cosmium/parsers/nosql/nosql.go
2024-02-12 00:55:07 +02:00

2717 lines
66 KiB
Go

// Code generated by pigeon; DO NOT EDIT.
package nosql
import (
"bytes"
"errors"
"fmt"
"io"
"math"
"os"
"sort"
"strconv"
"strings"
"sync"
"unicode"
"unicode/utf8"
"github.com/pikami/cosmium/parsers"
)
func makeSelectStmt(columns, table, whereClause interface{}) (parsers.SelectStmt, error) {
selectStmt := parsers.SelectStmt{
Columns: columns.([]parsers.FieldPath),
Table: table.(parsers.Table),
}
if filters, ok := whereClause.(parsers.ComparisonExpression); ok {
selectStmt.Filters = filters
} else if filters, ok := whereClause.(parsers.LogicalExpression); ok {
selectStmt.Filters = filters
}
return selectStmt, nil
}
func makeFieldPath(name interface{}, path interface{}, alias interface{}) (parsers.FieldPath, error) {
ps := path.([]interface{})
paths := make([]string, 1)
paths[0] = name.(string)
for _, p := range ps {
paths = append(paths, p.(string))
}
fieldPath := parsers.FieldPath{Path: paths}
if aliasValue, ok := alias.(string); ok {
fieldPath.Alias = aliasValue
}
return fieldPath, nil
}
func makeColumnList(column interface{}, other_columns interface{}) ([]parsers.FieldPath, error) {
collsAsArray := other_columns.([]interface{})
columnList := make([]parsers.FieldPath, len(collsAsArray)+1)
columnList[0] = column.(parsers.FieldPath)
for i, v := range collsAsArray {
if col, ok := v.(parsers.FieldPath); ok {
columnList[i+1] = col
}
}
return columnList, nil
}
func joinStrings(array []interface{}) string {
var stringsArray []string
for _, elem := range array {
str, ok := elem.(string)
if !ok {
continue
}
stringsArray = append(stringsArray, str)
}
return strings.Join(stringsArray, "")
}
func combineExpressions(ex1 interface{}, exs interface{}, operation parsers.LogicalExpressionType) (interface{}, error) {
if exs == nil || len(exs.([]interface{})) < 1 {
return ex1, nil
}
return parsers.LogicalExpression{
Expressions: append([]interface{}{ex1}, exs.([]interface{})...),
Operation: operation,
}, nil
}
var g = &grammar{
rules: []*rule{
{
name: "Input",
pos: position{line: 78, col: 1, offset: 1970},
expr: &actionExpr{
pos: position{line: 78, col: 10, offset: 1979},
run: (*parser).callonInput1,
expr: &labeledExpr{
pos: position{line: 78, col: 10, offset: 1979},
label: "selectStmt",
expr: &ruleRefExpr{
pos: position{line: 78, col: 21, offset: 1990},
name: "SelectStmt",
},
},
},
},
{
name: "SelectStmt",
pos: position{line: 82, col: 1, offset: 2033},
expr: &actionExpr{
pos: position{line: 82, col: 15, offset: 2047},
run: (*parser).callonSelectStmt1,
expr: &seqExpr{
pos: position{line: 82, col: 15, offset: 2047},
exprs: []any{
&ruleRefExpr{
pos: position{line: 82, col: 15, offset: 2047},
name: "Select",
},
&ruleRefExpr{
pos: position{line: 82, col: 22, offset: 2054},
name: "ws",
},
&labeledExpr{
pos: position{line: 82, col: 25, offset: 2057},
label: "columns",
expr: &ruleRefExpr{
pos: position{line: 82, col: 33, offset: 2065},
name: "ColumnList",
},
},
&ruleRefExpr{
pos: position{line: 82, col: 44, offset: 2076},
name: "ws",
},
&ruleRefExpr{
pos: position{line: 83, col: 5, offset: 2083},
name: "From",
},
&ruleRefExpr{
pos: position{line: 83, col: 10, offset: 2088},
name: "ws",
},
&labeledExpr{
pos: position{line: 83, col: 13, offset: 2091},
label: "table",
expr: &ruleRefExpr{
pos: position{line: 83, col: 19, offset: 2097},
name: "TableName",
},
},
&ruleRefExpr{
pos: position{line: 83, col: 29, offset: 2107},
name: "ws",
},
&labeledExpr{
pos: position{line: 84, col: 5, offset: 2114},
label: "whereClause",
expr: &zeroOrOneExpr{
pos: position{line: 84, col: 17, offset: 2126},
expr: &actionExpr{
pos: position{line: 84, col: 18, offset: 2127},
run: (*parser).callonSelectStmt15,
expr: &seqExpr{
pos: position{line: 84, col: 18, offset: 2127},
exprs: []any{
&ruleRefExpr{
pos: position{line: 84, col: 18, offset: 2127},
name: "ws",
},
&ruleRefExpr{
pos: position{line: 84, col: 21, offset: 2130},
name: "Where",
},
&ruleRefExpr{
pos: position{line: 84, col: 27, offset: 2136},
name: "ws",
},
&labeledExpr{
pos: position{line: 84, col: 30, offset: 2139},
label: "condition",
expr: &ruleRefExpr{
pos: position{line: 84, col: 40, offset: 2149},
name: "Condition",
},
},
},
},
},
},
},
},
},
},
},
{
name: "ColumnList",
pos: position{line: 88, col: 1, offset: 2247},
expr: &actionExpr{
pos: position{line: 88, col: 15, offset: 2261},
run: (*parser).callonColumnList1,
expr: &seqExpr{
