X-Git-Url: http://gitweb.michael.orlitzky.com/?a=blobdiff_plain;f=src%2FMain.hs;h=492f0ceeda96c27fbf1adec1b123cffe2edbde7e;hb=9b28d5ed0e1570f0890eae1ee01721688c20e266;hp=fc6d14b8a116aef185d0c7d362b2204c912ff135;hpb=3768238ca528db394e4777ac1e19b2091979e1c8;p=dead%2Fhtsn-import.git diff --git a/src/Main.hs b/src/Main.hs index fc6d14b..492f0ce 100644 --- a/src/Main.hs +++ b/src/Main.hs @@ -58,14 +58,21 @@ import qualified TSN.XML.InjuriesDetail as InjuriesDetail ( dtd, pickle_message ) import qualified TSN.XML.JFile as JFile ( dtd, pickle_message ) -import qualified TSN.XML.News as News ( dtd, pickle_message ) +import qualified TSN.XML.News as News ( + dtd, + has_only_single_sms, + pickle_message ) import qualified TSN.XML.Odds as Odds ( dtd, pickle_message ) import qualified TSN.XML.ScheduleChanges as ScheduleChanges ( dtd, pickle_message ) import qualified TSN.XML.Scores as Scores ( dtd, pickle_message ) import qualified TSN.XML.SportInfo as SportInfo ( dtds, parse_xml ) -import qualified TSN.XML.Weather as Weather ( dtd, pickle_message ) +import qualified TSN.XML.Weather as Weather ( + dtd, + is_type1, + pickle_message, + teams_are_normal ) import Xml ( DtdName(..), parse_opts ) @@ -194,8 +201,16 @@ import_file cfg path = do | dtd == JFile.dtd = go JFile.pickle_message - | dtd == News.dtd = go News.pickle_message - + | dtd == News.dtd = + -- Some of the newsxml docs are busted in predictable ways. + -- We want them to "succeed" so that they're deleted. + -- We already know we can't parse them. + if News.has_only_single_sms xml + then go News.pickle_message + else do + let msg = "Unsupported newsxml.dtd with multiple SMS " ++ + "(" ++ path ++ ")" + return $ ImportUnsupported msg | dtd == Odds.dtd = go Odds.pickle_message | dtd == ScheduleChanges.dtd = go ScheduleChanges.pickle_message @@ -203,7 +218,20 @@ import_file cfg path = do | dtd == Scores.dtd = go Scores.pickle_message -- SportInfo and GameInfo appear last in the guards - | dtd == Weather.dtd = go Weather.pickle_message + | dtd == Weather.dtd = + -- Some of the weatherxml docs are busted in predictable ways. + -- We want them to "succeed" so that they're deleted. + -- We already know we can't parse them. + if Weather.is_type1 xml + then if Weather.teams_are_normal xml + then go Weather.pickle_message + else do + let msg = "Teams in reverse order in weatherxml.dtd" ++ + " (" ++ path ++ ")" + return $ ImportUnsupported msg + else do + let msg = "Unsupported weatherxml.dtd type (" ++ path ++ ")" + return $ ImportUnsupported msg | dtd `elem` GameInfo.dtds = do let either_m = GameInfo.parse_xml dtd xml