git.iko.soy/typeable/bloodhound
on master help sign in
Merge pull request #53 from bermanjosh/scanType

Scan type

commit9e0179b9a4606ecd823f6617a4e26829d404d893 authorChris Allen <cma@bitemyapp.com> dateMon 17 Aug 2015, 23:57 (11 years ago) parentf185e9dc02b6c7
4 files changed+79−17
bloodhound.cabal+1−1 src/Database/Bloodhound/Client.hs+58−11 src/Database/Bloodhound/Types.hs+19−4 tests/tests.hs+1−1
bloodhound.cabal+1−1
@@ -1,5 +1,5 @@
11name: bloodhound
2−version: 0.7.0.1
2+version: 0.7.1.0
33synopsis: ElasticSearch client library for Haskell
44description: ElasticSearch made awesome for Haskell hackers
55homepage: https://github.com/bitemyapp/bloodhound
src/Database/Bloodhound/Client.hs+58−11
@@ -1,5 +1,5 @@
1−{-# LANGUAGE OverloadedStrings #-}
2−{-# LANGUAGE RecordWildCards #-}
1+{-# LANGUAGE OverloadedStrings #-}
2+{-# LANGUAGE RecordWildCards #-}
33
44-------------------------------------------------------------------------------
55-- |
@@ -39,6 +39,7 @@ module Database.Bloodhound.Client
3939 , searchAll
4040 , searchByIndex
4141 , searchByType
42+ , scanSearch
4243 , refreshIndex
4344 , mkSearch
4445 , mkAggregateSearch
@@ -171,6 +172,18 @@ joinPath ps = do
171172 Server s <- bhServer <$> getBHEnv
172173 return $ joinPath' (s:ps)
173174
175+appendSearchTypeParam :: Text -> SearchType -> Text
176+appendSearchTypeParam originalUrl st = addQuery [(keyEq, Just stParams)] originalUrl
177+ where keyEq = "search_type="
178+ stParams
179+ | st == SearchTypeDfsQueryThenFetch = "dfs_query_then_fetch"
180+ | st == SearchTypeCount = "count"
181+ | st == SearchTypeScan = "scan&scroll=1m"
182+ | st == SearchTypeQueryAndFetch = "query_and_fetch"
183+ | st == SearchTypeDfsQueryAndFetch = "dfs_query_and_fetch"
184+ -- used to catch 'SearchTypeQueryThenFetch', which is also the default
185+ | otherwise = "query_then_fetch"
186+
174187-- | Severely dumbed down query renderer. Assumes your data doesn't
175188-- need any encoding
176189addQuery :: [(Text, Maybe Text)] -> Text -> Text
@@ -180,11 +193,10 @@ addQuery q u = u <> rendered
180193 T.decodeUtf8 $ BB.toByteString $ NHTU.renderQueryText prependQuestionMark q
181194 prependQuestionMark = True
182195
183−
184196bindM2 :: (Applicative m, Monad m) => (a -> b -> m c) -> m a -> m b -> m c
185197bindM2 f ma mb = join (f <$> ma <*> mb)
186198
187−-- | Convenience function that sets up a mananager and BHEnv and runs
199+-- | Convenience function that sets up a manager and BHEnv and runs
188200-- the given set of bloodhound operations. Connections will be
189201-- pipelined automatically in accordance with the given manager
190202-- settings in IO. If you've got your own monad transformer stack, you
@@ -496,7 +508,8 @@ documentExists (IndexName indexName)
496508 where url = joinPath [indexName, mappingName, docId]
497509
498510dispatchSearch :: MonadBH m => Text -> Search -> m Reply
499−dispatchSearch url search = post url (Just (encode search))
511+dispatchSearch url search = post url' (Just (encode search))
512+ where url' = appendSearchTypeParam url (searchType search)
500513
501514-- | 'searchAll', given a 'Search', will perform that search against all indexes
502515-- on an Elasticsearch server. Try to avoid doing this if it can be helped.