pos: position{line: 88, col: 15, offset: 2261},
exprs: []any{
&labeledExpr{
pos: position{line: 88, col: 15, offset: 2261},
label: "column",
expr: &ruleRefExpr{
pos: position{line: 88, col: 22, offset: 2268},
name: "FieldPath",
},
},
&labeledExpr{
pos: position{line: 88, col: 32, offset: 2278},
label: "other_columns",
expr: &zeroOrMoreExpr{
pos: position{line: 88, col: 46, offset: 2292},
expr: &actionExpr{
pos: position{line: 88, col: 47, offset: 2293},
run: (*parser).callonColumnList7,
expr: &seqExpr{
pos: position{line: 88, col: 47, offset: 2293},
exprs: []any{
&ruleRefExpr{
pos: position{line: 88, col: 47, offset: 2293},
name: "ws",
},
&litMatcher{
pos: position{line: 88, col: 50, offset: 2296},
val: ",",
ignoreCase: false,
want: "\",\"",
},
&ruleRefExpr{
pos: position{line: 88, col: 54, offset: 2300},
name: "ws",
},
&labeledExpr{
pos: position{line: 88, col: 57, offset: 2303},
label: "coll",
expr: &ruleRefExpr{
pos: position{line: 88, col: 62, offset: 2308},
name: "FieldPath",
},
},
},
},
},
},
},
},
},
},
},
{
name: "TableName",
pos: position{line: 92, col: 1, offset: 2394},
expr: &actionExpr{
pos: position{line: 92, col: 14, offset: 2407},
run: (*parser).callonTableName1,
expr: &labeledExpr{
pos: position{line: 92, col: 14, offset: 2407},
label: "key",
expr: &ruleRefExpr{
pos: position{line: 92, col: 18, offset: 2411},
name: "Identifier",
},
},
},
},
{
name: "FieldPath",
pos: position{line: 96, col: 1, offset: 2478},
expr: &actionExpr{
pos: position{line: 96, col: 14, offset: 2491},
run: (*parser).callonFieldPath1,
expr: &seqExpr{
pos: position{line: 96, col: 14, offset: 2491},
exprs: []any{
&labeledExpr{
pos: position{line: 96, col: 14, offset: 2491},
label: "name",
expr: &ruleRefExpr{
pos: position{line: 96, col: 19, offset: 2496},
name: "Identifier",
},
},
&labeledExpr{
pos: position{line: 96, col: 30, offset: 2507},
label: "path",
expr: &zeroOrMoreExpr{
pos: position{line: 96, col: 35, offset: 2512},
expr: &choiceExpr{
pos: position{line: 96, col: 36, offset: 2513},
alternatives: []any{
&ruleRefExpr{
pos: position{line: 96, col: 36, offset: 2513},
name: "DotFieldAccess",
},
&ruleRefExpr{
pos: position{line: 96, col: 53, offset: 2530},
name: "ArrayFieldAccess",
},
},
},
},
},
&labeledExpr{
pos: position{line: 97, col: 5, offset: 2553},
label: "asClause",
expr: &zeroOrOneExpr{
pos: position{line: 97, col: 14, offset: 2562},
expr: &actionExpr{
pos: position{line: 97, col: 15, offset: 2563},
run: (*parser).callonFieldPath12,
expr: &seqExpr{
pos: position{line: 97, col: 15, offset: 2563},
exprs: []any{
&ruleRefExpr{
pos: position{line: 97, col: 15, offset: 2563},
name: "ws",
},
&litMatcher{
pos: position{line: 97, col: 18, offset: 2566},
val: "AS",
ignoreCase: false,
want: "\"AS\"",
},
&ruleRefExpr{
pos: position{line: 97, col: 23, offset: 2571},
name: "ws",
},
&labeledExpr{
pos: position{line: 97, col: 26, offset: 2574},
label: "alias",
expr: &ruleRefExpr{
pos: position{line: 97, col: 32, offset: 2580},
name: "Identifier",
},
},
},
},
},
},
},
},
},
},
},
{
name: "DotFieldAccess",
pos: position{line: 101, col: 1, offset: 2667},
expr: &actionExpr{
pos: position{line: 101, col: 19, offset: 2685},
run: (*parser).callonDotFieldAccess1,
expr: &seqExpr{
pos: position{line: 101, col: 19, offset: 2685},
exprs: []any{
&litMatcher{
pos: position{line: 101, col: 19, offset: 2685},
val: ".",
ignoreCase: false,
want: "\".\"",
},
&labeledExpr{
pos: position{line: 101, col: 23, offset: 2689},
label: "id",
expr: &ruleRefExpr{
pos: position{line: 101, col: 26, offset: 2692},
name: "Identifier",
},
},
},
},
},
},
{
name: "ArrayFieldAccess",
pos: position{line: 105, col: 1, offset: 2727},
expr: &actionExpr{
pos: position{line: 105, col: 21, offset: 2747},
run: (*parser).callonArrayFieldAccess1,
expr: &seqExpr{
pos: position{line: 105, col: 21, offset: 2747},
exprs: []any{
&litMatcher{
pos: position{line: 105, col: 21, offset: 2747},
val: "[\"",
ignoreCase: false,
want: "\"[\\\"\"",
},
&labeledExpr{
pos: position{line: 105, col: 27, offset: 2753},
label: "id",
expr: &ruleRefExpr{
pos: position{line: 105, col: 30, offset: 2756},
name: "Identifier",
},
},
&litMatcher{
pos: position{line: 105, col: 41, offset: 2767},
val: "\"]",
ignoreCase: false,
want: "\"\\\"]\"",
},
},
},
},
},
{
name: "Identifier",
pos: position{line: 109, col: 1, offset: 2797},
expr: &actionExpr{
pos: position{line: 109, col: 15, offset: 2811},
run: (*parser).callonIdentifier1,
expr: &seqExpr{
