2016-09-10 08:30:40 +00:00
|
|
|
# NimYAML - YAML implementation in Nim
|
|
|
|
# (c) Copyright 2015 Felix Krause
|
|
|
|
#
|
|
|
|
# See the file "copying.txt", included in this
|
|
|
|
# distribution, for details about the copyright.
|
|
|
|
|
2016-09-10 11:38:42 +00:00
|
|
|
import lexbase, streams, strutils, unicode
|
2016-09-10 15:19:37 +00:00
|
|
|
when defined(yamlDebug):
|
|
|
|
import terminal
|
|
|
|
export terminal
|
2016-09-10 08:30:40 +00:00
|
|
|
|
|
|
|
type
|
|
|
|
StringSource* = object
|
|
|
|
src: string
|
|
|
|
pos: int
|
|
|
|
line, lineStart: int
|
|
|
|
|
|
|
|
SourceProvider* = concept c
|
|
|
|
advance(c) is char
|
|
|
|
lexCR(c)
|
|
|
|
lexLF(c)
|
|
|
|
|
2016-09-10 10:38:03 +00:00
|
|
|
# YamlLexer*[T: SourceProvider] = ref object # not possible -> compiler bug
|
|
|
|
YamlLexer*[T] = ref object
|
2016-09-10 08:30:40 +00:00
|
|
|
source: T
|
|
|
|
inFlow: bool
|
|
|
|
literalEndIndent: int
|
2016-09-10 15:19:37 +00:00
|
|
|
nextImpl, lineStartImpl, inlineImpl: LexerState[T]
|
2016-09-10 08:30:40 +00:00
|
|
|
buf*: string not nil
|
|
|
|
indentation*: int
|
2016-09-10 15:19:37 +00:00
|
|
|
blockScalarIndent: int
|
|
|
|
moreIndented*, folded*: bool
|
|
|
|
chomp*: ChompType
|
2016-09-10 10:38:03 +00:00
|
|
|
c: char
|
|
|
|
|
|
|
|
LexerState[T] = proc(lex: YamlLexer[T], t: var LexerToken): bool
|
2016-09-10 08:30:40 +00:00
|
|
|
|
|
|
|
LexerToken* = enum
|
|
|
|
ltYamlDirective, ltYamlVersion, ltTagDirective, ltTagShorthand,
|
2016-09-10 15:19:37 +00:00
|
|
|
ltTagUri, ltUnknownDirective, ltUnknownDirectiveParams, ltEmptyLine,
|
2016-09-10 08:30:40 +00:00
|
|
|
ltDirectivesEnd, ltDocumentEnd, ltStreamEnd, ltIndentation, ltQuotedScalar,
|
2016-09-10 15:19:37 +00:00
|
|
|
ltScalarPart, ltBlockScalarHeader, ltSeqItemInd, ltMapKeyInd, ltMapValInd,
|
2016-09-10 08:30:40 +00:00
|
|
|
ltBraceOpen, ltBraceClose, ltBracketOpen, ltBracketClose, ltComma,
|
|
|
|
ltLiteralTag, ltTagSuffix, ltAnchor, ltAlias
|
|
|
|
|
|
|
|
YamlLexerError* = object of Exception
|
2016-09-10 15:19:37 +00:00
|
|
|
line*, column*: int
|
|
|
|
lineContent*: string
|
|
|
|
|
|
|
|
ChompType* = enum
|
|
|
|
ctKeep, ctClip, ctStrip
|
2016-09-10 08:30:40 +00:00
|
|
|
|
|
|
|
# templates
|
|
|
|
|
2016-09-10 11:38:42 +00:00
|
|
|
proc advance(lex: YamlLexer[BaseLexer], step: int = 1) {.inline.} =
|
2016-09-10 08:30:40 +00:00
|
|
|
lex.source.bufpos.inc(step)
|
|
|
|
lex.c = lex.source.buf[lex.source.bufpos]
|
|
|
|
|
2016-09-10 11:38:42 +00:00
|
|
|
proc advance(lex: YamlLexer[StringSource], step: int = 1) {.inline.} =
|
2016-09-10 08:30:40 +00:00
|
|
|
lex.source.pos.inc(step)
|
2016-09-10 11:38:42 +00:00
|
|
|
if lex.source.pos >= lex.source.src.len: lex.c = EndOfFile
|
|
|
|
else: lex.c = lex.source.src[lex.source.pos]
|
2016-09-10 08:30:40 +00:00
|
|
|
|
|
|
|
# lexer states
|
|
|
|
|
|
|
|
proc outsideDoc[T](lex: YamlLexer[T], t: var LexerToken): bool
|
|
|
|
proc yamlVersion[T](lex: YamlLexer[T], t: var LexerToken): bool
|
|
|
|
proc tagShorthand[T](lex: YamlLexer[T], t: var LexerToken): bool
|
|
|
|
proc tagUri[T](lex: YamlLexer[T], t: var LexerToken): bool
|
|
|
|
proc unknownDirParams[T](lex: YamlLexer[T], t: var LexerToken): bool
|
|
|
|
proc expectLineEnd[T](lex: YamlLexer[T], t: var LexerToken): bool
|
2016-09-10 15:19:37 +00:00
|
|
|
proc possibleDirectivesEnd[T](lex: YamlLexer[T], t: var LexerToken): bool
|
|
|
|
proc possibleDocumentEnd[T](lex: YamlLexer[T], t: var LexerToken): bool
|
|
|
|
proc afterSeqInd[T](lex: YamlLexer[T], t: var LexerToken): bool
|
2016-09-11 08:02:10 +00:00
|
|
|
proc insideDoc[T](lex: YamlLexer[T], t: var LexerToken): bool {.locks:0.}
|
|
|
|
proc insideLine[T](lex: YamlLexer[T], t: var LexerToken): bool
