12345678910111213141516171819202122232425262728293031323334353637383940414243444546474849505152535455565758596061626364656667686970717273747576777879808182838485868788899091929394959697989910010110210310410510610710810911011111211311411511611711811912012112212312412512612712812913013113213313413513613713813914014114214314414514614714814915015115215315415515615715815916016116216316416516616716816917017117217317417517617717817918018118218318418518618718818919019119219319419519619719819920020120220320420520620720820921021121221321421521621721821922022122222322422522622722822923023123223323423523623723823924024124224324424524624724824925025125225325425525625725825926026126226326426526626726826927027127227327427527627727827928028128228328428528628728828929029129229329429529629729829930030130230330430530630730830931031131231331431531631731831932032132232332432532632732832933033133233333433533633733833934034134234334434534634734834935035135235335435535635735835936036136236336436536636736836937037137237337437537637737837938038138238338438538638738838939039139239339439539639739839940040140240340440540640740840941041141241341441541641741841942042142242342442542642742842943043143243343443543643743843944044144244344444544644744844945045145245345445545645745845946046146246346446546646746846947047147247347447547647747847948048148248348448548648748848949049149249349449549649749849950050150250350450550650750850951051151251351451551651751851952052152252352452552652752852953053153253353453553653753853954054154254354454554654754854955055155255355455555655755855956056156256356456556656756856957057157257357457557657757857958058158258358458558658758858959059159259359459559659759859960060160260360460560660760860961061161261361461561661761861962062162262362462562662762862963063163263363463563663763863964064164264364464564664764864965065165265365465565665765865966066166266366466566666766866967067167267367467567667767867968068168268368468568668768868969069169269369469569669769869970070170270370470570670770870971071171271371471571671771871972072172272372472572672772872973073173273373473573673773873974074174274374474574674774874975075175275375475575675775875976076176276376476576676776876977077177277377477577677777877978078178278378478578678778878979079179279379479579679779879980080180280380480580680780880981081181281381481581681781881982082182282382482582682782882983083183283383483583683783883984084184284384484584684784884985085185285385485585685785885986086186286386486586686786886987087187287387487587687787887988088188288388488588688788888989089189289389489589689789889990090190290390490590690790890991091191291391491591691791891992092192292392492592692792892993093193293393493593693793893994094194294394494594694794894995095195295395495595695795895996096196296396496596696796896997097197297397497597697797897998098198298398498598698798898999099199299399499599699799899910001001100210031004100510061007100810091010101110121013101410151016101710181019102010211022102310241025102610271028102910301031103210331034103510361037103810391040104110421043104410451046104710481049105010511052105310541055105610571058105910601061106210631064106510661067106810691070107110721073107410751076107710781079108010811082108310841085108610871088108910901091109210931094109510961097109810991100110111021103110411051106110711081109111011111112111311141115111611171118111911201121112211231124112511261127112811291130113111321133113411351136113711381139114011411142114311441145114611471148114911501151115211531154115511561157115811591160116111621163116411651166116711681169117011711172117311741175117611771178117911801181118211831184118511861187118811891190119111921193119411951196119711981199120012011202120312041205120612071208120912101211121212131214121512161217121812191220122112221223122412251226122712281229123012311232123312341235123612371238123912401241124212431244124512461247124812491250125112521253125412551256125712581259126012611262126312641265126612671268126912701271127212731274127512761277127812791280128112821283128412851286128712881289129012911292129312941295129612971298129913001301130213031304130513061307130813091310131113121313131413151316131713181319132013211322132313241325132613271328132913301331133213331334133513361337133813391340134113421343134413451346134713481349135013511352135313541355135613571358135913601361136213631364136513661367136813691370137113721373137413751376137713781379138013811382138313841385138613871388138913901391139213931394139513961397139813991400140114021403140414051406140714081409141014111412141314141415141614171418141914201421142214231424142514261427142814291430143114321433143414351436143714381439144014411442144314441445144614471448144914501451145214531454145514561457145814591460146114621463146414651466146714681469147014711472147314741475147614771478147914801481148214831484148514861487148814891490149114921493149414951496149714981499150015011502150315041505150615071508150915101511151215131514151515161517151815191520152115221523152415251526152715281529153015311532153315341535153615371538153915401541154215431544154515461547154815491550155115521553155415551556155715581559156015611562156315641565156615671568156915701571157215731574157515761577157815791580158115821583158415851586158715881589159015911592159315941595159615971598159916001601160216031604 |
- // Code generated by pigeon; DO NOT EDIT.
