X-Git-Url: http://gitweb.michael.orlitzky.com/?p=dead%2Flwn-epub.git;a=blobdiff_plain;f=src%2FMain.hs;h=2700038032651efec82cf886161abdac603a6633;hp=f0e81f76478bab18ec4bf625a474c036ec29f467;hb=e9151ade502fa997be41b6e37fb9841cacd0f528;hpb=5b4118c3497d02fc7b533f6b3f5db3b13969f99a diff --git a/src/Main.hs b/src/Main.hs index f0e81f7..2700038 100644 --- a/src/Main.hs +++ b/src/Main.hs @@ -2,8 +2,9 @@ module Main where +import Control.Concurrent (threadDelay) import Control.Monad (when) -import Data.Maybe (fromJust) +import Data.Maybe (fromJust, isJust) import Prelude hiding (readFile) import System.Directory (doesFileExist) import System.IO ( @@ -19,8 +20,17 @@ import Test.HUnit (Assertion, assertEqual) import Test.Framework (Test, testGroup) import Test.Framework.Providers.HUnit (testCase) import Text.Regex.Posix ((=~)) -import Text.XML.HXT.Core hiding (when) - +import Text.XML.HXT.Core ( + IOStateArrow, + SysConfigList, + XmlTree, + no, + readString, + withParseHTML, + withValidate, + withWarnings, + yes + ) import CommandLine (show_help) import Configuration (Cfg(..), get_cfg, use_account) import LWN.HTTP (get_page, log_in, make_cookie_jar) @@ -38,6 +48,15 @@ my_read_opts = [ withValidate no, my_read :: String -> IOStateArrow s b XmlTree my_read = readString my_read_opts + +-- |A wrapper around threadDelay which takes seconds instead of +-- microseconds as its argument. +thread_sleep :: Int -> IO () +thread_sleep seconds = do + let microseconds = seconds * (10 ^ (6 :: Int)) + threadDelay microseconds + + -- | Try to parse the given article using HXT. We try a few different -- methods; if none of them work, we return 'Nothing'. get_xml_from_article :: Cfg -> IO (Maybe (IOStateArrow s b XmlTree)) @@ -56,23 +75,31 @@ get_xml_from_article cfg = do li_result <- log_in cj (fromJust $ username cfg) (fromJust $ password cfg) - when (not li_result) $ do - hPutStrLn stderr "Failed to log in." + + -- Without this, the cookie file is empty during + -- get_page. Whaaat? + thread_sleep 1 + + when (isJust li_result) $ do + let msg = "Failed to log in. " ++ (fromJust li_result) + hPutStrLn stderr msg html <- get_page (Just cj) my_article - print $ fromJust $ html - return $ - case html of - Nothing -> Nothing - Just h -> Just $ my_read h + + case html of + Left err -> do + let msg = "Failed to retrieve page. " ++ err + hPutStrLn stderr msg + return Nothing + Right h -> return $ Just $ my_read h else do html <- get_page Nothing my_article - putStrLn "Not logged in." - print $ fromJust $ html - return $ - case html of - Nothing -> Nothing - Just h -> Just $ my_read h + case html of + Left err -> do + let msg = "Failed to retrieve page. " ++ err + hPutStrLn stderr msg + return Nothing + Right h -> return $ Just $ my_read h -- | If we're given an empty path, return a handle to -- 'stdout'. Otherwise, open the given file and return a read/write