Skip to main content

Posts

Showing posts from April, 2018

What if parsing was ...? Aspirations to a world that is easier to parse.

Parsing is one of those activities that never goes away. It comes up in a surprisingly diverse range of programmer tasks, from compiler writing, to log analysis, to API communication. It's been studied ad nauseum by computer scientists and enormous numbers of important theoretical results are known about every corner of the discipline. And yet, the tools that we use on a day to day basis as practitioners seem strangely inflexible, weak and brittle. The most widely used parsing tools are regexps. And of course, regexps are great for what they do. If what you need is to recognise a regular language, or perhaps even extract some information from a regular language, then you've got the perfect tool for the job. Not only that, but most languages give you a fast and convenient way to do it. Beyond that however, thinks start getting fairly arcane. There are of course good tools out there, all of which make slightly different assumptions about what you might like to parse and how