| alexander.behm | c576c60 | 2012-07-06 02:41:15 +0000 | [diff] [blame] | 1 | /* |
| 2 | * Description : Tests whether an ngram_index is applied to optimize a selection query using the similarity-jaccard-check function on 3-gram tokens. |
| 3 | * Tests that the optimizer rule correctly drills through the let clauses. |
| 4 | * The index should be applied. |
| 5 | * Success : Yes |
| 6 | */ |
| 7 | |
| 8 | drop dataverse test if exists; |
| 9 | create dataverse test; |
| 10 | use dataverse test; |
| 11 | |
| 12 | create type DBLPType as closed { |
| 13 | id: int32, |
| 14 | dblpid: string, |
| 15 | title: string, |
| 16 | authors: string, |
| 17 | misc: string |
| 18 | } |
| 19 | |
| 20 | create dataset DBLP(DBLPType) partitioned by key id; |
| 21 | |
| 22 | load dataset DBLP |
| 23 | using "edu.uci.ics.asterix.external.dataset.adapter.NCFileSystemAdapter" |
| 24 | (("path"="nc1://data/dblp-small/dblp-small-id.txt"),("format"="delimited-text"),("delimiter"=":")) pre-sorted; |
| 25 | |
| 26 | create index ngram_index on DBLP(title) type ngram(3); |
| 27 | |
| 28 | write output to nc1:"rttest/inverted-index-complex_ngram-jaccard-check-let.adm"; |
| 29 | |
| 30 | for $o in dataset('DBLP') |
| 31 | let $jacc := similarity-jaccard-check(gram-tokens($o.title, 3, false), gram-tokens("Transactions for Cooperative Environments", 3, false), 0.5f) |
| 32 | where $jacc[0] |
| 33 | return $o |