- package grammar
- import (
- "bytes"
- "errors"
- "fmt"
- "io"
- "io/ioutil"
- "math"
- "os"
- "sort"
- "strconv"
- "strings"
- "sync"
- "unicode"
- "unicode/utf8"
- )
- var g = &grammar{
- rules: []*rule{
- {
- name: "Input",
- pos: position{line: 6, col: 1, offset: 22},
- expr: &oneOrMoreExpr{
- pos: position{line: 6, col: 9, offset: 30},
- expr: &choiceExpr{
- pos: position{line: 6, col: 10, offset: 31},
- alternatives: []interface{}{
- &ruleRefExpr{
- pos: position{line: 6, col: 10, offset: 31},
- name: "Std",
- },
- &ruleRefExpr{
- pos: position{line: 6, col: 16, offset: 37},
- name: "EOL",
- },
- },
- },
- },
- },
- {
- name: "LabelUpper",
- pos: position{line: 8, col: 1, offset: 44},
- expr: &actionExpr{
- pos: position{line: 8, col: 14, offset: 57},
- run: (*parser).callonLabelUpper1,
- expr: &oneOrMoreExpr{
- pos: position{line: 8, col: 14, offset: 57},
- expr: &charClassMatcher{
- pos: position{line: 8, col: 14, offset: 57},
- val: "[A-Z_]",
- chars: []rune{'_'},
- ranges: []rune{'A', 'Z'},
- ignoreCase: false,
- inverted: false,
- },
- },
- },
- },
- {
- name: "Std",
- pos: position{line: 13, col: 1, offset: 119},
- expr: &seqExpr{
- pos: position{line: 13, col: 7, offset: 125},
- exprs: []interface{}{
- &ruleRefExpr{
- pos: position{line: 13, col: 7, offset: 125},
- name: "LabelUpperLine",
- },
- &choiceExpr{
- pos: position{line: 13, col: 23, offset: 141},
- alternatives: []interface{}{
- &oneOrMoreExpr{
- pos: position{line: 13, col: 23, offset: 141},
- expr: &ruleRefExpr{
- pos: position{line: 13, col: 23, offset: 141},
- name: "LabelLine",
- },
- },
- &oneOrMoreExpr{
- pos: position{line: 13, col: 36, offset: 154},
- expr: &ruleRefExpr{
- pos: position{line: 13, col: 36, offset: 154},
- name: "UpperLabelLine",
- },
- },
- },
- },
- },
- },
- },
- {
- name: "LabelUpperLine",
- pos: position{line: 15, col: 1, offset: 172},
- expr: &seqExpr{
- pos: position{line: 15, col: 18, offset: 189},
- exprs: []interface{}{
- &ruleRefExpr{
- pos: position{line: 15, col: 18, offset: 189},
- name: "LabelUpper",
- },
- &ruleRefExpr{
- pos: position{line: 15, col: 29, offset: 200},
- name: "EOL",
- },
- },
- },
- },
- {
- name: "LabelLine",
- pos: position{line: 16, col: 1, offset: 204},
- expr: &seqExpr{
- pos: position{line: 16, col: 13, offset: 216},
- exprs: []interface{}{
- &charClassMatcher{
- pos: position{line: 16, col: 13, offset: 216},
- val: "[\\t]",
- chars: []rune{'\t'},
- ignoreCase: false,
- inverted: false,
- },
- &ruleRefExpr{
- pos: position{line: 16, col: 18, offset: 221},
- name: "Label",
- },
- &ruleRefExpr{
- pos: position{line: 16, col: 24, offset: 227},
- name: "EOL",
- },
- },
- },
- },
- {
- name: "UpperLabelLine",
- pos: position{line: 17, col: 1, offset: 231},
- expr: &seqExpr{
- pos: position{line: 17, col: 18, offset: 248},
- exprs: []interface{}{
- &charClassMatcher{
- pos: position{line: 17, col: 18, offset: 248},
- val: "[\\t]",
- chars: []rune{'\t'},
- ignoreCase: false,
- inverted: false,
- },
- &oneOrMoreExpr{
- pos: position{line: 17, col: 23, offset: 253},
- expr: &seqExpr{
- pos: position{line: 17, col: 24, offset: 254},
- exprs: []interface{}{
- &ruleRefExpr{
- pos: position{line: 17, col: 24, offset: 254},
- name: "LabelUpper",
- },
- &ruleRefExpr{
- pos: position{line: 17, col: 35, offset: 265},
- name: "_",
- },
- },
- },
- },
- &ruleRefExpr{
- pos: position{line: 17, col: 39, offset: 269},
- name: "UpperLabel",
- },
- &ruleRefExpr{
- pos: position{line: 17, col: 50, offset: 280},
- name: "EOL",
- },
- },
- },
- },
- {
- name: "Label",
- pos: position{line: 19, col: 1, offset: 285},
- expr: &actionExpr{
- pos: position{line: 19, col: 9, offset: 293},
- run: (*parser).callonLabel1,
- expr: &oneOrMoreExpr{
- pos: position{line: 19, col: 9, offset: 293},
- expr: &charClassMatcher{
- pos: position{line: 19, col: 9, offset: 293},
- val: "[a-z_]",
- chars: []rune{'_'},
- ranges: []rune{'a', 'z'},
- ignoreCase: false,
- inverted: false,
- },
- },
- },
- },
- {
- name: "_",
- pos: position{line: 25, col: 1, offset: 356},
- expr: &oneOrMoreExpr{
- pos: position{line: 25, col: 5, offset: 360},
- expr: &charClassMatcher{
- pos: position{line: 25, col: 5, offset: 360},
- val: "[ \\t]",
- chars: []rune{' ', '\t'},
- ignoreCase: false,
- inverted: false,
- },
- },
- },
- {
- name: "EOL",
- pos: position{line: 27, col: 1, offset: 368},
- expr: &seqExpr{
- pos: position{line: 27, col: 7, offset: 374},
- exprs: []interface{}{
- &zeroOrOneExpr{
- pos: position{line: 27, col: 7, offset: 374},
- expr: &ruleRefExpr{
- pos: position{line: 27, col: 7, offset: 374},
- name: "_",
- },
- },
- &zeroOrOneExpr{
- pos: position{line: 27, col: 10, offset: 377},
- expr: &ruleRefExpr{
- pos: position{line: 27, col: 10, offset: 377},
- name: "Comment",
- },
- },
- &choiceExpr{
- pos: position{line: 27, col: 20, offset: 387},
- alternatives: []interface{}{
- &litMatcher{
- pos: position{line: 27, col: 20, offset: 387},
- val: "\r\n",
- ignoreCase: false,
- want: "\"\\r\\n\"",
- },
- &litMatcher{
- pos: position{line: 27, col: 29, offset: 396},
- val: "\n\r",
- ignoreCase: false,
- want: "\"\\n\\r\"",
- },
- &litMatcher{
- pos: position{line: 27, col: 38, offset: 405},
- val: "\r",
- ignoreCase: false,
- want: "\"\\r\"",
- },
- &litMatcher{
- pos: position{line: 27, col: 45, offset: 412},
- val: "\n",
- ignoreCase: false,
- want: "\"\\n\"",
- },
- &ruleRefExpr{
- pos: position{line: 27, col: 52, offset: 419},
- name: "EOF",
- },
- },
- },
- },
- },
- },
- {
- name: "EOF",
- pos: position{line: 29, col: 1, offset: 425},
- expr: ¬Expr{
- pos: position{line: 29, col: 7, offset: 431},
- expr: &anyMatcher{
- line: 29, col: 8, offset: 432,
- },
- },
- },
- {
- name: "Comment",
- pos: position{line: 31, col: 1, offset: 435},
- expr: &seqExpr{
- pos: position{line: 31, col: 11, offset: 445},
- exprs: []interface{}{
- &litMatcher{
- pos: position{line: 31, col: 11, offset: 445},
- val: "//",
- ignoreCase: false,
- want: "\"//\"",
- },
- &zeroOrMoreExpr{
- pos: position{line: 31, col: 16, offset: 450},
- expr: &charClassMatcher{
- pos: position{line: 31, col: 16, offset: 450},
- val: "[^\\r\\n]",
- chars: []rune{'\r', '\n'},
- ignoreCase: false,
- inverted: true,
- },
- },
- },
- },
- },
- },
- }
- func (c *current) onLabelUpper1() (interface{}, error) {
- fmt.Println(string(c.text))
- return c.text, nil
- }
- func (p *parser) callonLabelUpper1() (interface{}, error) {
- stack := p.vstack[len(p.vstack)-1]
- _ = stack
- return p.cur.onLabelUpper1()
- }
- func (c *current) onLabel1() (interface{}, error) {
- fmt.Println(string(c.text))
- return c.text, nil
- }
- func (p *parser) callonLabel1() (interface{}, error) {
- stack := p.vstack[len(p.vstack)-1]
- _ = stack
- return p.cur.onLabel1()
- }
- var (
- // errNoRule is returned when the grammar to parse has no rule.