pos: position{line: 109, col: 15, offset: 2811},
exprs: []any{
&charClassMatcher{
pos: position{line: 109, col: 15, offset: 2811},
val: "[a-zA-Z_]",
chars: []rune{'_'},
ranges: []rune{'a', 'z', 'A', 'Z'},
ignoreCase: false,
inverted: false,
},
&zeroOrMoreExpr{
pos: position{line: 109, col: 24, offset: 2820},
expr: &charClassMatcher{
pos: position{line: 109, col: 24, offset: 2820},
val: "[a-zA-Z0-9_]",
chars: []rune{'_'},
ranges: []rune{'a', 'z', 'A', 'Z', '0', '9'},
ignoreCase: false,
inverted: false,
},
},
},
},
},
},
{
name: "Condition",
pos: position{line: 113, col: 1, offset: 2870},
expr: &actionExpr{
pos: position{line: 113, col: 14, offset: 2883},
run: (*parser).callonCondition1,
expr: &labeledExpr{
pos: position{line: 113, col: 14, offset: 2883},
label: "expression",
expr: &ruleRefExpr{
pos: position{line: 113, col: 25, offset: 2894},
name: "OrExpression",
},
},
},
},
{
name: "OrExpression",
pos: position{line: 117, col: 1, offset: 2939},
expr: &actionExpr{
pos: position{line: 117, col: 17, offset: 2955},
run: (*parser).callonOrExpression1,
expr: &seqExpr{
pos: position{line: 117, col: 17, offset: 2955},
exprs: []any{
&labeledExpr{
pos: position{line: 117, col: 17, offset: 2955},
label: "ex1",
expr: &ruleRefExpr{
pos: position{line: 117, col: 21, offset: 2959},
name: "AndExpression",
},
},
&labeledExpr{
pos: position{line: 117, col: 35, offset: 2973},
label: "ex2",
expr: &zeroOrMoreExpr{
pos: position{line: 117, col: 39, offset: 2977},
expr: &actionExpr{
pos: position{line: 117, col: 40, offset: 2978},
run: (*parser).callonOrExpression7,
expr: &seqExpr{
pos: position{line: 117, col: 40, offset: 2978},
exprs: []any{
&ruleRefExpr{
pos: position{line: 117, col: 40, offset: 2978},
name: "ws",
},
&litMatcher{
pos: position{line: 117, col: 43, offset: 2981},
val: "OR",
ignoreCase: false,
want: "\"OR\"",
},
&ruleRefExpr{
pos: position{line: 117, col: 48, offset: 2986},
name: "ws",
},
&labeledExpr{
pos: position{line: 117, col: 51, offset: 2989},
label: "ex",
expr: &ruleRefExpr{
pos: position{line: 117, col: 54, offset: 2992},
name: "AndExpression",
},
},
},
},
},
},
},
},
},
},
},
{
name: "AndExpression",
pos: position{line: 121, col: 1, offset: 3105},
expr: &actionExpr{
pos: position{line: 121, col: 18, offset: 3122},
run: (*parser).callonAndExpression1,
expr: &seqExpr{
pos: position{line: 121, col: 18, offset: 3122},
exprs: []any{
&labeledExpr{
pos: position{line: 121, col: 18, offset: 3122},
label: "ex1",
expr: &ruleRefExpr{
pos: position{line: 121, col: 22, offset: 3126},
name: "ComparisonExpression",
},
},
&labeledExpr{
pos: position{line: 121, col: 43, offset: 3147},
label: "ex2",
expr: &zeroOrMoreExpr{
pos: position{line: 121, col: 47, offset: 3151},
expr: &actionExpr{
pos: position{line: 121, col: 48, offset: 3152},
run: (*parser).callonAndExpression7,
expr: &seqExpr{
pos: position{line: 121, col: 48, offset: 3152},
exprs: []any{
&ruleRefExpr{
pos: position{line: 121, col: 48, offset: 3152},
name: "ws",
},
&litMatcher{
pos: position{line: 121, col: 51, offset: 3155},
val: "AND",
ignoreCase: false,
want: "\"AND\"",
},
&ruleRefExpr{
pos: position{line: 121, col: 57, offset: 3161},
name: "ws",
},
&labeledExpr{
pos: position{line: 121, col: 60, offset: 3164},
label: "ex",
expr: &ruleRefExpr{
pos: position{line: 121, col: 63, offset: 3167},
name: "ComparisonExpression",
},
},
},
},
},
},
},
},
},
},
},
{
name: "ComparisonExpression",
pos: position{line: 125, col: 1, offset: 3288},
expr: &actionExpr{
pos: position{line: 125, col: 25, offset: 3312},
run: (*parser).callonComparisonExpression1,
expr: &seqExpr{
pos: position{line: 125, col: 25, offset: 3312},
exprs: []any{
&labeledExpr{
pos: position{line: 125, col: 25, offset: 3312},
label: "left",
expr: &choiceExpr{
pos: position{line: 125, col: 31, offset: 3318},
alternatives: []any{
&ruleRefExpr{
pos: position{line: 125, col: 31, offset: 3318},
name: "Literal",
},
&ruleRefExpr{
pos: position{line: 125, col: 41, offset: 3328},
name: "FieldPath",
},
},
},
},
&ruleRefExpr{
pos: position{line: 125, col: 52, offset: 3339},
name: "ws",
},
&labeledExpr{
pos: position{line: 125, col: 55, offset: 3342},
label: "op",
expr: &ruleRefExpr{
pos: position{line: 125, col: 58, offset: 3345},
name: "ComparisonOperator",
},
},
&ruleRefExpr{
pos: position{line: 125, col: 77, offset: 3364},
name: "ws",
},
&labeledExpr{
pos: position{line: 125, col: 80, offset: 3367},
label: "right",
expr: &choiceExpr{
