scannerc.go

	if !hasTag {
		yaml_parser_set_scanner_tag_error(parser, directive,
			start_mark, "did not find expected tag URI")
		return false
	}
	*uri = s
	return true
}

// Decode an URI-escape sequence corresponding to a single UTF-8 character.
func yaml_parser_scan_uri_escapes(parser *yaml_parser_t, directive bool, start_mark yaml_mark_t, s *[]byte) bool {

	// Decode the required number of characters.
	w := 1024
	for w > 0 {
		// Check for a URI-escaped octet.
		if parser.unread < 3 && !yaml_parser_update_buffer(parser, 3) {
			return false
		}

		if !(parser.buffer[parser.buffer_pos] == '%' &&
			is_hex(parser.buffer, parser.buffer_pos+1) &&
			is_hex(parser.buffer, parser.buffer_pos+2)) {
			return yaml_parser_set_scanner_tag_error(parser, directive,
				start_mark, "did not find URI escaped octet")
		}

		// Get the octet.
		octet := byte((as_hex(parser.buffer, parser.buffer_pos+1) << 4) + as_hex(parser.buffer, parser.buffer_pos+2))

		// If it is the leading octet, determine the length of the UTF-8 sequence.
		if w == 1024 {
			w = width(octet)
			if w == 0 {
				return yaml_parser_set_scanner_tag_error(parser, directive,
					start_mark, "found an incorrect leading UTF-8 octet")
			}
		} else {
			// Check if the trailing octet is correct.
			if octet&0xC0 != 0x80 {
				return yaml_parser_set_scanner_tag_error(parser, directive,
					start_mark, "found an incorrect trailing UTF-8 octet")
			}
		}

		// Copy the octet and move the pointers.
		*s = append(*s, octet)
		skip(parser)
		skip(parser)
		skip(parser)
		w--
	}
	return true
}

// Scan a block scalar.
func yaml_parser_scan_block_scalar(parser *yaml_parser_t, token *yaml_token_t, literal bool) bool {
	// Eat the indicator '|' or '>'.
	start_mark := parser.mark
	skip(parser)

	// Scan the additional block scalar indicators.
	if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) {
		return false
	}

	// Check for a chomping indicator.
	var chomping, increment int
	if parser.buffer[parser.buffer_pos] == '+' || parser.buffer[parser.buffer_pos] == '-' {
		// Set the chomping method and eat the indicator.
		if parser.buffer[parser.buffer_pos] == '+' {
			chomping = +1
		} else {
			chomping = -1
		}
		skip(parser)

		// Check for an indentation indicator.
		if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) {
			return false
		}
		if is_digit(parser.buffer, parser.buffer_pos) {
			// Check that the indentation is greater than 0.
			if parser.buffer[parser.buffer_pos] == '0' {
				yaml_parser_set_scanner_error(parser, "while scanning a block scalar",
					start_mark, "found an indentation indicator equal to 0")
				return false
			}

			// Get the indentation level and eat the indicator.
			increment = as_digit(parser.buffer, parser.buffer_pos)
			skip(parser)
		}

	} else if is_digit(parser.buffer, parser.buffer_pos) {
		// Do the same as above, but in the opposite order.

		if parser.buffer[parser.buffer_pos] == '0' {
			yaml_parser_set_scanner_error(parser, "while scanning a block scalar",
				start_mark, "found an indentation indicator equal to 0")
			return false
		}
		increment = as_digit(parser.buffer, parser.buffer_pos)
		skip(parser)

		if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) {
			return false
		}
		if parser.buffer[parser.buffer_pos] == '+' || parser.buffer[parser.buffer_pos] == '-' {
			if parser.buffer[parser.buffer_pos] == '+' {
				chomping = +1
			} else {
				chomping = -1
			}
			skip(parser)
		}
	}

	// Eat whitespaces and comments to the end of the line.
	if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) {
		return false
	}
	for is_blank(parser.buffer, parser.buffer_pos) {
		skip(parser)
		if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) {
			return false
		}
	}
	if parser.buffer[parser.buffer_pos] == '#' {
		for !is_breakz(parser.buffer, parser.buffer_pos) {
			skip(parser)
			if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) {
				return false
			}
		}
	}