- errNoRule = errors.New("grammar has no rule")
- // errInvalidEntrypoint is returned when the specified entrypoint rule
- // does not exit.
- errInvalidEntrypoint = errors.New("invalid entrypoint")
- // errInvalidEncoding is returned when the source is not properly
- // utf8-encoded.
- errInvalidEncoding = errors.New("invalid encoding")
- // errMaxExprCnt is used to signal that the maximum number of
- // expressions have been parsed.
- errMaxExprCnt = errors.New("max number of expresssions parsed")
- )
- // Option is a function that can set an option on the parser. It returns
- // the previous setting as an Option.
- type Option func(*parser) Option
- // MaxExpressions creates an Option to stop parsing after the provided
- // number of expressions have been parsed, if the value is 0 then the parser will
- // parse for as many steps as needed (possibly an infinite number).
- //
- // The default for maxExprCnt is 0.
- func MaxExpressions(maxExprCnt uint64) Option {
- return func(p *parser) Option {
- oldMaxExprCnt := p.maxExprCnt
- p.maxExprCnt = maxExprCnt
- return MaxExpressions(oldMaxExprCnt)
- }
- }
- // Entrypoint creates an Option to set the rule name to use as entrypoint.
- // The rule name must have been specified in the -alternate-entrypoints
- // if generating the parser with the -optimize-grammar flag, otherwise
- // it may have been optimized out. Passing an empty string sets the
- // entrypoint to the first rule in the grammar.
- //
- // The default is to start parsing at the first rule in the grammar.
- func Entrypoint(ruleName string) Option {
- return func(p *parser) Option {
- oldEntrypoint := p.entrypoint
- p.entrypoint = ruleName
- if ruleName == "" {
- p.entrypoint = g.rules[0].name
- }
- return Entrypoint(oldEntrypoint)
- }
- }
- // Statistics adds a user provided Stats struct to the parser to allow
- // the user to process the results after the parsing has finished.
- // Also the key for the "no match" counter is set.
- //
- // Example usage:
- //
- // input := "input"
- // stats := Stats{}
- // _, err := Parse("input-file", []byte(input), Statistics(&stats, "no match"))
- // if err != nil {
- // log.Panicln(err)
- // }
- // b, err := json.MarshalIndent(stats.ChoiceAltCnt, "", " ")
- // if err != nil {
- // log.Panicln(err)
- // }
- // fmt.Println(string(b))
- func Statistics(stats *Stats, choiceNoMatch string) Option {
- return func(p *parser) Option {
- oldStats := p.Stats
- p.Stats = stats
- oldChoiceNoMatch := p.choiceNoMatch
- p.choiceNoMatch = choiceNoMatch
- if p.Stats.ChoiceAltCnt == nil {
- p.Stats.ChoiceAltCnt = make(map[string]map[string]int)
- }
- return Statistics(oldStats, oldChoiceNoMatch)
- }
- }
- // Debug creates an Option to set the debug flag to b. When set to true,
- // debugging information is printed to stdout while parsing.
- //
- // The default is false.
- func Debug(b bool) Option {
- return func(p *parser) Option {
- old := p.debug
- p.debug = b
- return Debug(old)
- }
- }
- // Memoize creates an Option to set the memoize flag to b. When set to true,
- // the parser will cache all results so each expression is evaluated only
- // once. This guarantees linear parsing time even for pathological cases,
- // at the expense of more memory and slower times for typical cases.
- //
- // The default is false.
- func Memoize(b bool) Option {
- return func(p *parser) Option {
- old := p.memoize
- p.memoize = b
- return Memoize(old)
- }
- }
- // AllowInvalidUTF8 creates an Option to allow invalid UTF-8 bytes.
- // Every invalid UTF-8 byte is treated as a utf8.RuneError (U+FFFD)
- // by character class matchers and is matched by the any matcher.
- // The returned matched value, c.text and c.offset are NOT affected.
- //
- // The default is false.
- func AllowInvalidUTF8(b bool) Option {
- return func(p *parser) Option {
- old := p.allowInvalidUTF8
- p.allowInvalidUTF8 = b
- return AllowInvalidUTF8(old)
- }
- }
- // Recover creates an Option to set the recover flag to b. When set to
- // true, this causes the parser to recover from panics and convert it
- // to an error. Setting it to false can be useful while debugging to
- // access the full stack trace.
- //
- // The default is true.
- func Recover(b bool) Option {
- return func(p *parser) Option {
- old := p.recover
- p.recover = b
- return Recover(old)
- }
- }
- // GlobalStore creates an Option to set a key to a certain value in
- // the globalStore.