pos: position{line: 125, col: 87, offset: 3374},
alternatives: []any{
&ruleRefExpr{
pos: position{line: 125, col: 87, offset: 3374},
name: "Literal",
},
&ruleRefExpr{
pos: position{line: 125, col: 97, offset: 3384},
name: "FieldPath",
},
},
},
},
},
},
},
},
{
name: "Select",
pos: position{line: 129, col: 1, offset: 3499},
expr: &choiceExpr{
pos: position{line: 129, col: 12, offset: 3510},
alternatives: []any{
&litMatcher{
pos: position{line: 129, col: 12, offset: 3510},
val: "select",
ignoreCase: false,
want: "\"select\"",
},
&litMatcher{
pos: position{line: 129, col: 23, offset: 3521},
val: "SELECT",
ignoreCase: false,
want: "\"SELECT\"",
},
},
},
},
{
name: "From",
pos: position{line: 131, col: 1, offset: 3532},
expr: &choiceExpr{
pos: position{line: 131, col: 10, offset: 3541},
alternatives: []any{
&litMatcher{
pos: position{line: 131, col: 10, offset: 3541},
val: "from",
ignoreCase: false,
want: "\"from\"",
},
&litMatcher{
pos: position{line: 131, col: 19, offset: 3550},
val: "FROM",
ignoreCase: false,
want: "\"FROM\"",
},
},
},
},
{
name: "Where",
pos: position{line: 133, col: 1, offset: 3559},
expr: &choiceExpr{
pos: position{line: 133, col: 11, offset: 3569},
alternatives: []any{
&litMatcher{
pos: position{line: 133, col: 11, offset: 3569},
val: "where",
ignoreCase: false,
want: "\"where\"",
},
&litMatcher{
pos: position{line: 133, col: 21, offset: 3579},
val: "WHERE",
ignoreCase: false,
want: "\"WHERE\"",
},
},
},
},
{
name: "ComparisonOperator",
pos: position{line: 135, col: 1, offset: 3589},
expr: &choiceExpr{
pos: position{line: 135, col: 23, offset: 3611},
alternatives: []any{
&litMatcher{
pos: position{line: 135, col: 23, offset: 3611},
val: "=",
ignoreCase: false,
want: "\"=\"",
},
&litMatcher{
pos: position{line: 135, col: 29, offset: 3617},
val: "!=",
ignoreCase: false,
want: "\"!=\"",
},
&litMatcher{
pos: position{line: 135, col: 36, offset: 3624},
val: "<",
ignoreCase: false,
want: "\"<\"",
},
&litMatcher{
pos: position{line: 135, col: 42, offset: 3630},
val: "<=",
ignoreCase: false,
want: "\"<=\"",
},
&litMatcher{
pos: position{line: 135, col: 49, offset: 3637},
val: ">",
ignoreCase: false,
want: "\">\"",
},
&actionExpr{
pos: position{line: 135, col: 55, offset: 3643},
run: (*parser).callonComparisonOperator7,
expr: &litMatcher{
pos: position{line: 135, col: 55, offset: 3643},
val: ">=",
ignoreCase: false,
want: "\">=\"",
},
},
},
},
},
{
name: "Literal",
pos: position{line: 139, col: 1, offset: 3684},
expr: &choiceExpr{
pos: position{line: 139, col: 12, offset: 3695},
alternatives: []any{
&ruleRefExpr{
pos: position{line: 139, col: 12, offset: 3695},
name: "FloatLiteral",
},
&ruleRefExpr{
pos: position{line: 139, col: 27, offset: 3710},
name: "IntegerLiteral",
},
&ruleRefExpr{
pos: position{line: 139, col: 44, offset: 3727},
name: "StringLiteral",
},
&ruleRefExpr{
pos: position{line: 139, col: 60, offset: 3743},
name: "BooleanLiteral",
},
},
},
},
{
name: "IntegerLiteral",
pos: position{line: 141, col: 1, offset: 3759},
expr: &actionExpr{
pos: position{line: 141, col: 19, offset: 3777},
run: (*parser).callonIntegerLiteral1,
expr: &oneOrMoreExpr{
pos: position{line: 141, col: 19, offset: 3777},
expr: &charClassMatcher{
pos: position{line: 141, col: 19, offset: 3777},
val: "[0-9]",
ranges: []rune{'0', '9'},
ignoreCase: false,
inverted: false,
},
},
},
},
{
name: "StringLiteral",
pos: position{line: 145, col: 1, offset: 3921},
expr: &actionExpr{
pos: position{line: 145, col: 18, offset: 3938},
run: (*parser).callonStringLiteral1,
expr: &seqExpr{
pos: position{line: 145, col: 18, offset: 3938},
exprs: []any{
&litMatcher{
pos: position{line: 145, col: 18, offset: 3938},
val: "\"",
ignoreCase: false,
want: "\"\\\"\"",
},
&labeledExpr{
pos: position{line: 145, col: 23, offset: 3943},
label: "chars",
expr: &zeroOrMoreExpr{
pos: position{line: 145, col: 29, offset: 3949},
expr: &ruleRefExpr{
pos: position{line: 145, col: 29, offset: 3949},
name: "StringCharacter",
},
},
},
&litMatcher{
pos: position{line: 145, col: 46, offset: 3966},
val: "\"",
ignoreCase: false,
want: "\"\\\"\"",
},
},
},
},
},
{
name: "FloatLiteral",
pos: position{line: 148, col: 1, offset: 4084},
expr: &actionExpr{
pos: position{line: 148, col: 17, offset: 4100},
run: (*parser).callonFloatLiteral1,
expr: &seqExpr{
pos: position{line: 148, col: 17, offset: 4100},
exprs: []any{
&oneOrMoreExpr{
pos: position{line: 148, col: 17, offset: 4100},
expr: &charClassMatcher{
