1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349
|
package ini
import (
"fmt"
"io"
)
// ParseState represents the current state of the parser.
type ParseState uint
// State enums for the parse table
const (
InvalidState ParseState = iota
// stmt -> value stmt'
StatementState
// stmt' -> MarkComplete | op stmt
StatementPrimeState
// value -> number | string | boolean | quoted_string
ValueState
// section -> [ section'
OpenScopeState
// section' -> value section_close
SectionState
// section_close -> ]
CloseScopeState
// SkipState will skip (NL WS)+
SkipState
// SkipTokenState will skip any token and push the previous
// state onto the stack.
SkipTokenState
// comment -> # comment' | ; comment'
// comment' -> MarkComplete | value
CommentState
// MarkComplete state will complete statements and move that
// to the completed AST list
MarkCompleteState
// TerminalState signifies that the tokens have been fully parsed
TerminalState
)
// parseTable is a state machine to dictate the grammar above.
var parseTable = map[ASTKind]map[TokenType]ParseState{
ASTKindStart: {
TokenLit: StatementState,
TokenSep: OpenScopeState,
TokenWS: SkipTokenState,
TokenNL: SkipTokenState,
TokenComment: CommentState,
TokenNone: TerminalState,
},
ASTKindCommentStatement: {
TokenLit: StatementState,
TokenSep: OpenScopeState,
TokenWS: SkipTokenState,
TokenNL: SkipTokenState,
TokenComment: CommentState,
TokenNone: MarkCompleteState,
},
ASTKindExpr: {
TokenOp: StatementPrimeState,
TokenLit: ValueState,
TokenSep: OpenScopeState,
TokenWS: ValueState,
TokenNL: SkipState,
TokenComment: CommentState,
TokenNone: MarkCompleteState,
},
ASTKindEqualExpr: {
TokenLit: ValueState,
TokenSep: ValueState,
TokenOp: ValueState,
TokenWS: SkipTokenState,
TokenNL: SkipState,
},
ASTKindStatement: {
TokenLit: SectionState,
TokenSep: CloseScopeState,
TokenWS: SkipTokenState,
TokenNL: SkipTokenState,
TokenComment: CommentState,
TokenNone: MarkCompleteState,
},
ASTKindExprStatement: {
TokenLit: ValueState,
TokenSep: ValueState,
TokenOp: ValueState,
TokenWS: ValueState,
TokenNL: MarkCompleteState,
TokenComment: CommentState,
TokenNone: TerminalState,
TokenComma: SkipState,
},
ASTKindSectionStatement: {
TokenLit: SectionState,
TokenOp: SectionState,
TokenSep: CloseScopeState,
TokenWS: SectionState,
TokenNL: SkipTokenState,
},
ASTKindCompletedSectionStatement: {
TokenWS: SkipTokenState,
TokenNL: SkipTokenState,
TokenLit: StatementState,
TokenSep: OpenScopeState,
TokenComment: CommentState,
TokenNone: MarkCompleteState,
},
ASTKindSkipStatement: {
TokenLit: StatementState,
TokenSep: OpenScopeState,
TokenWS: SkipTokenState,
TokenNL: SkipTokenState,
TokenComment: CommentState,
TokenNone: TerminalState,
},
}
// ParseAST will parse input from an io.Reader using
// an LL(1) parser.
func ParseAST(r io.Reader) ([]AST, error) {
lexer := iniLexer{}
tokens, err := lexer.Tokenize(r)
if err != nil {
return []AST{}, err
}
return parse(tokens)
}
// ParseASTBytes will parse input from a byte slice using
// an LL(1) parser.
func ParseASTBytes(b []byte) ([]AST, error) {
lexer := iniLexer{}
tokens, err := lexer.tokenize(b)
if err != nil {
return []AST{}, err
}
return parse(tokens)
}
func parse(tokens []Token) ([]AST, error) {
start := Start
stack := newParseStack(3, len(tokens))
stack.Push(start)
s := newSkipper()
loop:
for stack.Len() > 0 {
k := stack.Pop()
var tok Token
if len(tokens) == 0 {
// this occurs when all the tokens have been processed
// but reduction of what's left on the stack needs to
// occur.
tok = emptyToken
} else {
tok = tokens[0]
}
step := parseTable[k.Kind][tok.Type()]
if s.ShouldSkip(tok) {
// being in a skip state with no tokens will break out of
// the parse loop since there is nothing left to process.
if len(tokens) == 0 {
break loop
}
// if should skip is true, we skip the tokens until should skip is set to false.
step = SkipTokenState
}
switch step {
case TerminalState:
// Finished parsing. Push what should be the last
// statement to the stack. If there is anything left
// on the stack, an error in parsing has occurred.
if k.Kind != ASTKindStart {
stack.MarkComplete(k)
}
break loop
case SkipTokenState:
// When skipping a token, the previous state was popped off the stack.
