Text-Tokenize-Indented As part of the Decl language project (the windmill I've been tilting at since 2010), I end up working with text a lot that is structured by indentation. Finally, I think, this module provides a solid underpinning to working with that kind of text, in that it provides as convenient a tokenizer as possible. It's based on L, meaning that it (1) does a lazy tokenization of a list passed into it, and (2) provides a peek and unget so that you can easily chain tokenizers; if a given piece that has already been identified turns out to break into multiple tokens, you simply tokenize it and push the subpieces back into the stream for later retrieval as individual tokens. This allows very nice compartmentalization of the details of parsing, leaving you a lot less to debug when parsing more difficult items. You use it like this: use Text::Tokenize::Indented; my $tok = Text::Tokenize::Indented ({tab => 4}, < 8}, $trailing_iterator) text text text text text EOF (For instance.) This then returns the following token stream: [0, 'text'] [0, 'text'] [3, 'text'] [3, 'text'] [-1] [0, 'text'] (whatever the trailing iterator returns) We might then chain another tokenizer onto this one which would tokenize the individual lines into more meaningful things. Note that blank lines officially have an indentation of -1.