X-Git-Url: http://gitweb.michael.orlitzky.com/?a=blobdiff_plain;f=src%2FMain.hs;h=e7f9e9c6213ae8df2b65214d4ff46b69e9b294f5;hb=e88453d996ef8c94d581122309669c5c2aa87c9d;hp=38f393080f9d7f128f1833b32c18199b1888163c;hpb=5b051a15402e5ebbe4e388265dbf8cef6c31df6f;p=dead%2Fhtsn-import.git

diff --git a/src/Main.hs b/src/Main.hs
index 38f3930..e7f9e9c 100644
--- a/src/Main.hs
+++ b/src/Main.hs
@@ -45,24 +45,44 @@ import Network.Services.TSN.Report (
   report_info,
   report_error )
 import TSN.DbImport ( DbImport(..), ImportResult(..) )
+import TSN.Parse ( format_parse_error )
+import qualified TSN.XML.AutoRacingDriverList as AutoRacingDriverList (
+  dtd,
+  pickle_message )
 import qualified TSN.XML.AutoRacingResults as AutoRacingResults (
   dtd,
   pickle_message )
 import qualified TSN.XML.AutoRacingSchedule as AutoRacingSchedule (
   dtd,
   pickle_message )
+import qualified TSN.XML.EarlyLine as EarlyLine (
+  dtd,
+  pickle_message )
 import qualified TSN.XML.GameInfo as GameInfo ( dtds, parse_xml )
 import qualified TSN.XML.Heartbeat as Heartbeat ( dtd, verify )
 import qualified TSN.XML.Injuries as Injuries ( dtd, pickle_message )
 import qualified TSN.XML.InjuriesDetail as InjuriesDetail (
   dtd,
   pickle_message )
+import qualified TSN.XML.MLBEarlyLine as MLBEarlyLine (
+  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 )
 
 
@@ -93,7 +113,7 @@ import_file cfg path = do
       report_error $ "Unable to determine DTD for file " ++ path ++ "."
       return False
     (ImportFailed errmsg:_) -> do
-      report_error errmsg
+      report_error $ errmsg ++ " (" ++ path ++ ")"
       return False
     (ImportSkipped infomsg:_) -> do
       -- We processed the message but didn't import anything. Return
@@ -178,12 +198,18 @@ import_file cfg path = do
                    (unpickleDoc f xml)
 
           importer
+            | dtd == AutoRacingDriverList.dtd =
+                go AutoRacingDriverList.pickle_message
+
             | dtd == AutoRacingResults.dtd =
                 go AutoRacingResults.pickle_message
 
             | dtd == AutoRacingSchedule.dtd =
                 go AutoRacingSchedule.pickle_message
 
+            | dtd == EarlyLine.dtd =
+                go EarlyLine.pickle_message
+
             -- GameInfo and SportInfo appear last in the guards
             | dtd == Injuries.dtd = go Injuries.pickle_message
 
@@ -191,21 +217,47 @@ import_file cfg path = do
 
             | dtd == JFile.dtd = go JFile.pickle_message
 
-            | dtd == News.dtd = go News.pickle_message
-
+            | dtd == MLBEarlyLine.dtd =
+                go MLBEarlyLine.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
+
             | 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
                 case either_m of
                   -- This might give us a slightly better error
                   -- message than the default 'errmsg'.
-                  Left err -> return $ ImportFailed err
+                  Left err -> return $ ImportFailed (format_parse_error err)
                   Right m     -> migrate_and_import m
 
             | dtd `elem` SportInfo.dtds = do
@@ -213,7 +265,7 @@ import_file cfg path = do
                 case either_m of
                   -- This might give us a slightly better error
                   -- message than the default 'errmsg'.
-                  Left err -> return $ ImportFailed err
+                  Left err -> return $ ImportFailed (format_parse_error err)
                   Right m     -> migrate_and_import m
 
             | otherwise = do