clerk/journal/parser/parser.go (view raw)
Oleksandr Smirnov
Oleksandr Smirnov
olexsmir@gmail.com journal: parse the conversion value, 1 month ago
olexsmir@gmail.com journal: parse the conversion value, 1 month ago
| 1 | package parser |
| 2 | |
| 3 | import ( |
| 4 | "fmt" |
| 5 | "strconv" |
| 6 | "strings" |
| 7 | |
| 8 | "olexsmir.xyz/clerk/internal/decimal" |
| 9 | "olexsmir.xyz/clerk/journal/ast" |
| 10 | "olexsmir.xyz/clerk/journal/lexer" |
| 11 | "olexsmir.xyz/clerk/journal/token" |
| 12 | ) |
| 13 | |
| 14 | type Parser struct { |
| 15 | lexer *lexer.Lexer |
| 16 | errors []*ast.ParseError |
| 17 | cur token.Token |
| 18 | peek token.Token |
| 19 | } |
| 20 | |
| 21 | func New(lex *lexer.Lexer) *Parser { |
| 22 | p := &Parser{lexer: lex} |
| 23 | p.advance() // populate .peek |
| 24 | p.advance() // populate .cur |
| 25 | return p |
| 26 | } |
| 27 | |
| 28 | func (p *Parser) ParseJournal() *ast.Journal { |
| 29 | f := &ast.Journal{} |
| 30 | for p.cur.Type != token.EOF { |
| 31 | if e := p.parseEntry(); e != nil { |
| 32 | f.Entries = append(f.Entries, e) |
| 33 | } |
| 34 | } |
| 35 | f.Errors = p.errors |
| 36 | return f |
| 37 | } |
| 38 | |
| 39 | func isDirectiveKeyword(t token.Type) bool { |
| 40 | switch t { |
| 41 | case token.COMMENTKW, token.ACCOUNT, token.COMMODITY, token.INCLUDE, |
| 42 | token.ALIAS, token.PAYEE, token.TAG, token.APPLY, token.END, |
| 43 | token.YEAR, token.DECIMALMARK, token.D, token.P, token.N, token.C: |
| 44 | return true |
| 45 | } |
| 46 | return false |
| 47 | } |
| 48 | |
| 49 | func (p *Parser) parseEntry() ast.Entry { |
| 50 | if p.got(token.BANG) || p.got(token.AT) { |
| 51 | if isDirectiveKeyword(p.peek.Type) { |
| 52 | p.advance() // consume prefix |
| 53 | } |
| 54 | } |
| 55 | switch p.cur.Type { |
| 56 | case token.ILLEGAL: |
| 57 | p.errorf("illegal character %q", p.cur.Literal) |
| 58 | p.advance() |
| 59 | return nil |
| 60 | case token.INDENT: |
| 61 | p.errorf("unexpected indent") |
| 62 | p.syncToNextline() |
| 63 | return nil |
| 64 | case token.DATE: |
| 65 | return p.parseTransaction() |
| 66 | case token.TILDE: |
| 67 | return p.parsePeriodicTransaction() |
| 68 | case token.EQ: |
| 69 | return p.parseAutomatedTransaction() |
| 70 | case token.NEWLINE: |
| 71 | return p.parseBlankLine() |
| 72 | case token.SEMICOLON, token.HASH, token.PERCENT, token.STAR: |
| 73 | return p.parseComment() |
| 74 | case token.ACCOUNT: |
| 75 | return p.parseAccountDirective() |
| 76 | case token.COMMODITY: |
| 77 | return p.parseCommodityDirective() |
| 78 | case token.INCLUDE: |
| 79 | return p.parseIncludeDirective() |
| 80 | case token.ALIAS: |
| 81 | return p.parseAliasDirective() |
| 82 | case token.PAYEE: |
| 83 | return p.parsePayeeDirective() |
| 84 | case token.TAG: |
| 85 | return p.parseTagDirective() |
| 86 | case token.YEAR: |
| 87 | return p.parseYearDirective() |
| 88 | case token.DECIMALMARK: |
| 89 | return p.parseDecimalMarkDirective() |
| 90 | case token.D: |
| 91 | return p.parseDefaultCommodityDirective() |
| 92 | case token.P: |
| 93 | return p.parseMarketPriceDirective() |
| 94 | case token.N: |
| 95 | return p.parseIgnoredDirective() |
| 96 | case token.C: |
| 97 | return p.parseConversionDirective() |
| 98 | case token.APPLY: |
| 99 | return p.parseApplyDirective() |
| 100 | case token.END: |
| 101 | return p.parseEndDirective() |
| 102 | case token.COMMENTKW: |
| 103 | return p.parseCommentBlockDirective() |
| 104 | default: |
| 105 | p.errorf("unexpected token %s", p.cur.Type) |
| 106 | p.sync() |
| 107 | return nil |
| 108 | } |
| 109 | } |
| 110 | |
| 111 | func (p *Parser) parseTransaction() *ast.Transaction { |
| 112 | s := p.cur.Span |
| 113 | tx := &ast.Transaction{} |
| 114 | |
| 115 | tx.Date = p.parseDate() |
| 116 | |
| 117 | p.skipWhitespace() |
| 118 | |
| 119 | // optional secondary date |
| 120 | if p.got(token.EQ) { |
| 121 | p.advance() |
| 122 | p.skipWhitespace() |
| 123 | d := p.parseDate() |
| 124 | tx.SecondDate = &d |
| 125 | } |
| 126 | |
| 127 | p.skipWhitespace() |
| 128 | |
| 129 | // optional status |
| 130 | tx.Status = p.parseStatus() |
| 131 | |
| 132 | // optional code |
| 133 | if p.got(token.LPAREN) { |
| 134 | p.advance() |
| 135 | var code strings.Builder |
| 136 | for p.cur.Type != token.RPAREN { |
| 137 | _, _ = code.WriteString(p.cur.Literal) |
| 138 | p.advance() |
| 139 | } |
| 140 | tx.Code = new(code.String()) |
| 141 | p.advance() // TODO: why? |
| 142 | p.skipWhitespace() |
| 143 | } |
| 144 | |
| 145 | // optional payee | note |
| 146 | if p.got(token.TEXT) || p.got(token.STRING) { |
| 147 | tx.Payee = p.parsePayee() |
| 148 | |
| 149 | // check for | separator |
| 150 | if p.got(token.WHITESPACE) { |
| 151 | p.skipWhitespace() |
| 152 | } |
| 153 | |
| 154 | if p.got(token.PIPE) { |
| 155 | p.advance() |
| 156 | if p.got(token.TEXT) { |
| 157 | n := p.cur.Literal |
| 158 | p.advance() |
| 159 | tx.Note = &n |
| 160 | } |
| 161 | } |
| 162 | } |
| 163 | |
| 164 | tx.Comment = p.parseOptInlineComment() |
| 165 | p.expectNewline() |
| 166 | |
| 167 | // header comments — indented ; lines before first posting |
| 168 | for p.got(token.INDENT) && p.willGet(token.SEMICOLON) { |
| 169 | p.advance() // consume indent |
