Class: Psych::Pure::Parser

Inherits:

StringScanner

• Object
• StringScanner
• Psych::Pure::Parser

Defined in:

lib/psych/pure.rb

Overview

The parser is responsible for taking a YAML string and converting it into a series of events that can be used by the consumer.

Defined Under Namespace

Classes: Context, Debug

Instance Method Summary

• #initialize(handler) ⇒ Parser constructor

  Initialize a new parser with the given source string.

• #parse(yaml, filename = yaml.respond_to?(:path) ? yaml.path : "<unknown>", comments: false) ⇒ Object

  Top-level parse function that starts the parsing process.

Constructor Details

#initialize(handler) ⇒ Parser

Initialize a new parser with the given source string.

# File 'lib/psych/pure.rb', line 1029

def initialize(handler)
  super("")

  # These are used to track the current state of the parser.
  @filename = nil
  @source = nil

  # The handler is the consumer of the events generated by the parser.
  @handler = handler

  # This functions as a list of temporary lists of events that may be
  # flushed into the handler if current context is matched.
  @events_cache = []
  @events_cache_marks = []
  @events_cache_depth = 0

  # Document start/end state. These are deferred and flushed lazily —
  # the document start is emitted when the first content event arrives,
  # and the document end is emitted when the next document starts.
  @doc_start_pos = nil
  @doc_start_version = nil
  @doc_start_implicit = true
  @doc_end_pos = nil
  @doc_end_implicit = true

  # Each document gets its own set of tags. This is a mapping of tag
  # handles to tag prefixes.
  @tag_directives = nil

  # When a tag property is parsed, it is stored here until it can be
  # flushed into the next event.
  @tag = nil

  # When a tag handle is parsed, it is stored here until the tag prefix
  # is parsed and the full tag can be resolved.
  @tag_handle = nil

  # When an anchor is parsed, it is stored here until it can be flushed
  # into the next event.
  @anchor = nil

  # In a bare document, explicit document starts (---) and ends (...) are
  # disallowed. In that case we need to check for those delimiters.
  @in_bare_document = false
  @check_forbidden = false

  # In a literal or folded scalar, we need to track that state in order to
  # insert the correct plain text prefix.
  @in_scalar = false
  @text_prefix = +""

  # This parser can optionally parse comments and attach them to the
  # resulting tree, if the option is passed.
  @comments = nil

  # The context of the parser at any given time, which is used to decorate
  # error messages to make it easier to find the specific location where
  # they occurred.
  @context = Context.new
end

Instance Method Details

#parse(yaml, filename = yaml.respond_to?(:path) ? yaml.path : "<unknown>", comments: false) ⇒ Object

Top-level parse function that starts the parsing process.

# File 'lib/psych/pure.rb', line 1091

def parse(yaml, filename = yaml.respond_to?(:path) ? yaml.path : "<unknown>", comments: false)
  if yaml.respond_to?(:read)
    yaml = yaml.read
  elsif !yaml.is_a?(String)
    raise TypeError, "Expected an IO or a String, got #{yaml.class}"
  end

  # This parser only supports UTF-8 encoding at the moment. This is
  # largely due to the performance impact of having to convert all of the
  # strings and regular expressions into compatible encodings. We do not
  # attempt to transcode, as the locations would all be off at that point.
  if yaml.encoding != Encoding::UTF_8
    raise ArgumentError, "Expected UTF-8 encoding, got #{yaml.encoding}"
  end

  yaml += "\n" if !yaml.empty? && !yaml.end_with?("\n")

  # Set StringScanner's source (used by skip/match/eos?) and keep a
  # direct reference for raw byte access (getbyte/byteslice) which
  # bypasses StringScanner for performance-critical paths.
  self.string = yaml
  @string = yaml
  @filename = filename
  @source = Source.new(yaml)
  @comments = {} if comments

  # Precompute positions where --- or ... appear at start of a line
  # followed by whitespace or end of string. These are forbidden content
  # positions in bare documents.
  @forbidden_content = {}
  yaml.scan(/^(?:---|\.\.\.)(?=[\s]|\z)/m) { @forbidden_content[$~.begin(0)] = true }
  @has_forbidden_content = !@forbidden_content.empty?

  parse_l_yaml_stream
  @comments = nil if comments
  true
end