recommit generated code

This commit is contained in:
Megan Marsh 2019-06-26 11:06:06 -07:00
parent ded67a29b3
commit 823e2e260f
1 changed files with 468 additions and 53 deletions

View File

@ -1,3 +1,5 @@
// Code generated by pigeon; DO NOT EDIT.
package bootcommand package bootcommand
import ( import (
@ -6,7 +8,9 @@ import (
"fmt" "fmt"
"io" "io"
"io/ioutil" "io/ioutil"
"math"
"os" "os"
"sort"
"strconv" "strconv"
"strings" "strings"
"time" "time"
@ -787,18 +791,85 @@ var (
// errNoRule is returned when the grammar to parse has no rule. // errNoRule is returned when the grammar to parse has no rule.
errNoRule = errors.New("grammar 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 // errInvalidEncoding is returned when the source is not properly
// utf8-encoded. // utf8-encoded.
errInvalidEncoding = errors.New("invalid encoding") errInvalidEncoding = errors.New("invalid encoding")
// errNoMatch is returned if no match could be found. // errMaxExprCnt is used to signal that the maximum number of
errNoMatch = errors.New("no match found") // 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 // Option is a function that can set an option on the parser. It returns
// the previous setting as an Option. // the previous setting as an Option.
type Option func(*parser) 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, // Debug creates an Option to set the debug flag to b. When set to true,
// debugging information is printed to stdout while parsing. // debugging information is printed to stdout while parsing.
// //
@ -825,6 +896,20 @@ func Memoize(b bool) Option {
} }
} }
// 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 // 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 // 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 // to an error. Setting it to false can be useful while debugging to
@ -839,13 +924,37 @@ func Recover(b bool) Option {
} }
} }
// 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. // ParseFile parses the file identified by filename.
func ParseFile(filename string, opts ...Option) (interface{}, error) { func ParseFile(filename string, opts ...Option) (i interface{}, err error) {
f, err := os.Open(filename) f, err := os.Open(filename)
if err != nil { if err != nil {
return nil, err return nil, err
} }
defer f.Close() defer func() {
if closeErr := f.Close(); closeErr != nil {
err = closeErr
}
}()
return ParseReader(filename, f, opts...) return ParseReader(filename, f, opts...)
} }
@ -886,8 +995,22 @@ type savepoint struct {
type current struct { type current struct {
pos position // start position of the match pos position // start position of the match
text []byte // raw text 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... // the AST types...
type grammar struct { type grammar struct {
@ -913,11 +1036,23 @@ type actionExpr struct {
run func(*parser) (interface{}, error) run func(*parser) (interface{}, error)
} }
type recoveryExpr struct {
pos position
expr interface{}
recoverExpr interface{}
failureLabel []string
}
type seqExpr struct { type seqExpr struct {
pos position pos position
exprs []interface{} exprs []interface{}
} }
type throwExpr struct {
pos position
label string
}
type labeledExpr struct { type labeledExpr struct {
pos position pos position
label string label string
@ -940,6 +1075,11 @@ type ruleRefExpr struct {
name string name string
} }
type stateCodeExpr struct {
pos position
run func(*parser) error
}
type andCodeExpr struct { type andCodeExpr struct {
pos position pos position
run func(*parser) (bool, error) run func(*parser) (bool, error)
@ -957,13 +1097,14 @@ type litMatcher struct {
} }
type charClassMatcher struct { type charClassMatcher struct {
pos position pos position
val string val string
chars []rune basicLatinChars [128]bool
ranges []rune chars []rune
classes []*unicode.RangeTable ranges []rune
ignoreCase bool classes []*unicode.RangeTable
inverted bool ignoreCase bool
inverted bool
} }
type anyMatcher position type anyMatcher position
@ -1017,9 +1158,10 @@ func (e errList) Error() string {
// parserError wraps an error with a prefix indicating the rule in which // parserError wraps an error with a prefix indicating the rule in which
// the error occurred. The original error is stored in the Inner field. // the error occurred. The original error is stored in the Inner field.
type parserError struct { type parserError struct {
Inner error Inner error
pos position pos position
prefix string prefix string
expected []string
} }
// Error returns the error message. // Error returns the error message.
