From 500caaeda74dd9c660279036293f4b2997cf0b03 Mon Sep 17 00:00:00 2001 From: Dimitri Sokolyuk Date: Sat, 9 Sep 2017 09:42:37 +0200 Subject: Add vendor --- vendor/github.com/llgcode/ps/scanner.go | 572 ++++++++++++++++++++++++++++++++ 1 file changed, 572 insertions(+) create mode 100644 vendor/github.com/llgcode/ps/scanner.go (limited to 'vendor/github.com/llgcode/ps/scanner.go') diff --git a/vendor/github.com/llgcode/ps/scanner.go b/vendor/github.com/llgcode/ps/scanner.go new file mode 100644 index 0000000..fb7cab5 --- /dev/null +++ b/vendor/github.com/llgcode/ps/scanner.go @@ -0,0 +1,572 @@ +// Copyright 2009 The Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. +// postscript scanner derived form the scanner package of go sources +package ps + +import ( + "bytes" + "fmt" + "io" + "os" + "unicode" + "unicode/utf8" +) + +// A source position is represented by a Position value. +// A position is valid if Line > 0. +type Position struct { + Filename string // filename, if any + Offset int // byte offset, starting at 0 + Line int // line number, starting at 1 + Column int // column number, starting at 0 (character count per line) +} + +// IsValid returns true if the position is valid. +func (pos *Position) IsValid() bool { return pos.Line > 0 } + +func (pos Position) String() string { + s := pos.Filename + if pos.IsValid() { + if s != "" { + s += ":" + } + s += fmt.Sprintf("%d:%d", pos.Line, pos.Column) + } + if s == "" { + s = "???" + } + return s +} + +// Predefined mode bits to control recognition of tokens. For instance, +// to configure a Scanner such that it only recognizes (Go) identifiers, +// integers, and skips comments, set the Scanner's Mode field to: +// +// ScanIdents | ScanInts | SkipComments +// +const ( + ScanIdents = 1 << -Ident + ScanInts = 1 << -Int + ScanFloats = 1 << -Float // includes Ints + ScanChars = 1 << -Char + ScanStrings = 1 << -String + ScanRawStrings = 1 << -RawString + ScanComments = 1 << -Comment + SkipComments = 1 << -skipComment // if set with ScanComments, comments become white space + GoTokens = ScanIdents | ScanFloats | ScanChars | ScanStrings | ScanRawStrings | ScanComments | SkipComments +) + +// The result of Scan is one of the following tokens or a Unicode character. +const ( + EOF = -(iota + 1) + Ident + Int + Float + Char + String + RawString + Comment + skipComment +) + +var tokenString = map[int]string{ + EOF: "EOF", + Ident: "Ident", + Int: "Int", + Float: "Float", + Char: "Char", + String: "String", + RawString: "RawString", + Comment: "Comment", +} + +// TokenString returns a (visible) string for a token or Unicode character. +func TokenString(tok int) string { + if s, found := tokenString[tok]; found { + return s + } + return fmt.Sprintf("U+%04X", tok) +} + +// GoWhitespace is the default value for the Scanner's Whitespace field. +// Its value selects Go's white space characters. +const GoWhitespace = 1<<'\t' | 1<<'\n' | 1<<'\r' | 1<<' ' + +const bufLen = 1024 // at least utf8.UTFMax + +// A Scanner implements reading of Unicode characters and tokens from an io.Reader. +type Scanner struct { + // Input + src io.Reader + + // Source buffer + srcBuf [bufLen + 1]byte // +1 for sentinel for common case of s.next() + srcPos int // reading position (srcBuf index) + srcEnd int // source end (srcBuf index) + + // Source position + srcBufOffset int // byte offset of srcBuf[0] in source + line int // newline count + 1 + column int // character count on line + + // Token text buffer + // Typically, token text is stored completely in srcBuf, but in general + // the token text's head may be buffered in tokBuf while the token text's + // tail is stored in srcBuf. + tokBuf bytes.Buffer // token text head that is not in srcBuf anymore + tokPos int // token text tail position (srcBuf index) + tokEnd int // token text tail end (srcBuf index) + + // One character look-ahead + ch rune // character