| alexander.behm | a9faa61 | 2012-11-24 22:38:05 +0000 | [diff] [blame] | 1 | drop dataverse test if exists; |
| 2 | create dataverse test; |
| 3 | use dataverse test; |
| 4 | |
| 5 | create type DBLPType as closed { |
| 6 | id: int32, |
| 7 | dblpid: string, |
| 8 | title: string, |
| 9 | authors: string, |
| 10 | misc: string |
| 11 | } |
| 12 | |
| 13 | create nodegroup group1 if not exists on nc1, nc2; |
| 14 | |
| 15 | create dataset DBLP(DBLPType) |
| 16 | partitioned by key id on group1; |
| 17 | |
| 18 | load dataset DBLP |
| 19 | using "edu.uci.ics.asterix.external.dataset.adapter.NCFileSystemAdapter" |
| 20 | (("path"="nc1://data/dblp-small/dblp-small-id.txt"),("format"="delimited-text"),("delimiter"=":")) pre-sorted; |
| 21 | |
| 22 | create index ngram_index on DBLP(authors) type fuzzy ngram(3); |
| 23 | |
| 24 | write output to nc1:"rttest/index-selection_fuzzy-inverted-index-ngram-edit-distance-panic.adm"; |
| 25 | |
| 26 | for $o in dataset('DBLP') |
| 27 | let $ed := edit-distance-check($o.authors, "Amihay Motro", 5) |
| 28 | where $ed[0] |
| 29 | return $o |