kowey

<!-- guess what this Mediawiki translates into in HTML -->
Mixed markup redistribution <em>p1

p2</em>

kowey

use XML::Simple;
use File::Slurp;

my $foo = read_file "foo.xml";
my $xml = eval { XMLin($foo); };
for my $patch_name (keys %{ $xml->{patch} }) {
  print "saw: $patch_name \n";
}

# $xml structure returned looks like this: 

kowey

-- ----------------------------------------------------------------------
-- Simplify wiki text
-- ----------------------------------------------------------------------

dropEndOfArticle = takeWhile (not . end)
 where
  end (Header _ [ Str "See", Space, Str "also"]) = True
  end (Header _ [ Str "External", Space, Str "links"]) = True
  end (Header _ [ Str "References"]) = True
  end _ = False

kowey

tokenise :: CleanHillEntry -> String -> String
tokenise hill str = unwords (render toks)
 where
  toks = case parse parseChunks "" str of
            Left e   -> error $ "bug (chunking should always work): " ++ show e
            Right cs -> cs
  --
  render [] = []
  render (Word x : ts)  = x : render ts
  render (Desig x : ts) = x : render ts

kowey

diff -ru mediawiki-0.2.4/MediaWiki/Util/Fetch.hs /home/eykk10/tmp/haskell/mediawiki-0.2.4-kowey/MediaWiki/Util/Fetch.hs
--- mediawiki-0.2.4/MediaWiki/Util/Fetch.hs	2009-01-04 04:15:55.000000000 +0000
+++ /home/eykk10/tmp/haskell/mediawiki-0.2.4-kowey/MediaWiki/Util/Fetch.hs	2009-08-25 18:40:09.000000000 +0100
@@ -49,7 +49,7 @@
       Just ur -> return (defaultGETRequest ur)

     -- don't like doing this, but HTTP is awfully chatty re: cookie handling..

   let nullHandler _ = return ()

kowey

go :: LlrType -> FilePath -> IO ()
go unitType d =
 do enFiles <- filter (\f -> takeExtension f == ".en") `fmap` getDirectoryContents d
    let bnames = map takeBaseName enFiles
    hPutStrLn stderr "Reading input files... (step 1)"
    pairs <- case unitType of
              SentenceLevel -> concat `fmap` mapM (readSentencePairs d) bnames
              DocumentLevel -> mapM (readDocPair d) bnames
    rnf pairs `seq` hPutStrLn stderr "Computing frequencies... (step 2)"
    let freqs = frequencies pairs

kowey

c381 6c 76 61 72 6f 20 64 65                           Álvaro de
20 4d 65 6e 64 61 c3b1 61 20                            Mendaña 
64 65 20 4e 65 69 72 61 20                             de Neira 
6f 72 20 4e 65 79 72 61 20                             or Neyra 
28 31 35 34 32 20 2d 20 4f                             (1542 - O
63 74 6f 62 65 72 20 31 35                             ctober 15
39 35 29 20 77 61 73 20 61                             95) was a
20 53 70 61 6e 69 73 68 20                              Spanish 
6e 61 76 69 67 61 74 6f 72                             navigator
2e 20 42 6f 72 6e 20 69 6e                             . Born in

kowey

#!/bin/bash

if [ $# -ne 5 ]; then
  echo "Usage: $0 dir-in dir-out ext-in ext-out cmd" >&2
  echo "Eg: $0 foo-d bar-d .xml .txt foo2text" >&2
  exit 1
fi

DIR_IN=$1
DIR_OUT=$2

kowey

# Attribution lines
color quoted green black # quoted text
color body brightred   black "(^[> ]*)-([^-])" # what I really want to is match such lines
                                               # but only colour the -/+ symbol
                                               # was hoping mutt by convention ignores groups 1 & 3
                                               # but I can live with the line being red
color body brightgreen black "(^[> ]*)[+]" # brightgreen seems really garish for the whole
                                           # line, just the plus please

kowey

color header white red    "^X-GTD:" # for GTD next-actions
set display_filter="list-actions"

# use the /tmp/gtd folder to automatically retrieve messages with an action/WF
macro index \Cr "!email-get-msgids mairix^Mc/tmp/gtd^M"
folder-hook /tmp/gtd    'set index_format="%Z %{%b %d} %-25.25L  %-35.35i %s"'