before current srcPos + + // Error is called for each error encountered. If no Error + // function is set, the error is reported to os.Stderr. + Error func(s *Scanner, msg string) + + // ErrorCount is incremented by one for each error encountered. + ErrorCount int + + // The Mode field controls which tokens are recognized. For instance, + // to recognize Ints, set the ScanInts bit in Mode. The field may be + // changed at any time. + Mode uint + + // The Whitespace field controls which characters are recognized + // as white space. To recognize a character ch <= ' ' as white space, + // set the ch'th bit in Whitespace (the Scanner's behavior is undefined + // for values ch > ' '). The field may be changed at any time. + Whitespace uint64 + + // Current token position. The Offset, Line, and Column fields + // are set by Scan(); the Filename field is left untouched by the + // Scanner. + Position +} + +// Init initializes a Scanner with a new source and returns itself. +// Error is set to nil, ErrorCount is set to 0, Mode is set to GoTokens, +// and Whitespace is set to GoWhitespace. +func (s *Scanner) Init(src io.Reader) *Scanner { + s.src = src + + // initialize source buffer + s.srcBuf[0] = utf8.RuneSelf // sentinel + s.srcPos = 0 + s.srcEnd = 0 + + // initialize source position + s.srcBufOffset = 0 + s.line = 1 + s.column = 0 + + // initialize token text buffer + s.tokPos = -1 + + // initialize one character look-ahead + s.ch = s.next() + + // initialize public fields + s.Error = nil + s.ErrorCount = 0 + s.Mode = GoTokens + s.Whitespace = GoWhitespace + + return s +} + +// next reads and returns the next Unicode character. It is designed such +// that only a minimal amount of work needs to be done in the common ASCII +// case (one test to check for both ASCII and end-of-buffer, and one test +// to check for newlines). +func (s *Scanner) next() rune { + ch := rune(s.srcBuf[s.srcPos]) + + if ch >= utf8.RuneSelf { + // uncommon case: not ASCII or not enough bytes + for s.srcPos+utf8.UTFMax > s.srcEnd && !utf8.FullRune(s.srcBuf[s.srcPos:s.srcEnd]) { + // not enough bytes: read some more, but first + // save away token text if any + if s.tokPos >= 0 { + s.tokBuf.Write(s.srcBuf[s.tokPos:s.srcPos]) + s.tokPos = 0 + } + // move unread bytes to beginning of buffer + copy(s.srcBuf[0:], s.srcBuf[s.srcPos:s.srcEnd]) + s.srcBufOffset += s.srcPos + // read more bytes + i := s.srcEnd - s.srcPos + n, err := s.src.Read(s.srcBuf[i:bufLen]) + s.srcEnd = i + n + s.srcPos = 0 + s.srcBuf[s.srcEnd] = utf8.RuneSelf // sentinel + if err != nil { + if s.srcEnd == 0 { + return EOF + } + if err != io.EOF { + s.error(err.Error()) + break + } + } + } + // at least one byte + ch = rune(s.srcBuf[s.srcPos]) + if ch >= utf8.RuneSelf { + // uncommon case: not ASCII + var width int + ch, width = utf8.DecodeRune(s.srcBuf[s.srcPos:s.srcEnd]) + if ch == utf8.RuneError && width == 1 { + s.error("illegal UTF-8 encoding") + } + s.srcPos += width - 1 + } + } + + s.srcPos++ + s.column++ + switch ch { + case 0: + // implementation restriction for compatibility with other tools + s.error("illegal character NUL") + case '\n': + s.line++ + s.column = 0 + } + + return ch +} + +// Next reads and returns the next Unicode character. +// It returns EOF at the end of the source. It reports +// a read error by calling s.Error, if set, or else +// prints an error message to os.Stderr. Next does not +// update the Scanner's Position field; use Pos() to +// get the current position. +func (s *Scanner) Next() rune { + s.tokPos = -1 // don't collect token text + ch := s.ch + s.ch = s.next() + return ch +} + +// Peek returns the next Unicode character in the source without advancing +// the scanner. It returns EOF if the scanner's position is at the last +// character of the source. +func (s *Scanner) Peek() rune { + return s.ch +} + +func (s *Scanner) error(msg string) { + s.ErrorCount++ + if s.Error != nil { + s.Error(s, msg) + return + } + fmt.Fprintf(os.Stderr, "%s: %s", s.Position, msg) +} + +func (s *Scanner) scanIdentifier() rune { + ch := s.next() // read character after first '_' or letter + for ch == '_' || unicode.IsLetter(ch) || unicode.IsDigit(ch) || ch == '.' || ch == '-' || ch == '`' { + ch = s.next() + } + return ch +} + +func digitVal(ch rune) rune { + switch { + case '0' <= ch && ch <= '9': + return ch - '0' + case 'a' <= ch && ch <= 'f': + return ch - 'a' + 10 + case 'A' <= ch && ch <= 'F': + return ch - 'A' + 10 + } + return 16 // larger than any legal digit val +} + +func isDecimal(ch rune) bool { return '0' <= ch && ch <= '9' } + +func (s *Scanner) scanMantissa(ch rune) rune { + for isDecimal(ch) { + ch = s.next() + } + return ch +} + +func (s *Scanner) scanFraction(ch rune) rune { + if ch == '.' { + ch = s.scanMantissa(s.next()) + } + return ch +} + +func (s *Scanner) scanExponent(ch rune) rune { + if ch == 'e' || ch == 'E' { + ch = s.next() + if ch == '-' || ch == '+' { + ch = s.next() + } + ch = s.scanMantissa(ch) + } + return ch +} + +func (s *Scanner) scanNumber(ch rune) (int, rune) { + // isDecimal(ch) + if ch == '0' { + // int or float + ch = s.next() + if ch == 'x' || ch == 'X' { + // hexadecimal int + ch = s.next() + for digitVal(ch) < 16 { + ch = s.next() + } + } else { + // octal int or float + seenDecimalDigit := false + for isDecimal(ch) { + if ch > '7' { + seenDecimalDigit = true + } + ch = s.next() + } + if s.Mode&ScanFloats != 0 && (ch == '.' || ch == 'e' || ch == 'E') { + // float + ch = s.scanFraction(ch) + ch = s.scanExponent(ch) + return Float, ch + } + // octal int + if seenDecimalDigit { + s.error("illegal octal number") + } + } + return Int, ch + } + // decimal int or float + ch = s.scanMantissa(ch) + if s.Mode&ScanFloats != 0 && (ch == '.' || ch == 'e' || ch == 'E') { + // float + ch = s.scanFraction(ch) + ch = s.scanExponent(ch) + return Float, ch + } + return Int, ch +} + +func (s *Scanner) scanDigits(ch rune, base, n int) rune { + for n > 0 && int(digitVal(ch)) < base { + ch = s.next() + n-- + } + if n > 0 { + s.error("illegal char escape") + } + return ch +} + +func (s *Scanner) scanEscape(quote rune) rune { + ch := s.next() // read character after '/' + switch ch { + case 'a', 'b', 'f', 'n', 'r', 't', 'v', '\\', quote: + // nothing to do + ch = s.next() + case '0', '1', '2', '3', '4', '5', '6', '7': + ch = s.scanDigits(ch, 8, 3) + case 'x': + ch = s.scanDigits(s.next(), 16, 2) + case 'u': + ch = s.scanDigits(s.next(), 16, 4) + case 'U': + ch = s.scanDigits(s.next(), 16, 8) + default: + s.error("illegal char escape") + } + return ch +} + +func (s *Scanner) scanString(quote rune) (n int) { + ch := s.next() // read character after quote + for ch != quote { + if ch == '\n' || ch < 0 { + s.error("literal not terminated") + return + } + if ch == '\\' { + ch = s.scanEscape(quote) + } else { + ch = s.next() + } + n++ + } + return +} + +func (s *Scanner) scanRawString() { + ch := s.next() // read character after '`' + for ch != '`' { + if ch < 0 { + s.error("literal not terminated") + return + } + ch = s.next() + } +} + +func (s *Scanner) scanLineComment() { + ch := s.next() // read character after "//" + for ch != '\n' { + if ch < 0 { + s.error("comment not terminated") + return + } + ch = s.next() + } +} + +func (s *Scanner) scanComment(ch rune) { + s.scanLineComment() +} + +// Scan reads the next token or Unicode character from source and returns it. +// It only recognizes tokens t for which the respective Mode bit (1<<-t) is set. +// It returns EOF at the end of the source. It reports scanner errors (read and +// token errors) by calling s.Error, if set; otherwise it prints an error message +// to os.Stderr. +func (s *Scanner) Scan() int { + ch := s.ch + + // reset token text position + s.tokPos = -1 + +redo: + // skip white space + for s.Whitespace&(1<