| 170 | c := p.parseComment() |
| 171 | tx.HeaderComments = append(tx.HeaderComments, c) |
| 172 | } |
| 173 | |
| 174 | // postings |
| 175 | for p.got(token.INDENT) { |
| 176 | if p := p.parsePosting(); p != nil { |
| 177 | tx.Postings = append(tx.Postings, p) |
| 178 | } |
| 179 | } |
| 180 | |
| 181 | tx.Span = p.span(s) |
| 182 | return tx |
| 183 | } |
| 184 | |
| 185 | func unquote(s string) string { |
| 186 | if len(s) >= 2 && ((s[0] == '"' && s[len(s)-1] == '"') || (s[0] == '\'' && s[len(s)-1] == '\'')) { |
| 187 | return s[1 : len(s)-1] |
| 188 | } |
| 189 | return s |
| 190 | } |
| 191 | |
| 192 | func (p *Parser) parsePayee() *ast.Payee { |
| 193 | s := p.cur.Span |
| 194 | |
| 195 | if p.got(token.STRING) { |
| 196 | name := unquote(p.cur.Literal) |
| 197 | p.advance() |
| 198 | return &ast.Payee{Name: name, Span: p.span(s)} |
| 199 | } |
| 200 | |
| 201 | // keep spaces/tags between text tokens; stop before trailing whitespace |
| 202 | var name strings.Builder |
| 203 | for p.got(token.TEXT) || p.got(token.INT) || p.got(token.DECIMAL) || (p.got(token.WHITESPACE) && (p.willGet(token.TEXT) || p.willGet(token.INT) || p.willGet(token.DECIMAL))) { |
| 204 | _, _ = name.WriteString(p.cur.Literal) |
| 205 | p.advance() |
| 206 | } |
| 207 | return &ast.Payee{Name: unquote(name.String()), Span: p.span(s)} |
| 208 | } |
| 209 | |
| 210 | func (p *Parser) parsePeriodicTransaction() *ast.PeriodicTransaction { |
| 211 | s := p.cur.Span |
| 212 | p.expect(token.TILDE) |
| 213 | p.skipWhitespace() |
| 214 | |
| 215 | pt := &ast.PeriodicTransaction{} |
| 216 | |
| 217 | pt.Span = p.span(s) |
| 218 | pt.Period = p.parsePeriod() |
| 219 | |
| 220 | if desc := p.parseOptPeriodicDescription(); desc != "" { |
| 221 | pt.Description = &desc |
| 222 | } |
| 223 | |
| 224 | comment := p.parseOptInlineComment() |
| 225 | p.expectNewline() |
| 226 | |
| 227 | var headerComments []*ast.Comment |
| 228 | var postings []*ast.Posting |
| 229 | for p.got(token.INDENT) || p.got(token.SEMICOLON) { |
| 230 | if p.got(token.SEMICOLON) { |
| 231 | c := p.parseComment() |
| 232 | headerComments = append(headerComments, c) |
| 233 | continue |
| 234 | } |
| 235 | posting := p.parsePosting() |
| 236 | if posting != nil { |
| 237 | postings = append(postings, posting) |
| 238 | } |
| 239 | } |
| 240 | |
| 241 | pt.HeaderComments = headerComments |
| 242 | pt.Postings = postings |
| 243 | pt.Comment = comment |
| 244 | return pt |
| 245 | } |
| 246 | |
| 247 | func (p *Parser) parseAutomatedTransaction() *ast.AutomatedTransaction { |
| 248 | s := p.cur.Span |
| 249 | p.expect(token.EQ) |
| 250 | p.skipWhitespace() |
| 251 | |
| 252 | at := &ast.AutomatedTransaction{} |
| 253 | at.Span = p.span(s) |
| 254 | |
| 255 | at.Expr = p.parseDirectiveExpr() |
| 256 | at.Comment = p.parseOptInlineComment() |
| 257 | p.expectNewline() |
| 258 | |
| 259 | // header comments |
| 260 | for p.got(token.INDENT) && p.willGet(token.SEMICOLON) { |
| 261 | p.advance() |
| 262 | at.HeaderComments = append(at.HeaderComments, p.parseComment()) |
| 263 | } |
| 264 | |
| 265 | // postings |
| 266 | for p.got(token.INDENT) { |
| 267 | if p := p.parsePosting(); p != nil { |
| 268 | at.Postings = append(at.Postings, p) |
| 269 | } |
| 270 | } |
| 271 | |
| 272 | return at |
| 273 | } |
| 274 | |
| 275 | func (p *Parser) parsePeriod() ast.Period { |
| 276 | s := p.cur.Span |
| 277 | |
| 278 | var periodBuf strings.Builder |
| 279 | |
| 280 | for !p.got(token.NEWLINE) && !p.got(token.EOF) && |
| 281 | !p.got(token.SEMICOLON) && !p.got(token.HASH) && !p.got(token.PERCENT) && !p.got(token.STAR) { |
| 282 | |
| 283 | if p.got(token.WHITESPACE) { |
| 284 | if len(p.cur.Literal) >= 2 { |
| 285 | break |
| 286 | } |
| 287 | if p.willGet(token.NEWLINE) || p.willGet(token.EOF) || |
| 288 | p.willGet(token.SEMICOLON) || p.willGet(token.HASH) || |
| 289 | p.willGet(token.PERCENT) || p.willGet(token.STAR) { |
| 290 | p.advance() |
| 291 | continue |
| 292 | } |
| 293 | } |
| 294 | |
| 295 | periodBuf.WriteString(p.cur.Literal) |
| 296 | p.advance() |
| 297 | } |
| 298 | |
| 299 | str := periodBuf.String() |
| 300 | period := ast.Period{Raw: str, Span: p.span(s)} |
| 301 | |
| 302 | if _, after, ok := strings.Cut(str, " from "); ok { |
| 303 | end := strings.Index(after, " ") |
| 304 | dateStr := after |
| 305 | if end >= 0 { |
| 306 | dateStr = after[:end] |
| 307 | } |
| 308 | if d := parseSimpleDate(dateStr); d.Year > 0 { |
| 309 | period.From = &d |
| 310 | rest := after |
| 311 | if end >= 0 { |
| 312 | rest = after[end:] |
| 313 | } |
| 314 | if _, toAfter, ok := strings.Cut(rest, " to "); ok { |
| 315 | if toEnd := strings.Index(toAfter, " "); toEnd >= 0 { |
| 316 | toAfter = toAfter[:toEnd] |
| 317 | } |
| 318 | if d := parseSimpleDate(toAfter); d.Year > 0 { |
| 319 | period.To = &d |
| 320 | } |
| 321 | } |
| 322 | } |
| 323 | } |
| 324 | return period |
| 325 | } |
| 326 | |
| 327 | func (p *Parser) parseComment() *ast.Comment { |
| 328 | s := p.cur.Span |
| 329 | marker := p.cur.Literal[0] |
| 330 | p.advance() |
| 331 | p.skipWhitespace() |
| 332 | |
| 333 | var text string |
| 334 | if p.got(token.TEXT) { |
| 335 | text = p.cur.Literal |
| 336 | p.advance() |
| 337 | } |
