On 08/29/2013 08:49 AM, Tsantilas Christos wrote: > This is similar with what you are doing now. We are using NextToken, > RegexToken and NextQuotedOrToEol, so your suggestion is to add a new > NextEqSeperatedToken() and merge all these methods to NextToken.
Please note that RegexToken() and NextQuotedOrToEol() are there to support legacy syntax. In strict mode, they become NextToken() convenience wrappers. > Implement a NextEqSeparatedToken() looks that required to keep > compatibility with legacy parser... Same here. NextKeyValue() or similar is a good idea, but, in a strict mode, it has to be a wrapper around two or three NextToken() calls. NextKeyValue() should not define tokenization syntax in strict mode. It only validates that the input (tokenized by NextToken()) is what the caller expects. HTH, Alex.