	// Check if we are at the end of the line.
	if !is_breakz(parser.buffer, parser.buffer_pos) {
		yaml_parser_set_scanner_error(parser, "while scanning a block scalar",
			start_mark, "did not find expected comment or line break")
		return false
	}

	// Eat a line break.
	if is_break(parser.buffer, parser.buffer_pos) {
		if parser.unread < 2 && !yaml_parser_update_buffer(parser, 2) {
			return false
		}
		skip_line(parser)
	}

	end_mark := parser.mark

	// Set the indentation level if it was specified.
	var indent int
	if increment > 0 {
		if parser.indent >= 0 {
			indent = parser.indent + increment
		} else {
			indent = increment
		}
	}

	// Scan the leading line breaks and determine the indentation level if needed.
	var s, leading_break, trailing_breaks []byte
	if !yaml_parser_scan_block_scalar_breaks(parser, &indent, &trailing_breaks, start_mark, &end_mark) {
		return false
	}

	// Scan the block scalar content.
	if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) {
		return false
	}
	var leading_blank, trailing_blank bool
	for parser.mark.column == indent && !is_z(parser.buffer, parser.buffer_pos) {
		// We are at the beginning of a non-empty line.

		// Is it a trailing whitespace?
		trailing_blank = is_blank(parser.buffer, parser.buffer_pos)

		// Check if we need to fold the leading line break.
		if !literal && !leading_blank && !trailing_blank && len(leading_break) > 0 && leading_break[0] == '\n' {
			// Do we need to join the lines by space?
			if len(trailing_breaks) == 0 {
				s = append(s, ' ')
			}
		} else {
			s = append(s, leading_break...)
		}
		leading_break = leading_break[:0]

		// Append the remaining line breaks.
		s = append(s, trailing_breaks...)
		trailing_breaks = trailing_breaks[:0]

		// Is it a leading whitespace?
		leading_blank = is_blank(parser.buffer, parser.buffer_pos)

		// Consume the current line.
		for !is_breakz(parser.buffer, parser.buffer_pos) {
			s = read(parser, s)
			if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) {
				return false
			}
		}

		// Consume the line break.
		if parser.unread < 2 && !yaml_parser_update_buffer(parser, 2) {
			return false
		}

		leading_break = read_line(parser, leading_break)

		// Eat the following indentation spaces and line breaks.
		if !yaml_parser_scan_block_scalar_breaks(parser, &indent, &trailing_breaks, start_mark, &end_mark) {
			return false
		}
	}

	// Chomp the tail.
	if chomping != -1 {
		s = append(s, leading_break...)
	}
	if chomping == 1 {
		s = append(s, trailing_breaks...)
	}

	// Create a token.
	*token = yaml_token_t{
		typ:        yaml_SCALAR_TOKEN,
		start_mark: start_mark,
		end_mark:   end_mark,
		value:      s,
		style:      yaml_LITERAL_SCALAR_STYLE,
	}
	if !literal {
		token.style = yaml_FOLDED_SCALAR_STYLE
	}
	return true
}

// Scan indentation spaces and line breaks for a block scalar.  Determine the
// indentation level if needed.
func yaml_parser_scan_block_scalar_breaks(parser *yaml_parser_t, indent *int, breaks *[]byte, start_mark yaml_mark_t, end_mark *yaml_mark_t) bool {
	*end_mark = parser.mark

	// Eat the indentation spaces and line breaks.
	max_indent := 0
	for {
		// Eat the indentation spaces.
		if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) {
			return false
		}
		for (*indent == 0 || parser.mark.column < *indent) && is_space(parser.buffer, parser.buffer_pos) {
			skip(parser)
			if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) {
				return false
			}
		}
		if parser.mark.column > max_indent {
			max_indent = parser.mark.column
		}

		// Check for a tab character messing the indentation.
		if (*indent == 0 || parser.mark.column < *indent) && is_tab(parser.buffer, parser.buffer_pos) {
			return yaml_parser_set_scanner_error(parser, "while scanning a block scalar",
				start_mark, "found a tab character where an indentation space is expected")
		}

		// Have we found a non-empty line?
		if !is_break(parser.buffer, parser.buffer_pos) {
			break
		}

		// Consume the line break.
		if parser.unread < 2 && !yaml_parser_update_buffer(parser, 2) {
			return false
		}
		// [Go] Should really be returning breaks instead.
		*breaks = read_line(parser, *breaks)
		*end_mark = parser.mark
	}

