sxreader - a reader for symbolic expressions

This packages provides a reader to transform the textual representation of an symbolic expression into one or more internal sx.Objects.

sxreader.Reader uses a value that implements the io.Reader interface to read the textual representation.

Currently, the reader supports these representations:

• ( ... ) is transformed into a sx.Pair. Before the last element, a dot . is allowed to signal an improper list. A dot outside a list is not allowed.
• " ... " is transformed into a sx.String. Escape sequences, like \", \n, \x2a, \u2a2a, or \U2a2a2a, are allowed to specify special Unicode code points.
• A sequence of digits 0 ... 9, optional starting with a plus + or a minus - character is transformed into a sx.Number. Currently, sx.Int64 is the only type supported. This will change.
• 'OBJ is transformed into (quote OBJ), `OBJ into (quasiquote OBJ), ,OBJ into (unquote OBJ), and ,@OBJ into (unquote-splicing OBJ). Other read macros are not supported.
• ; ... is ignored until the end of the current line. Therefore, this works as a comment to the human reader.
• A printable sequence of other Unicode code points is transformed into a sx.Symbol.

Textual representations starting with # must not be used as a sx.Symbol, because such representation may encode other sx.Objects in the future.

When creating a sxreader.Reader, some options can be specified to customize the behaviour of the reader:

• WithListLimit(uint) specifies a maximum list length. It should be used, if the reader reads from external sources to mitigate DOS attacks. If not specified, the value of sxreader.DefaultListLimit is used.
• WithNestingLimit(uint) specifies the maximum nesting of lists, for the same reasons. Since many functions works on nested lists recursively, you should use a value that does not exceed the available stack space. If not specified, the value of sxreader.DefaultNestingLimit is used.