|
2016-09-10 08:30:40 +00:00
|
|
|
proc plainScalarPart[T](lex: YamlLexer[T], t: var LexerToken): bool
|
2016-09-10 15:19:37 +00:00
|
|
|
proc blockScalarHeader[T](lex: YamlLexer[T], t: var LexerToken): bool
|
|
|
|
proc blockScalar[T](lex: YamlLexer[T], t: var LexerToken): bool
|
2016-09-10 08:30:40 +00:00
|
|
|
proc streamEnd[T](lex: YamlLexer[T], t: var LexerToken): bool
|
|
|
|
|
|
|
|
# interface
|
|
|
|
|
|
|
|
proc newYamlLexer*(source: Stream): YamlLexer[BaseLexer] =
|
2016-09-10 10:38:03 +00:00
|
|
|
result = YamlLexer[BaseLexer](source: BaseLexer(), inFlow: false, buf: "")
|
2016-09-10 08:30:40 +00:00
|
|
|
result.source.open(source)
|
|
|
|
result.c = result.source.buf[result.source.bufpos]
|
|
|
|
|
2016-09-10 10:38:03 +00:00
|
|
|
proc newYamlLexer*(source: string, startAt: int = 0):
|
|
|
|
YamlLexer[StringSource] =
|
|
|
|
result = YamlLexer[StringSource](buf: "", source:
|
2016-09-10 08:30:40 +00:00
|
|
|
StringSource(src: source, pos: startAt, lineStart: startAt, line: 1),
|
|
|
|
inFlow: false, c: source[startAt])
|
|
|
|
|
2016-09-10 10:38:03 +00:00
|
|
|
proc init*[T](lex: YamlLexer[T]) =
|
|
|
|
lex.nextImpl = outsideDoc[T]
|
2016-09-10 15:19:37 +00:00
|
|
|
lex.lineStartImpl = outsideDoc[T]
|
2016-09-11 08:02:10 +00:00
|
|
|
lex.inlineImpl = insideLine[T]
|
2016-09-10 10:38:03 +00:00
|
|
|
|
2016-09-10 08:30:40 +00:00
|
|
|
proc next*(lex: YamlLexer): LexerToken =
|
2016-09-10 10:38:03 +00:00
|
|
|
while not lex.nextImpl(lex, result): discard
|
2016-09-10 08:30:40 +00:00
|
|
|
|
2016-09-10 15:19:37 +00:00
|
|
|
proc setFlow*[T](lex: YamlLexer[T], value: bool) =
|
|
|
|
lex.inFlow = value
|
|
|
|
|
|
|
|
proc endBlockScalar*[T](lex: YamlLexer[T]) =
|
|
|
|
assert lex.nextImpl == blockScalar[T], "Expected blockScalar, got " & lex.nextImpl.repr
|
2016-09-11 08:02:10 +00:00
|
|
|
lex.inlineImpl = insideLine[T]
|
|
|
|
lex.nextImpl = insideLine[T]
|
2016-09-10 15:19:37 +00:00
|
|
|
|
2016-09-10 08:30:40 +00:00
|
|
|
# implementation
|
|
|
|
|
|
|
|
const
|
|
|
|
space = {' ', '\t'}
|
|
|
|
lineEnd = {'\l', '\c', EndOfFile}
|
|
|
|
spaceOrLineEnd = {' ', '\t', '\l', '\c', EndOfFile}
|
|
|
|
digits = {'0'..'9'}
|
|
|
|
flowIndicators = {'[', ']', '{', '}', ','}
|
2016-09-10 11:38:42 +00:00
|
|
|
|
|
|
|
UTF8NextLine = toUTF8(0x85.Rune)
|
|
|
|
UTF8NonBreakingSpace = toUTF8(0xA0.Rune)
|
|
|
|
UTF8LineSeparator = toUTF8(0x2028.Rune)
|
|
|
|
UTF8ParagraphSeparator = toUTF8(0x2029.Rune)
|
2016-09-10 15:19:37 +00:00
|
|
|
|
|
|
|
UnknownIndentation = int.low
|
2016-09-10 08:30:40 +00:00
|
|
|
|
|
|
|
template debug(message: string) {.dirty.} =
|
|
|
|
when defined(yamlDebug):
|
|
|
|
try: styledWriteLine(stdout, fgBlue, message)
|
|
|
|
except IOError: discard
|
|
|
|
|
2016-09-10 10:38:03 +00:00
|
|
|
template lexCR(lex: YamlLexer[BaseLexer]) =
|
2016-09-10 08:30:40 +00:00
|
|
|
lex.source.bufpos = lex.source.handleCR(lex.source.bufpos)
|
2016-09-10 11:38:42 +00:00
|
|
|
lex.c = lex.source.buf[lex.source.bufpos]
|
2016-09-10 08:30:40 +00:00
|
|
|
|
2016-09-10 10:38:03 +00:00
|
|
|
template lexCR(lex: YamlLexer[StringSource]) =
|
2016-09-10 08:30:40 +00:00
|
|
|
lex.source.pos.inc()
|
|
|
|
if lex.source.src[lex.source.pos] == '\l': lex.source.pos.inc()
|
|
|
|
lex.source.lineStart = lex.source.pos
|
2016-09-10 10:38:03 +00:00
|
|
|
lex.source.line.inc()
|
2016-09-10 11:38:42 +00:00
|
|
|
lex.c = lex.source.src[lex.source.pos]
|
2016-09-10 08:30:40 +00:00
|
|
|
|
2016-09-10 10:38:03 +00:00
|
|
|
template lexLF(lex: YamlLexer[BaseLexer]) =
|
2016-09-10 08:30:40 +00:00
|
|
|
lex.source.bufpos = lex.source.handleLF(lex.source.bufpos)
|
2016-09-10 11:38:42 +00:00
|
|
|
