Lentil - frugal issue tracker (v. 1.2)

Show and Tell
#1

I am happy to announce the release of lentil 1.2!

What is lentil?

lentil is a minimum effort issue tracker; if you litter your code with:

-- FIXME: does outPop work with b/w terminals? [ansi]

lentil can help you listing, sorting, exporting, tagging slicing and dicing those todos.
To see it in action just type:

cabal new-install lentil     # binaries available too (see homepage)
cd ~/your-project-folder/
lentil .

Any repo will do, not only Haskell ones!
For further information & manual, check project homepage.

What’s new in 1.2?

  • Switched to megaparsec, speeding things up by 35%;
  • Added custom flagwords;
  • Added a plethora of new languages support (Erlang, OCaml, Rust, SML, Scala, …);
  • Added basic support for org-mode files (TODO and list-like items) and markdown files;
  • Added more export options (XML);
  • Added view to lists tags by popularity.

I want to thank everyone who contributed: your patches/reports/suggestions helped the tremendously!

10 Likes
#2

Wohoo, this is very useful. I didn’t realise I have that many TODOs in the codebase. Thanks!

1 Like
#3

Cool, nice program! :grin:

… but can you please add support for PureScript (http://www.purescript.org)!? :wink:

1 Like
#4

Sure! I just uploaded lentil-1.2.3.0, it auto-scans for .purs file!

#5

Awesome, thanks! Added to my tool-belt :grin:

1 Like
#6

Since this is not on github/gitlab/etc., is there a way to see a list of bugs/feature requests? Otherwise I’ll ask via email, but nowadays seems somewhat opaque.

Thanks again!

1 Like
#7

True to dogfooding, I write every reported Issue/request in a TODO and run lentil periodically to update this list.

Simon Michael suggested a web-like interface to display them in a WWW friendly manner. It is a good idea and I have some handwritten notes on how to implement it (tentative name: chickpea), but never got around to make it!

#8

Ah, I see, didn’t think to check.

Anyway, here’s my suggestions:

  • a markdown output would be useful, as then it can be transformed via pandoc into whatever; I’d use this to put up an HTML up somewhere; I don’t need it to be serving (in real time), just to be able to render a HTML, and the markdown→pandoc route seems best
  • scanner for yesod templates (hamlet/julius/casius), which are a bit difficult since they are html/js/css + embedded Haskell

Both are very “would scratch my itch” requests, of course :slight_smile:

1 Like
#9

Thanks, added both to the issue-list!

For the first request: I use markdown very little. Say you have:

doc/dev/issues.txt
  11  check --output pretty, there are trailing white spaces [bug] [u:1]

what would be a reasonable .md output?

#10

Maybe it would be easier to provide some JSON output. JSON is easily consumed by most tools, and writing a small script to convert JSON into pandoc’s internal format (and from there to HTML/Markdown/Org-mode/…) is reasonably simple.

1 Like
#11

I’ve delayed responding to this because, I realised, not what to generate as markdown is the question, but is is rather how should a web report look like. Backtracking from that to markdown is rather easy.

And what should the HTML look like I’m not so sure about, I realised. Surely all the tags in the file list would link to a summary section that is basically what -f tagpop does today, but beyond that, not sure.

1 Like