Should we maybe use a PEG parser?

[nawarian]

Many of the formats we implement and want to implement have structures that can easily be parsed using Regular Expressions. This approach is very repetitive and, given citation formats have little variation, can lead to lots of duplicate code and poor abstractions.

Using a proper lexer could make the implementation process much faster. One option is to use the PEG format, commonly used for compilers. This way we can implement new parser/creator/templates just by implementing a grammar file.

See https://github.com/hafriedlander/php-peg/.