lex.c = lex.source.buf[lex.source.bufpos]
|
2016-09-10 08:30:40 +00:00
|
|
|
|
2016-09-10 10:38:03 +00:00
|
|
|
template lexLF(lex: YamlLexer[StringSource]) =
|
2016-09-10 08:30:40 +00:00
|
|
|
lex.source.pos.inc()
|
|
|
|
lex.source.lineStart = lex.source.pos
|
2016-09-10 10:38:03 +00:00
|
|
|
lex.source.line.inc()
|
2016-09-10 11:38:42 +00:00
|
|
|
lex.c = lex.source.src[lex.source.pos]
|
2016-09-10 10:38:03 +00:00
|
|
|
|
|
|
|
template lineNumber(lex: YamlLexer[BaseLexer]): int =
|
|
|
|
lex.source.lineNumber
|
|
|
|
|
|
|
|
template lineNumber(lex: YamlLexer[StringSource]): int =
|
|
|
|
lex.source.line
|
|
|
|
|
|
|
|
template columnNumber(lex: YamlLexer[BaseLexer]): int =
|
|
|
|
lex.source.getColNumber() + 1
|
|
|
|
|
|
|
|
template columnNumber(lex: YamlLexer[StringSource]): int =
|
|
|
|
lex.source.pos - lex.source.lineStart + 1
|
|
|
|
|
|
|
|
template currentLine(lex: YamlLexer[BaseLexer]): string =
|
|
|
|
lex.source.getCurrentLine(true)
|
|
|
|
|
|
|
|
template currentLine(lex: YamlLexer[StringSource]): string =
|
|
|
|
var result = ""
|
|
|
|
var i = lex.source.lineStart
|
|
|
|
while lex.source.src[i] notin lineEnd:
|
|
|
|
result.add(lex.source.src[i])
|
|
|
|
inc(i)
|
2016-09-10 15:19:37 +00:00
|
|
|
result.add("\n" & spaces(lex.columnNumber - 1) & "^\n")
|
2016-09-10 10:38:03 +00:00
|
|
|
result
|
|
|
|
|
|
|
|
proc generateError(lex: YamlLexer, message: string):
|
|
|
|
ref YamlLexerError {.raises: [].} =
|
|
|
|
result = newException(YamlLexerError, message)
|
|
|
|
result.line = lex.lineNumber
|
|
|
|
result.column = lex.columnNumber
|
|
|
|
result.lineContent = lex.currentLine
|
2016-09-10 08:30:40 +00:00
|
|
|
|
|
|
|
proc directiveName(lex: YamlLexer) =
|
|
|
|
while lex.c notin spaceOrLineEnd:
|
|
|
|
lex.buf.add(lex.c)
|
|
|
|
lex.advance()
|
|
|
|
|
|
|
|
proc yamlVersion[T](lex: YamlLexer[T], t: var LexerToken): bool =
|
|
|
|
debug("lex: yamlVersion")
|
2016-09-10 10:38:03 +00:00
|
|
|
while lex.c in space: lex.advance()
|
2016-09-10 08:30:40 +00:00
|
|
|
if lex.c notin digits: raise lex.generateError("Invalid YAML version number")
|
|
|
|
lex.buf.add(lex.c)
|
|
|
|
lex.advance()
|
|
|
|
while lex.c in digits:
|
|
|
|
lex.buf.add(lex.c)
|
|
|
|
lex.advance()
|
|
|
|
if lex.c != '.': raise lex.generateError("Invalid YAML version number")
|
|
|
|
lex.buf.add('.')
|
|
|
|
lex.advance()
|
|
|
|
if lex.c notin digits: raise lex.generateError("Invalid YAML version number")
|
|
|
|
lex.buf.add(lex.c)
|
|
|
|
lex.advance()
|
|
|
|
while lex.c in digits:
|
|
|
|
lex.buf.add(lex.c)
|
|
|
|
lex.advance()
|
|
|
|
if lex.c notin spaceOrLineEnd:
|
|
|
|
raise lex.generateError("Invalid YAML version number")
|
|
|
|
t = ltYamlVersion
|
|
|
|
result = true
|
|
|
|
lex.nextImpl = expectLineEnd[T]
|
|
|
|
|
|
|
|
template nextIsPlainSafe(lex: YamlLexer[BaseLexer], inFlow: bool): bool =
|
|
|
|
case lex.source.buf[lex.source.bufpos + 1]
|
|
|
|
of spaceOrLineEnd: result = false
|
|
|
|
of flowIndicators: result = not inFlow
|
|
|
|
else: result = true
|
|
|
|
|
|
|
|
template nextIsPlainSafe(lex: YamlLexer[StringSource], inFlow: bool): bool =
|
2016-09-10 10:38:03 +00:00
|
|
|
var result: bool
|
2016-09-10 08:30:40 +00:00
|
|
|
case lex.source.src[lex.source.pos + 1]
|
|
|
|
of spaceOrLineEnd: result = false
|
|
|
|
of flowIndicators: result = not inFlow
|
|
|
|
else: result = true
|
2016-09-10 10:38:03 +00:00
|
|
|
result
|
2016-09-10 08:30:40 +00:00
|
|
|
|
|
|
|
proc tagShorthand[T](lex: YamlLexer[T], t: var LexerToken): bool =
|
|
|
|
debug("lex: tagShorthand")
|
|
|
|
while lex.c in space: lex.advance()
|
|
|
|
if lex.c != '!': raise lex.generateError("Tag shorthand must start with a '!'")