@ -1029,14 +1171,32 @@ func (p *parserError) Error() string {
// newParser creates a parser with the specified input source and options. // newParser creates a parser with the specified input source and options.
func newParser(filename string, b []byte, opts ...Option) *parser { func newParser(filename string, b []byte, opts ...Option) *parser {
stats := Stats{
ChoiceAltCnt: make(map[string]map[string]int),
}
p := &parser{ p := &parser{
filename: filename, filename: filename,
errs: new(errList), errs: new(errList),
data: b, data: b,
pt: savepoint{position: position{line: 1}}, pt: savepoint{position: position{line: 1}},
recover: true, 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) p.setOptions(opts)
if p.maxExprCnt == 0 {
p.maxExprCnt = math.MaxUint64
}
return p return p
} }
@ -1053,6 +1213,30 @@ type resultTuple struct {
end savepoint 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 { type parser struct {
filename string filename string
pt savepoint pt savepoint
@ -1061,9 +1245,9 @@ type parser struct {
data []byte data []byte
errs *errList errs *errList
depth int
recover bool recover bool
debug bool debug bool
depth int
memoize bool memoize bool
// memoization table for the packrat algorithm: // memoization table for the packrat algorithm:
@ -1077,8 +1261,23 @@ type parser struct {
// rule stack, allows identification of the current rule in errors // rule stack, allows identification of the current rule in errors
rstack []*rule rstack []*rule
// stats // parse fail
exprCnt int 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. // push a variable set on the vstack.
@ -1113,6 +1312,31 @@ func (p *parser) popV() {
p.vstack = p.vstack[:len(p.vstack)-1] 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 { func (p *parser) print(prefix, s string) string {
if !p.debug { if !p.debug {
return s return s
@ -1134,10 +1358,10 @@ func (p *parser) out(s string) string {
} }
func (p *parser) addErr(err error) { func (p *parser) addErr(err error) {
p.addErrAt(err, p.pt.position) p.addErrAt(err, p.pt.position, []string{})
} }
func (p *parser) addErrAt(err error, pos position) { func (p *parser) addErrAt(err error, pos position, expected []string) {
var buf bytes.Buffer var buf bytes.Buffer
if p.filename != "" { if p.filename != "" {
buf.WriteString(p.filename) buf.WriteString(p.filename)
@ -1157,10 +1381,29 @@ func (p *parser) addErrAt(err error, pos position) {
buf.WriteString("rule " + rule.name) buf.WriteString("rule " + rule.name)
} }
} }
pe := &parserError{Inner: err, pos: pos, prefix: buf.String()} pe := &parserError{Inner: err, pos: pos, prefix: buf.String(), expected: expected}
p.errs.add(pe) 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. // read advances the parser to the next rune.
func (p *parser) read() { func (p *parser) read() {
p.pt.offset += p.pt.w p.pt.offset += p.pt.w
@ -1173,8 +1416,8 @@ func (p *parser) read() {
p.pt.col = 0 p.pt.col = 0
} }
if rn == utf8.RuneError { if rn == utf8.RuneError && n == 1 { // see utf8.DecodeRune
if n == 1 { if !p.allowInvalidUTF8 {
p.addErr(errInvalidEncoding) p.addErr(errInvalidEncoding)
} }
} }
@ -1191,6 +1434,43 @@ func (p *parser) restore(pt savepoint) {
p.pt = pt 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{}
}
// clone and return parser current state.
func (p *parser) cloneState() storeDict {
if p.debug {
defer p.out(p.in("cloneState"))
}
state := make(storeDict, len(p.cur.state))
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 = state
}
// get the slice of bytes from the savepoint start to the current position. // get the slice of bytes from the savepoint start to the current position.