- func GlobalStore(key string, value interface{}) Option {
- return func(p *parser) Option {
- old := p.cur.globalStore[key]
- p.cur.globalStore[key] = value
- return GlobalStore(key, old)
- }
- }
- // InitState creates an Option to set a key to a certain value in
- // the global "state" store.
- func InitState(key string, value interface{}) Option {
- return func(p *parser) Option {
- old := p.cur.state[key]
- p.cur.state[key] = value
- return InitState(key, old)
- }
- }
- // ParseFile parses the file identified by filename.
- func ParseFile(filename string, opts ...Option) (i interface{}, err error) {
- f, err := os.Open(filename)
- if err != nil {
- return nil, err
- }
- defer func() {
- if closeErr := f.Close(); closeErr != nil {
- err = closeErr
- }
- }()
- return ParseReader(filename, f, opts...)
- }
- // ParseReader parses the data from r using filename as information in the
- // error messages.
- func ParseReader(filename string, r io.Reader, opts ...Option) (interface{}, error) {
- b, err := ioutil.ReadAll(r)
- if err != nil {
- return nil, err
- }
- return Parse(filename, b, opts...)
- }
- // Parse parses the data from b using filename as information in the
- // error messages.
- func Parse(filename string, b []byte, opts ...Option) (interface{}, error) {
- return newParser(filename, b, opts...).parse(g)
- }
- // position records a position in the text.
- type position struct {
- line, col, offset int
- }
- func (p position) String() string {
- return strconv.Itoa(p.line) + ":" + strconv.Itoa(p.col) + " [" + strconv.Itoa(p.offset) + "]"
- }
- // savepoint stores all state required to go back to this point in the
- // parser.
- type savepoint struct {
- position
- rn rune
- w int
- }
- type current struct {
- pos position // start position of the match
- text []byte // raw text of the match
- // state is a store for arbitrary key,value pairs that the user wants to be
- // tied to the backtracking of the parser.
- // This is always rolled back if a parsing rule fails.
- state storeDict
- // globalStore is a general store for the user to store arbitrary key-value
- // pairs that they need to manage and that they do not want tied to the
- // backtracking of the parser. This is only modified by the user and never
- // rolled back by the parser. It is always up to the user to keep this in a
- // consistent state.
- globalStore storeDict
- }
- type storeDict map[string]interface{}
- // the AST types...
- type grammar struct {
- pos position
- rules []*rule
- }
- type rule struct {
- pos position
- name string
- displayName string
- expr interface{}
- }
- type choiceExpr struct {
- pos position
- alternatives []interface{}
- }
- type actionExpr struct {
- pos position
- expr interface{}
- run func(*parser) (interface{}, error)
- }
- type recoveryExpr struct {
- pos position
- expr interface{}
- recoverExpr interface{}
- failureLabel []string
- }
- type seqExpr struct {
- pos position
- exprs []interface{}
- }
- type throwExpr struct {
- pos position
- label string
- }
- type labeledExpr struct {
- pos position
- label string
- expr interface{}
- }
- type expr struct {
- pos position
- expr interface{}
- }
- type andExpr expr
- type notExpr expr
- type zeroOrOneExpr expr
- type zeroOrMoreExpr expr
- type oneOrMoreExpr expr
- type ruleRefExpr struct {
- pos position
- name string
- }
- type stateCodeExpr struct {
- pos position
- run func(*parser) error
- }
- type andCodeExpr struct {
- pos position
- run func(*parser) (bool, error)
- }
- type notCodeExpr struct {
- pos position
- run func(*parser) (bool, error)
- }
- type litMatcher struct {
- pos position
- val string
- ignoreCase bool
- want string
- }
- type charClassMatcher struct {
- pos position
- val string
- basicLatinChars [128]bool
- chars []rune
- ranges []rune
- classes []*unicode.RangeTable
- ignoreCase bool
- inverted bool
- }
- type anyMatcher position
- // errList cumulates the errors found by the parser.
- type errList []error
- func (e *errList) add(err error) {
- *e = append(*e, err)
- }
- func (e errList) err() error {
- if len(e) == 0 {
- return nil
- }
- e.dedupe()
- return e
- }
- func (e *errList) dedupe() {
- var cleaned []error
- set := make(map[string]bool)
- for _, err := range *e {
- if msg := err.Error(); !set[msg] {
- set[msg] = true
- cleaned = append(cleaned, err)
- }
- }
- *e = cleaned
- }
- func (e errList) Error() string {
- switch len(e) {
- case 0:
- return ""
- case 1:
- return e[0].Error()
- default:
- var buf bytes.Buffer
- for i, err := range e {
- if i > 0 {
- buf.WriteRune('\n')
- }
- buf.WriteString(err.Error())
- }
- return buf.String()
- }
- }
- // parserError wraps an error with a prefix indicating the rule in which
- // the error occurred. The original error is stored in the Inner field.
- type parserError struct {
- Inner error
- pos position
- prefix string
- expected []string
- }
- // Error returns the error message.
- func (p *parserError) Error() string {
- return p.prefix + ": " + p.Inner.Error()
- }
- // newParser creates a parser with the specified input source and options.
- func newParser(filename string, b []byte, opts ...Option) *parser {
- stats := Stats{
- ChoiceAltCnt: make(map[string]map[string]int),
- }
- p := &parser{
- filename: filename,
- errs: new(errList),
- data: b,
- pt: savepoint{position: position{line: 1}},
- recover: true,
- cur: current{
- state: make(storeDict),
- globalStore: make(storeDict),
- },
- maxFailPos: position{col: 1, line: 1},
- maxFailExpected: make([]string, 0, 20),
- Stats: &stats,
- // start rule is rule [0] unless an alternate entrypoint is specified
- entrypoint: g.rules[0].name,
- }
- p.setOptions(opts)
- if p.maxExprCnt == 0 {
- p.maxExprCnt = math.MaxUint64
- }
- return p
- }
- // setOptions applies the options to the parser.
- func (p *parser) setOptions(opts []Option) {
- for _, opt := range opts {
- opt(p)
- }
- }
- type resultTuple struct {
- v interface{}
- b bool
- end savepoint
- }
- const choiceNoMatch = -1
- // Stats stores some statistics, gathered during parsing
- type Stats struct {
- // ExprCnt counts the number of expressions processed during parsing
- // This value is compared to the maximum number of expressions allowed
- // (set by the MaxExpressions option).