| 338 | |
| 339 | p.expectNewline() |
| 340 | |
| 341 | return &ast.Comment{ |
| 342 | Marker: marker, |
| 343 | Text: text, |
| 344 | Span: p.span(s), |
| 345 | } |
| 346 | } |
| 347 | |
| 348 | func (p *Parser) parseAccountDirective() *ast.AccountDirective { |
| 349 | s := p.cur.Span |
| 350 | p.expect(token.ACCOUNT) |
| 351 | p.skipWhitespace() |
| 352 | |
| 353 | account := p.parseAccount() |
| 354 | comment := p.parseOptInlineComment() |
| 355 | p.expectNewline() |
| 356 | |
| 357 | for p.got(token.INDENT) { |
| 358 | p.advance() |
| 359 | for !p.got(token.NEWLINE) && !p.got(token.EOF) { |
| 360 | p.advance() |
| 361 | } |
| 362 | p.expectNewline() |
| 363 | } |
| 364 | |
| 365 | return &ast.AccountDirective{ |
| 366 | Account: account, |
| 367 | Comment: comment, |
| 368 | Span: p.span(s), |
| 369 | } |
| 370 | } |
| 371 | |
| 372 | func (p *Parser) parseCommodityDirective() *ast.CommodityDirective { |
| 373 | s := p.cur.Span |
| 374 | p.expect(token.COMMODITY) |
| 375 | p.skipWhitespace() |
| 376 | |
| 377 | var commodity string |
| 378 | var format *ast.Amount |
| 379 | |
| 380 | switch p.cur.Type { |
| 381 | case token.COMMODITYMARK, token.TEXT, token.STRING: |
| 382 | commodity = p.cur.Literal |
| 383 | if p.got(token.STRING) { |
| 384 | commodity = unquote(commodity) |
| 385 | } |
| 386 | p.advance() |
| 387 | hadSpace := p.got(token.WHITESPACE) |
| 388 | p.skipWhitespace() |
| 389 | if p.got(token.INT) || p.got(token.DECIMAL) || p.got(token.TEXT) { |
| 390 | format = p.parseAmount() |
| 391 | format.Commodity = commodity |
| 392 | format.CommodityPos = ast.CommodityBefore |
| 393 | format.HasSpace = hadSpace |
| 394 | } |
| 395 | case token.INT, token.DECIMAL: |
| 396 | format = p.parseAmount() |
| 397 | commodity = format.Commodity |
| 398 | default: |
| 399 | p.errorf("expected commodity name or amount, got %s", p.cur.Type) |
| 400 | } |
| 401 | |
| 402 | if commodity == "" { |
| 403 | p.errorf("expected commodity name, got %s", p.cur.Type) |
| 404 | } |
| 405 | |
| 406 | comment := p.parseOptInlineComment() |
| 407 | p.expectNewline() |
| 408 | |
| 409 | for p.got(token.INDENT) { |
| 410 | p.advance() |
| 411 | p.skipWhitespace() |
| 412 | if p.got(token.COMMODITYMARK) && p.cur.Literal == "format" { |
| 413 | p.advance() |
| 414 | p.skipWhitespace() |
| 415 | format = p.parseAmount() |
| 416 | } else { |
| 417 | for !p.got(token.NEWLINE) && !p.got(token.EOF) { |
| 418 | p.advance() |
| 419 | } |
| 420 | } |
| 421 | p.expectNewline() |
| 422 | } |
| 423 | |
| 424 | cd := &ast.CommodityDirective{ |
| 425 | Commodity: commodity, |
| 426 | Comment: comment, |
| 427 | Span: p.span(s), |
| 428 | } |
| 429 | if format != nil { |
| 430 | cd.Format = *format |
| 431 | } |
| 432 | return cd |
| 433 | } |
| 434 | |
| 435 | func (p *Parser) parseIncludeDirective() *ast.IncludeDirective { |
| 436 | s := p.cur.Span |
| 437 | p.expect(token.INCLUDE) |
| 438 | p.skipWhitespace() |
| 439 | |
| 440 | id := &ast.IncludeDirective{} |
| 441 | |
| 442 | if p.got(token.TEXT) { |
| 443 | id.Path = p.cur.Literal |
| 444 | p.advance() |
| 445 | } else { |
| 446 | p.errorf("expected file path, got %s", p.cur.Type) |
| 447 | } |
| 448 | |
| 449 | p.skipWhitespace() |
| 450 | id.Comment = p.parseOptInlineComment() |
| 451 | p.expectNewline() |
| 452 | id.Span = p.span(s) |
| 453 | return id |
| 454 | } |
| 455 | |
| 456 | func (p *Parser) parseAliasDirective() *ast.AliasDirective { |
| 457 | s := p.cur.Span |
| 458 | alias := &ast.AliasDirective{} |
| 459 | p.expect(token.ALIAS) |
| 460 | p.skipWhitespace() |
| 461 | alias.From = p.parseAccount() |
| 462 | p.skipWhitespace() |
| 463 | p.expect(token.EQ) |
| 464 | p.skipWhitespace() |
| 465 | alias.To = p.parseAccount() |
| 466 | p.skipWhitespace() |
| 467 | alias.Comment = p.parseOptInlineComment() |
| 468 | p.expectNewline() |
| 469 | alias.Span = p.span(s) |
| 470 | return alias |
| 471 | } |
| 472 | |
| 473 | func (p *Parser) parsePayeeDirective() *ast.PayeeDirective { |
| 474 | s := p.cur.Span |
| 475 | p.expect(token.PAYEE) |
| 476 | p.skipWhitespace() |
| 477 | |
| 478 | name := "" |
| 479 | if p.got(token.TEXT) || p.got(token.STRING) { |
| 480 | name = p.parsePayee().Name |
| 481 | } |
| 482 | |
| 483 | comment := p.parseOptInlineComment() |
| 484 | p.expectNewline() |
| 485 | |
| 486 | return &ast.PayeeDirective{ |
| 487 | Name: name, |
| 488 | Comment: comment, |
| 489 | Span: p.span(s), |
| 490 | } |
| 491 | } |
| 492 | |
| 493 | func (p *Parser) parseTagDirective() *ast.TagDirective { |
| 494 | s := p.cur.Span |
| 495 | p.expect(token.TAG) |
| 496 | p.skipWhitespace() |
| 497 | |
| 498 | name := "" |
| 499 | if p.got(token.TEXT) { |
| 500 | name = p.cur.Literal |
| 501 | p.advance() |
| 502 | } |
| 503 | |
| 504 | comment := p.parseOptInlineComment() |
| 505 | p.expectNewline() |
| 506 | |
| 507 | return &ast.TagDirective{ |
| 508 | Name: name, |
| 509 | Comment: comment, |
| 510 | Span: p.span(s), |
| 511 | } |
| 512 | } |
| 513 | |
| 514 | func (p *Parser) parseYearDirective() *ast.YearDirective { |
| 515 | s := p.cur.Span |
| 516 | year := &ast.YearDirective{} |
| 517 | p.expect(token.YEAR) |
| 518 | p.skipWhitespace() |
| 519 | |
| 520 | if p.got(token.INT) { |
