RethinkDB calls a WHERE clause “filter”, and you can write it in several forms.<\/p>\n
One is the “equals” form:<\/p>\n
\nr.db('performance')\n .table('query_timings')\n .filter({ 'Duration': 0 })\n<\/pre>\nIf you can actually use this form, you can put many filters in the one function call, although they are all equality comparisons, so they aren’t that useful:<\/p>\n
\nr.db('performance')\n .table('query_timings')\n .filter({ 'Duration': 0, 'Day': 0 })\n<\/pre>\nAccording to the documentation, you can use an index if you completely rewrite this form:<\/p>\n
\nr.table(\"performance\").index_create(\"Duration\").run(conn)\n\nr.table(\"users\")\n .get_all(\"0\", index=\"Duration\")\n<\/pre>\nThe more useful form of a filter is to just write a function. However, beware that you have to use functions like “ne” for not equals (and eq, add, etc), as they aren’t able to send Javascript operators across.<\/p>\n
\nr.db('performance')\n .table('query_timings')\n .filter(function(row) { return row('Duration').ne(null) })\n<\/pre>\nAlternately, if you absolutely must use normal Javascript syntax, or are writing an Orm, you can pass the Javascript as a string, which is eval’ed on the server side. Comparing this with alternate implementations would be an interesting possibility here, to test the performance of various implementations of the same query.<\/p>\n
\nr.db('performance')\n .table('query_timings')\n .filter(r.js('(function (row) { return row.Duration != null; })'))\n<\/pre>\nThey do offer one final shortcut method, which is to use “match”, which presumably returns a lamdba (note this uses the somewhat unexpected “r.row” syntax, so this is operating as some sort of static method):<\/p>\n
\nr.db('performance')\n .table('query_timings').filter(\n r.row['Query'].match(\"^test.*\")}\n )\n<\/pre>\nThis is likely better for performance than building you’re own regex in a function, as (hopefully) this is compiled prior to creating the lambda, rather than on each call to the function.<\/p>\n","protected":false},"excerpt":{"rendered":"
RethinkDB calls a WHERE clause “filter”, and you can write it in several forms. One is the “equals” form: r.db(‘performance’) .table(‘query_timings’) .filter({ ‘Duration’: 0 }) If you can actually use this form, you can put many filters in the one function call, although they are all equality comparisons, so they aren’t that useful: r.db(‘performance’) .table(‘query_timings’) … <\/p>\n