pos: position{line: 148, col: 17, offset: 4100},
val: "[0-9]",
ranges: []rune{'0', '9'},
ignoreCase: false,
inverted: false,
},
},
&litMatcher{
pos: position{line: 148, col: 23, offset: 4106},
val: ".",
ignoreCase: false,
want: "\".\"",
},
&oneOrMoreExpr{
pos: position{line: 148, col: 26, offset: 4109},
expr: &charClassMatcher{
pos: position{line: 148, col: 26, offset: 4109},
val: "[0-9]",
ranges: []rune{'0', '9'},
ignoreCase: false,
inverted: false,
},
},
},
},
},
},
{
name: "BooleanLiteral",
pos: position{line: 152, col: 1, offset: 4265},
expr: &actionExpr{
pos: position{line: 152, col: 19, offset: 4283},
run: (*parser).callonBooleanLiteral1,
expr: &choiceExpr{
pos: position{line: 152, col: 20, offset: 4284},
alternatives: []any{
&litMatcher{
pos: position{line: 152, col: 20, offset: 4284},
val: "true",
ignoreCase: false,
want: "\"true\"",
},
&litMatcher{
pos: position{line: 152, col: 29, offset: 4293},
val: "false",
ignoreCase: false,
want: "\"false\"",
},
},
},
},
},
{
name: "StringCharacter",
pos: position{line: 157, col: 1, offset: 4447},
expr: &choiceExpr{
pos: position{line: 157, col: 20, offset: 4466},
alternatives: []any{
&actionExpr{
pos: position{line: 157, col: 20, offset: 4466},
run: (*parser).callonStringCharacter2,
expr: &seqExpr{
pos: position{line: 157, col: 20, offset: 4466},
exprs: []any{
&notExpr{
pos: position{line: 157, col: 20, offset: 4466},
expr: &choiceExpr{
pos: position{line: 157, col: 22, offset: 4468},
alternatives: []any{
&litMatcher{
pos: position{line: 157, col: 22, offset: 4468},
val: "\"",
ignoreCase: false,
want: "\"\\\"\"",
},
&litMatcher{
pos: position{line: 157, col: 28, offset: 4474},
val: "\\",
ignoreCase: false,
want: "\"\\\\\"",
},
},
},
},
&anyMatcher{
line: 157, col: 34, offset: 4480,
},
},
},
},
&actionExpr{
pos: position{line: 158, col: 5, offset: 4517},
run: (*parser).callonStringCharacter9,
expr: &seqExpr{
pos: position{line: 158, col: 5, offset: 4517},
exprs: []any{
&litMatcher{
pos: position{line: 158, col: 5, offset: 4517},
val: "\\",
ignoreCase: false,
want: "\"\\\\\"",
},
&labeledExpr{
pos: position{line: 158, col: 10, offset: 4522},
label: "seq",
expr: &ruleRefExpr{
pos: position{line: 158, col: 14, offset: 4526},
name: "EscapeSequenceCharacter",
},
},
},
},
},
},
},
},
{
name: "EscapeSequenceCharacter",
pos: position{line: 160, col: 1, offset: 4571},
expr: &labeledExpr{
pos: position{line: 160, col: 28, offset: 4598},
label: "char",
expr: &ruleRefExpr{
pos: position{line: 160, col: 33, offset: 4603},
name: "EscapeCharacter",
},
},
},
{
name: "EscapeCharacter",
pos: position{line: 162, col: 1, offset: 4620},
expr: &choiceExpr{
pos: position{line: 162, col: 20, offset: 4639},
alternatives: []any{
&litMatcher{
pos: position{line: 162, col: 20, offset: 4639},
val: "'",
ignoreCase: false,
want: "\"'\"",
},
&litMatcher{
pos: position{line: 163, col: 5, offset: 4647},
val: "\"",
ignoreCase: false,
want: "\"\\\"\"",
},
&litMatcher{
pos: position{line: 164, col: 5, offset: 4655},
val: "\\",
ignoreCase: false,
want: "\"\\\\\"",
},
&actionExpr{
pos: position{line: 165, col: 5, offset: 4664},
run: (*parser).callonEscapeCharacter5,
expr: &litMatcher{
pos: position{line: 165, col: 5, offset: 4664},
val: "b",
ignoreCase: false,
want: "\"b\"",
},
},
&actionExpr{
pos: position{line: 166, col: 5, offset: 4693},
run: (*parser).callonEscapeCharacter7,
expr: &litMatcher{
pos: position{line: 166, col: 5, offset: 4693},
val: "f",
ignoreCase: false,
want: "\"f\"",
},
},
&actionExpr{
pos: position{line: 167, col: 5, offset: 4722},
run: (*parser).callonEscapeCharacter9,
expr: &litMatcher{
pos: position{line: 167, col: 5, offset: 4722},
val: "n",
ignoreCase: false,
want: "\"n\"",
},
},
&actionExpr{
pos: position{line: 168, col: 5, offset: 4751},
run: (*parser).callonEscapeCharacter11,
expr: &litMatcher{
pos: position{line: 168, col: 5, offset: 4751},
val: "r",
ignoreCase: false,
want: "\"r\"",
},
},
&actionExpr{
pos: position{line: 169, col: 5, offset: 4780},
run: (*parser).callonEscapeCharacter13,
expr: &litMatcher{
pos: position{line: 169, col: 5, offset: 4780},
val: "t",
ignoreCase: false,
want: "\"t\"",
},
},
},
},
},
{
name: "non_escape_character",
pos: position{line: 171, col: 1, offset: 4806},
expr: &actionExpr{
pos: position{line: 171, col: 25, offset: 4830},
run: (*parser).callonnon_escape_character1,
expr: &seqExpr{
pos: position{line: 171, col: 25, offset: 4830},
exprs: []any{
&notExpr{