- ExprCnt uint64
- // ChoiceAltCnt is used to count for each ordered choice expression,
- // which alternative is used how may times.
- // These numbers allow to optimize the order of the ordered choice expression
- // to increase the performance of the parser
- //
- // The outer key of ChoiceAltCnt is composed of the name of the rule as well
- // as the line and the column of the ordered choice.
- // The inner key of ChoiceAltCnt is the number (one-based) of the matching alternative.
- // For each alternative the number of matches are counted. If an ordered choice does not
- // match, a special counter is incremented. The name of this counter is set with
- // the parser option Statistics.
- // For an alternative to be included in ChoiceAltCnt, it has to match at least once.
- ChoiceAltCnt map[string]map[string]int
- }
- type parser struct {
- filename string
- pt savepoint
- cur current
- data []byte
- errs *errList
- depth int
- recover bool
- debug bool
- memoize bool
- // memoization table for the packrat algorithm:
- // map[offset in source] map[expression or rule] {value, match}
- memo map[int]map[interface{}]resultTuple
- // rules table, maps the rule identifier to the rule node
- rules map[string]*rule
- // variables stack, map of label to value
- vstack []map[string]interface{}
- // rule stack, allows identification of the current rule in errors
- rstack []*rule
- // parse fail
- maxFailPos position
- maxFailExpected []string
- maxFailInvertExpected bool
- // max number of expressions to be parsed
- maxExprCnt uint64
- // entrypoint for the parser
- entrypoint string
- allowInvalidUTF8 bool
- *Stats
- choiceNoMatch string
- // recovery expression stack, keeps track of the currently available recovery expression, these are traversed in reverse
- recoveryStack []map[string]interface{}
- }
- // push a variable set on the vstack.
- func (p *parser) pushV() {
- if cap(p.vstack) == len(p.vstack) {
- // create new empty slot in the stack
- p.vstack = append(p.vstack, nil)
- } else {
- // slice to 1 more
- p.vstack = p.vstack[:len(p.vstack)+1]
- }
- // get the last args set
- m := p.vstack[len(p.vstack)-1]
- if m != nil && len(m) == 0 {
- // empty map, all good
- return
- }
- m = make(map[string]interface{})
- p.vstack[len(p.vstack)-1] = m
- }
- // pop a variable set from the vstack.
- func (p *parser) popV() {
- // if the map is not empty, clear it
- m := p.vstack[len(p.vstack)-1]
- if len(m) > 0 {
- // GC that map
- p.vstack[len(p.vstack)-1] = nil
- }
- p.vstack = p.vstack[:len(p.vstack)-1]
- }
- // push a recovery expression with its labels to the recoveryStack
- func (p *parser) pushRecovery(labels []string, expr interface{}) {
- if cap(p.recoveryStack) == len(p.recoveryStack) {
- // create new empty slot in the stack
- p.recoveryStack = append(p.recoveryStack, nil)
- } else {
- // slice to 1 more
- p.recoveryStack = p.recoveryStack[:len(p.recoveryStack)+1]
- }
- m := make(map[string]interface{}, len(labels))
- for _, fl := range labels {
- m[fl] = expr
- }
- p.recoveryStack[len(p.recoveryStack)-1] = m
- }
- // pop a recovery expression from the recoveryStack
- func (p *parser) popRecovery() {
- // GC that map
- p.recoveryStack[len(p.recoveryStack)-1] = nil
- p.recoveryStack = p.recoveryStack[:len(p.recoveryStack)-1]
- }
- func (p *parser) print(prefix, s string) string {
- if !p.debug {
- return s
- }
- fmt.Printf("%s %d:%d:%d: %s [%#U]\n",
- prefix, p.pt.line, p.pt.col, p.pt.offset, s, p.pt.rn)
- return s
- }
- func (p *parser) in(s string) string {
- p.depth++
- return p.print(strings.Repeat(" ", p.depth)+">", s)
- }
- func (p *parser) out(s string) string {
- p.depth--
- return p.print(strings.Repeat(" ", p.depth)+"<", s)
- }
- func (p *parser) addErr(err error) {
- p.addErrAt(err, p.pt.position, []string{})
- }
- func (p *parser) addErrAt(err error, pos position, expected []string) {
- var buf bytes.Buffer
- if p.filename != "" {
- buf.WriteString(p.filename)
- }
- if buf.Len() > 0 {
- buf.WriteString(":")
- }
- buf.WriteString(fmt.Sprintf("%d:%d (%d)", pos.line, pos.col, pos.offset))
- if len(p.rstack) > 0 {
- if buf.Len() > 0 {
- buf.WriteString(": ")
- }
- rule := p.rstack[len(p.rstack)-1]
- if rule.displayName != "" {
- buf.WriteString("rule " + rule.displayName)
- } else {
- buf.WriteString("rule " + rule.name)
- }
- }
- pe := &parserError{Inner: err, pos: pos, prefix: buf.String(), expected: expected}
- p.errs.add(pe)
- }
- func (p *parser) failAt(fail bool, pos position, want string) {
- // process fail if parsing fails and not inverted or parsing succeeds and invert is set
- if fail == p.maxFailInvertExpected {
- if pos.offset < p.maxFailPos.offset {
- return
- }
- if pos.offset > p.maxFailPos.offset {
- p.maxFailPos = pos
- p.maxFailExpected = p.maxFailExpected[:0]
- }
- if p.maxFailInvertExpected {
- want = "!" + want
- }
- p.maxFailExpected = append(p.maxFailExpected, want)
- }
- }
- // read advances the parser to the next rune.
- func (p *parser) read() {
- p.pt.offset += p.pt.w
- rn, n := utf8.DecodeRune(p.data[p.pt.offset:])
- p.pt.rn = rn
- p.pt.w = n
- p.pt.col++
- if rn == '\n' {
- p.pt.line++
- p.pt.col = 0
- }
- if rn == utf8.RuneError && n == 1 { // see utf8.DecodeRune
- if !p.allowInvalidUTF8 {
- p.addErr(errInvalidEncoding)
- }
- }
- }
- // restore parser position to the savepoint pt.