func (p *parser) sliceFrom(start savepoint) []byte { func (p *parser) sliceFrom(start savepoint) []byte {
return p.data[start.position.offset:p.pt.position.offset] return p.data[start.position.offset:p.pt.position.offset]
@ -1256,19 +1536,54 @@ func (p *parser) parse(g *grammar) (val interface{}, err error) {
}() }()
} }
// start rule is rule [0] startRule, ok := p.rules[p.entrypoint]
if !ok {
p.addErr(errInvalidEntrypoint)
return nil, p.errs.err()
}
p.read() // advance to first rune p.read() // advance to first rune
val, ok := p.parseRule(g.rules[0]) val, ok = p.parseRule(startRule)
if !ok { if !ok {
if len(*p.errs) == 0 { if len(*p.errs) == 0 {
// make sure this doesn't go out silently // If parsing fails, but no errors have been recorded, the expected values
p.addErr(errNoMatch) // 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 nil, p.errs.err()
} }
return val, 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 fmt.Sprintf("%s %s %s", strings.Join(list[:len(list)-1], sep), lastSep, list[len(list)-1])
}
}
func (p *parser) parseRule(rule *rule) (interface{}, bool) { func (p *parser) parseRule(rule *rule) (interface{}, bool) {
if p.debug { if p.debug {
defer p.out(p.in("parseRule " + rule.name)) defer p.out(p.in("parseRule " + rule.name))
@ -1300,7 +1615,6 @@ func (p *parser) parseRule(rule *rule) (interface{}, bool) {
func (p *parser) parseExpr(expr interface{}) (interface{}, bool) { func (p *parser) parseExpr(expr interface{}) (interface{}, bool) {
var pt savepoint var pt savepoint
var ok bool
if p.memoize { if p.memoize {
res, ok := p.getMemoized(expr) res, ok := p.getMemoized(expr)
@ -1311,8 +1625,13 @@ func (p *parser) parseExpr(expr interface{}) (interface{}, bool) {
pt = p.pt pt = p.pt
} }
p.exprCnt++ p.ExprCnt++
if p.ExprCnt > p.maxExprCnt {
panic(errMaxExprCnt)
}
var val interface{} var val interface{}
var ok bool
switch expr := expr.(type) { switch expr := expr.(type) {
case *actionExpr: case *actionExpr:
val, ok = p.parseActionExpr(expr) val, ok = p.parseActionExpr(expr)
@ -1336,10 +1655,16 @@ func (p *parser) parseExpr(expr interface{}) (interface{}, bool) {
val, ok = p.parseNotExpr(expr) val, ok = p.parseNotExpr(expr)
case *oneOrMoreExpr: case *oneOrMoreExpr:
val, ok = p.parseOneOrMoreExpr(expr) val, ok = p.parseOneOrMoreExpr(expr)
case *recoveryExpr:
val, ok = p.parseRecoveryExpr(expr)
case *ruleRefExpr: case *ruleRefExpr:
val, ok = p.parseRuleRefExpr(expr) val, ok = p.parseRuleRefExpr(expr)
case *seqExpr: case *seqExpr:
val, ok = p.parseSeqExpr(expr) val, ok = p.parseSeqExpr(expr)
case *stateCodeExpr:
val, ok = p.parseStateCodeExpr(expr)
case *throwExpr:
val, ok = p.parseThrowExpr(expr)
case *zeroOrMoreExpr: case *zeroOrMoreExpr:
val, ok = p.parseZeroOrMoreExpr(expr) val, ok = p.parseZeroOrMoreExpr(expr)
case *zeroOrOneExpr: case *zeroOrOneExpr:
@ -1363,10 +1688,13 @@ func (p *parser) parseActionExpr(act *actionExpr) (interface{}, bool) {
if ok { if ok {
p.cur.pos = start.position p.cur.pos = start.position
p.cur.text = p.sliceFrom(start) p.cur.text = p.sliceFrom(start)
state := p.cloneState()
actVal, err := act.run(p) actVal, err := act.run(p)
if err != nil { if err != nil {
p.addErrAt(err, start.position) p.addErrAt(err, start.position, []string{})
} }
p.restoreState(state)
val = actVal val = actVal
} }
if ok && p.debug { if ok && p.debug {
@ -1380,10 +1708,14 @@ func (p *parser) parseAndCodeExpr(and *andCodeExpr) (interface{}, bool) {
defer p.out(p.in("parseAndCodeExpr")) defer p.out(p.in("parseAndCodeExpr"))
} }
state := p.cloneState()
ok, err := and.run(p) ok, err := and.run(p)
if err != nil { if err != nil {
p.addErr(err) p.addErr(err)
} }
p.restoreState(state)
return nil, ok return nil, ok
} }
@ -1393,10 +1725,13 @@ func (p *parser) parseAndExpr(and *andExpr) (interface{}, bool) {
} }
pt := p.pt pt := p.pt
state := p.cloneState()
p.pushV() p.pushV()
_, ok := p.parseExpr(and.expr) _, ok := p.parseExpr(and.expr)
p.popV() p.popV()
p.restoreState(state)
p.restore(pt) p.restore(pt)
return nil, ok return nil, ok
} }
@ -1405,12 +1740,15 @@ func (p *parser) parseAnyMatcher(any *anyMatcher) (interface{}, bool) {
defer p.out(p.in("parseAnyMatcher")) defer p.out(p.in("parseAnyMatcher"))
} }
if p.pt.rn != utf8.RuneError { if p.pt.rn == utf8.RuneError && p.pt.w == 0 {
start := p.pt // EOF - see utf8.DecodeRune
p.read() p.failAt(false, p.pt.position, ".")