	// Determine the indentation level if needed.
	if *indent == 0 {
		*indent = max_indent
		if *indent < parser.indent+1 {
			*indent = parser.indent + 1
		}
		if *indent < 1 {
			*indent = 1
		}
	}
	return true
}

// Scan a quoted scalar.
func yaml_parser_scan_flow_scalar(parser *yaml_parser_t, token *yaml_token_t, single bool) bool {
	// Eat the left quote.
	start_mark := parser.mark
	skip(parser)

	// Consume the content of the quoted scalar.
	var s, leading_break, trailing_breaks, whitespaces []byte
	for {
		// Check that there are no document indicators at the beginning of the line.
		if parser.unread < 4 && !yaml_parser_update_buffer(parser, 4) {
			return false
		}

		if parser.mark.column == 0 &&
			((parser.buffer[parser.buffer_pos+0] == '-' &&
				parser.buffer[parser.buffer_pos+1] == '-' &&
				parser.buffer[parser.buffer_pos+2] == '-') ||
				(parser.buffer[parser.buffer_pos+0] == '.' &&
					parser.buffer[parser.buffer_pos+1] == '.' &&
					parser.buffer[parser.buffer_pos+2] == '.')) &&
			is_blankz(parser.buffer, parser.buffer_pos+3) {
			yaml_parser_set_scanner_error(parser, "while scanning a quoted scalar",
				start_mark, "found unexpected document indicator")
			return false
		}

		// Check for EOF.
		if is_z(parser.buffer, parser.buffer_pos) {
			yaml_parser_set_scanner_error(parser, "while scanning a quoted scalar",
				start_mark, "found unexpected end of stream")
			return false
		}

		// Consume non-blank characters.
		leading_blanks := false
		for !is_blankz(parser.buffer, parser.buffer_pos) {
			if single && parser.buffer[parser.buffer_pos] == '\'' && parser.buffer[parser.buffer_pos+1] == '\'' {
				// Is is an escaped single quote.
				s = append(s, '\'')
				skip(parser)
				skip(parser)

			} else if single && parser.buffer[parser.buffer_pos] == '\'' {
				// It is a right single quote.
				break
			} else if !single && parser.buffer[parser.buffer_pos] == '"' {
				// It is a right double quote.
				break

			} else if !single && parser.buffer[parser.buffer_pos] == '\\' && is_break(parser.buffer, parser.buffer_pos+1) {
				// It is an escaped line break.
				if parser.unread < 3 && !yaml_parser_update_buffer(parser, 3) {
					return false
				}
				skip(parser)
				skip_line(parser)
				leading_blanks = true
				break

			} else if !single && parser.buffer[parser.buffer_pos] == '\\' {
				// It is an escape sequence.
				code_length := 0

				// Check the escape character.
				switch parser.buffer[parser.buffer_pos+1] {
				case '0':
					s = append(s, 0)
				case 'a':
					s = append(s, '\x07')
				case 'b':
					s = append(s, '\x08')
				case 't', '\t':
					s = append(s, '\x09')
				case 'n':
					s = append(s, '\x0A')
				case 'v':
					s = append(s, '\x0B')
				case 'f':
					s = append(s, '\x0C')
				case 'r':
					s = append(s, '\x0D')
				case 'e':
					s = append(s, '\x1B')
				case ' ':
					s = append(s, '\x20')
				case '"':
					s = append(s, '"')
				case '\'':
					s = append(s, '\'')
				case '\\':
					s = append(s, '\\')
				case 'N': // NEL (#x85)
					s = append(s, '\xC2')
					s = append(s, '\x85')
				case '_': // #xA0
					s = append(s, '\xC2')
					s = append(s, '\xA0')
				case 'L': // LS (#x2028)
					s = append(s, '\xE2')
					s = append(s, '\x80')
					s = append(s, '\xA8')
				case 'P': // PS (#x2029)
					s = append(s, '\xE2')
					s = append(s, '\x80')
					s = append(s, '\xA9')
				case 'x':
					code_length = 2
				case 'u':
					code_length = 4
				case 'U':
					code_length = 8
				default:
					yaml_parser_set_scanner_error(parser, "while parsing a quoted scalar",
						start_mark, "found unknown escape character")
					return false
				}

				skip(parser)
				skip(parser)