pos: position{line: 171, col: 25, offset: 4830},
expr: &ruleRefExpr{
pos: position{line: 171, col: 27, offset: 4832},
name: "escape_character",
},
},
&labeledExpr{
pos: position{line: 171, col: 45, offset: 4850},
label: "char",
expr: &anyMatcher{
line: 171, col: 50, offset: 4855,
},
},
},
},
},
},
{
name: "ws",
pos: position{line: 174, col: 1, offset: 4894},
expr: &zeroOrMoreExpr{
pos: position{line: 174, col: 7, offset: 4900},
expr: &charClassMatcher{
pos: position{line: 174, col: 7, offset: 4900},
val: "[ \\t\\n\\r]",
chars: []rune{' ', '\t', '\n', '\r'},
ignoreCase: false,
inverted: false,
},
},
},
{
name: "EOF",
pos: position{line: 176, col: 1, offset: 4912},
expr: &notExpr{
pos: position{line: 176, col: 8, offset: 4919},
expr: &anyMatcher{
line: 176, col: 9, offset: 4920,
},
},
},
},
}
func (c *current) onInput1(selectStmt any) (any, error) {
return selectStmt, nil
}
func (p *parser) callonInput1() (any, error) {
stack := p.vstack[len(p.vstack)-1]
_ = stack
return p.cur.onInput1(stack["selectStmt"])
}
func (c *current) onSelectStmt15(condition any) (any, error) {
return condition, nil
}
func (p *parser) callonSelectStmt15() (any, error) {
stack := p.vstack[len(p.vstack)-1]
_ = stack
return p.cur.onSelectStmt15(stack["condition"])
}
func (c *current) onSelectStmt1(columns, table, whereClause any) (any, error) {
return makeSelectStmt(columns, table, whereClause)
}
func (p *parser) callonSelectStmt1() (any, error) {
stack := p.vstack[len(p.vstack)-1]
_ = stack
return p.cur.onSelectStmt1(stack["columns"], stack["table"], stack["whereClause"])
}
func (c *current) onColumnList7(coll any) (any, error) {
return coll, nil
}
func (p *parser) callonColumnList7() (any, error) {
stack := p.vstack[len(p.vstack)-1]
_ = stack
return p.cur.onColumnList7(stack["coll"])
}
func (c *current) onColumnList1(column, other_columns any) (any, error) {
return makeColumnList(column, other_columns)
}
func (p *parser) callonColumnList1() (any, error) {
stack := p.vstack[len(p.vstack)-1]
_ = stack
return p.cur.onColumnList1(stack["column"], stack["other_columns"])
}
func (c *current) onTableName1(key any) (any, error) {
return parsers.Table{Value: key.(string)}, nil
}
func (p *parser) callonTableName1() (any, error) {
stack := p.vstack[len(p.vstack)-1]
_ = stack
return p.cur.onTableName1(stack["key"])
}
func (c *current) onFieldPath12(alias any) (any, error) {
return alias, nil
}
func (p *parser) callonFieldPath12() (any, error) {
stack := p.vstack[len(p.vstack)-1]
_ = stack
return p.cur.onFieldPath12(stack["alias"])
}
func (c *current) onFieldPath1(name, path, asClause any) (any, error) {
return makeFieldPath(name, path, asClause)
}
func (p *parser) callonFieldPath1() (any, error) {
stack := p.vstack[len(p.vstack)-1]
_ = stack
return p.cur.onFieldPath1(stack["name"], stack["path"], stack["asClause"])
}
func (c *current) onDotFieldAccess1(id any) (any, error) {
return id, nil
}
func (p *parser) callonDotFieldAccess1() (any, error) {
stack := p.vstack[len(p.vstack)-1]
_ = stack
return p.cur.onDotFieldAccess1(stack["id"])
}
func (c *current) onArrayFieldAccess1(id any) (any, error) {
return id, nil
}
func (p *parser) callonArrayFieldAccess1() (any, error) {
stack := p.vstack[len(p.vstack)-1]
_ = stack
return p.cur.onArrayFieldAccess1(stack["id"])
}
func (c *current) onIdentifier1() (any, error) {
return string(c.text), nil
}
func (p *parser) callonIdentifier1() (any, error) {
stack := p.vstack[len(p.vstack)-1]
_ = stack
return p.cur.onIdentifier1()
}
func (c *current) onCondition1(expression any) (any, error) {
return expression, nil
}
func (p *parser) callonCondition1() (any, error) {
stack := p.vstack[len(p.vstack)-1]
_ = stack
return p.cur.onCondition1(stack["expression"])
}
func (c *current) onOrExpression7(ex any) (any, error) {
return ex, nil
}
func (p *parser) callonOrExpression7() (any, error) {
stack := p.vstack[len(p.vstack)-1]
_ = stack
return p.cur.onOrExpression7(stack["ex"])
}
func (c *current) onOrExpression1(ex1, ex2 any) (any, error) {
return combineExpressions(ex1, ex2, parsers.LogicalExpressionTypeOr)
}
func (p *parser) callonOrExpression1() (any, error) {
stack := p.vstack[len(p.vstack)-1]
_ = stack
return p.cur.onOrExpression1(stack["ex1"], stack["ex2"])
}
func (c *current) onAndExpression7(ex any) (any, error) {
return ex, nil
}
func (p *parser) callonAndExpression7() (any, error) {
stack := p.vstack[len(p.vstack)-1]
_ = stack
return p.cur.onAndExpression7(stack["ex"])