|
|
|
|
lex.buf.add(lex.c)
|
|
|
|
lex.advance()
|
|
|
|
|
|
|
|
if lex.c in spaceOrLineEnd: discard
|
|
|
|
else:
|
|
|
|
while lex.c != '!':
|
|
|
|
case lex.c
|
|
|
|
of 'a' .. 'z', 'A' .. 'Z', '0' .. '9', '-':
|
|
|
|
lex.buf.add(lex.c)
|
|
|
|
lex.advance()
|
|
|
|
else: raise lex.generateError("Illegal character in tag shorthand")
|
|
|
|
lex.buf.add(lex.c)
|
|
|
|
lex.advance()
|
|
|
|
if lex.c notin spaceOrLineEnd:
|
|
|
|
raise lex.generateError("Missing space after tag shorthand")
|
2016-09-10 15:19:37 +00:00
|
|
|
t = ltTagShorthand
|
|
|
|
result = true
|
2016-09-10 08:30:40 +00:00
|
|
|
lex.nextImpl = tagUri[T]
|
|
|
|
|
|
|
|
proc tagUri[T](lex: YamlLexer[T], t: var LexerToken): bool =
|
|
|
|
debug("lex: tagUri")
|
|
|
|
while lex.c in space: lex.advance()
|
|
|
|
if lex.c == '!':
|
|
|
|
lex.buf.add(lex.c)
|
2016-09-10 10:38:03 +00:00
|
|
|
lex.advance()
|
2016-09-10 08:30:40 +00:00
|
|
|
while true:
|
|
|
|
case lex.c
|
|
|
|
of spaceOrLineEnd: break
|
|
|
|
of 'a' .. 'z', 'A' .. 'Z', '0' .. '9', '#', ';', '/', '?', ':', '@', '&',
|
|
|
|
'-', '=', '+', '$', ',', '_', '.', '~', '*', '\'', '(', ')':
|
|
|
|
lex.buf.add(lex.c)
|
|
|
|
lex.advance()
|
|
|
|
else: raise lex.generateError("Invalid character in tag uri: " &
|
|
|
|
escape("" & lex.c))
|
2016-09-10 15:19:37 +00:00
|
|
|
t = ltTagUri
|
|
|
|
result = true
|
|
|
|
lex.nextImpl = expectLineEnd[T]
|
2016-09-10 08:30:40 +00:00
|
|
|
|
|
|
|
proc unknownDirParams[T](lex: YamlLexer[T], t: var LexerToken): bool =
|
2016-09-10 15:19:37 +00:00
|
|
|
debug("lex: unknownDirParams")
|
|
|
|
while lex.c in space: lex.advance()
|
|
|
|
while lex.c notin lineEnd + {'#'}:
|
|
|
|
lex.buf.add(lex.c)
|
|
|
|
lex.advance()
|
2016-09-10 08:30:40 +00:00
|
|
|
t = ltUnknownDirectiveParams
|
|
|
|
result = true
|
|
|
|
lex.nextImpl = expectLineEnd[T]
|
|
|
|
|
|
|
|
proc expectLineEnd[T](lex: YamlLexer[T], t: var LexerToken): bool =
|
|
|
|
debug("lex: expectLineEnd")
|
|
|
|
result = false
|
2016-09-10 10:38:03 +00:00
|
|
|
while lex.c in space: lex.advance()
|
2016-09-10 08:30:40 +00:00
|
|
|
while true:
|
|
|
|
case lex.c
|
|
|
|
of '#':
|
|
|
|
lex.advance()
|
|
|
|
while lex.c notin lineEnd: lex.advance()
|
|
|
|
of EndOfFile:
|
|
|
|
lex.nextImpl = streamEnd[T]
|
|
|
|
break
|
|
|
|
of '\l':
|
2016-09-10 10:38:03 +00:00
|
|
|
lex.lexLF()
|
2016-09-10 15:19:37 +00:00
|
|
|
lex.nextImpl = lex.lineStartImpl
|
2016-09-10 08:30:40 +00:00
|
|
|
break
|
|
|
|
of '\c':
|
2016-09-10 10:38:03 +00:00
|
|
|
lex.lexCR()
|
2016-09-10 15:19:37 +00:00
|
|
|
lex.nextImpl = lex.lineStartImpl
|
2016-09-10 08:30:40 +00:00
|
|
|
break
|
|
|
|
else:
|
|
|
|
raise lex.generateError("Unexpected character (expected line end): " &
|
|
|
|
escape("" & lex.c))
|
|
|
|
|
2016-09-10 15:19:37 +00:00
|
|
|
proc possibleDirectivesEnd[T](lex: YamlLexer[T], t: var LexerToken): bool =
|
|
|
|
debug("lex: possibleDirectivesEnd")
|
2016-09-11 08:02:10 +00:00
|
|
|
lex.lineStartImpl = insideDoc[T]
|
2016-09-10 08:30:40 +00:00
|
|
|
lex.advance()
|
|
|
|
if lex.c == '-':
|
|
|
|
lex.advance()
|
|
|
|
if lex.c == '-':
|
|
|
|
lex.advance()
|
|
|
|
if lex.c in spaceOrLineEnd:
|
|
|
|
t = ltDirectivesEnd
|
2016-09-11 08:02:10 +00:00
|
|
|
lex.nextImpl = insideLine[T]
|
2016-09-10 15:19:37 +00:00
|
|
|
return true
|
2016-09-10 08:30:40 +00:00
|
|
|
lex.buf.add('-')
|
|
|
|
lex.buf.add('-')
|
|
|
|
elif lex.c in spaceOrLineEnd:
|
2016-09-10 15:19:37 +00:00
|
|
|
lex.indentation = 0
|
|
|
|
t = ltIndentation
|
|
|
|
lex.nextImpl = afterSeqInd[T]
|
|
|
|
return true
|
2016-09-10 08:30:40 +00:00
|
|
|
lex.buf.add('-')
|
|
|
|
lex.nextImpl = plainScalarPart[T]
|
2016-09-10 15:19:37 +00:00
|
|
|
result = false
|
2016-09-10 08:30:40 +00:00
|
|
|
|
2016-09-10 15:19:37 +00:00
|
|
|
proc afterSeqInd[T](lex: YamlLexer[T], t: var LexerToken): bool =
|
|
|
|
result = true
|
|
|
|
t = ltSeqItemInd
|
|
|
|
if lex.c notin lineEnd: lex.advance()
|
2016-09-11 08:02:10 +00:00
|
|
|
lex.nextImpl = insideLine[T]
|
2016-09-10 15:19:37 +00:00
|
|
|
|
|
|
|
proc possibleDocumentEnd[T](lex: YamlLexer[T], t: var LexerToken): bool =
|
|
|
|
debug("lex: possibleDocumentEnd")
|
2016-09-10 08:30:40 +00:00
|
|
|
lex.advance()
|
|
|
|
if lex.c == '.':
|
|
|
|
lex.advance()
|
|
|
|
if lex.c == '.':
|
|
|
|
lex.advance()
|
|
|
|
if lex.c in spaceOrLineEnd:
|
|
|
|
t = ltDocumentEnd
|
|
|
|
lex.nextImpl = expectLineEnd[T]
|
2016-09-10 15:19:37 +00:00
|
|
|
lex.lineStartImpl = outsideDoc[T]
|
|
|
|
return true
|
2016-09-10 08:30:40 +00:00
|
|
|
lex.buf.add('.')