				// Consume an arbitrary escape code.
				if code_length > 0 {
					var value int

					// Scan the character value.
					if parser.unread < code_length && !yaml_parser_update_buffer(parser, code_length) {
						return false
					}
					for k := 0; k < code_length; k++ {
						if !is_hex(parser.buffer, parser.buffer_pos+k) {
							yaml_parser_set_scanner_error(parser, "while parsing a quoted scalar",
								start_mark, "did not find expected hexdecimal number")
							return false
						}
						value = (value << 4) + as_hex(parser.buffer, parser.buffer_pos+k)
					}

					// Check the value and write the character.
					if (value >= 0xD800 && value <= 0xDFFF) || value > 0x10FFFF {
						yaml_parser_set_scanner_error(parser, "while parsing a quoted scalar",
							start_mark, "found invalid Unicode character escape code")
						return false
					}
					if value <= 0x7F {
						s = append(s, byte(value))
					} else if value <= 0x7FF {
						s = append(s, byte(0xC0+(value>>6)))
						s = append(s, byte(0x80+(value&0x3F)))
					} else if value <= 0xFFFF {
						s = append(s, byte(0xE0+(value>>12)))
						s = append(s, byte(0x80+((value>>6)&0x3F)))
						s = append(s, byte(0x80+(value&0x3F)))
					} else {
						s = append(s, byte(0xF0+(value>>18)))
						s = append(s, byte(0x80+((value>>12)&0x3F)))
						s = append(s, byte(0x80+((value>>6)&0x3F)))
						s = append(s, byte(0x80+(value&0x3F)))
					}

					// Advance the pointer.
					for k := 0; k < code_length; k++ {
						skip(parser)
					}
				}
			} else {
				// It is a non-escaped non-blank character.
				s = read(parser, s)
			}
			if parser.unread < 2 && !yaml_parser_update_buffer(parser, 2) {
				return false
			}
		}

		if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) {
			return false
		}

		// Check if we are at the end of the scalar.
		if single {
			if parser.buffer[parser.buffer_pos] == '\'' {
				break
			}
		} else {
			if parser.buffer[parser.buffer_pos] == '"' {
				break
			}
		}

		// Consume blank characters.
		for is_blank(parser.buffer, parser.buffer_pos) || is_break(parser.buffer, parser.buffer_pos) {
			if is_blank(parser.buffer, parser.buffer_pos) {
				// Consume a space or a tab character.
				if !leading_blanks {
					whitespaces = read(parser, whitespaces)
				} else {
					skip(parser)
				}
			} else {
				if parser.unread < 2 && !yaml_parser_update_buffer(parser, 2) {
					return false
				}

				// Check if it is a first line break.
				if !leading_blanks {
					whitespaces = whitespaces[:0]
					leading_break = read_line(parser, leading_break)
					leading_blanks = true
				} else {
					trailing_breaks = read_line(parser, trailing_breaks)
				}
			}
			if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) {
				return false
			}
		}

		// Join the whitespaces or fold line breaks.
		if leading_blanks {
			// Do we need to fold line breaks?
			if len(leading_break) > 0 && leading_break[0] == '\n' {
				if len(trailing_breaks) == 0 {
					s = append(s, ' ')
				} else {
					s = append(s, trailing_breaks...)
				}
			} else {
				s = append(s, leading_break...)
				s = append(s, trailing_breaks...)
			}
			trailing_breaks = trailing_breaks[:0]
			leading_break = leading_break[:0]
		} else {
			s = append(s, whitespaces...)
			whitespaces = whitespaces[:0]
		}
	}

	// Eat the right quote.
	skip(parser)
	end_mark := parser.mark

	// Create a token.
	*token = yaml_token_t{
		typ:        yaml_SCALAR_TOKEN,
		start_mark: start_mark,
		end_mark:   end_mark,
		value:      s,
		style:      yaml_SINGLE_QUOTED_SCALAR_STYLE,
	}
	if !single {
		token.style = yaml_DOUBLE_QUOTED_SCALAR_STYLE
	}
	return true
}

