X-Git-Url: http://gitweb.michael.orlitzky.com/?a=blobdiff_plain;f=src%2FMain.hs;h=568f0fe99f55c48b105ab31e04fd5e99d12dd08b;hb=f8a623928407ceaaa8c28b60316e9123a4f3821b;hp=950a2a117a04915821042e94979b01f381a6e817;hpb=c99d184584e014aff4953fa8f90c9b3b6dc65229;p=dead%2Fhtsn-import.git diff --git a/src/Main.hs b/src/Main.hs index 950a2a1..568f0fe 100644 --- a/src/Main.hs +++ b/src/Main.hs @@ -1,54 +1,145 @@ -{-# LANGUAGE ScopedTypeVariables #-} -{-# LANGUAGE OverloadedStrings #-} - +{-# LANGUAGE NoMonomorphismRestriction #-} module Main where -import Control.Monad.IO.Class (liftIO) -import Data.Maybe ( listToMaybe ) -import Database.Persist ( insert ) -import Database.Persist.Sql ( runMigration ) -import Database.Persist.Sqlite ( runSqlite ) -import Text.Show.Pretty ( ppShow ) +import Control.Arrow ( (&&&), arr, returnA ) +import Control.Monad ( when ) +import Control.Monad.IO.Class ( MonadIO, liftIO ) +import Database.Groundhog.Generic ( runDbConn ) +import Database.Groundhog.Sqlite ( + withSqliteConn ) +import Database.Groundhog.Postgresql ( + withPostgresqlConn ) +import Data.Monoid ( (<>) ) +import Network.Services.TSN.Logging ( init_logging ) +import System.Console.CmdArgs ( def ) +import System.Exit ( exitWith, ExitCode (ExitFailure) ) +import System.IO.Error ( catchIOError ) import Text.XML.HXT.Core ( - SysConfigList, - XmlPickler, - no, - runX, - withPreserveComment, - withRemoveWS, - withValidate, - xpickle, - xunpickleDocument, - yes ) - -import TSN.Injuries ( InjuriesMessage, migrate_injuries ) - - -parse_opts :: SysConfigList -parse_opts = - [ withPreserveComment no, - withRemoveWS yes, - withValidate no ] - -parse_file :: XmlPickler a => FilePath -> IO (Maybe a) -parse_file path = - fmap listToMaybe $ - runX ( xunpickleDocument xpickle parse_opts path ) - -main_sql :: IO () -main_sql = - runSqlite "foo.sqlite3" $ do - runMigration migrate_injuries - msg :: Maybe InjuriesMessage <- liftIO $ parse_file - "test/xml/injuriesxml.xml" - case msg of - Nothing -> return () - Just m -> do - msg_id <- insert m - return () + ArrowXml, + IOStateArrow, + XmlTree, + (>>>), + (/>), + getAttrl, + getText, + hasName, + readDocument, + runX ) + +import Backend ( Backend(..) ) +import CommandLine ( get_args ) +import Configuration ( Configuration(..), merge_optional ) +import ConnectionString ( ConnectionString(..) ) +import ExitCodes ( exit_no_xml_files ) +import qualified OptionalConfiguration as OC ( + OptionalConfiguration ( xml_files ), + from_rc ) +import Network.Services.TSN.Report ( + report_info, + report_error ) +import TSN.DbImport +import qualified TSN.Injuries as Injuries ( Listing ) +import qualified TSN.InjuriesDetail as InjuriesDetail ( PlayerListing ) +import qualified TSN.News as News ( Message ) +import Xml ( parse_opts ) + + + +import_file :: Configuration -> FilePath -> IO () +import_file cfg path = do + results <- catchIOError + parse_and_import + (\e -> do + report_error (show e) + report_error $ "Failed to import file " ++ path ++ "." + -- Return a nonempty list so we don't claim incorrectly that + -- we couldn't parse the DTD. + return [ Nothing ] ) + + case results of + -- If results' is empty, one of the arrows return "nothing." + [] -> report_error $ "Unable to determine DTD for file " ++ path ++ "." + (r:_) -> + case r of + Nothing -> return () + Just cnt -> report_info $ "Successfully imported " ++ + (show cnt) ++ + " records from " ++ path ++ "." + where + -- | An arrow that reads a document into an 'XmlTree'. + readA :: IOStateArrow s a XmlTree + readA = readDocument parse_opts path + + -- | An arrow which parses the doctype "SYSTEM" of an 'XmlTree'. + -- We use these to determine the parser to use. + doctypeA :: ArrowXml a => a XmlTree String + doctypeA = getAttrl >>> hasName "doctype-SYSTEM" /> getText + + -- | Combine the arrows above as well as the function below + -- (arrowized with 'arr') into an IO action that does everything + -- (parses and then runs the import on what was parsed). + -- + -- The result of runX has type IO [IO (Maybe Int)]. We thus use + -- bind (>>=) and sequence to combine all of the IOs into one + -- big one outside of the list. + parse_and_import :: IO [Maybe Int] + parse_and_import = + runX (readA >>> (doctypeA &&& returnA) >>> (arr import_with_dtd)) + >>= + sequence + + -- | Takes a 'Doctype', 'XmlTree' pair and uses the 'Doctype' to + -- determine which function to call on the 'XmlTree'. + import_with_dtd :: (String, XmlTree) -> IO (Maybe Int) + import_with_dtd (dtd,xml) = + -- We need NoMonomorphismRestriction here. + if backend cfg == Postgres + then withPostgresqlConn cs $ runDbConn $ importer xml + else withSqliteConn cs $ runDbConn $ importer xml + where + -- | Pull the real connection String out of the configuration. + cs :: String + cs = get_connection_string $ connection_string cfg + + importer + | dtd == "injuriesxml.dtd" = + dbimport (undefined :: Injuries.Listing) + + | dtd == "Injuries_Detail_XML.dtd" = + dbimport (undefined :: InjuriesDetail.PlayerListing) + + | dtd == "newsxml.dtd" = + dbimport (undefined :: News.Message) + + | otherwise = \_ -> do -- Dummy arg simplifies the other cases. + let errmsg = "Unrecognized DTD in " ++ path ++ ": " ++ dtd ++ "." + liftIO $ report_info errmsg + return Nothing + + main :: IO () main = do - msg :: Maybe InjuriesMessage <- parse_file "test/xml/injuriesxml.xml" - putStr $ ppShow msg + rc_cfg <- OC.from_rc + cmd_cfg <- get_args + + -- Merge the config file options with the command-line ones, + -- prefering the command-line ones. + let opt_config = rc_cfg <> cmd_cfg + + -- Update a default config with any options that have been set in + -- either the config file or on the command-line. We initialize + -- logging before the missing parameter checks below so that we can + -- log the errors. + let cfg = (def :: Configuration) `merge_optional` opt_config + init_logging (log_file cfg) (log_level cfg) (syslog cfg) + + -- Check the optional config for missing required options. + when (null $ OC.xml_files opt_config) $ do + report_error "No XML files given." + exitWith (ExitFailure exit_no_xml_files) + + -- We don't do this in parallel (for now?) to keep the error + -- messages nice and linear. + mapM_ (import_file cfg) (OC.xml_files opt_config)