| 521 | year.Year, _ = strconv.Atoi(p.cur.Literal) |
| 522 | p.advance() |
| 523 | } else { |
| 524 | p.errorf("expected year, got %s", p.cur.Type) |
| 525 | } |
| 526 | |
| 527 | p.skipWhitespace() |
| 528 | year.Comment = p.parseOptInlineComment() |
| 529 | p.expectNewline() |
| 530 | year.Span = p.span(s) |
| 531 | return year |
| 532 | } |
| 533 | |
| 534 | func (p *Parser) parseDecimalMarkDirective() *ast.DecimalMarkDirective { |
| 535 | s := p.cur.Span |
| 536 | mark := &ast.DecimalMarkDirective{} |
| 537 | p.expect(token.DECIMALMARK) |
| 538 | p.skipWhitespace() |
| 539 | |
| 540 | mark.Mark = byte('.') |
| 541 | if p.got(token.TEXT) { |
| 542 | if len(p.cur.Literal) > 0 { |
| 543 | mark.Mark = p.cur.Literal[0] |
| 544 | } |
| 545 | p.advance() |
| 546 | } |
| 547 | |
| 548 | p.skipWhitespace() |
| 549 | mark.Comment = p.parseOptInlineComment() |
| 550 | p.expectNewline() |
| 551 | mark.Span = p.span(s) |
| 552 | return mark |
| 553 | } |
| 554 | |
| 555 | func (p *Parser) parseDefaultCommodityDirective() *ast.DefaultCommodityDirective { |
| 556 | s := p.cur.Span |
| 557 | com := &ast.DefaultCommodityDirective{} |
| 558 | p.expect(token.D) |
| 559 | p.skipWhitespace() |
| 560 | com.Amount = *p.parseAmount() |
| 561 | p.skipWhitespace() |
| 562 | com.Comment = p.parseOptInlineComment() |
| 563 | p.expectNewline() |
| 564 | com.Span = p.span(s) |
| 565 | return com |
| 566 | } |
| 567 | |
| 568 | func (p *Parser) parseConversionDirective() *ast.ConversionDirective { |
| 569 | s := p.cur.Span |
| 570 | cd := &ast.ConversionDirective{} |
| 571 | p.expect(token.C) |
| 572 | p.skipWhitespace() |
| 573 | |
| 574 | if p.isAmountStart() { |
| 575 | cd.From = *p.parseAmount() |
| 576 | } else { |
| 577 | p.errorf("expected amount, got %s", p.cur.Type) |
| 578 | } |
| 579 | |
| 580 | p.skipWhitespace() |
| 581 | if p.got(token.EQ) { |
| 582 | p.advance() |
| 583 | p.skipWhitespace() |
| 584 | if p.isAmountStart() { |
| 585 | cd.To = *p.parseAmount() |
| 586 | } else { |
| 587 | p.errorf("expected amount, got %s", p.cur.Type) |
| 588 | } |
| 589 | } |
| 590 | |
| 591 | p.skipWhitespace() |
| 592 | cd.Comment = p.parseOptInlineComment() |
| 593 | p.expectNewline() |
| 594 | cd.Span = p.span(s) |
| 595 | return cd |
| 596 | } |
| 597 | |
| 598 | func (p *Parser) parseIgnoredDirective() *ast.IgnoredDirective { |
| 599 | s := p.cur.Span |
| 600 | p.expect(token.N) |
| 601 | p.skipWhitespace() |
| 602 | |
| 603 | id := &ast.IgnoredDirective{} |
| 604 | if p.got(token.TEXT) || p.got(token.COMMODITYMARK) { |
| 605 | id.Text = p.cur.Literal |
| 606 | p.advance() |
| 607 | } |
| 608 | p.skipWhitespace() |
| 609 | id.Comment = p.parseOptInlineComment() |
| 610 | |
| 611 | p.expectNewline() |
| 612 | id.Span = p.span(s) |
| 613 | return id |
| 614 | } |
| 615 | |
| 616 | func (p *Parser) parseMarketPriceDirective() *ast.MarketPriceDirective { |
| 617 | s := p.cur.Span |
| 618 | p.expect(token.P) |
| 619 | p.skipWhitespace() |
| 620 | |
| 621 | mp := &ast.MarketPriceDirective{} |
| 622 | mp.DateTime.Date = p.parseDate() |
| 623 | p.skipWhitespace() |
| 624 | |
| 625 | if p.got(token.TIME) { |
| 626 | mp.DateTime.Time = new(p.parseTime()) |
| 627 | p.skipWhitespace() |
| 628 | } |
| 629 | |
| 630 | tok, _ := p.expect(token.COMMODITYMARK) |
| 631 | mp.Commodity = tok.Literal |
| 632 | p.advance() |
| 633 | |
| 634 | mp.Amount = *p.parseAmount() |
| 635 | |
| 636 | p.skipWhitespace() |
| 637 | mp.Comment = p.parseOptInlineComment() |
| 638 | |
| 639 | p.expectNewline() |
| 640 | mp.Span = p.span(s) |
| 641 | return mp |
| 642 | } |
| 643 | |
| 644 | func (p *Parser) parseTime() ast.Time { |
| 645 | s := p.cur.Span |
| 646 | tok, _ := p.expect(token.TIME) |
| 647 | lit := tok.Literal |
| 648 | |
| 649 | parts := strings.Split(lit, ":") |
| 650 | if len(parts) < 2 { |
| 651 | p.errorf("invalid time format: %q", lit) |
| 652 | return ast.Time{Span: p.span(s)} |
| 653 | } |
| 654 | |
| 655 | hour, _ := strconv.Atoi(parts[0]) |
| 656 | minute, _ := strconv.Atoi(parts[1]) |
| 657 | second := 0 |
| 658 | if len(parts) > 2 { |
| 659 | second, _ = strconv.Atoi(parts[2]) |
| 660 | } |
| 661 | |
| 662 | if hour < 0 || hour > 23 { |
| 663 | p.errorf("invalid hour %d in time %q", hour, lit) |
| 664 | } |
| 665 | if minute < 0 || minute > 59 { |
| 666 | p.errorf("invalid minute %d in time %q", minute, lit) |
| 667 | } |
| 668 | if second < 0 || second > 59 { |
| 669 | p.errorf("invalid second %d in time %q", second, lit) |
| 670 | } |
| 671 | |
| 672 | return ast.Time{ |
| 673 | Hour: hour, |
| 674 | Minute: minute, |
| 675 | Second: second, |
| 676 | Span: p.span(s), |
| 677 | } |
| 678 | } |
| 679 | |
| 680 | func (p *Parser) parseApplyDirective() *ast.ApplyDirective { |
| 681 | s := p.cur.Span |
| 682 | p.expect(token.APPLY) |
| 683 | p.skipWhitespace() |
| 684 | |
| 685 | expr := p.parseDirectiveExpr() |
| 686 | comment := p.parseOptInlineComment() |
| 687 | p.expectNewline() |
| 688 | |
| 689 | return &ast.ApplyDirective{ |
| 690 | Expr: expr, |
| 691 | Comment: comment, |
| 692 | Span: p.span(s), |
| 693 | } |
| 694 | } |
| 695 | |
| 696 | func (p *Parser) parseEndDirective() *ast.EndDirective { |
| 697 | s := p.cur.Span |
| 698 | p.expect(token.END) |
