In today’s Programming Praxis exercise our goal is to replicate a script Phil wrote to generate chronological and reverse chronological lists of all of his posts. He did it in 24 lines of AWK, so let’s see how Haskell measures up.
import Data.List import Data.List.Split import Text.Printf import Text.Regex.Posix
We need a function to display the name of a month.
toMonth :: Int -> String toMonth m = chunk 3 "JanFebMarAprMayJunJulAugSepOctNovDec" !! (m - 1)
Generating the html for a post is roughly the same as in his version, save for the fact that I removed a bit of duplication in the links.
item :: [[String]] -> String item xs = printf "<tr><td>%s</td><td>%02s %s %s</td><td>%s: %s</td>\ \<td>%s%s<a href=\"http://programmingpraxis.codepad.org/%s\">\ \codepad</a></td></tr>" (g "number") (g "pubday") (toMonth . read $ g "pubmon") (g "pubyear") (link "" (g "title")) (g "blurb") (link "" "exercise") (link ("/" ++ g "soln") "solution") (g "codepad") where g x = maybe "" last $ find ((== x) . head) xs link :: String -> String -> String link = printf "<a href=\"/%s/%02s/%02s/%s%s/\">%s</a>" (g "pubyear") (g "pubmon") (g "pubday") (g "file")
Generating a list of items is pretty self-explanatory: separate the blocks, filter out the posts, parse the properties and generate the necessary html.
items :: String -> [String] items = map (item . map (splitOn "\t") . lines) . filter (=~ "^number\t[1-9][0-9]*$") . splitOn "\n\n"
All that’s left to do is sort the items as required and put them in a table. Like the original implementation, this version requires that the file containing the list is sorted chronologically.
listing :: ([String] -> [String]) -> String -> String listing f xs = "<table cellpadding=\"10\">" ++ concat (f $ items xs) ++ "</table>"
Let’s see if everything works:
main :: IO () main = do x <- readFile "praxis.info" putStrLn $ listing id x putStrLn $ listing reverse x
Yup. And at 15 lines, I think I’ll continue to use Haskell for my text munging needs.