Functional programming languages are uniquely suitable for writing parsers for grammars, their whole feature set (discriminated unions, pattern matching etc.) is very suited to turning text into ASTs. It's not often that one needs to parse a custom DSL grammar.

That said, I'd say it's a pretty rare occasion to have that need, and other languages have great libraries for accomplishing the same, I'm partial to Chevrotain for Javascript (https://chevrotain.io/docs/), which has a small DSL for describing grammars.

Parser combinators are both less efficient and less featureful than bespoke solutions, which is a problem when you're writing a parser for your custom language, and you want to add extra features, like resilience, or rich error reporting.

To add a bit of snark, functional languages seem to best suited for writing parsers/compilers for other functional programming languages, which is pretty telling of the inward-facing functional culture.

> Parser combinators are both less efficient and less featureful than bespoke solutions, which is a problem when you're writing a parser for your custom language, and you want to add extra features, like resilience, or rich error reporting.

This is untrue. PCs are the bespoke solutions. The article shows how to introduce a feature like state. Parser generators are the technique which just dump fixed-functionality source code on you.