// To maintain the correct state, the previous state will be pushed
// onto the stack.
stack.Push(k)
case StatementState:
if k.Kind != ASTKindStart {
stack.MarkComplete(k)
}
expr := newExpression(tok)
stack.Push(expr)
case StatementPrimeState:
if tok.Type() != TokenOp {
stack.MarkComplete(k)
continue
}
if k.Kind != ASTKindExpr {
return nil, NewParseError(
fmt.Sprintf("invalid expression: expected Expr type, but found %T type", k),
)
}
k = trimSpaces(k)
expr := newEqualExpr(k, tok)
stack.Push(expr)
case ValueState:
// ValueState requires the previous state to either be an equal expression
// or an expression statement.
switch k.Kind {
case ASTKindEqualExpr:
// assigning a value to some key
k.AppendChild(newExpression(tok))
stack.Push(newExprStatement(k))
case ASTKindExpr:
k.Root.raw = append(k.Root.raw, tok.Raw()...)
stack.Push(k)
case ASTKindExprStatement:
root := k.GetRoot()
children := root.GetChildren()
if len(children) == 0 {
return nil, NewParseError(
fmt.Sprintf("invalid expression: AST contains no children %s", k.Kind),
)
}
rhs := children[len(children)-1]
if rhs.Root.ValueType != QuotedStringType {
rhs.Root.ValueType = StringType
rhs.Root.raw = append(rhs.Root.raw, tok.Raw()...)
}
children[len(children)-1] = rhs
root.SetChildren(children)
stack.Push(k)
}
case OpenScopeState:
if !runeCompare(tok.Raw(), openBrace) {
return nil, NewParseError("expected '['")
}
// If OpenScopeState is not at the start, we must mark the previous ast as complete
//
// for example: if previous ast was a skip statement;
// we should mark it as complete before we create a new statement
if k.Kind != ASTKindStart {
stack.MarkComplete(k)
}
stmt := newStatement()
stack.Push(stmt)
case CloseScopeState:
if !runeCompare(tok.Raw(), closeBrace) {
return nil, NewParseError("expected ']'")
}
k = trimSpaces(k)
stack.Push(newCompletedSectionStatement(k))
case SectionState:
var stmt AST
switch k.Kind {
case ASTKindStatement:
// If there are multiple literals inside of a scope declaration,
// then the current token's raw value will be appended to the Name.
//
// This handles cases like [ profile default ]
//
// k will represent a SectionStatement with the children representing
// the label of the section
stmt = newSectionStatement(tok)
case ASTKindSectionStatement:
k.Root.raw = append(k.Root.raw, tok.Raw()...)
stmt = k
default:
return nil, NewParseError(
fmt.Sprintf("invalid statement: expected statement: %v", k.Kind),
)
}
stack.Push(stmt)
case MarkCompleteState:
if k.Kind != ASTKindStart {
stack.MarkComplete(k)
}
if stack.Len() == 0 {
stack.Push(start)
}
case SkipState:
stack.Push(newSkipStatement(k))
s.Skip()
case CommentState:
if k.Kind == ASTKindStart {
stack.Push(k)
} else {
stack.MarkComplete(k)
}
stmt := newCommentStatement(tok)
stack.Push(stmt)
default:
return nil, NewParseError(
fmt.Sprintf("invalid state with ASTKind %v and TokenType %v",
k.Kind, tok.Type()))
}
if len(tokens) > 0 {
tokens = tokens[1:]
}
}
// this occurs when a statement has not been completed
if stack.top > 1 {
return nil, NewParseError(fmt.Sprintf("incomplete ini expression"))
}
// returns a sublist which exludes the start symbol
return stack.List(), nil
}
// trimSpaces will trim spaces on the left and right hand side of
// the literal.
func trimSpaces(k AST) AST {
// trim left hand side of spaces
for i := 0; i < len(k.Root.raw); i++ {
if !isWhitespace(k.Root.raw[i]) {
break
}
k.Root.raw = k.Root.raw[1:]
i--
}
// trim right hand side of spaces
for i := len(k.Root.raw) - 1; i >= 0; i-- {
if !isWhitespace(k.Root.raw[i]) {
break
}
k.Root.raw = k.Root.raw[:len(k.Root.raw)-1]
}
return k
}
|