| 699 | p.skipWhitespace() |
| 700 | |
| 701 | expr := p.parseDirectiveExpr() |
| 702 | comment := p.parseOptInlineComment() |
| 703 | p.expectNewline() |
| 704 | |
| 705 | return &ast.EndDirective{ |
| 706 | Expr: expr, |
| 707 | Comment: comment, |
| 708 | Span: p.span(s), |
| 709 | } |
| 710 | } |
| 711 | |
| 712 | func (p *Parser) parseCommentBlockDirective() *ast.CommentBlockDirective { |
| 713 | start := p.cur.Span |
| 714 | p.expect(token.COMMENTKW) |
| 715 | p.skipWhitespace() |
| 716 | |
| 717 | header := p.parseDirectiveExpr() |
| 718 | comment := p.parseOptInlineComment() |
| 719 | p.expectNewline() |
| 720 | |
| 721 | var content strings.Builder |
| 722 | for p.cur.Type != token.EOF { |
| 723 | if p.got(token.END) { |
| 724 | if p.willGet(token.NEWLINE) || p.willGet(token.EOF) { |
| 725 | p.advance() |
| 726 | p.expectNewline() |
| 727 | break |
| 728 | } |
| 729 | if p.willGet(token.WHITESPACE) { |
| 730 | endTok := p.cur |
| 731 | p.advance() |
| 732 | wsTok := p.cur |
| 733 | p.advance() |
| 734 | if p.got(token.TEXT) && p.cur.Literal == "comment" { // todo: this should check if it's an actual COMMENTKW token |
| 735 | p.advance() |
| 736 | p.parseDirectiveExpr() |
| 737 | p.parseOptInlineComment() |
| 738 | p.expectNewline() |
| 739 | break |
| 740 | } |
| 741 | content.WriteString(endTok.Literal) |
| 742 | content.WriteString(wsTok.Literal) |
| 743 | continue |
| 744 | } |
| 745 | } |
| 746 | content.WriteString(p.cur.Literal) |
| 747 | p.advance() |
| 748 | } |
| 749 | |
| 750 | return &ast.CommentBlockDirective{ |
| 751 | Header: header, |
| 752 | Content: content.String(), |
| 753 | Comment: comment, |
| 754 | Span: p.span(start), |
| 755 | } |
| 756 | } |
| 757 | |
| 758 | func (p *Parser) parseStatus() *ast.Status { |
| 759 | if p.got(token.STAR) || p.got(token.BANG) { |
| 760 | status := ast.StatusPending |
| 761 | if p.cur.Literal[0] == '*' { |
| 762 | status = ast.StatusCleared |
| 763 | } |
| 764 | st := &ast.Status{Value: status, Span: p.cur.Span} |
| 765 | p.advance() |
| 766 | p.skipWhitespace() |
| 767 | return st |
| 768 | } |
| 769 | return nil |
| 770 | } |
| 771 | |
| 772 | func (p *Parser) isAmountStart() bool { |
| 773 | switch p.cur.Type { |
| 774 | default: |
| 775 | return false |
| 776 | case token.COMMODITYMARK, token.STRING, token.INT, token.DECIMAL, token.MINUS, token.PLUS, token.PARENEXPR: |
| 777 | return true |
| 778 | case token.TEXT: |
| 779 | return len(p.cur.Literal) > 0 && p.cur.Literal[0] >= '0' && p.cur.Literal[0] <= '9' |
| 780 | } |
| 781 | } |
| 782 | |
| 783 | func (p *Parser) parseAmount() *ast.Amount { |
| 784 | s := p.cur.Span |
| 785 | amt := &ast.Amount{ |
| 786 | QuantityFmt: ast.QuantityFormat{Decimal: '.'}, |
| 787 | Span: p.span(s), |
| 788 | } |
| 789 | |
| 790 | // commodity before quantity: $10.00, eur 10.00 |
| 791 | if p.got(token.COMMODITYMARK) || p.got(token.TEXT) || p.got(token.STRING) { |
| 792 | amt.Commodity = unquote(p.cur.Literal) |
| 793 | amt.CommodityPos = ast.CommodityBefore |
| 794 | p.advance() |
| 795 | if p.got(token.WHITESPACE) { |
| 796 | amt.HasSpace = true |
| 797 | p.skipWhitespace() |
| 798 | } |
| 799 | switch p.cur.Type { |
| 800 | case token.MINUS: |
| 801 | amt.IsNegative = true |
| 802 | p.advance() |
| 803 | case token.PLUS: |
| 804 | p.advance() |
| 805 | } |
| 806 | p.parseQuantityInto(amt) |
| 807 | } else { |
| 808 | // optional sign |
| 809 | switch p.cur.Type { |
| 810 | case token.MINUS: |
| 811 | amt.IsNegative = true |
| 812 | p.advance() |
| 813 | case token.PLUS: |
| 814 | p.advance() |
| 815 | } |
| 816 | |
| 817 | // commodity before quantity: -$120, -eur 120: |
| 818 | if p.got(token.COMMODITYMARK) || p.got(token.TEXT) || p.got(token.STRING) { |
| 819 | amt.Commodity = unquote(p.cur.Literal) |
| 820 | amt.CommodityPos = ast.CommodityBefore |
| 821 | p.advance() |
| 822 | if p.got(token.WHITESPACE) { |
| 823 | amt.HasSpace = true |
| 824 | p.skipWhitespace() |
| 825 | } |
| 826 | } |
| 827 | |
| 828 | p.parseQuantityInto(amt) |
| 829 | |
| 830 | // commodity after quantity: 10.00 UAH, 10.00 "EUR" (only if not set) |
| 831 | if amt.Commodity == "" { |
| 832 | switch p.cur.Type { |
| 833 | case token.WHITESPACE: |
| 834 | p.skipWhitespace() |
| 835 | if p.got(token.COMMODITYMARK) || p.got(token.TEXT) || p.got(token.STRING) { |
| 836 | amt.HasSpace = true |
| 837 | amt.Commodity = unquote(p.cur.Literal) |
| 838 | amt.CommodityPos = ast.CommodityAfter |
| 839 | p.advance() |
| 840 | } |
| 841 | case token.COMMODITYMARK, token.TEXT, token.STRING: |
| 842 | amt.Commodity = unquote(p.cur.Literal) |
| 843 | amt.CommodityPos = ast.CommodityAfter |
| 844 | p.advance() |
| 845 | } |
| 846 | } |
| 847 | } |
| 848 | |
| 849 | return amt |
| 850 | } |
| 851 | |
| 852 | func (p *Parser) parseAmountWithOptExpr() *ast.Amount { |
| 853 | if p.got(token.STAR) { |
| 854 | p.advance() |
| 855 | p.skipWhitespace() |
| 856 | amt := p.parseAmount() |
| 857 | if amt != nil { |
| 858 | amt.IsExpr = true |
| 859 | } |
| 860 | return amt |
| 861 | } |
| 862 | if p.got(token.PARENEXPR) { |
| 863 | lit := p.cur.Literal |
| 864 | amt := &ast.Amount{ |
| 865 | IsExpr: true, |
| 866 | QuantityFmt: ast.QuantityFormat{Decimal: '.'}, |
| 867 | } |
