0e0", ""},
  1004  		{token.FLOAT, "123e+0", "123e+0", ""},
  1005  		{token.FLOAT, "0123E-1", "0123E-1", ""},
  1006  
  1007  		{token.FLOAT, "0.e+1", "0.e+1", ""},
  1008  		{token.FLOAT, "123.E-10", "123.E-10", ""},
  1009  		{token.FLOAT, "0123.e123", "0123.e123", ""},
  1010  
  1011  		{token.FLOAT, ".0e-1", ".0e-1", ""},
  1012  		{token.FLOAT, ".123E+10", ".123E+10", ""},
  1013  		{token.FLOAT, ".0123E123", ".0123E123", ""},
  1014  
  1015  		{token.FLOAT, "0.0e1", "0.0e1", ""},
  1016  		{token.FLOAT, "123.123E-10", "123.123E-10", ""},
  1017  		{token.FLOAT, "0123.0123e+456", "0123.0123e+456", ""},
  1018  
  1019  		{token.FLOAT, "0e", "0e", "exponent has no digits"},
  1020  		{token.FLOAT, "0E+", "0E+", "exponent has no digits"},
  1021  		{token.FLOAT, "1e+f", "1e+ f", "exponent has no digits"},
  1022  		{token.FLOAT, "0p0", "0p0", "'p' exponent requires hexadecimal mantissa"},
  1023  		{token.FLOAT, "1.0P-1", "1.0P-1", "'P' exponent requires hexadecimal mantissa"},
  1024  
  1025  		{token.IMAG, "0.i", "0.i", ""},
  1026  		{token.IMAG, ".123i", ".123i", ""},
  1027  		{token.IMAG, "123.123i", "123.123i", ""},
  1028  		{token.IMAG, "123e+0i", "123e+0i", ""},
  1029  		{token.IMAG, "123.E-10i", "123.E-10i", ""},
  1030  		{token.IMAG, ".123E+10i", ".123E+10i", ""},
  1031  
  1032  		// hexadecimals
  1033  		{token.INT, "0x0", "0x0", ""},
  1034  		{token.INT, "0x1234", "0x1234", ""},
  1035  		{token.INT, "0xcafef00d", "0xcafef00d", ""},
  1036  		{token.INT, "0XCAFEF00D", "0XCAFEF00D", ""},
  1037  
  1038  		{token.INT, "0x", "0x", "hexadecimal literal has no digits"},
  1039  		{token.INT, "0x1g", "0x1 g", ""},
  1040  
  1041  		{token.IMAG, "0xf00i", "0xf00i", ""},
  1042  
  1043  		// hexadecimal floats
  1044  		{token.FLOAT, "0x0p0", "0x0p0", ""},
  1045  		{token.FLOAT, "0x12efp-123", "0x12efp-123", ""},
  1046  		{token.FLOAT, "0xABCD.p+0", "0xABCD.p+0", ""},
  1047  		{token.FLOAT, "0x.0189P-0", "0x.0189P-0", ""},
  1048  		{token.FLOAT, "0x1.ffffp+1023", "0x1.ffffp+1023", ""},
  1049  
  1050  		{token.FLOAT, "0x.", "0x.", "hexadecimal literal has no digits"},
  1051  		{token.FLOAT, "0x0.", "0x0.", "hexadecimal mantissa requires a 'p' exponent"},
  1052  		{token.FLOAT, "0x.0", "0x.0", "hexadecimal mantissa requires a 'p' exponent"},
  1053  		{token.FLOAT, "0x1.1", "0x1.1", "hexadecimal mantissa requires a 'p' exponent"},
  1054  		{token.FLOAT, "0x1.1e0", "0x1.1e0", "hexadecimal mantissa requires a 'p' exponent"},
  1055  		{token.FLOAT, "0x1.2gp1a", "0x1.2 gp1a", "hexadecimal mantissa requires a 'p' exponent"},
  1056  		{token.FLOAT, "0x0p", "0x0p", "exponent has no digits"},
  1057  		{token.FLOAT, "0xeP-", "0xeP-", "exponent has no digits"},
  1058  		{token.FLOAT, "0x1234PAB", "0x1234P AB", "exponent has no digits"},
  1059  		{token.FLOAT, "0x1.2p1a", "0x1.2p1 a", ""},
  1060  
  1061  		{token.IMAG, "0xf00.bap+12i", "0xf00.bap+12i", ""},
  1062  
  1063  		// separators
  1064  		{token.INT, "0b_1000_0001", "0b_1000_0001", ""},
  1065  		{token.INT, "0o_600", "0o_600", ""},
  1066  		{token.INT, "0_466", "0_466", ""},
  1067  		{token.INT, "1_000", "1_000", ""},
  1068  		{token.FLOAT, "1_000.000_1", "1_000.000_1", ""},
  1069  		{token.IMAG, "10e+1_2_3i", "10e+1_2_3i", ""},
  1070  		{token.INT, "0x_f00d", "0x_f00d", ""},
  1071  		{token.FLOAT, "0x_f00d.0p1_2", "0x_f00d.0p1_2", ""},
  1072  
  1073  		{token.INT, "0b__1000", "0b__1000", "'_' must separate successive digits"},
  1074  		{token.INT, "0o60___0", "0o60___0", "'_' must separate successive digits"},
  1075  		{token.INT, "0466_", "0466_", "'_' must separate successive digits"},
  1076  		{token.FLOAT, "1_.", "1_.", "'_' must separate successive digits"},
  1077  		{token.FLOAT, "0._1", "0._1", "'_' must separate successive digits"},
  1078  		{token.FLOAT, "2.7_e0", "2.7_e0", "'_' must separate successive digits"},
  1079  		{token.IMAG, "10e+12_i", "10e+12_i", "'_' must separate successive digits"},
  1080  		{token.INT, "0x___0", "0x___0", "'_' must separate successive digits"},
  1081  		{token.FLOAT, "0x1.0_p0", "0x1.0_p0", "'_' must separate successive digits"},
  1082  	} {
  1083  		var s Scanner
  1084  		var err string
  1085  		s.Init(fset.AddFile("", fset.Base(), len(test.src)), []byte(test.src), func(_ token.Position, msg string) {
  1086  			if err == "" {
  1087  				err = msg
  1088  			}
  1089  		}, 0)
  1090  		for i, want := range strings.Split(test.tokens, " ") {
  1091  			err = ""
  1092  			_, tok, lit := s.Scan()
  1093  
  1094  			// compute lit where for tokens where lit is not defined
  1095  			switch tok {
  1096  			case token.PERIOD:
  1097  				lit = "."
  1098  			case token.ADD:
  1099  				lit = "+"
  1100  			case token.SUB:
  1101  				lit = "-"
  1102  			}
  1103  
  1104  			if i == 0 {
  1105  				if tok != test.tok {
  1106  					t.Errorf("%q: got token %s; want %s", test.src, tok, test.tok)
  1107  				}
  1108  				if err != test.err {
  1109  					t.Errorf("%q: got error %q; want %q", test.src, err, test.err)
  1110  				}
  1111  			}
  1112  
  1113  			if lit != want {
  1114  				t.Errorf("%q: got literal %q (%s); want %s", test.src, lit, tok, want)
  1115  			}
  1116  		}
  1117  
  1118  		// make sure we read all
  1119  		_, tok, _ := s.Scan()
  1120  		if tok == token.SEMICOLON {
  1121  			_, tok, _ = s.Scan()
  1122  		}
  1123  		if tok != token.EOF {
  1124  			t.Errorf("%q: got %s; want EOF", test.src, tok)
  1125  		}
  1126  	}
  1127  }
  1128  

View as plain text