A scuttlebot plugin for querying data. With [map-filter-reduce](https://github.com/dominictarr/map-filter-reduce) you can write pretty flexible queries, similar to SQL, but more javascripty.
npm install ssb-queryA scuttlebot plugin for querying data.
With map-filter-reduce you can write
pretty flexible queries, similar to SQL, but more javascripty.
ssb-query is just a thin layer of glue,
giving access to flumeview-query
with secure-scuttlebutt data.
ssb-query is included in the ssb-server distribution by default.
see plugins documentation
if you are running ssb-server, run the following queries from another tab on the same machine.
``
ssb-server query.read --query '{MFR_QUERY}' options...
notice the json is inside single quotes ''. this is necessary, because " part of JSON but is handled
specially on the command line.
see read api documentation for options
using ssb-client,
connect to a locally running sbot and call query.read, which returns a pull-stream
`
require('ssb-client')(function (err, sbot) {
if(err) throw err
pull(
sbot.query.read({
query: MFR_QUERY,
...other options
//limit: 10, reverse: true
}),
pull.collect(function (err, ary) {
console.log(ary)
})
)
})
perform a query. query is a map-filter-reduce query.limit,reverse,old and live are standard options supported by most ssb database stream apis,
see createLogStream
returns internal information about what index will be used my ssb-query.
the name comes from the SQL "EXPLAIN" command
output might look like this:
`
{
gte: [...], lte: [...] //if this is present, then the query will use an index.
scan: true | false, //if scan is true, the entire database will be examined. this means a very slow query
live, old, //wether to include new and old records
sync: false//include a {sync: true} message after the old records have finished.
}
see flumeview-query for more information.
[{
"$filter": {
value: {
content: {channel: "solarpunk"}
}
}
}]
$3
[{
"$filter": {
value: {
content: {root: "%"}
}
}
}]
$3
[{
"$filter": {
value: {
author: "@",
content: {
type: ""
}
}
}
}]
$3
most recently published channels, with timestamp and message count.
[
{"$filter": {"value": {"content":{ "channel": {"$is": "string"}, "type": "post"}}}},
{"$reduce": {
"channel": ["value", "content", "channel"],
"count": {"$count": true},
"timestamp": {"$max": ["value", "timestamp"]}
}},
{"$sort": [["timestamp"], ["count"]]}
]sample output:
{
"channel": "heropunch",
"count": 83,
"timestamp": 1537465741596
}
{
"channel": "walkaway",
"count": 43,
"timestamp": 1537471373721
}
{
"channel": "music",
"count": 635,
"timestamp": 1537474414933
}
$3
indexes make your queries much faster.
read about flumeview-query indexes
currently supported indexes:
var indexes = [
{key: 'log', value: ['timestamp']},
{key: 'clk', value: [['value', 'author'], ['value', 'sequence']] },
{key: 'typ', value: [['value', 'content', 'type'], ['timestamp']] },
{key: 'tya', value: [['value', 'content', 'type'], ['value', 'timestamp']] },
{key: 'cha', value: [['value', 'content', 'channel'], ['timestamp']] },
{key: 'aty', value: [['value', 'author'], ['value', 'content', 'type'], ['timestamp']]},
{key: 'ata', value: [['value', 'author'], ['value', 'content', 'type'], ['value', 'timestamp']]},
{key: 'art', value: [['value', 'content', 'root'], ['value', 'timestamp']]},
{key: 'lor', value: [['rts']]}
]
so, because the of index, a query for value.content.channel and timestamp` would be quick.MIT