- func (p *parser) restore(pt savepoint) {
- if p.debug {
- defer p.out(p.in("restore"))
- }
- if pt.offset == p.pt.offset {
- return
- }
- p.pt = pt
- }
- // Cloner is implemented by any value that has a Clone method, which returns a
- // copy of the value. This is mainly used for types which are not passed by
- // value (e.g map, slice, chan) or structs that contain such types.
- //
- // This is used in conjunction with the global state feature to create proper
- // copies of the state to allow the parser to properly restore the state in
- // the case of backtracking.
- type Cloner interface {
- Clone() interface{}
- }
- var statePool = &sync.Pool{
- New: func() interface{} { return make(storeDict) },
- }
- func (sd storeDict) Discard() {
- for k := range sd {
- delete(sd, k)
- }
- statePool.Put(sd)
- }
- // clone and return parser current state.
- func (p *parser) cloneState() storeDict {
- if p.debug {
- defer p.out(p.in("cloneState"))
- }
- state := statePool.Get().(storeDict)
- for k, v := range p.cur.state {
- if c, ok := v.(Cloner); ok {
- state[k] = c.Clone()
- } else {
- state[k] = v
- }
- }
- return state
- }
- // restore parser current state to the state storeDict.
- // every restoreState should applied only one time for every cloned state
- func (p *parser) restoreState(state storeDict) {
- if p.debug {
- defer p.out(p.in("restoreState"))
- }
- p.cur.state.Discard()
- p.cur.state = state
- }
- // get the slice of bytes from the savepoint start to the current position.
- func (p *parser) sliceFrom(start savepoint) []byte {
- return p.data[start.position.offset:p.pt.position.offset]
- }
- func (p *parser) getMemoized(node interface{}) (resultTuple, bool) {
- if len(p.memo) == 0 {
- return resultTuple{}, false
- }
- m := p.memo[p.pt.offset]
- if len(m) == 0 {
- return resultTuple{}, false
- }
- res, ok := m[node]
- return res, ok
- }
- func (p *parser) setMemoized(pt savepoint, node interface{}, tuple resultTuple) {
- if p.memo == nil {
- p.memo = make(map[int]map[interface{}]resultTuple)
- }
- m := p.memo[pt.offset]
- if m == nil {
- m = make(map[interface{}]resultTuple)
- p.memo[pt.offset] = m
- }
- m[node] = tuple
- }
- func (p *parser) buildRulesTable(g *grammar) {
- p.rules = make(map[string]*rule, len(g.rules))
- for _, r := range g.rules {
- p.rules[r.name] = r
- }
- }
- func (p *parser) parse(g *grammar) (val interface{}, err error) {
- if len(g.rules) == 0 {
- p.addErr(errNoRule)
- return nil, p.errs.err()
- }
- // TODO : not super critical but this could be generated
- p.buildRulesTable(g)
- if p.recover {
- // panic can be used in action code to stop parsing immediately
- // and return the panic as an error.
- defer func() {
- if e := recover(); e != nil {
- if p.debug {
- defer p.out(p.in("panic handler"))
- }
- val = nil
- switch e := e.(type) {
- case error:
- p.addErr(e)
- default:
- p.addErr(fmt.Errorf("%v", e))
- }
- err = p.errs.err()
- }
- }()
- }
- startRule, ok := p.rules[p.entrypoint]
- if !ok {
- p.addErr(errInvalidEntrypoint)
- return nil, p.errs.err()
- }
- p.read() // advance to first rune
- val, ok = p.parseRule(startRule)
- if !ok {
- if len(*p.errs) == 0 {
- // If parsing fails, but no errors have been recorded, the expected values
- // for the farthest parser position are returned as error.
- maxFailExpectedMap := make(map[string]struct{}, len(p.maxFailExpected))
- for _, v := range p.maxFailExpected {
- maxFailExpectedMap[v] = struct{}{}
- }
- expected := make([]string, 0, len(maxFailExpectedMap))
- eof := false
- if _, ok := maxFailExpectedMap["!."]; ok {
- delete(maxFailExpectedMap, "!.")
- eof = true
- }
- for k := range maxFailExpectedMap {
- expected = append(expected, k)
- }
- sort.Strings(expected)
- if eof {
- expected = append(expected, "EOF")
- }
- p.addErrAt(errors.New("no match found, expected: "+listJoin(expected, ", ", "or")), p.maxFailPos, expected)
- }
- return nil, p.errs.err()
- }
- return val, p.errs.err()
- }
- func listJoin(list []string, sep string, lastSep string) string {
- switch len(list) {
- case 0:
- return ""
- case 1:
- return list[0]
- default:
- return strings.Join(list[:len(list)-1], sep) + " " + lastSep + " " + list[len(list)-1]
- }
- }
- func (p *parser) parseRule(rule *rule) (interface{}, bool) {
- if p.debug {
- defer p.out(p.in("parseRule " + rule.name))
- }
- if p.memoize {
- res, ok := p.getMemoized(rule)
- if ok {
- p.restore(res.end)
- return res.v, res.b
- }
- }
- start := p.pt
- p.rstack = append(p.rstack, rule)
- p.pushV()
- val, ok := p.parseExpr(rule.expr)
- p.popV()
- p.rstack = p.rstack[:len(p.rstack)-1]
- if ok && p.debug {
- p.print(strings.Repeat(" ", p.depth)+"MATCH", string(p.sliceFrom(start)))
- }
- if p.memoize {
- p.setMemoized(start, rule, resultTuple{val, ok, p.pt})
