-{-# LANGUAGE ScopedTypeVariables, RecordWildCards #-}
+{-# LANGUAGE DoAndIfThenElse #-}
module Main
where
+import Data.Maybe (fromJust)
import Prelude hiding (readFile)
import System.Directory (doesFileExist)
import System.IO (
Handle,
IOMode (WriteMode),
+ hPutStrLn,
openBinaryFile,
- stdout
- )
+ stderr,
+ stdout)
import System.IO.UTF8 (readFile)
-import Text.Regex.Posix ((=~))
-import Text.XML.HXT.Core
+import Text.XML.HXT.Core (
+ IOStateArrow,
+ XmlTree)
import CommandLine (show_help)
-import Configuration (Cfg(..), get_cfg)
-import LWN.Page
-import LWN.URI (is_lwn_url, make_absolute_url)
-import Misc (contains)
+import Configuration (Cfg(..), get_cfg, use_account)
+import LWN.Article (real_article_path)
+import LWN.HTTP (get_page, log_in, make_cookie_jar)
+import LWN.Page (epublish, parse)
+import LWN.XHTML (parse_lwn)
-my_read_opts :: SysConfigList
-my_read_opts = [ withValidate no,
- withParseHTML yes,
- withWarnings no ]
-
--- | My version of HandsomeSoup's parseHTML.
-my_read :: String -> IOStateArrow s b XmlTree
-my_read = readString my_read_opts
-- | 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 :: String -> IO (Maybe (IOStateArrow s b XmlTree))
-get_xml_from_article s = do
- article <- real_article_path s
- is_file <- doesFileExist article
+get_xml_from_article :: Cfg -> IO (Maybe (IOStateArrow s b XmlTree))
+get_xml_from_article cfg = do
+ my_article <- real_article_path (article cfg)
+ is_file <- doesFileExist my_article
case is_file of
True -> do
- contents <- readFile article
- return $ Just $ my_read contents
+ contents <- readFile my_article
+ return $ Just $ parse_lwn contents
False -> do
-- Download the URL and try to parse it.
- return Nothing
+ if use_account cfg then do
+ -- use_account would be false if these fromJusts would fail.
+ cj <- make_cookie_jar
+ li_result <- log_in cj
+ (fromJust $ username cfg)
+ (fromJust $ password cfg)
+
+ case li_result of
+ Left err -> do
+ let msg = "Failed to log in. " ++ err
+ hPutStrLn stderr msg
+ Right response_body -> do
+ hPutStrLn stderr response_body
+
+ html <- get_page (Just cj) my_article
+
+ case html of
+ Left err -> do
+ let msg = "Failed to retrieve page. " ++ err
+ hPutStrLn stderr msg
+ return Nothing
+ Right h -> return $ Just $ parse_lwn h
+ else do
+ html <- get_page Nothing my_article
+ case html of
+ Left err -> do
+ let msg = "Failed to retrieve page. " ++ err
+ hPutStrLn stderr msg
+ return Nothing
+ Right h -> return $ Just $ parse_lwn h
+
-- | If we're given an empty path, return a handle to
-- 'stdout'. Otherwise, open the given file and return a read/write
openBinaryFile path WriteMode
-
--- | Convert the given article to either a URL or a filesystem
--- path. If the given article exists on the filesystem, we assume
--- it's a file. Otherwise, we check to see if it's a URL. Failing
--- that, we try to construct a URL from what we're given and do our
--- best.
-real_article_path :: String -> IO String
-real_article_path s = do
- is_file <- doesFileExist s
- return $ if is_file then s else check_cases
- where
- abs_current =
- case make_absolute_url "current" of
- Nothing -> s
- Just ac -> ac
- abs_s =
- case make_absolute_url s of
- Nothing -> s
- Just as -> as
-
- check_cases :: String
- check_cases
- | is_lwn_url s = s
- | s `contains` "current" = abs_current
- | s =~ "^[0-9]+$" = abs_s
- | otherwise = s -- Give up
-
main :: IO ()
main = do
- Cfg{..} <- get_cfg
- output_handle <- get_output_handle output
- maybe_html <- get_xml_from_article article
+ cfg <- get_cfg
+ output_handle <- get_output_handle (output cfg)
+ maybe_html <- get_xml_from_article cfg
case maybe_html of
-
Just html -> do
result <- parse html
case result of