return p.sliceFrom(start), true return nil, false
} }
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) { func (p *parser) parseCharClassMatcher(chr *charClassMatcher) (interface{}, bool) {
@ -1419,11 +1757,14 @@ func (p *parser) parseCharClassMatcher(chr *charClassMatcher) (interface{}, bool
} }
cur := p.pt.rn cur := p.pt.rn
start := p.pt
// can't match EOF // can't match EOF
if cur == utf8.RuneError { if cur == utf8.RuneError && p.pt.w == 0 { // see utf8.DecodeRune
p.failAt(false, start.position, chr.val)
return nil, false return nil, false
} }
start := p.pt
if chr.ignoreCase { if chr.ignoreCase {
cur = unicode.ToLower(cur) cur = unicode.ToLower(cur)
} }
@ -1432,9 +1773,11 @@ func (p *parser) parseCharClassMatcher(chr *charClassMatcher) (interface{}, bool
for _, rn := range chr.chars { for _, rn := range chr.chars {
if rn == cur { if rn == cur {
if chr.inverted { if chr.inverted {
p.failAt(false, start.position, chr.val)
return nil, false return nil, false
} }
p.read() p.read()
p.failAt(true, start.position, chr.val)
return p.sliceFrom(start), true return p.sliceFrom(start), true
} }
} }
@ -1443,9 +1786,11 @@ func (p *parser) parseCharClassMatcher(chr *charClassMatcher) (interface{}, bool
for i := 0; i < len(chr.ranges); i += 2 { for i := 0; i < len(chr.ranges); i += 2 {
if cur >= chr.ranges[i] && cur <= chr.ranges[i+1] { if cur >= chr.ranges[i] && cur <= chr.ranges[i+1] {
if chr.inverted { if chr.inverted {
p.failAt(false, start.position, chr.val)
return nil, false return nil, false
} }
p.read() p.read()
p.failAt(true, start.position, chr.val)
return p.sliceFrom(start), true return p.sliceFrom(start), true
} }
} }
@ -1454,33 +1799,60 @@ func (p *parser) parseCharClassMatcher(chr *charClassMatcher) (interface{}, bool
for _, cl := range chr.classes { for _, cl := range chr.classes {
if unicode.Is(cl, cur) { if unicode.Is(cl, cur) {
if chr.inverted { if chr.inverted {
p.failAt(false, start.position, chr.val)
return nil, false return nil, false
} }
p.read() p.read()
p.failAt(true, start.position, chr.val)
return p.sliceFrom(start), true return p.sliceFrom(start), true
} }
} }
if chr.inverted { if chr.inverted {
p.read() p.read()
p.failAt(true, start.position, chr.val)
return p.sliceFrom(start), true return p.sliceFrom(start), true
} }
p.failAt(false, start.position, chr.val)
return nil, false 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) { func (p *parser) parseChoiceExpr(ch *choiceExpr) (interface{}, bool) {
if p.debug { if p.debug {
defer p.out(p.in("parseChoiceExpr")) defer p.out(p.in("parseChoiceExpr"))
} }
for _, alt := range ch.alternatives { for altI, alt := range ch.alternatives {
// dummy assignment to prevent compile error if optimized
_ = altI
state := p.cloneState()
p.pushV() p.pushV()
val, ok := p.parseExpr(alt) val, ok := p.parseExpr(alt)
p.popV() p.popV()
if ok { if ok {
p.incChoiceAltCnt(ch, altI)
return val, ok return val, ok
} }
p.restoreState(state)
} }
p.incChoiceAltCnt(ch, choiceNoMatch)
return nil, false return nil, false
} }
@ -1504,6 +1876,11 @@ func (p *parser) parseLitMatcher(lit *litMatcher) (interface{}, bool) {
defer p.out(p.in("parseLitMatcher")) defer p.out(p.in("parseLitMatcher"))
} }