}
func (c *current) onAndExpression1(ex1, ex2 any) (any, error) {
return combineExpressions(ex1, ex2, parsers.LogicalExpressionTypeAnd)
}
func (p *parser) callonAndExpression1() (any, error) {
stack := p.vstack[len(p.vstack)-1]
_ = stack
return p.cur.onAndExpression1(stack["ex1"], stack["ex2"])
}
func (c *current) onComparisonExpression1(left, op, right any) (any, error) {
return parsers.ComparisonExpression{Left: left, Right: right, Operation: string(op.([]uint8))}, nil
}
func (p *parser) callonComparisonExpression1() (any, error) {
stack := p.vstack[len(p.vstack)-1]
_ = stack
return p.cur.onComparisonExpression1(stack["left"], stack["op"], stack["right"])
}
func (c *current) onComparisonOperator7() (any, error) {
return string(c.text), nil
}
func (p *parser) callonComparisonOperator7() (any, error) {
stack := p.vstack[len(p.vstack)-1]
_ = stack
return p.cur.onComparisonOperator7()
}
func (c *current) onIntegerLiteral1() (any, error) {
intValue, _ := strconv.Atoi(string(c.text))
return parsers.Constant{Type: parsers.ConstantTypeInteger, Value: intValue}, nil
}
func (p *parser) callonIntegerLiteral1() (any, error) {
stack := p.vstack[len(p.vstack)-1]
_ = stack
return p.cur.onIntegerLiteral1()
}
func (c *current) onStringLiteral1(chars any) (any, error) {
return parsers.Constant{Type: parsers.ConstantTypeString, Value: joinStrings(chars.([]interface{}))}, nil
}
func (p *parser) callonStringLiteral1() (any, error) {
stack := p.vstack[len(p.vstack)-1]
_ = stack
return p.cur.onStringLiteral1(stack["chars"])
}
func (c *current) onFloatLiteral1() (any, error) {
floatValue, _ := strconv.ParseFloat(string(c.text), 64)
return parsers.Constant{Type: parsers.ConstantTypeFloat, Value: floatValue}, nil
}
func (p *parser) callonFloatLiteral1() (any, error) {
stack := p.vstack[len(p.vstack)-1]
_ = stack
return p.cur.onFloatLiteral1()
}
func (c *current) onBooleanLiteral1() (any, error) {
boolValue, _ := strconv.ParseBool(string(c.text))
return parsers.Constant{Type: parsers.ConstantTypeBoolean, Value: boolValue}, nil
}
func (p *parser) callonBooleanLiteral1() (any, error) {
stack := p.vstack[len(p.vstack)-1]
_ = stack
return p.cur.onBooleanLiteral1()
}
func (c *current) onStringCharacter2() (any, error) {
return string(c.text), nil
}
func (p *parser) callonStringCharacter2() (any, error) {
stack := p.vstack[len(p.vstack)-1]
_ = stack
return p.cur.onStringCharacter2()
}
func (c *current) onStringCharacter9(seq any) (any, error) {
return seq, nil
}
func (p *parser) callonStringCharacter9() (any, error) {
stack := p.vstack[len(p.vstack)-1]
_ = stack
return p.cur.onStringCharacter9(stack["seq"])
}
func (c *current) onEscapeCharacter5() (any, error) {
return "\b", nil
}
func (p *parser) callonEscapeCharacter5() (any, error) {
stack := p.vstack[len(p.vstack)-1]
_ = stack
return p.cur.onEscapeCharacter5()
}
func (c *current) onEscapeCharacter7() (any, error) {
return "\f", nil
}
func (p *parser) callonEscapeCharacter7() (any, error) {
stack := p.vstack[len(p.vstack)-1]
_ = stack
return p.cur.onEscapeCharacter7()
}
func (c *current) onEscapeCharacter9() (any, error) {
return "\n", nil
}
func (p *parser) callonEscapeCharacter9() (any, error) {
stack := p.vstack[len(p.vstack)-1]
_ = stack
return p.cur.onEscapeCharacter9()
}
func (c *current) onEscapeCharacter11() (any, error) {
return "\r", nil
}
func (p *parser) callonEscapeCharacter11() (any, error) {
stack := p.vstack[len(p.vstack)-1]
_ = stack
return p.cur.onEscapeCharacter11()
}
func (c *current) onEscapeCharacter13() (any, error) {
return "\t", nil
}
func (p *parser) callonEscapeCharacter13() (any, error) {
stack := p.vstack[len(p.vstack)-1]
_ = stack
return p.cur.onEscapeCharacter13()
}
func (c *current) onnon_escape_character1(char any) (any, error) {
return string(c.text), nil
}
func (p *parser) callonnon_escape_character1() (any, error) {
stack := p.vstack[len(p.vstack)-1]
_ = stack
return p.cur.onnon_escape_character1(stack["char"])
}
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 expressions 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 any) 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 any) 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 any, 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) (any, error) {
b, err := io.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) (any, 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]any
// the AST types...