| 868 | if len(lit) >= 2 && lit[0] == '(' && lit[len(lit)-1] == ')' { |
| 869 | inner := lit[1 : len(lit)-1] |
| 870 | i := 0 |
| 871 | for i < len(inner) && (inner[i] == ' ' || inner[i] == '\t') { |
| 872 | i++ |
| 873 | } |
| 874 | j := len(inner) |
| 875 | for j > i && (inner[j-1] == ' ' || inner[j-1] == '\t') { |
| 876 | j-- |
| 877 | } |
| 878 | amt.Expr = inner[i:j] |
| 879 | } |
| 880 | amt.Span = p.cur.Span |
| 881 | p.advance() |
| 882 | return amt |
| 883 | } |
| 884 | return p.parseAmount() |
| 885 | } |
| 886 | |
| 887 | func (p *Parser) parsePosting() *ast.Posting { |
| 888 | s := p.cur.Span |
| 889 | posting := &ast.Posting{} |
| 890 | p.expect(token.INDENT) |
| 891 | |
| 892 | // exit if it's empty line |
| 893 | if p.got(token.NEWLINE) || p.got(token.EOF) { |
| 894 | p.syncToNextline() |
| 895 | return nil |
| 896 | } |
| 897 | |
| 898 | // optional status, outside of brackets, '! (account)' |
| 899 | posting.Status = p.parseStatus() |
| 900 | |
| 901 | // detect virtual posting brackets |
| 902 | switch p.cur.Type { |
| 903 | case token.LPAREN: |
| 904 | posting.Type = ast.PostingVirtualUnbalanced |
| 905 | p.advance() |
| 906 | case token.LBRACKET: |
| 907 | posting.Type = ast.PostingVirtualBalanced |
| 908 | p.advance() |
| 909 | } |
| 910 | |
| 911 | // optional status, inside of brackets, '(* account)' |
| 912 | if p.got(token.STAR) || p.got(token.BANG) { |
| 913 | posting.Status = p.parseStatus() |
| 914 | } |
| 915 | |
| 916 | // validate, must be account text |
| 917 | if p.cur.Type != token.TEXT { |
| 918 | p.errorf("expected account name, got %s", p.cur.Type) |
| 919 | p.syncToNextline() |
| 920 | return nil |
| 921 | } |
| 922 | |
| 923 | posting.Account = p.parseAccount() |
| 924 | |
| 925 | // consume closing bracket |
| 926 | switch p.cur.Type { |
| 927 | case token.RPAREN: |
| 928 | p.advance() |
| 929 | case token.RBRACKET: |
| 930 | p.advance() |
| 931 | } |
| 932 | |
| 933 | // optional amount - after two spaces |
| 934 | if p.got(token.WHITESPACE) { |
| 935 | p.skipWhitespace() |
| 936 | if p.isAmountStart() || p.got(token.STAR) { |
| 937 | posting.Amount = p.parseAmountWithOptExpr() |
| 938 | } |
| 939 | } |
| 940 | |
| 941 | // optional cost '@' or '@@' |
| 942 | if p.got(token.WHITESPACE) { |
| 943 | p.skipWhitespace() |
| 944 | } |
| 945 | if p.got(token.AT) || p.got(token.ATAT) { |
| 946 | posting.Cost = p.parseCost() |
| 947 | } |
| 948 | |
| 949 | // optional balance assertion |
| 950 | if p.got(token.WHITESPACE) { |
| 951 | p.skipWhitespace() |
| 952 | } |
| 953 | if p.got(token.EQ) || p.got(token.EQEQ) || p.got(token.EQEQEQ) { |
| 954 | posting.Balance = p.parseBalanceAssertion() |
| 955 | } |
| 956 | |
| 957 | posting.Comment = p.parseOptInlineComment() |
| 958 | p.expectNewline() |
| 959 | |
| 960 | // continuation comments |
| 961 | for p.got(token.INDENT) && p.willGet(token.SEMICOLON) { |
| 962 | p.advance() |
| 963 | c := p.parseComment() |
| 964 | posting.Comments = append(posting.Comments, *c) |
| 965 | } |
| 966 | |
| 967 | posting.Span = p.span(s) |
| 968 | return posting |
| 969 | } |
| 970 | |
| 971 | func (p *Parser) parseCost() *ast.Cost { |
| 972 | s := p.cur.Span |
| 973 | isTotal := p.got(token.ATAT) |
| 974 | p.advance() // consume '@' '@@' |
| 975 | p.skipWhitespace() |
| 976 | return &ast.Cost{ |
| 977 | IsTotal: isTotal, |
| 978 | Amount: *p.parseAmount(), |
| 979 | Span: p.span(s), |
| 980 | } |
| 981 | } |
| 982 | |
| 983 | func (p *Parser) parseBalanceAssertion() *ast.BalanceAssertion { |
| 984 | s := p.cur.Span |
| 985 | |
| 986 | ba := &ast.BalanceAssertion{} |
| 987 | switch p.cur.Type { |
| 988 | case token.EQ: // basic assertion |
| 989 | case token.EQEQ: |
| 990 | ba.IsStrict = true |
| 991 | case token.EQEQEQ: |
| 992 | ba.IsStrict = true |
| 993 | ba.IsInclusive = true |
| 994 | } |
| 995 | p.advance() |
| 996 | p.skipWhitespace() |
| 997 | |
| 998 | ba.Amount = *p.parseAmount() |
| 999 | p.skipWhitespace() |
| 1000 | if p.got(token.AT) || p.got(token.ATAT) { |
| 1001 | c := p.parseCost() |
| 1002 | ba.Cost = c |
| 1003 | } |
| 1004 | ba.Span = p.span(s) |
| 1005 | return ba |
| 1006 | } |
| 1007 | |
| 1008 | func (p *Parser) readAccountSegment() (ast.SubAccount, bool) { |
| 1009 | switch p.cur.Type { |
| 1010 | case token.TEXT: |
| 1011 | sub := ast.SubAccount{Name: p.cur.Literal, Span: p.cur.Span} |
| 1012 | p.advance() |
| 1013 | |
| 1014 | // handle multi work segment, e.g: "credit card" |
| 1015 | if p.got(token.WHITESPACE) && p.willGet(token.TEXT) && len(p.peek.Literal) > 0 && p.peek.Literal[0] != '(' { |
| 1016 | sub.Name += " " |
| 1017 | p.advance() |
| 1018 | sub.Name += p.cur.Literal |
| 1019 | p.advance() |
| 1020 | } |
| 1021 | return sub, true |
| 1022 | |
| 1023 | case token.COMMODITYMARK: |
| 1024 | sub := ast.SubAccount{Name: p.cur.Literal, Span: p.cur.Span} |
| 1025 | p.advance() |
| 1026 | // merge "EUR" + "-HRK" to "EUR-HRK" |
| 1027 | for p.got(token.TEXT) { |
| 1028 | sub.Name += p.cur.Literal |
| 1029 | p.advance() |
| 1030 | } |
| 1031 | return sub, true |
| 1032 | |
| 1033 | default: |
| 1034 | return ast.SubAccount{}, false |
| 1035 | } |
| 1036 | } |
| 1037 | |
| 1038 | func (p *Parser) parseAccount() ast.Account { |
| 1039 | s := p.cur.Span |