|
|
|
|
lex.buf.add('.')
|
|
|
|
lex.buf.add('.')
|
|
|
|
lex.nextImpl = plainScalarPart[T]
|
2016-09-10 15:19:37 +00:00
|
|
|
result = false
|
2016-09-10 08:30:40 +00:00
|
|
|
|
|
|
|
proc outsideDoc[T](lex: YamlLexer[T], t: var LexerToken): bool =
|
2016-09-10 15:19:37 +00:00
|
|
|
debug("lex: outsideDoc")
|
2016-09-10 08:30:40 +00:00
|
|
|
case lex.c
|
|
|
|
of '%':
|
|
|
|
lex.advance()
|
|
|
|
lex.directiveName()
|
|
|
|
case lex.buf
|
2016-09-10 10:38:03 +00:00
|
|
|
of "YAML":
|
|
|
|
t = ltYamlDirective
|
2016-09-10 15:19:37 +00:00
|
|
|
lex.buf.setLen(0)
|
2016-09-10 10:38:03 +00:00
|
|
|
lex.nextImpl = yamlVersion[T]
|
|
|
|
of "TAG":
|
2016-09-10 15:19:37 +00:00
|
|
|
lex.buf.setLen(0)
|
2016-09-10 10:38:03 +00:00
|
|
|
t = ltTagDirective
|
|
|
|
lex.nextImpl = tagShorthand[T]
|
|
|
|
else:
|
|
|
|
t = ltUnknownDirective
|
|
|
|
lex.nextImpl = unknownDirParams[T]
|
|
|
|
return true
|
2016-09-10 08:30:40 +00:00
|
|
|
of '-':
|
2016-09-10 15:19:37 +00:00
|
|
|
lex.nextImpl = possibleDirectivesEnd[T]
|
|
|
|
return false
|
2016-09-10 08:30:40 +00:00
|
|
|
of '.':
|
2016-09-10 15:19:37 +00:00
|
|
|
lex.indentation = 0
|
|
|
|
lex.nextImpl = possibleDocumentEnd[T]
|
2016-09-10 08:30:40 +00:00
|
|
|
of spaceOrLineEnd + {'#'}:
|
|
|
|
lex.indentation = 0
|
|
|
|
while lex.c == ' ':
|
|
|
|
lex.indentation.inc()
|
|
|
|
lex.advance()
|
|
|
|
if lex.c in spaceOrLineEnd + {'#'}:
|
|
|
|
lex.nextImpl = expectLineEnd[T]
|
|
|
|
return false
|
2016-09-11 08:02:10 +00:00
|
|
|
lex.nextImpl = insideLine[T]
|
2016-09-10 15:19:37 +00:00
|
|
|
else:
|
|
|
|
lex.indentation = 0
|
2016-09-11 08:02:10 +00:00
|
|
|
lex.nextImpl = insideLine[T]
|
|
|
|
lex.lineStartImpl = insideDoc[T]
|
2016-09-10 08:30:40 +00:00
|
|
|
t = ltIndentation
|
|
|
|
result = true
|
|
|
|
|
2016-09-11 08:02:10 +00:00
|
|
|
proc insideDoc[T](lex: YamlLexer[T], t: var LexerToken): bool =
|
|
|
|
debug("lex: insideDoc")
|
2016-09-10 15:19:37 +00:00
|
|
|
lex.indentation = 0
|
2016-09-10 08:30:40 +00:00
|
|
|
case lex.c
|
|
|
|
of '-':
|
2016-09-10 15:19:37 +00:00
|
|
|
lex.nextImpl = possibleDirectivesEnd[T]
|
|
|
|
return false
|
|
|
|
of '.': lex.nextImpl = possibleDocumentEnd[T]
|
|
|
|
of spaceOrLineEnd:
|
2016-09-10 08:30:40 +00:00
|
|
|
while lex.c == ' ':
|
|
|
|
lex.indentation.inc()
|
|
|
|
lex.advance()
|
2016-09-10 15:19:37 +00:00
|
|
|
if lex.c in spaceOrLineEnd:
|
2016-09-10 08:30:40 +00:00
|
|
|
t = ltEmptyLine
|
2016-09-10 15:19:37 +00:00
|
|
|
lex.nextImpl = expectLineEnd[T]
|
2016-09-10 08:30:40 +00:00
|
|
|
return true
|
2016-09-10 15:19:37 +00:00
|
|
|
else:
|
|
|
|
lex.nextImpl = lex.inlineImpl
|
|
|
|
else: lex.nextImpl = lex.inlineImpl
|
2016-09-10 08:30:40 +00:00
|
|
|
t = ltIndentation
|
|
|
|
result = true
|
|
|
|
|
|
|
|
proc possibleIndicatorChar[T](lex: YamlLexer[T], indicator: LexerToken,
|
|
|
|
t: var LexerToken): bool =
|
|
|
|
if lex.nextIsPlainSafe(false):
|
|
|
|
lex.nextImpl = plainScalarPart[T]
|
|
|
|
result = false
|
|
|
|
else:
|
|
|
|
t = indicator
|
|
|
|
result = true
|
|
|
|
lex.advance()
|
|
|
|
while lex.c in space: lex.advance()
|
|
|
|
if lex.c in lineEnd:
|
|
|
|
lex.nextImpl = expectLineEnd[T]
|
|
|
|
|
|
|
|
proc flowIndicator[T](lex: YamlLexer[T], indicator: LexerToken,
|
2016-09-10 15:33:58 +00:00