- }
- return val, ok
- }
- func (p *parser) parseExpr(expr interface{}) (interface{}, bool) {
- var pt savepoint
- if p.memoize {
- res, ok := p.getMemoized(expr)
- if ok {
- p.restore(res.end)
- return res.v, res.b
- }
- pt = p.pt
- }
- p.ExprCnt++
- if p.ExprCnt > p.maxExprCnt {
- panic(errMaxExprCnt)
- }
- var val interface{}
- var ok bool
- switch expr := expr.(type) {
- case *actionExpr:
- val, ok = p.parseActionExpr(expr)
- case *andCodeExpr:
- val, ok = p.parseAndCodeExpr(expr)
- case *andExpr:
- val, ok = p.parseAndExpr(expr)
- case *anyMatcher:
- val, ok = p.parseAnyMatcher(expr)
- case *charClassMatcher:
- val, ok = p.parseCharClassMatcher(expr)
- case *choiceExpr:
- val, ok = p.parseChoiceExpr(expr)
- case *labeledExpr:
- val, ok = p.parseLabeledExpr(expr)
- case *litMatcher:
- val, ok = p.parseLitMatcher(expr)
- case *notCodeExpr:
- val, ok = p.parseNotCodeExpr(expr)
- case *notExpr:
- val, ok = p.parseNotExpr(expr)
- case *oneOrMoreExpr:
- val, ok = p.parseOneOrMoreExpr(expr)
- case *recoveryExpr:
- val, ok = p.parseRecoveryExpr(expr)
- case *ruleRefExpr:
- val, ok = p.parseRuleRefExpr(expr)
- case *seqExpr:
- val, ok = p.parseSeqExpr(expr)
- case *stateCodeExpr:
- val, ok = p.parseStateCodeExpr(expr)
- case *throwExpr:
- val, ok = p.parseThrowExpr(expr)
- case *zeroOrMoreExpr:
- val, ok = p.parseZeroOrMoreExpr(expr)
- case *zeroOrOneExpr:
- val, ok = p.parseZeroOrOneExpr(expr)
- default:
- panic(fmt.Sprintf("unknown expression type %T", expr))
- }
- if p.memoize {
- p.setMemoized(pt, expr, resultTuple{val, ok, p.pt})
- }
- return val, ok
- }
- func (p *parser) parseActionExpr(act *actionExpr) (interface{}, bool) {
- if p.debug {
- defer p.out(p.in("parseActionExpr"))
- }
- start := p.pt
- val, ok := p.parseExpr(act.expr)
- if ok {
- p.cur.pos = start.position
- p.cur.text = p.sliceFrom(start)
- state := p.cloneState()
- actVal, err := act.run(p)
- if err != nil {
- p.addErrAt(err, start.position, []string{})
- }
- p.restoreState(state)
- val = actVal
- }
- if ok && p.debug {
- p.print(strings.Repeat(" ", p.depth)+"MATCH", string(p.sliceFrom(start)))
- }
- return val, ok
- }
- func (p *parser) parseAndCodeExpr(and *andCodeExpr) (interface{}, bool) {
- if p.debug {
- defer p.out(p.in("parseAndCodeExpr"))
- }
- state := p.cloneState()
- ok, err := and.run(p)
- if err != nil {
- p.addErr(err)
- }
- p.restoreState(state)
- return nil, ok
- }
- func (p *parser) parseAndExpr(and *andExpr) (interface{}, bool) {
- if p.debug {
- defer p.out(p.in("parseAndExpr"))
- }
- pt := p.pt
- state := p.cloneState()
- p.pushV()
- _, ok := p.parseExpr(and.expr)
- p.popV()
- p.restoreState(state)
- p.restore(pt)
- return nil, ok
- }
- func (p *parser) parseAnyMatcher(any *anyMatcher) (interface{}, bool) {
- if p.debug {
- defer p.out(p.in("parseAnyMatcher"))
- }
- if p.pt.rn == utf8.RuneError && p.pt.w == 0 {
- // EOF - see utf8.DecodeRune
- p.failAt(false, p.pt.position, ".")
- return nil, false
- }
- start := p.pt
- p.read()
- p.failAt(true, start.position, ".")
- return p.sliceFrom(start), true
- }
- func (p *parser) parseCharClassMatcher(chr *charClassMatcher) (interface{}, bool) {
- if p.debug {
- defer p.out(p.in("parseCharClassMatcher"))
- }
- cur := p.pt.rn
- start := p.pt
- // can't match EOF
- if cur == utf8.RuneError && p.pt.w == 0 { // see utf8.DecodeRune
- p.failAt(false, start.position, chr.val)
- return nil, false
- }
- if chr.ignoreCase {
- cur = unicode.ToLower(cur)
- }
- // try to match in the list of available chars
- for _, rn := range chr.chars {
- if rn == cur {
- if chr.inverted {
- p.failAt(false, start.position, chr.val)
- return nil, false
- }
- p.read()
- p.failAt(true, start.position, chr.val)
- return p.sliceFrom(start), true
- }
- }
- // try to match in the list of ranges
- for i := 0; i < len(chr.ranges); i += 2 {
- if cur >= chr.ranges[i] && cur <= chr.ranges[i+1] {
- if chr.inverted {
- p.failAt(false, start.position, chr.val)
- return nil, false
- }
- p.read()
- p.failAt(true, start.position, chr.val)
- return p.sliceFrom(start), true
- }
- }
- // try to match in the list of Unicode classes
- for _, cl := range chr.classes {
- if unicode.Is(cl, cur) {
- if chr.inverted {
- p.failAt(false, start.position, chr.val)
- return nil, false
- }
- p.read()
- p.failAt(true, start.position, chr.val)
- return p.sliceFrom(start), true
- }
- }
- if chr.inverted {
- p.read()
- p.failAt(true, start.position, chr.val)
- return p.sliceFrom(start), true
- }
- p.failAt(false, start.position, chr.val)
- return nil, false
- }
- func (p *parser) incChoiceAltCnt(ch *choiceExpr, altI int) {
- choiceIdent := fmt.Sprintf("%s %d:%d", p.rstack[len(p.rstack)-1].name, ch.pos.line, ch.pos.col)
- m := p.ChoiceAltCnt[choiceIdent]
- if m == nil {
- m = make(map[string]int)
