// code generated by "translate.go.pss" a pep script // http://bumble.sf.net/books/pars/tr/ // s.HasPrefix can be used instead of strings.HasPrefix package main import ( "fmt" "bufio" "strings" "strconv" "unicode" "io" "os" "unicode/utf8" ) // an alias for Println for brevity var pr = fmt.Println /* a machine for parsing */ type machine struct { SIZE int // how many elements in stack/tape/marks eof bool charsRead int linesRead int escape rune delimiter rune counter int work string stack []string cell int tape []string marks []string peep rune reader *bufio.Reader } // there is no special init for structures func newMachine(size int) *machine { mm := machine{SIZE: size} mm.eof = false // end of stream reached? mm.charsRead = 0 // how many chars already read mm.linesRead = 1 // how many lines already read mm.escape = '\\' mm.delimiter = '*' // push/pop delimiter (default "*") mm.counter = 0 // a counter for anything mm.work = "" // the workspace mm.stack = make([]string, 0, mm.SIZE) // stack for parse tokens mm.cell = 0 // current tape cell // slices not arrays mm.tape = make([]string, mm.SIZE, mm.SIZE) // a list of attribute for tokens mm.marks = make([]string, mm.SIZE, mm.SIZE) // marked tape cells // or dont initialse peep until "parse()" calls "setInput()" // check! this is not so simple mm.reader = bufio.NewReader(os.Stdin) var err error mm.peep, _, err = mm.reader.ReadRune() if err == io.EOF { mm.eof = true } else if err != nil { fmt.Fprintln(os.Stderr, "error:", err) os.Exit(1) } return &mm } // method syntax. // func (v * vertex) abs() float64 { ... } // multiline strings are ok ? func (mm *machine) setInput(newInput string) { print("to be implemented") } // read one utf8 character from the input stream and // update the machine. func (mm *machine) read() { var err error if mm.eof { os.Exit(0) } mm.charsRead += 1 // increment lines if mm.peep == '\n' { mm.linesRead += 1 } mm.work += string(mm.peep) // check! mm.peep, _, err = mm.reader.ReadRune() if err == io.EOF { mm.eof = true } else if err != nil { fmt.Fprintln(os.Stderr, "error:", err) os.Exit(1) } } // remove escape character: trivial method ? // check the python code for this, and the c code in machine.interp.c func (mm *machine) unescapeChar(c string) { // if mm.work = "" { return } mm.work = strings.Replace(mm.work, "\\"+c, c, -1) } // add escape character : trivial func (mm *machine) escapeChar(c string) { mm.work = strings.Replace(mm.work, c, "\\"+c, -1) } /** a helper function to count trailing escapes */ func (mm *machine) countEscapes(suffix string) int { count := 0 ss := "" if strings.HasSuffix(mm.work, suffix) { ss = strings.TrimSuffix(mm.work, suffix) } for (strings.HasSuffix(ss, string(mm.escape))) { ss = strings.TrimSuffix(ss, string(mm.escape)) count++ } return count } // reads the input stream until the workspace ends with the // given character or text, ignoring escaped characters func (mm *machine) until(suffix string) { if mm.eof { return; } // read at least one character mm.read() for true { if mm.eof { return; } // we need to count the mm.Escape chars preceding suffix // if odd, keep reading, if even, stop if strings.HasSuffix(mm.work, suffix) { if (mm.countEscapes(suffix) % 2 == 0) { return } } mm.read() } } /* increment the tape pointer (command ++) and grow the tape and marks arrays if necessary */ func (mm *machine) increment() { mm.cell++ if mm.cell >= len(mm.tape) { mm.tape = append(mm.tape, "") mm.marks = append(mm.marks, "") mm.SIZE++ } } /* pop the last token from the stack into the workspace */ func (mm *machine) pop() bool { if len(mm.stack) == 0 { return false } // no, get last element of stack // a[len(a)-1] mm.work = mm.stack[len(mm.stack)-1] + mm.work // a = a[:len(a)-1] mm.stack = mm.stack[:len(mm.stack)-1] if mm.cell > 0 { mm.cell -= 1 } return true } // push the first token from the workspace to the stack func (mm *machine) push() bool { // dont increment the tape pointer on an empty push if mm.work == "" { return false } // push first token, or else whole string if no delimiter aa := strings.SplitN(mm.work, string(mm.delimiter), 2) if len(aa) == 1 { mm.stack = append(mm.stack, mm.work) mm.work = "" } else { mm.stack = append(mm.stack, aa[0]+string(mm.delimiter)) mm.work = aa[1] } mm.increment() return true } // func (mm *machine) printState() { fmt.Printf("Stack %v Work[%s] Peep[%c] \n", mm.stack, mm.work, mm.peep) fmt.Printf("Acc:%v Esc:%c Delim:%c Chars:%v", mm.counter, mm.escape, mm.delimiter, mm.charsRead) fmt.Printf(" Lines:%v Cell:%v EOF:%v \n", mm.linesRead, mm.cell, mm.eof) for ii, vv := range mm.tape { fmt.Printf("%v [%s] \n", ii, vv) if ii > 4 { return; } } } func (mm *machine) goToMark(mark string) { markFound := false for ii := range mm.marks { if mm.marks[ii] == mark { mm.cell = ii; markFound = true; break } } if markFound == false { fmt.Printf("badmark '%s'", mark) os.Exit(1) } } // this is where the actual parsing/compiling code should go // so that it can be used by other go classes/objects. Also // should have a stream argument. func (mm *machine) parse(s string) { } /* adapt for clop and clip */ func trimLastChar(s string) string { r, size := utf8.DecodeLastRuneInString(s) if r == utf8.RuneError && (size == 0 || size == 1) { size = 0 } return s[:len(s)-size] } func (mm *machine) clip() { cc, _ := utf8.DecodeLastRuneInString(mm.work) mm.work = strings.TrimSuffix(mm.work, string(cc)) } func (mm *machine) clop() { _, size := utf8.DecodeRuneInString(mm.work) mm.work = mm.work[size:] } type fn func(rune) bool // eg unicode.IsLetter('x') /* check whether the string s only contains runes of type determined by the typeFn function */ func isInClass(typeFn fn, s string) bool { if s == "" { return false; } for _, rr := range s { //if !unicode.IsLetter(rr) { if !typeFn(rr) { return false } } return true } /* range in format 'a,z' */ func isInRange(start rune, end rune, s string) bool { if s == "" { return false; } for _, rr := range s { if (rr < start) || (rr > end) { return false } } return true } /* list of runes (unicode chars ) */ func isInList(list string, s string) bool { return strings.ContainsAny(s, list) } func main() { // This size needs to be big for some applications. Eg // calculating big palindromes. Really // it should be dynamically allocated. var size = 30000 var mm = newMachine(size); var restart = false; // the go compiler complains when modules are imported but // not used, also if vars are not used. if restart {}; unicode.IsDigit('0'); strconv.Itoa(0); for !mm.eof { /* lex block */ for true { mm.read() /* read */ // ignore \r if (isInList("\r", mm.work)) { mm.work = "" // clear if (mm.eof) { break } restart = true; break // restart } // dont tokenize non-leading space. One space will be printed // between each word. if (isInList(" \t\f", mm.work)) { /* while */ for isInList(" \t\f", string(mm.peep)) { if mm.eof { break } mm.read() } mm.work = "" // clear if (mm.eof) { break } restart = true; break // restart } if (isInList("\n", mm.work)) { // make character count relative to line. mm.charsRead = 0 /* nochars */ // save the leading space in the nl* token /* while */ for isInClass(unicode.IsSpace, string(mm.peep)) { if mm.eof { break } mm.read() } mm.tape[mm.cell] = mm.work /* put */ mm.work = "" // clear mm.work += "nl*" mm.push(); break } // everything else is a word if (mm.work != "") { /* whilenot */ for !isInClass(unicode.IsSpace, string(mm.peep)) { if mm.eof { break; } mm.read() } mm.tape[mm.cell] = mm.work /* put */ mm.work = "" // clear mm.work += "word*" mm.push(); break } break } if restart { restart = false; continue; } // parse block for true { // to visualise parse token reductions mm.work += "line " mm.work += strconv.Itoa(mm.linesRead) /* lines */ mm.work += " char " mm.work += strconv.Itoa(mm.charsRead) /* chars */ mm.work += ": " fmt.Printf("%s", mm.work) // print mm.work = "" // clear for mm.pop() {} /* unstack */ mm.work += "\n" fmt.Printf("%s", mm.work) // print mm.clip() for mm.push() {} /* stack */ //------- // 1 token mm.pop(); //------- // 2 tokens mm.pop(); // I want to recognise 2 word structures, so need to separate // the text*word* reduction from the word*word* rule. // is there any need for a file* token, link token etc? if (mm.work == "word*word*" || mm.work == "text*word*") { mm.work = "" // clear mm.work += mm.tape[mm.cell] /* get */ mm.work += " " mm.increment() /* ++ */ mm.work += mm.tape[mm.cell] /* get */ if mm.cell > 0 { mm.cell-- } /* -- */ mm.tape[mm.cell] = mm.work /* put */ mm.work = "" // clear mm.work += "text*" mm.push(); continue } if (mm.work == "word*nl*" || mm.work == "text*nl*") { mm.work = "" // clear mm.work += mm.tape[mm.cell] /* get */ mm.increment() /* ++ */ mm.work += mm.tape[mm.cell] /* get */ if mm.cell > 0 { mm.cell-- } /* -- */ mm.tape[mm.cell] = mm.work /* put */ mm.work = "" // clear mm.work += "line*" mm.push(); continue } if (mm.work == "line*line*" || mm.work == "lineset*line*") { mm.work = "" // clear mm.work += mm.tape[mm.cell] /* get */ mm.increment() /* ++ */ mm.work += mm.tape[mm.cell] /* get */ if mm.cell > 0 { mm.cell-- } /* -- */ mm.tape[mm.cell] = mm.work /* put */ mm.work = "" // clear mm.work += "lineset*" mm.push(); continue } mm.push(); mm.push(); if (mm.eof) { mm.pop(); if (mm.work == "word*" || mm.work == "text*" || mm.work == "line*" || mm.work == "lineset*") { mm.work = "" // clear mm.work += mm.tape[mm.cell] /* get */ mm.work += "\n" fmt.Printf("%s", mm.work) // print mm.work = "" // clear os.Exit(0) } } break } // parse } } // end of generated 'go' code