@@ -530,6 +543,40 @@ searchByType (IndexName indexName)
530543 (MappingName mappingName) = bindM2 dispatchSearch url . return
531544 where url = joinPath [indexName, mappingName, "_search"]
532545
546+scanSearch' :: MonadBH m => Search -> m (Maybe ScrollId)
547+scanSearch' search = do
548+ let url = joinPath ["_search"]
549+ search' = search { searchType = SearchTypeScan }
550+ resp' <- bindM2 dispatchSearch url (return search')
551+ let msr = decode' $ responseBody resp' :: Maybe (SearchResult ())
552+ msid = maybe Nothing scrollId msr
553+ return msid
554+
555+scroll' :: (FromJSON a, MonadBH m) => Maybe ScrollId -> m ([Hit a], Maybe ScrollId)
556+scroll' Nothing = return ([], Nothing)
557+scroll' (Just sid) = do
558+ url <- joinPath ["_search/scroll?scroll=1m"]
559+ resp' <- post url (Just . L.fromStrict $ T.encodeUtf8 sid)
560+ let msr = decode' $ responseBody resp' :: FromJSON a => Maybe (SearchResult a)
561+ resp = case msr of
562+ Just sr -> (hits $ searchHits sr, scrollId sr)
563+ _ -> ([], Nothing)
564+ return resp
565+
566+simpleAccumilator :: (FromJSON a, MonadBH m) => [Hit a] -> ([Hit a], Maybe ScrollId) -> m ([Hit a], Maybe ScrollId)
567+simpleAccumilator oldHits (newHits, Nothing) = return (oldHits ++ newHits, Nothing)
568+simpleAccumilator oldHits ([], _) = return (oldHits, Nothing)
569+simpleAccumilator oldHits (newHits, msid) = do
570+ (newHits', msid') <- scroll' msid
571+ simpleAccumilator (oldHits ++ newHits) (newHits', msid')
572+
573+scanSearch :: (FromJSON a, MonadBH m) => Search -> m [Hit a]
574+scanSearch search = do
575+ msid <- scanSearch' search
576+ (hits, msid') <- scroll' msid
577+ (totalHits, _) <- simpleAccumilator [] (hits, msid')
578+ return totalHits
579+
533580-- | 'mkSearch' is a helper function for defaulting additional fields of a 'Search'
534581-- to Nothing in case you only care about your 'Query' and 'Filter'. Use record update
535582-- syntax if you want to add things like aggregations or highlights while still using
@@ -537,9 +584,9 @@ searchByType (IndexName indexName)
537584--
538585-- >>> let query = TermQuery (Term "user" "bitemyapp") Nothing
539586-- >>> mkSearch (Just query) Nothing
540−-- Search {queryBody = Just (TermQuery (Term {termField = "user", termValue = "bitemyapp"}) Nothing), filterBody = Nothing, sortBody = Nothing, aggBody = Nothing, highlight = Nothing, trackSortScores = False, from = From 0, size = Size 10, fields = Nothing}
587+-- Search {queryBody = Just (TermQuery (Term {termField = "user", termValue = "bitemyapp"}) Nothing), filterBody = Nothing, sortBody = Nothing, aggBody = Nothing, highlight = Nothing, trackSortScores = False, from = From 0, size = Size 10, searchType = SearchTypeQueryThenFetch, fields = Nothing}
541588mkSearch :: Maybe Query -> Maybe Filter -> Search
542−mkSearch query filter = Search query filter Nothing Nothing Nothing False (From 0) (Size 10) Nothing
589+mkSearch query filter = Search query filter Nothing Nothing Nothing False (From 0) (Size 10) SearchTypeQueryThenFetch Nothing
543590
544591-- | 'mkAggregateSearch' is a helper function that defaults everything in a 'Search' except for
545592-- the 'Query' and the 'Aggregation'.
