X-Git-Url: http://gitweb.michael.orlitzky.com/?p=dead%2Flwn-epub.git;a=blobdiff_plain;f=src%2FLWN%2FPage.hs;h=97171c641be08a6761a1a1800002190c7f10b9ed;hp=3027eaef6c83e76067a744acf98be3a8b6b8579f;hb=aad40cd8e1e8c84c5fc294674a7159bb40838440;hpb=2953924e2016393a1ffb9e2e82b4c90c8c57dfd3 diff --git a/src/LWN/Page.hs b/src/LWN/Page.hs index 3027eae..97171c6 100644 --- a/src/LWN/Page.hs +++ b/src/LWN/Page.hs @@ -9,9 +9,7 @@ import qualified Data.ByteString.Lazy as B (ByteString, hPut) import Data.String.Utils (split, strip) import Data.Maybe (catMaybes, fromJust, isNothing) import Prelude hiding (readFile) -import System.Directory (doesFileExist) -import System.IO (Handle, hClose, hFlush, hPutStrLn, stderr) -import System.IO.UTF8 (readFile) +import System.IO (Handle, hClose, hFlush) import Test.HUnit (Assertion, assertEqual) import Test.Framework (Test, testGroup) import Test.Framework.Providers.HUnit (testCase) @@ -24,92 +22,40 @@ import Text.Pandoc ( import Text.XML.HXT.Core ( ArrowXml, IOSArrow, - IOStateArrow, XmlTree, (>>>), (/>), (//>), changeAttrValue, - getAttrValue, getChildren, getText, - hasAttrValue, hasName, - isElem, - mkName, - none, processAttrl, processTopDown, runX, - setElemName, xshow, when) import Text.HandsomeSoup (css, parseHtml) -import Configuration (Cfg, password, use_account, username) +import Configuration (Cfg) import LWN.Article import LWN.HTTP ( ImageMap, download_image_urls, - get_page, - log_in, - make_cookie_jar) -import LWN.URI (URL, try_make_absolute_url) -import LWN.XHTML (XHTML, parse_lwn, to_xhtml) -import Misc (contains) - - --- | 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 -> URL -> IO (Maybe (IOStateArrow s b XmlTree)) -get_xml_from_article cfg article_name = do - my_article <- real_article_path article_name - is_file <- doesFileExist my_article - case is_file of - True -> do - contents <- readFile my_article - return $ Just $ parse_lwn contents - False -> do - -- Download the URL and try to parse it. - 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 + get_article_contents) +import LWN.URI (URL) +import LWN.XHTML ( + XHTML, + image_srcs, + is_image, + preprocess, + remove_byline, + remove_title, + to_xhtml, + xml_from_contents) --- Should be called *after* preprocessing. -download_images :: IOSArrow XmlTree XmlTree -> IO ImageMap -download_images xml = do - image_urls <- runX $ xml >>> image_srcs - download_image_urls image_urls - data Page = -- | An LWN page with one article on it. @@ -158,43 +104,19 @@ instance XHTML Page where page_from_url :: Cfg -> URL -> IO (Maybe Page) page_from_url cfg url = do - maybe_html <- get_xml_from_article cfg url - case maybe_html of + contents <- get_article_contents cfg url + case (xml_from_contents contents) of Just html -> parse html Nothing -> return Nothing -is_link :: (ArrowXml a) => a XmlTree XmlTree -is_link = - isElem >>> hasName "a" - - -remove_comment_links :: (ArrowXml a) => a XmlTree XmlTree -remove_comment_links = - processTopDown $ kill_comments `when` is_link - where - is_comment_link = - hasAttrValue "href" (contains "#Comments") - - kill_comments = - none `when` is_comment_link - -replace_links_with_spans :: (ArrowXml a) => a XmlTree XmlTree -replace_links_with_spans = - processTopDown $ (make_span >>> remove_attrs) `when` is_link - where - make_span = setElemName $ mkName "span" - remove_attrs = processAttrl none +-- Should be called *after* preprocessing. +download_images :: IOSArrow XmlTree XmlTree -> IO ImageMap +download_images xml = do + image_urls <- runX $ xml >>> image_srcs + download_image_urls image_urls --- | Preprocessing common to both page types. -preprocess :: (ArrowXml a) => a XmlTree XmlTree -preprocess = - make_image_srcs_absolute - >>> - remove_comment_links - >>> - replace_links_with_spans replace_remote_img_srcs :: (ArrowXml a) => ImageMap -> a XmlTree XmlTree @@ -324,33 +246,6 @@ fp_parse_article_title xml = do -is_title :: (ArrowXml a) => a XmlTree XmlTree -is_title = - (hasName "h2") - >>> - (hasAttrValue "class" (== "SummaryHL")) - - -is_byline :: (ArrowXml a) => a XmlTree XmlTree -is_byline = - (hasName "div") - >>> - (hasAttrValue "class" (== "FeatureByLine")) - - -is_image :: (ArrowXml a) => a XmlTree XmlTree -is_image = isElem >>> hasName "img" - -remove_title :: (ArrowXml a) => a XmlTree XmlTree -remove_title = - processTopDown ((none) `when` is_title) - - -remove_byline :: (ArrowXml a) => a XmlTree XmlTree -remove_byline = - processTopDown ((none) `when` is_byline) - - fp_parse_article_body :: IOSArrow XmlTree XmlTree -> IO (Maybe String) fp_parse_article_body xml = do @@ -457,30 +352,6 @@ xhtml_to_epub epmd = read_html = readHtml defaultParserState - --- --- Misc --- - -image_srcs :: (ArrowXml a) => a XmlTree URL -image_srcs = - css "img" - >>> - getAttrValue "src" - -make_image_srcs_absolute :: (ArrowXml a) => a XmlTree XmlTree -make_image_srcs_absolute = - processTopDown (make_srcs_absolute `when` is_image) - where - change_src :: (ArrowXml a) => a XmlTree XmlTree - change_src = - changeAttrValue try_make_absolute_url - - make_srcs_absolute :: (ArrowXml a) => a XmlTree XmlTree - make_srcs_absolute = - processAttrl $ change_src `when` hasName "src" - - -- -- Tests --