|
|
|
t: var LexerToken): bool {.inline.} =
|
2016-09-10 08:30:40 +00:00
|
|
|
t = indicator
|
|
|
|
lex.advance()
|
|
|
|
while lex.c in space: lex.advance()
|
|
|
|
if lex.c in lineEnd:
|
|
|
|
lex.nextImpl = expectLineEnd[T]
|
2016-09-10 15:33:58 +00:00
|
|
|
result = true
|
2016-09-10 08:30:40 +00:00
|
|
|
|
2016-09-10 11:38:42 +00:00
|
|
|
proc addMultiple(s: var string, c: char, num: int) {.raises: [], inline.} =
|
2016-09-10 15:19:37 +00:00
|
|
|
for i in 1..num: s.add(c)
|
2016-09-10 11:38:42 +00:00
|
|
|
|
|
|
|
proc processQuotedWhitespace(lex: YamlLexer, newlines: var int) =
|
|
|
|
block outer:
|
|
|
|
let beforeSpace = lex.buf.len
|
|
|
|
while true:
|
|
|
|
case lex.c
|
|
|
|
of ' ', '\t': lex.buf.add(lex.c)
|
|
|
|
of '\l':
|
|
|
|
lex.lexLF()
|
|
|
|
break
|
|
|
|
of '\c':
|
|
|
|
lex.lexCR()
|
|
|
|
break
|
|
|
|
else: break outer
|
|
|
|
lex.advance()
|
|
|
|
lex.buf.setLen(beforeSpace)
|
|
|
|
while true:
|
|
|
|
case lex.c
|
|
|
|
of ' ', '\t': discard
|
|
|
|
of '\l':
|
|
|
|
lex.lexLF()
|
|
|
|
newlines.inc()
|
|
|
|
continue
|
|
|
|
of '\c':
|
|
|
|
lex.lexCR()
|
|
|
|
newlines.inc()
|
|
|
|
continue
|
|
|
|
else:
|
|
|
|
if newlines == 0: discard
|
|
|
|
elif newlines == 1: lex.buf.add(' ')
|
|
|
|
else: lex.buf.addMultiple('\l', newlines - 1)
|
|
|
|
break
|
|
|
|
lex.advance()
|
|
|
|
|
2016-09-10 10:38:03 +00:00
|
|
|
proc singleQuotedScalar[T](lex: YamlLexer[T]) =
|
2016-09-10 11:38:42 +00:00
|
|
|
debug("lex: singleQuotedScalar")
|
|
|
|
lex.advance()
|
|
|
|
while true:
|
|
|
|
case lex.c
|
|
|
|
of '\'':
|
|
|
|
lex.advance()
|
|
|
|
if lex.c == '\'': lex.buf.add('\'')
|
|
|
|
else: break
|
|
|
|
of EndOfFile: raise lex.generateError("Unfinished single quoted string")
|
|
|
|
of '\l', '\c', '\t', ' ':
|
|
|
|
var newlines = 1
|
|
|
|
lex.processQuotedWhitespace(newlines)
|
|
|
|
continue
|
|
|
|
else: lex.buf.add(lex.c)
|
|
|
|
lex.advance()
|
|
|
|
|
|
|
|
proc unicodeSequence(lex: YamlLexer, length: int) =
|
|
|
|
debug("lex: unicodeSequence")
|
|
|
|
var unicodeChar = 0.int
|
|
|
|
for i in countup(0, length - 1):
|
|
|
|
lex.advance()
|
|
|
|
let digitPosition = length - i - 1
|
|
|
|
case lex.c
|
|
|
|
of EndOFFile, '\l', '\c':
|
|
|
|
raise lex.generateError("Unfinished unicode escape sequence")
|
|
|
|
of '0' .. '9':
|
|
|
|
unicodeChar = unicodechar or (int(lex.c) - 0x30) shl (digitPosition * 4)
|
|
|
|
of 'A' .. 'F':
|
|
|
|
unicodeChar = unicodechar or (int(lex.c) - 0x37) shl (digitPosition * 4)
|
|
|
|
of 'a' .. 'f':
|
|
|
|
unicodeChar = unicodechar or (int(lex.c) - 0x57) shl (digitPosition * 4)
|
|
|
|
else:
|
|
|
|
raise lex.generateError(
|
|
|
|
"Invalid character in unicode escape sequence: " &
|
|
|
|
escape("" & lex.c))
|
|
|
|
lex.buf.add(toUTF8(Rune(unicodeChar)))
|
|
|
|
|
2016-09-10 10:38:03 +00:00
|
|
|
proc doubleQuotedScalar[T](lex: YamlLexer[T]) =
|
2016-09-10 11:38:42 +00:00
|
|
|
debug("lex: doubleQuotedScalar")
|
|
|
|
lex.advance()
|
|
|
|
while true:
|
|
|
|
case lex.c
|
|
|
|
of EndOfFile:
|
|
|
|
raise lex.generateError("Unfinished double quoted string")
|
|
|
|
of '\\':
|
|
|
|
lex.advance()
|
|
|
|
case lex.c
|
|
|
|
of EndOfFile:
|
|
|
|
raise lex.generateError("Unfinished escape sequence")
|
|
|
|
of '0': lex.buf.add('\0')
|
|
|
|