@@ -549,7 +596,7 @@ mkSearch query filter = Search query filter Nothing Nothing Nothing False (From
549596-- TermsAgg (TermsAggregation {term = Left "user", termInclude = Nothing, termExclude = Nothing, termOrder = Nothing, termMinDocCount = Nothing, termSize = Nothing, termShardSize = Nothing, termCollectMode = Just BreadthFirst, termExecutionHint = Nothing, termAggs = Nothing})
550597-- >>> let myAggregation = mkAggregateSearch Nothing $ mkAggregations "users" terms
551598mkAggregateSearch :: Maybe Query -> Aggregations -> Search
552−mkAggregateSearch query mkSearchAggs = Search query Nothing Nothing (Just mkSearchAggs) Nothing False (From 0) (Size 0) Nothing
599+mkAggregateSearch query mkSearchAggs = Search query Nothing Nothing (Just mkSearchAggs) Nothing False (From 0) (Size 0) SearchTypeQueryThenFetch Nothing
553600
554601-- | 'mkHighlightSearch' is a helper function that defaults everything in a 'Search' except for
555602-- the 'Query' and the 'Aggregation'.
@@ -558,7 +605,7 @@ mkAggregateSearch query mkSearchAggs = Search query Nothing Nothing (Just mkSear
558605-- >>> let testHighlight = Highlights Nothing [FieldHighlight (FieldName "message") Nothing]
559606-- >>> let search = mkHighlightSearch (Just query) testHighlight
560607mkHighlightSearch :: Maybe Query -> Highlights -> Search
561−mkHighlightSearch query searchHighlights = Search query Nothing Nothing Nothing (Just searchHighlights) False (From 0) (Size 10) Nothing
608+mkHighlightSearch query searchHighlights = Search query Nothing Nothing Nothing (Just searchHighlights) False (From 0) (Size 10) SearchTypeQueryThenFetch Nothing
562609
563610-- | 'pageSearch' is a helper function that takes a search and assigns the from
564611-- and size fields for the search. The from parameter defines the offset
@@ -568,9 +615,9 @@ mkHighlightSearch query searchHighlights = Search query Nothing Nothing Nothing
568615-- >>> let query = QueryMatchQuery $ mkMatchQuery (FieldName "_all") (QueryString "haskell")
569616-- >>> let search = mkSearch (Just query) Nothing
570617-- >>> search
571−-- Search {queryBody = Just (QueryMatchQuery (MatchQuery {matchQueryField = FieldName "_all", matchQueryQueryString = QueryString "haskell", matchQueryOperator = Or, matchQueryZeroTerms = ZeroTermsNone, matchQueryCutoffFrequency = Nothing, matchQueryMatchType = Nothing, matchQueryAnalyzer = Nothing, matchQueryMaxExpansions = Nothing, matchQueryLenient = Nothing})), filterBody = Nothing, sortBody = Nothing, aggBody = Nothing, highlight = Nothing, trackSortScores = False, from = From 0, size = Size 10, fields = Nothing}
618+-- Search {queryBody = Just (QueryMatchQuery (MatchQuery {matchQueryField = FieldName "_all", matchQueryQueryString = QueryString "haskell", matchQueryOperator = Or, matchQueryZeroTerms = ZeroTermsNone, matchQueryCutoffFrequency = Nothing, matchQueryMatchType = Nothing, matchQueryAnalyzer = Nothing, matchQueryMaxExpansions = Nothing, matchQueryLenient = Nothing})), filterBody = Nothing, sortBody = Nothing, aggBody = Nothing, highlight = Nothing, trackSortScores = False, from = From 0, size = Size 10, searchType = SearchTypeQueryThenFetch, fields = Nothing}
572619-- >>> pageSearch (From 10) (Size 100) search
573−-- Search {queryBody = Just (QueryMatchQuery (MatchQuery {matchQueryField = FieldName "_all", matchQueryQueryString = QueryString "haskell", matchQueryOperator = Or, matchQueryZeroTerms = ZeroTermsNone, matchQueryCutoffFrequency = Nothing, matchQueryMatchType = Nothing, matchQueryAnalyzer = Nothing, matchQueryMaxExpansions = Nothing, matchQueryLenient = Nothing})), filterBody = Nothing, sortBody = Nothing, aggBody = Nothing, highlight = Nothing, trackSortScores = False, from = From 10, size = Size 100, fields = Nothing}
620+-- Search {queryBody = Just (QueryMatchQuery (MatchQuery {matchQueryField = FieldName "_all", matchQueryQueryString = QueryString "haskell", matchQueryOperator = Or, matchQueryZeroTerms = ZeroTermsNone, matchQueryCutoffFrequency = Nothing, matchQueryMatchType = Nothing, matchQueryAnalyzer = Nothing, matchQueryMaxExpansions = Nothing, matchQueryLenient = Nothing})), filterBody = Nothing, sortBody = Nothing, aggBody = Nothing, highlight = Nothing, trackSortScores = False, from = From 10, size = Size 100, searchType = SearchTypeQueryThenFetch, fields = Nothing}
574621pageSearch :: From -- ^ The result offset
575622 -> Size -- ^ The number of results to return
576623 -> Search -- ^ The current seach
src/Database/Bloodhound/Types.hs+19−4
@@ -68,7 +68,9 @@ module Database.Bloodhound.Types
6868 , IndexDocumentSettings(..)