| 1040 | acc := ast.Account{} |
| 1041 | |
| 1042 | sub, ok := p.readAccountSegment() |
| 1043 | if !ok { |
| 1044 | p.errorf("expected account, got %s", p.cur.Type) |
| 1045 | return ast.Account{} |
| 1046 | } |
| 1047 | acc.Name = append(acc.Name, sub) |
| 1048 | |
| 1049 | for p.got(token.COLON) { |
| 1050 | p.advance() |
| 1051 | sub, ok := p.readAccountSegment() |
| 1052 | if !ok { |
| 1053 | break |
| 1054 | } |
| 1055 | acc.Name = append(acc.Name, sub) |
| 1056 | } |
| 1057 | |
| 1058 | acc.Span = p.span(s) |
| 1059 | return acc |
| 1060 | } |
| 1061 | |
| 1062 | func (p *Parser) parseDate() ast.Date { |
| 1063 | s := p.cur.Span |
| 1064 | tok, ok := p.expect(token.DATE) |
| 1065 | if !ok { |
| 1066 | return ast.Date{Span: p.span(s)} |
| 1067 | } |
| 1068 | |
| 1069 | sep := byte(0) |
| 1070 | lit := tok.Literal |
| 1071 | for i := 0; i < len(lit); i++ { |
| 1072 | if lit[i] == '/' || lit[i] == '-' || lit[i] == '.' { |
| 1073 | sep = lit[i] |
| 1074 | break |
| 1075 | } |
| 1076 | } |
| 1077 | if sep == 0 { |
| 1078 | p.errorf("invalid date format: %q", lit) |
| 1079 | return ast.Date{Span: p.span(s)} |
| 1080 | } |
| 1081 | |
| 1082 | parts := strings.Split(lit, string(sep)) |
| 1083 | |
| 1084 | // M/D or MM/DD (year inferred) |
| 1085 | if len(parts) == 2 { |
| 1086 | month, err := strconv.Atoi(parts[0]) |
| 1087 | day, err2 := strconv.Atoi(parts[1]) |
| 1088 | if err != nil || err2 != nil { |
| 1089 | p.errorf("invalid date literal: %q", lit) |
| 1090 | return ast.Date{Span: p.span(s)} |
| 1091 | } |
| 1092 | if month < 1 || month > 12 { |
| 1093 | p.errorf("invalid month %d in %q", month, lit) |
| 1094 | return ast.Date{Span: p.span(s)} |
| 1095 | } |
| 1096 | if day < 1 || day > 31 { |
| 1097 | p.errorf("invalid day %d in %q", day, lit) |
| 1098 | return ast.Date{Span: p.span(s)} |
| 1099 | } |
| 1100 | return ast.Date{Month: month, Day: day, Sep: sep, Span: p.span(s)} |
| 1101 | } |
| 1102 | |
| 1103 | if len(parts) != 3 { |
| 1104 | p.errorf("invalid date format: %q", lit) |
| 1105 | return ast.Date{Span: p.span(s)} |
| 1106 | } |
| 1107 | |
| 1108 | year, err := strconv.Atoi(parts[0]) |
| 1109 | month, err2 := strconv.Atoi(parts[1]) |
| 1110 | day, err3 := strconv.Atoi(parts[2]) |
| 1111 | if err != nil || err2 != nil || err3 != nil { |
| 1112 | p.errorf("invalid date literal: %q", lit) |
| 1113 | return ast.Date{Span: p.span(s)} |
| 1114 | } |
| 1115 | if month < 1 || month > 12 { |
| 1116 | p.errorf("invalid month %d in %q", month, lit) |
| 1117 | return ast.Date{Span: p.span(s)} |
| 1118 | } |
| 1119 | if day < 1 || day > 31 { |
| 1120 | p.errorf("invalid day %d in %q", day, lit) |
| 1121 | return ast.Date{Span: p.span(s)} |
| 1122 | } |
| 1123 | |
| 1124 | return ast.Date{ |
| 1125 | Year: year, |
| 1126 | Month: month, |
| 1127 | Day: day, |
| 1128 | Sep: sep, |
| 1129 | Span: p.span(s), |
| 1130 | } |
| 1131 | } |
| 1132 | |
| 1133 | func (p *Parser) parseOptInlineComment() *ast.Comment { |
| 1134 | p.skipWhitespace() // todo: |
| 1135 | if p.cur.Type != token.SEMICOLON { |
| 1136 | return nil |
| 1137 | } |
| 1138 | |
| 1139 | s := p.cur.Span |
| 1140 | marker := p.cur.Literal[0] |
| 1141 | p.advance() // consume marker |
| 1142 | p.skipWhitespace() |
| 1143 | |
| 1144 | text := "" |
| 1145 | if p.got(token.TEXT) { |
| 1146 | text = p.cur.Literal |
| 1147 | p.advance() |
| 1148 | } |
| 1149 | |
| 1150 | return &ast.Comment{ |
| 1151 | Marker: marker, |
| 1152 | Text: text, |
| 1153 | Span: p.span(s), |
| 1154 | } |
| 1155 | } |
| 1156 | |
| 1157 | func (p *Parser) parseOptPeriodicDescription() string { |
| 1158 | if p.cur.Type != token.WHITESPACE || len(p.cur.Literal) < 2 { |
| 1159 | return "" |
| 1160 | } |
| 1161 | |
| 1162 | p.skipWhitespace() |
| 1163 | |
| 1164 | if p.cur.Type != token.TEXT { |
| 1165 | return "" |
| 1166 | } |
| 1167 | |
| 1168 | return p.parseDescription() |
| 1169 | } |
| 1170 | |
| 1171 | func (p *Parser) parseDescription() string { |
| 1172 | var desc strings.Builder |
| 1173 | for p.got(token.TEXT) || (p.got(token.WHITESPACE) && p.willGet(token.TEXT)) { |
| 1174 | _, _ = desc.WriteString(p.cur.Literal) |
| 1175 | p.advance() |
| 1176 | } |
| 1177 | return desc.String() |
| 1178 | } |
| 1179 | |
| 1180 | func (p *Parser) parseDirectiveExpr() string { |
| 1181 | var b strings.Builder |
| 1182 | for p.cur.Type != token.NEWLINE && p.cur.Type != token.EOF && p.cur.Type != token.SEMICOLON { |
| 1183 | _, _ = b.WriteString(p.cur.Literal) |
| 1184 | p.advance() |
| 1185 | } |
| 1186 | return b.String() |
| 1187 | } |
| 1188 | |
| 1189 | func (p *Parser) parseQuantityInto(amt *ast.Amount) { |
| 1190 | if p.cur.Type != token.INT && p.cur.Type != token.DECIMAL && p.cur.Type != token.TEXT { |
| 1191 | p.errorf("expected quantity, got %s", p.cur.Type) |
| 1192 | return |
| 1193 | } |
| 1194 | |
| 1195 | lit := p.cur.Literal |
| 1196 | p.advance() |
| 1197 | |
| 1198 | // detect format metadata before normalizing |
| 1199 | amt.QuantityFmt = detectFormat(lit) |
| 1200 | |
| 1201 | // normalize for decimal.NewFromString |
| 1202 | // remove thousands separators, replace decimal mark with '.' |
| 1203 | normalized := normalizeLiteral(lit, amt.QuantityFmt.Thousands, amt.QuantityFmt.Decimal) |