of 'a': lex.buf.add('\x07')
|
|
|
|
of 'b': lex.buf.add('\x08')
|
|
|
|
of '\t', 't': lex.buf.add('\t')
|
|
|
|
of 'n': lex.buf.add('\l')
|
|
|
|
of 'v': lex.buf.add('\v')
|
|
|
|
of 'f': lex.buf.add('\f')
|
|
|
|
of 'r': lex.buf.add('\c')
|
|
|
|
of 'e': lex.buf.add('\e')
|
|
|
|
of ' ': lex.buf.add(' ')
|
|
|
|
of '"': lex.buf.add('"')
|
|
|
|
of '/': lex.buf.add('/')
|
|
|
|
of '\\': lex.buf.add('\\')
|
|
|
|
of 'N': lex.buf.add(UTF8NextLine)
|
|
|
|
of '_': lex.buf.add(UTF8NonBreakingSpace)
|
|
|
|
of 'L': lex.buf.add(UTF8LineSeparator)
|
|
|
|
of 'P': lex.buf.add(UTF8ParagraphSeparator)
|
|
|
|
of 'x': lex.unicodeSequence(2)
|
|
|
|
of 'u': lex.unicodeSequence(4)
|
|
|
|
of 'U': lex.unicodeSequence(8)
|
|
|
|
of '\l', '\c':
|
|
|
|
var newlines = 0
|
|
|
|
lex.processQuotedWhitespace(newlines)
|
|
|
|
continue
|
|
|
|
else: raise lex.generateError("Illegal character in escape sequence")
|
|
|
|
of '"':
|
|
|
|
lex.advance()
|
|
|
|
break
|
|
|
|
of '\l', '\c', '\t', ' ':
|
|
|
|
var newlines = 1
|
|
|
|
lex.processQuotedWhitespace(newlines)
|
|
|
|
continue
|
|
|
|
else: lex.buf.add(lex.c)
|
|
|
|
lex.advance()
|
2016-09-10 10:38:03 +00:00
|
|
|
|
2016-09-11 08:02:10 +00:00
|
|
|
proc insideLine[T](lex: YamlLexer[T], t: var LexerToken): bool =
|
|
|
|
debug("lex: insideLine")
|
2016-09-10 08:30:40 +00:00
|
|
|
case lex.c
|
|
|
|
of ':': result = lex.possibleIndicatorChar(ltMapValInd, t)
|
|
|
|
of '?': result = lex.possibleIndicatorChar(ltMapKeyInd, t)
|
|
|
|
of '-': result = lex.possibleIndicatorChar(ltSeqItemInd, t)
|
|
|
|
of lineEnd + {'#'}:
|
|
|
|
result = false
|
|
|
|
lex.nextImpl = expectLineEnd[T]
|
|
|
|
of '\"':
|
|
|
|
lex.doubleQuotedScalar()
|
|
|
|
t = ltQuotedScalar
|
|
|
|
result = true
|
|
|
|
of '\'':
|
|
|
|
lex.singleQuotedScalar()
|
|
|
|
t = ltQuotedScalar
|
|
|
|
result = true
|
|
|
|
of '>', '|':
|
2016-09-10 15:19:37 +00:00
|
|
|
if lex.inFlow: lex.nextImpl = plainScalarPart[T]
|
|
|
|
else: lex.nextImpl = blockScalarHeader[T]
|
|
|
|
result = false
|
2016-09-10 15:33:58 +00:00
|
|
|
of '{': result = lex.flowIndicator(ltBraceOpen, t)
|
|
|
|
of '}': result = lex.flowIndicator(ltBraceClose, t)
|
|
|
|
of '[': result = lex.flowIndicator(ltBracketOpen, t)
|
|
|
|
of ']': result = lex.flowIndicator(ltBracketClose, t)
|
|
|
|
of ',': result = lex.flowIndicator(ltComma, t)
|
2016-09-10 08:30:40 +00:00
|
|
|
else:
|
|
|
|
lex.nextImpl = plainScalarPart[T]
|
|
|
|
result = false
|
|
|
|
|
|
|
|
proc plainScalarPart[T](lex: YamlLexer[T], t: var LexerToken): bool =
|
2016-09-10 15:19:37 +00:00
|
|
|
debug("lex: plainScalarPart")
|
2016-09-10 08:30:40 +00:00
|
|
|
block outer:
|
|
|
|
while true:
|
|
|
|
lex.buf.add(lex.c)
|
|
|
|
lex.advance()
|
|
|
|
case lex.c
|
|
|
|
of space:
|
|
|
|
let lenBeforeSpace = lex.buf.len()
|
|
|
|
while true:
|
|
|
|
lex.buf.add(lex.c)
|
|
|
|
lex.advance()
|
2016-09-10 10:38:03 +00:00
|
|
|
case lex.c
|
2016-09-10 08:30:40 +00:00
|
|
|
of lineEnd + {'#'}:
|
|
|
|
lex.buf.setLen(lenBeforeSpace)
|
|
|
|
lex.nextImpl = expectLineEnd[T]
|
|
|
|
break outer
|
|
|
|
of ':':
|
2016-09-10 10:38:03 +00:00
|
|
|
if lex.nextIsPlainSafe(lex.inFlow): break
|
2016-09-10 08:30:40 +00:00
|
|
|
else:
|
|
|
|
lex.buf.setLen(lenBeforeSpace)
|
2016-09-11 08:02:10 +00:00
|
|
|