6969 , Query(..)
7070 , Search(..)
71+ , SearchType(..)
7172 , SearchResult(..)
73+ , ScrollId
7274 , SearchHits(..)
7375 , TrackSortScores
7476 , From(..)
@@ -731,9 +733,18 @@ data Search = Search { queryBody :: Maybe Query
731733 -- default False
732734 , trackSortScores :: TrackSortScores
733735 , from :: From
734− , size :: Size
736+ , size :: Size
737+ , searchType :: SearchType
735738 , fields :: Maybe [FieldName] } deriving (Eq, Show)
736739
740+data SearchType = SearchTypeQueryThenFetch
741+ | SearchTypeDfsQueryThenFetch
742+ | SearchTypeCount
743+ | SearchTypeScan
744+ | SearchTypeQueryAndFetch
745+ | SearchTypeDfsQueryAndFetch
746+ deriving (Eq, Show)
747+
737748data Highlights = Highlights { globalsettings :: Maybe HighlightSettings
738749 , highlightFields :: [FieldHighlight]
739750 } deriving (Show, Eq)
@@ -1245,7 +1256,10 @@ data SearchResult a =
12451256 , timedOut :: Bool
12461257 , shards :: ShardResult
12471258 , searchHits :: SearchHits a
1248− , aggregations :: Maybe AggregationResults } deriving (Eq, Show)
1259+ , aggregations :: Maybe AggregationResults
1260+ , scrollId :: Maybe ScrollId } deriving (Eq, Show)
1261+
1262+type ScrollId = Text -- Fixme: Newtype
12491263
12501264type Score = Maybe Double
12511265
@@ -2094,7 +2108,7 @@ instance (FromJSON a) => FromJSON (EsResultFound a) where
20942108 parseJSON _ = empty
20952109
20962110instance ToJSON Search where
2097− toJSON (Search query sFilter sort searchAggs highlight sTrackSortScores sFrom sSize sFields) =
2111+ toJSON (Search query sFilter sort searchAggs highlight sTrackSortScores sFrom sSize _ sFields) =
20982112 omitNulls [ "query" .= query
20992113 , "filter" .= sFilter
21002114 , "sort" .= sort
@@ -2323,7 +2337,8 @@ instance (FromJSON a) => FromJSON (SearchResult a) where
23232337 v .: "timed_out" <*>
23242338 v .: "_shards" <*>
23252339 v .: "hits" <*>
2326− v .:? "aggregations"
2340+ v .:? "aggregations" <*>
2341+ v .:? "_scroll_id"
23272342 parseJSON _ = empty
23282343
23292344instance (FromJSON a) => FromJSON (SearchHits a) where
tests/tests.hs+1−1
@@ -435,7 +435,7 @@ main = hspec $ do
435435 let sortSpec = DefaultSortSpec $ mkSort (FieldName "age") Ascending
436436 let search = Search Nothing
437437 (Just IdentityFilter) (Just [sortSpec]) Nothing Nothing
438− False (From 0) (Size 10) Nothing
438+ False (From 0) (Size 10) SearchTypeQueryThenFetch Nothing
439439 result <- searchTweets search
440440 let myTweet = fmap (hitSource . head . hits . searchHits) result
441441 liftIO $