- p.ChoiceAltCnt[choiceIdent] = m
- }
- // We increment altI by 1, so the keys do not start at 0
- alt := strconv.Itoa(altI + 1)
- if altI == choiceNoMatch {
- alt = p.choiceNoMatch
- }
- m[alt]++
- }
- func (p *parser) parseChoiceExpr(ch *choiceExpr) (interface{}, bool) {
- if p.debug {
- defer p.out(p.in("parseChoiceExpr"))
- }
- for altI, alt := range ch.alternatives {
- // dummy assignment to prevent compile error if optimized
- _ = altI
- state := p.cloneState()
- p.pushV()
- val, ok := p.parseExpr(alt)
- p.popV()
- if ok {
- p.incChoiceAltCnt(ch, altI)
- return val, ok
- }
- p.restoreState(state)
- }
- p.incChoiceAltCnt(ch, choiceNoMatch)
- return nil, false
- }
- func (p *parser) parseLabeledExpr(lab *labeledExpr) (interface{}, bool) {
- if p.debug {
- defer p.out(p.in("parseLabeledExpr"))
- }
- p.pushV()
- val, ok := p.parseExpr(lab.expr)
- p.popV()
- if ok && lab.label != "" {
- m := p.vstack[len(p.vstack)-1]
- m[lab.label] = val
- }
- return val, ok
- }
- func (p *parser) parseLitMatcher(lit *litMatcher) (interface{}, bool) {
- if p.debug {
- defer p.out(p.in("parseLitMatcher"))
- }
- start := p.pt
- for _, want := range lit.val {
- cur := p.pt.rn
- if lit.ignoreCase {
- cur = unicode.ToLower(cur)
- }
- if cur != want {
- p.failAt(false, start.position, lit.want)
- p.restore(start)
- return nil, false
- }
- p.read()
- }
- p.failAt(true, start.position, lit.want)
- return p.sliceFrom(start), true
- }
- func (p *parser) parseNotCodeExpr(not *notCodeExpr) (interface{}, bool) {
- if p.debug {
- defer p.out(p.in("parseNotCodeExpr"))
- }
- state := p.cloneState()
- ok, err := not.run(p)
- if err != nil {
- p.addErr(err)
- }
- p.restoreState(state)
- return nil, !ok
- }
- func (p *parser) parseNotExpr(not *notExpr) (interface{}, bool) {
- if p.debug {
- defer p.out(p.in("parseNotExpr"))
- }
- pt := p.pt
- state := p.cloneState()
- p.pushV()
- p.maxFailInvertExpected = !p.maxFailInvertExpected
- _, ok := p.parseExpr(not.expr)
- p.maxFailInvertExpected = !p.maxFailInvertExpected
- p.popV()
- p.restoreState(state)
- p.restore(pt)
- return nil, !ok
- }
- func (p *parser) parseOneOrMoreExpr(expr *oneOrMoreExpr) (interface{}, bool) {
- if p.debug {
- defer p.out(p.in("parseOneOrMoreExpr"))
- }
- var vals []interface{}
- for {
- p.pushV()
- val, ok := p.parseExpr(expr.expr)
- p.popV()
- if !ok {
- if len(vals) == 0 {
- // did not match once, no match
- return nil, false
- }
- return vals, true
- }
- vals = append(vals, val)
- }
- }
- func (p *parser) parseRecoveryExpr(recover *recoveryExpr) (interface{}, bool) {
- if p.debug {
- defer p.out(p.in("parseRecoveryExpr (" + strings.Join(recover.failureLabel, ",") + ")"))
- }
- p.pushRecovery(recover.failureLabel, recover.recoverExpr)
- val, ok := p.parseExpr(recover.expr)
- p.popRecovery()
- return val, ok
- }
- func (p *parser) parseRuleRefExpr(ref *ruleRefExpr) (interface{}, bool) {
- if p.debug {
- defer p.out(p.in("parseRuleRefExpr " + ref.name))
- }
- if ref.name == "" {
- panic(fmt.Sprintf("%s: invalid rule: missing name", ref.pos))
- }
- rule := p.rules[ref.name]
- if rule == nil {
- p.addErr(fmt.Errorf("undefined rule: %s", ref.name))
- return nil, false
- }
- return p.parseRule(rule)
- }
- func (p *parser) parseSeqExpr(seq *seqExpr) (interface{}, bool) {
- if p.debug {
- defer p.out(p.in("parseSeqExpr"))
- }
- vals := make([]interface{}, 0, len(seq.exprs))
- pt := p.pt
- state := p.cloneState()
- for _, expr := range seq.exprs {
- val, ok := p.parseExpr(expr)
- if !ok {
- p.restoreState(state)
- p.restore(pt)
- return nil, false
- }
- vals = append(vals, val)
- }
- return vals, true
- }
- func (p *parser) parseStateCodeExpr(state *stateCodeExpr) (interface{}, bool) {
- if p.debug {
- defer p.out(p.in("parseStateCodeExpr"))
- }
- err := state.run(p)
- if err != nil {
- p.addErr(err)
- }
- return nil, true
- }
- func (p *parser) parseThrowExpr(expr *throwExpr) (interface{}, bool) {
- if p.debug {
- defer p.out(p.in("parseThrowExpr"))
- }
- for i := len(p.recoveryStack) - 1; i >= 0; i-- {
- if recoverExpr, ok := p.recoveryStack[i][expr.label]; ok {
- if val, ok := p.parseExpr(recoverExpr); ok {
- return val, ok
- }
- }
- }
- return nil, false
- }
- func (p *parser) parseZeroOrMoreExpr(expr *zeroOrMoreExpr) (interface{}, bool) {
- if p.debug {
- defer p.out(p.in("parseZeroOrMoreExpr"))
- }
- var vals []interface{}
- for {
- p.pushV()
- val, ok := p.parseExpr(expr.expr)
- p.popV()
- if !ok {
- return vals, true
- }
- vals = append(vals, val)
- }
- }
- func (p *parser) parseZeroOrOneExpr(expr *zeroOrOneExpr) (interface{}, bool) {
- if p.debug {
- defer p.out(p.in("parseZeroOrOneExpr"))
- }
- p.pushV()
- val, _ := p.parseExpr(expr.expr)
- p.popV()
- // whether it matched or not, consider it a match
- return val, true
- }
|