lex.nextImpl = insideLine[T]
|
2016-09-10 08:30:40 +00:00
|
|
|
break outer
|
|
|
|
of flowIndicators:
|
|
|
|
if lex.inFlow:
|
|
|
|
lex.buf.setLen(lenBeforeSpace)
|
2016-09-11 08:02:10 +00:00
|
|
|
lex.nextImpl = insideLine[T]
|
2016-09-10 08:30:40 +00:00
|
|
|
break outer
|
|
|
|
else:
|
|
|
|
lex.buf.add(lex.c)
|
|
|
|
lex.advance()
|
|
|
|
break
|
|
|
|
of space: discard
|
|
|
|
else: break
|
2016-09-10 11:38:42 +00:00
|
|
|
of lineEnd:
|
|
|
|
lex.nextImpl = expectLineEnd[T]
|
|
|
|
break
|
2016-09-10 08:30:40 +00:00
|
|
|
of flowIndicators:
|
|
|
|
if lex.inFlow:
|
2016-09-11 08:02:10 +00:00
|
|
|
lex.nextImpl = insideLine[T]
|
2016-09-10 08:30:40 +00:00
|
|
|
break
|
|
|
|
of ':':
|
2016-09-10 11:38:42 +00:00
|
|
|
if not lex.nextIsPlainSafe(lex.inFlow):
|
2016-09-11 08:02:10 +00:00
|
|
|
lex.nextImpl = insideLine[T]
|
2016-09-10 11:38:42 +00:00
|
|
|
break outer
|
2016-09-10 08:30:40 +00:00
|
|
|
else: discard
|
|
|
|
t = ltScalarPart
|
|
|
|
result = true
|
|
|
|
|
2016-09-10 15:19:37 +00:00
|
|
|
proc blockScalarHeader[T](lex: YamlLexer[T], t: var LexerToken): bool =
|
|
|
|
debug("lex: blockScalarHeader")
|
|
|
|
lex.chomp = ctClip
|
|
|
|
lex.blockScalarIndent = UnknownIndentation
|
|
|
|
lex.folded = lex.c == '>'
|
|
|
|
while true:
|
|
|
|
lex.advance()
|
|
|
|
case lex.c
|
|
|
|
of '+':
|
|
|
|
if lex.chomp != ctClip:
|
|
|
|
raise lex.generateError("Only one chomping indicator is allowed")
|
|
|
|
lex.chomp = ctKeep
|
|
|
|
of '-':
|
|
|
|
if lex.chomp != ctClip:
|
|
|
|
raise lex.generateError("Only one chomping indicator is allowed")
|
|
|
|
lex.chomp = ctStrip
|
|
|
|
of '1'..'9':
|
|
|
|
if lex.blockScalarIndent != UnknownIndentation:
|
|
|
|
raise lex.generateError("Only one indentation indicator is allowed")
|
|
|
|
lex.blockScalarIndent = lex.indentation + ord(lex.c) - ord('\x30')
|
|
|
|
of spaceOrLineEnd: break
|
|
|
|
else:
|
|
|
|
raise lex.generateError(
|
|
|
|
"Illegal character in block scalar header: '" & escape("" & lex.c) &
|
|
|
|
'\'')
|
|
|
|
lex.nextImpl = expectLineEnd[T]
|
|
|
|
lex.inlineImpl = blockScalar[T]
|
|
|
|
t = ltBlockScalarHeader
|
|
|
|
result = true
|
|
|
|
|
|
|
|
proc blockScalar[T](lex: YamlLexer[T], t: var LexerToken): bool =
|
|
|
|
debug("lex: blockScalarLine")
|
2016-09-10 08:30:40 +00:00
|
|
|
result = false
|
2016-09-10 15:19:37 +00:00
|
|
|
if lex.blockScalarIndent == UnknownIndentation:
|
|
|
|
lex.blockScalarIndent = lex.indentation
|
|
|
|
elif lex.c == '#':
|
|
|
|
lex.nextImpl = expectLineEnd[T]
|
|
|
|
return false
|
|
|
|
elif lex.indentation < lex.blockScalarIndent:
|
|
|
|
raise lex.generateError("Too little indentation in block scalar")
|
|
|
|
elif lex.indentation > lex.blockScalarIndent or lex.c == '\t':
|
|
|
|
lex.moreIndented = true
|
|
|
|
lex.buf.addMultiple(' ', lex.indentation - lex.blockScalarIndent)
|
|
|
|
else: lex.moreIndented = false
|
|
|
|
while lex.c notin lineEnd:
|
|
|
|
lex.buf.add(lex.c)
|
|
|
|
lex.advance()
|
|
|
|
t = ltScalarPart
|
|
|
|
result = true
|
|
|
|
lex.nextImpl = expectLineEnd[T]
|
2016-09-10 08:30:40 +00:00
|
|
|
|
|
|
|
proc streamEnd[T](lex: YamlLexer[T], t: var LexerToken): bool =
|
2016-09-10 15:19:37 +00:00
|
|
|
debug("lex: streamEnd")
|
2016-09-10 08:30:40 +00:00
|
|
|
t = ltStreamEnd
|
|
|
|
result = true
|