type grammar struct {
pos position
rules []*rule
}
type rule struct {
pos position
name string
displayName string
expr any
}
type choiceExpr struct {
pos position
alternatives []any
}
type actionExpr struct {
pos position
expr any
run func(*parser) (any, error)
}
type recoveryExpr struct {
pos position
expr any
recoverExpr any
failureLabel []string
}
type seqExpr struct {
pos position
exprs []any
}
type throwExpr struct {
pos position
label string
}
type labeledExpr struct {
pos position
label string
expr any
}
type expr struct {
pos position
expr any
}
type (
andExpr expr
notExpr expr
zeroOrOneExpr expr
zeroOrMoreExpr expr
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 any
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[any]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]any
// 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]any
}
// 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]any)
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 any) {
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]any, 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) printIndent(mark string, s string) string {
return p.print(strings.Repeat(" ", p.depth)+mark, s)
}
func (p *parser) in(s string) string {
res := p.printIndent(">", s)
p.depth++
return res
}
func (p *parser) out(s string) string {
p.depth--
return p.printIndent("<", 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() any
}
var statePool = &sync.Pool{
New: func() any { 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 any) (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 any, tuple resultTuple) {
if p.memo == nil {
p.memo = make(map[int]map[any]resultTuple)
}
m := p.memo[pt.offset]
if m == nil {
m = make(map[any]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 any, 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.parseRuleWrap(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) parseRuleMemoize(rule *rule) (any, bool) {
res, ok := p.getMemoized(rule)
if ok {
p.restore(res.end)
return res.v, res.b
}
startMark := p.pt
val, ok := p.parseRule(rule)
p.setMemoized(startMark, rule, resultTuple{val, ok, p.pt})
return val, ok
}
func (p *parser) parseRuleWrap(rule *rule) (any, bool) {
if p.debug {
defer p.out(p.in("parseRule " + rule.name))
}
var (
val any
ok bool
startMark = p.pt
)
if p.memoize {
val, ok = p.parseRuleMemoize(rule)
} else {
val, ok = p.parseRule(rule)
}
if ok && p.debug {
p.printIndent("MATCH", string(p.sliceFrom(startMark)))
}
return val, ok
}
func (p *parser) parseRule(rule *rule) (any, bool) {
p.rstack = append(p.rstack, rule)
p.pushV()
val, ok := p.parseExprWrap(rule.expr)
p.popV()
p.rstack = p.rstack[:len(p.rstack)-1]
return val, ok
}
func (p *parser) parseExprWrap(expr any) (any, 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
}
val, ok := p.parseExpr(expr)
if p.memoize {
p.setMemoized(pt, expr, resultTuple{val, ok, p.pt})
}
return val, ok
}
func (p *parser) parseExpr(expr any) (any, bool) {
p.ExprCnt++
if p.ExprCnt > p.maxExprCnt {
panic(errMaxExprCnt)
}
var val any
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))
}
return val, ok
}
func (p *parser) parseActionExpr(act *actionExpr) (any, bool) {
if p.debug {
defer p.out(p.in("parseActionExpr"))
}
start := p.pt
val, ok := p.parseExprWrap(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.printIndent("MATCH", string(p.sliceFrom(start)))
}
return val, ok
}
func (p *parser) parseAndCodeExpr(and *andCodeExpr) (any, 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) (any, bool) {
if p.debug {
defer p.out(p.in("parseAndExpr"))
}
pt := p.pt
state := p.cloneState()
p.pushV()
_, ok := p.parseExprWrap(and.expr)
p.popV()
p.restoreState(state)
p.restore(pt)
return nil, ok
}
func (p *parser) parseAnyMatcher(any *anyMatcher) (any, 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) (any, 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) (any, 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.parseExprWrap(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) (any, bool) {
if p.debug {
defer p.out(p.in("parseLabeledExpr"))
}
p.pushV()
val, ok := p.parseExprWrap(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) (any, 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) (any, 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) (any, bool) {
if p.debug {
defer p.out(p.in("parseNotExpr"))
}
pt := p.pt
state := p.cloneState()
p.pushV()
p.maxFailInvertExpected = !p.maxFailInvertExpected
_, ok := p.parseExprWrap(not.expr)
p.maxFailInvertExpected = !p.maxFailInvertExpected
p.popV()
p.restoreState(state)
p.restore(pt)
return nil, !ok
}
func (p *parser) parseOneOrMoreExpr(expr *oneOrMoreExpr) (any, bool) {
if p.debug {
defer p.out(p.in("parseOneOrMoreExpr"))
}
var vals []any
for {
p.pushV()
val, ok := p.parseExprWrap(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) (any, bool) {
if p.debug {
defer p.out(p.in("parseRecoveryExpr (" + strings.Join(recover.failureLabel, ",") + ")"))
}
p.pushRecovery(recover.failureLabel, recover.recoverExpr)
val, ok := p.parseExprWrap(recover.expr)
p.popRecovery()
return val, ok
}
func (p *parser) parseRuleRefExpr(ref *ruleRefExpr) (any, 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.parseRuleWrap(rule)
}
func (p *parser) parseSeqExpr(seq *seqExpr) (any, bool) {
if p.debug {
defer p.out(p.in("parseSeqExpr"))
}
vals := make([]any, 0, len(seq.exprs))
pt := p.pt
state := p.cloneState()
for _, expr := range seq.exprs {
val, ok := p.parseExprWrap(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) (any, 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) (any, 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.parseExprWrap(recoverExpr); ok {
return val, ok
}
}
}
return nil, false
}
func (p *parser) parseZeroOrMoreExpr(expr *zeroOrMoreExpr) (any, bool) {
if p.debug {
defer p.out(p.in("parseZeroOrMoreExpr"))
}
var vals []any
for {
p.pushV()
val, ok := p.parseExprWrap(expr.expr)
p.popV()
if !ok {
return vals, true
}
vals = append(vals, val)
}
}
func (p *parser) parseZeroOrOneExpr(expr *zeroOrOneExpr) (any, bool) {
if p.debug {
defer p.out(p.in("parseZeroOrOneExpr"))
}
p.pushV()
val, _ := p.parseExprWrap(expr.expr)
p.popV()
// whether it matched or not, consider it a match
return val, true
}