| 1204 | |
| 1205 | q, err := decimal.FromString(normalized) |
| 1206 | if err != nil { |
| 1207 | p.errorf("invalid quantity %q: %v", lit, err) |
| 1208 | return |
| 1209 | } |
| 1210 | |
| 1211 | if amt.IsNegative { |
| 1212 | q = q.Neg() |
| 1213 | } |
| 1214 | amt.Quantity = q |
| 1215 | } |
| 1216 | |
| 1217 | func (p *Parser) parseBlankLine() *ast.BlankLine { |
| 1218 | s := p.cur.Span |
| 1219 | p.expectNewline() |
| 1220 | return &ast.BlankLine{Span: s} |
| 1221 | } |
| 1222 | |
| 1223 | func (p *Parser) expectNewline() { |
| 1224 | if p.got(token.NEWLINE) || p.got(token.EOF) { |
| 1225 | if p.got(token.NEWLINE) { |
| 1226 | p.advance() |
| 1227 | } |
| 1228 | return |
| 1229 | } |
| 1230 | p.errorf("expected %s, got %s", token.NEWLINE, p.cur.Type) |
| 1231 | } |
| 1232 | |
| 1233 | func (p *Parser) advance() token.Token { |
| 1234 | prev := p.cur |
| 1235 | p.cur = p.peek |
| 1236 | p.peek = p.lexer.Next() |
| 1237 | return prev |
| 1238 | } |
| 1239 | |
| 1240 | func (p *Parser) got(kind token.Type) bool { return p.cur.Type == kind } |
| 1241 | func (p *Parser) willGet(kind token.Type) bool { return p.peek.Type == kind } |
| 1242 | |
| 1243 | func (p *Parser) expect(kind token.Type) (token.Token, bool) { |
| 1244 | if p.got(kind) { |
| 1245 | return p.advance(), true |
| 1246 | } |
| 1247 | p.errorf("expected %s, got %s", kind, p.cur.Type) |
| 1248 | return p.cur, false |
| 1249 | } |
| 1250 | |
| 1251 | func (p *Parser) errorf(format string, args ...any) { |
| 1252 | p.errors = append(p.errors, &ast.ParseError{ |
| 1253 | Span: p.cur.Span, |
| 1254 | Message: fmt.Sprintf(format, args...), |
| 1255 | }) |
| 1256 | } |
| 1257 | |
| 1258 | func (p *Parser) sync() { |
| 1259 | for { |
| 1260 | switch p.cur.Type { |
| 1261 | case token.EOF: |
| 1262 | return |
| 1263 | case token.NEWLINE: |
| 1264 | p.advance() |
| 1265 | switch p.cur.Type { |
| 1266 | case token.DATE, token.ACCOUNT, token.COMMODITY, |
| 1267 | token.INCLUDE, token.ALIAS, token.PAYEE, |
| 1268 | token.TAG, token.YEAR, token.D, token.P, |
| 1269 | token.APPLY, token.END, token.COMMENTKW, |
| 1270 | token.DECIMALMARK, token.TILDE, token.N, token.EQ: |
| 1271 | return |
| 1272 | } |
| 1273 | default: |
| 1274 | p.advance() |
| 1275 | } |
| 1276 | } |
| 1277 | } |
| 1278 | |
| 1279 | func (p *Parser) syncToNextline() { |
| 1280 | for p.cur.Type != token.NEWLINE && p.cur.Type != token.EOF { |
| 1281 | p.advance() |
| 1282 | } |
| 1283 | if p.got(token.NEWLINE) { |
| 1284 | p.advance() |
| 1285 | } |
| 1286 | } |
| 1287 | |
| 1288 | func (p *Parser) skipWhitespace() { |
| 1289 | for p.got(token.WHITESPACE) { |
| 1290 | p.advance() |
| 1291 | } |
| 1292 | } |
| 1293 | |
| 1294 | func (p *Parser) span(s token.Span) token.Span { |
| 1295 | return token.Span{Start: s.Start, End: p.cur.Span.Start} |
| 1296 | } |
| 1297 | |
| 1298 | func normalizeLiteral(lit string, thousands, decimal byte) string { |
| 1299 | var b strings.Builder |
| 1300 | for _, ch := range []byte(lit) { |
| 1301 | if thousands != 0 && ch == thousands { |
| 1302 | continue // skip thousands separator |
| 1303 | } |
| 1304 | if ch == decimal { |
| 1305 | b.WriteByte('.') |
| 1306 | } else { |
| 1307 | b.WriteByte(ch) |
| 1308 | } |
| 1309 | } |
| 1310 | return b.String() |
| 1311 | } |
| 1312 | |
| 1313 | func detectFormat(lit string) ast.QuantityFormat { |
| 1314 | var separators []int |
| 1315 | for i, ch := range []byte(lit) { |
| 1316 | if ch == '.' || ch == ',' || ch == ' ' || ch == '_' || ch == '\'' { |
| 1317 | separators = append(separators, i) |
| 1318 | } |
| 1319 | } |
| 1320 | |
| 1321 | if len(separators) == 0 { |
| 1322 | return ast.QuantityFormat{Decimal: '.', Thousands: 0, Precision: 0} |
| 1323 | } |
| 1324 | |
| 1325 | var decimal byte |
| 1326 | thousands := byte(0) |
| 1327 | precision := 0 |
| 1328 | |
| 1329 | if len(separators) == 1 { |
| 1330 | pos := separators[0] |
| 1331 | sepChar := lit[pos] |
| 1332 | if sepChar == ' ' || sepChar == '_' || sepChar == '\'' { |
| 1333 | thousands = sepChar |
| 1334 | decimal = '.' // default |
| 1335 | precision = 0 |
| 1336 | } else { |
| 1337 | decimal = sepChar |
| 1338 | precision = len(lit) - pos - 1 |
| 1339 | } |
| 1340 | } else { |
| 1341 | last := separators[len(separators)-1] |
| 1342 | decimal = lit[last] |
| 1343 | thousands = lit[separators[0]] |
| 1344 | precision = len(lit) - last - 1 |
| 1345 | } |
| 1346 | |
| 1347 | return ast.QuantityFormat{ |
| 1348 | Decimal: decimal, |
| 1349 | Thousands: thousands, |
| 1350 | Precision: precision, |
| 1351 | } |
| 1352 | } |
| 1353 | |
| 1354 | func parseSimpleDate(s string) ast.Date { |
| 1355 | if len(s) < 8 { |
| 1356 | return ast.Date{} |
| 1357 | } |
| 1358 | sep := byte('-') |
| 1359 | if strings.Contains(s, "/") { |
| 1360 | sep = byte('/') |
| 1361 | } else if strings.Contains(s, ".") { |
| 1362 | sep = byte('.') |
| 1363 | } |
| 1364 | parts := strings.Split(s, string(sep)) |
| 1365 | if len(parts) != 3 { |
| 1366 | return ast.Date{} |
| 1367 | } |
| 1368 | year, _ := strconv.Atoi(parts[0]) |
| 1369 | month, _ := strconv.Atoi(parts[1]) |
| 1370 | day, _ := strconv.Atoi(parts[2]) |
| 1371 | return ast.Date{Year: year, Month: month, Day: day, Sep: sep} |
| 1372 | } |