// Scan a plain scalar.
func yaml_parser_scan_plain_scalar(parser *yaml_parser_t, token *yaml_token_t) bool {

	var s, leading_break, trailing_breaks, whitespaces []byte
	var leading_blanks bool
	var indent = parser.indent + 1

	start_mark := parser.mark
	end_mark := parser.mark

	// Consume the content of the plain scalar.
	for {
		// Check for a document indicator.
		if parser.unread < 4 && !yaml_parser_update_buffer(parser, 4) {
			return false
		}
		if parser.mark.column == 0 &&
			((parser.buffer[parser.buffer_pos+0] == '-' &&
				parser.buffer[parser.buffer_pos+1] == '-' &&
				parser.buffer[parser.buffer_pos+2] == '-') ||
				(parser.buffer[parser.buffer_pos+0] == '.' &&
					parser.buffer[parser.buffer_pos+1] == '.' &&
					parser.buffer[parser.buffer_pos+2] == '.')) &&
			is_blankz(parser.buffer, parser.buffer_pos+3) {
			break
		}

		// Check for a comment.
		if parser.buffer[parser.buffer_pos] == '#' {
			break
		}

		// Consume non-blank characters.
		for !is_blankz(parser.buffer, parser.buffer_pos) {

			// Check for indicators that may end a plain scalar.
			if (parser.buffer[parser.buffer_pos] == ':' && is_blankz(parser.buffer, parser.buffer_pos+1)) ||
				(parser.flow_level > 0 &&
					(parser.buffer[parser.buffer_pos] == ',' ||
						parser.buffer[parser.buffer_pos] == '?' || parser.buffer[parser.buffer_pos] == '[' ||
						parser.buffer[parser.buffer_pos] == ']' || parser.buffer[parser.buffer_pos] == '{' ||
						parser.buffer[parser.buffer_pos] == '}')) {
				break
			}

			// Check if we need to join whitespaces and breaks.
			if leading_blanks || len(whitespaces) > 0 {
				if leading_blanks {
					// Do we need to fold line breaks?
					if leading_break[0] == '\n' {
						if len(trailing_breaks) == 0 {
							s = append(s, ' ')
						} else {
							s = append(s, trailing_breaks...)
						}
					} else {
						s = append(s, leading_break...)
						s = append(s, trailing_breaks...)
					}
					trailing_breaks = trailing_breaks[:0]
					leading_break = leading_break[:0]
					leading_blanks = false
				} else {
					s = append(s, whitespaces...)
					whitespaces = whitespaces[:0]
				}
			}

			// Copy the character.
			s = read(parser, s)

			end_mark = parser.mark
			if parser.unread < 2 && !yaml_parser_update_buffer(parser, 2) {
				return false
			}
		}

		// Is it the end?
		if !(is_blank(parser.buffer, parser.buffer_pos) || is_break(parser.buffer, parser.buffer_pos)) {
			break
		}

		// Consume blank characters.
		if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) {
			return false
		}

		for is_blank(parser.buffer, parser.buffer_pos) || is_break(parser.buffer, parser.buffer_pos) {
			if is_blank(parser.buffer, parser.buffer_pos) {

				// Check for tab characters that abuse indentation.
				if leading_blanks && parser.mark.column < indent && is_tab(parser.buffer, parser.buffer_pos) {
					yaml_parser_set_scanner_error(parser, "while scanning a plain scalar",
						start_mark, "found a tab character that violates indentation")
					return false
				}

				// Consume a space or a tab character.
				if !leading_blanks {
					whitespaces = read(parser, whitespaces)
				} else {
					skip(parser)
				}
			} else {
				if parser.unread < 2 && !yaml_parser_update_buffer(parser, 2) {
					return false
				}

				// Check if it is a first line break.
				if !leading_blanks {
					whitespaces = whitespaces[:0]
					leading_break = read_line(parser, leading_break)
					leading_blanks = true
				} else {
					trailing_breaks = read_line(parser, trailing_breaks)
				}
			}
			if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) {
				return false
			}
		}

		// Check indentation level.
		if parser.flow_level == 0 && parser.mark.column < indent {
			break
		}
	}

	// Create a token.
	*token = yaml_token_t{
		typ:        yaml_SCALAR_TOKEN,
		start_mark: start_mark,
		end_mark:   end_mark,
		value:      s,
		style:      yaml_PLAIN_SCALAR_STYLE,
	}

	// Note that we change the 'simple_key_allowed' flag.
	if leading_blanks {
		parser.simple_key_allowed = true
	}
	return true
}