ignoreCase := ""
if lit.ignoreCase {
ignoreCase = "i"
}
val := fmt.Sprintf("%q%s", lit.val, ignoreCase)
start := p.pt start := p.pt
for _, want := range lit.val { for _, want := range lit.val {
cur := p.pt.rn cur := p.pt.rn
@ -1511,11 +1888,13 @@ func (p *parser) parseLitMatcher(lit *litMatcher) (interface{}, bool) {
cur = unicode.ToLower(cur) cur = unicode.ToLower(cur)
} }
if cur != want { if cur != want {
p.failAt(false, start.position, val)
p.restore(start) p.restore(start)
return nil, false return nil, false
} }
p.read() p.read()
} }
p.failAt(true, start.position, val)
return p.sliceFrom(start), true return p.sliceFrom(start), true
} }
@ -1524,10 +1903,14 @@ func (p *parser) parseNotCodeExpr(not *notCodeExpr) (interface{}, bool) {
defer p.out(p.in("parseNotCodeExpr")) defer p.out(p.in("parseNotCodeExpr"))
} }
state := p.cloneState()
ok, err := not.run(p) ok, err := not.run(p)
if err != nil { if err != nil {
p.addErr(err) p.addErr(err)
} }
p.restoreState(state)
return nil, !ok return nil, !ok
} }
@ -1537,10 +1920,15 @@ func (p *parser) parseNotExpr(not *notExpr) (interface{}, bool) {
} }
pt := p.pt pt := p.pt
state := p.cloneState()
p.pushV() p.pushV()
p.maxFailInvertExpected = !p.maxFailInvertExpected
_, ok := p.parseExpr(not.expr) _, ok := p.parseExpr(not.expr)
p.maxFailInvertExpected = !p.maxFailInvertExpected
p.popV() p.popV()
p.restoreState(state)
p.restore(pt) p.restore(pt)
return nil, !ok return nil, !ok
} }
@ -1566,6 +1954,18 @@ func (p *parser) parseOneOrMoreExpr(expr *oneOrMoreExpr) (interface{}, bool) {
} }
} }
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) { func (p *parser) parseRuleRefExpr(ref *ruleRefExpr) (interface{}, bool) {
if p.debug { if p.debug {
defer p.out(p.in("parseRuleRefExpr " + ref.name)) defer p.out(p.in("parseRuleRefExpr " + ref.name))
@ -1588,12 +1988,14 @@ func (p *parser) parseSeqExpr(seq *seqExpr) (interface{}, bool) {
defer p.out(p.in("parseSeqExpr")) defer p.out(p.in("parseSeqExpr"))
} }
var vals []interface{} vals := make([]interface{}, 0, len(seq.exprs))
pt := p.pt pt := p.pt
state := p.cloneState()
for _, expr := range seq.exprs { for _, expr := range seq.exprs {
val, ok := p.parseExpr(expr) val, ok := p.parseExpr(expr)
if !ok { if !ok {
p.restoreState(state)
p.restore(pt) p.restore(pt)
return nil, false return nil, false
} }
@ -1602,6 +2004,34 @@ func (p *parser) parseSeqExpr(seq *seqExpr) (interface{}, bool) {
return vals, true 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) { func (p *parser) parseZeroOrMoreExpr(expr *zeroOrMoreExpr) (interface{}, bool) {
if p.debug { if p.debug {
defer p.out(p.in("parseZeroOrMoreExpr")) defer p.out(p.in("parseZeroOrMoreExpr"))
@ -1631,18 +2061,3 @@ func (p *parser) parseZeroOrOneExpr(expr *zeroOrOneExpr) (interface{}, bool) {
// whether it matched or not, consider it a match // whether it matched or not, consider it a match
return val, true return val, true
} }
func rangeTable(class string) *unicode.RangeTable {
if rt, ok := unicode.Categories[class]; ok {
return rt
}
if rt, ok := unicode.Properties[class]; ok {
return rt
}
if rt, ok := unicode.Scripts[class]; ok {
return rt
}
// cannot happen
panic(fmt.Sprintf("invalid Unicode class: %s", class))
}