Cursors

QueryCursor

Cursor defaults are access(1), batchSize(100), grant(null), maxTimeMs(2000), skipAcl(false).

access(level) batchSize(sz) close() count() expand(...props) forEach(fn) grant(level = null) hasNext() include(...props) isClosed() limit(count) locale(identifier) map(fn) maxTimeMS(ms) next() passthru() pathPrefix(path = null) pathRead(path, options) paths(...props) reduce(fn, memo) skip(count) skipAcl(bool = true) sort(object) stream(fnEach, fnMap) where(filter) toArray() toList()

return org.objects.account
  .find()
  .skipAcl()
  .grant(consts.accessLevels.read)
  .paths('_id', 'name.first')
  .sort({'name.first': 1})
  .limit(50)
  .toList() // use paging with hasMore

QueryCursor.access(level)

Sets the minimum context access level the caller needs to return results. To return only results to which the caller has update access, for example, call cursor.access(consts.accessLevels.update).

Arguments

  • level (Number)

Returns

this

QueryCursor.batchSize(sz)

Sets the internal cursor batch size for fetching results. Normally, the default is optimal but in cases where a large number of very small documents is being fetched, a larger batch size may be slightly more performant.

Arguments

  • sz (Number) The internal cursor batch size (between 1 and 1000).

Returns

this

QueryCursor.close()

Releases the cursor resources. It's not necessary to close cursors as this is done automatically when a script ends or is exhausted, but in some cases a script may need to free up a cursor mid-stream in order to open up another one.

Returns

this

QueryCursor.count()

Executes a count using the current match filter and cursor options.

Returns

Number the number of matched documents.

QueryCursor.expand(...props)

Sets the references that should be expanded in the returned documents. Reference properties can also be selectively retrieved using the paths option.

Arguments

  • ...props (String) Property paths to expand.

Returns

this

QueryCursor.forEach(fn)

Executes the cursor, calling fn with each document.

Arguments

  • fn (Function) a function to call for each element in the form of (document) => {}.

Returns

undefined

QueryCursor.grant(level = null)

Sets the property grant level for the operation.

Arguments

  • level (Number)

Returns

this

QueryCursor.hasNext()

Executes the cursor, returning true if next() can be safely called to fetch the next result.

Returns

bool

QueryCursor.include(...props)

Sets the optional properties to include in results. Optional properties can also be explicitly retrieved using the paths option. Some properties are marked as optional, meaning they are not included in results by default (e.g. certain List properties).

Arguments

  • ...props (String) Property paths to include (e.g. include('connections'));

Returns

this

QueryCursor.isClosed()

Returns true if a cursor has been opened and has been subsequently closed or exhausted.

Returns

Boolean

QueryCursor.limit(count)

Limits the number of query results to count. By default, the script toList() limit is 10, though cursors have no default limit.

Arguments

  • count (Number) The limit.

Returns

this

QueryCursor.locale(identifier)

Specifies the locale of the cursor for any localized strings.

Arguments

  • identifier (String) The locale identifier, which consists of a 2 or 3 letter language tag optionally followed by a region separated by _. For example, en, en_US, en_UK, etc.

Returns

this

QueryCursor.map(fn)

Exhausts the cursor, calling fn with each document, and returning an array of results. Because map() iterates through the cursor, the caller should set a sane limit().

Arguments

  • fn (Function) a function to call for each result in the form of (document) => { return document; }.

Returns

Object[]

QueryCursor.maxTimeMs(ms)

Set the maximum time a query is allowed to run.

Arguments

  • ms (Number) A number between 10 and 10000.

Returns

this

QueryCursor.next()

Returns the next result. Throws a RangeError is there are no more results. Use while ( cursor.hasNext() ) { cursor.next() } or for result of cursor loops to iterate cursors.

Returns

Object

QueryCursor.passthru()

Opens a cursor (without fetching any results) which can be returned from a route script and streamed directly through the api. When returned, the script is detached and the cursor remains open until it's exhausted, incurring no further script ops or serialization costs. This is ideal for returning query results that do not require any post-query in-script processing. - Without passthru: return org.objects.c_example.find().toList() (script->api->script->api->response) - With passthru: return org.objects.c_example.find().limit(10).passthru() (script->api->response)

Returns

Object

QueryCursor.pathPrefix(path = null)

Sets the query prefix for retrieving a nested List property. All cursor options are passed through the path down to the list (including grant and skipAcl). Compatible with next(), toArray(), and with passthru() (to retrieve a streamable cursor). For example, org.objects.c_grandparent.find().pathPrefix( '5a2c44bc8664e438e45387d5.c_parents.5a2c44f88664e438e45387d6.c_children' ).where( {c_prop: value} ) passes the where() option down through 2 list properties to the c_parents.c_children.

Arguments

  • path (String) Path to a list property.

Returns

Object

QueryCursor.pathRead(path, options)

Returns the first matching document single path base on the current cursor options.

Arguments

  • path (String) Property path to read (Reading through references and lists is supported. eg "/creator/name/first").

  • options (Object)

    • throwNotFound (Boolean) Defaults to true. If false, will return null if no match is found.

Returns

Object

QueryCursor.paths(...props)

Limits the result properties to ...props. Only properties specified here (and those include()/expand()) are included in the results. Expandable reference properties and optional properties can also be listed here instead, limiting the output of those properties. For example, cursor.paths('c_parent.owner.name') jumps through the c_parent reference and it's owner reference to return the name property. Path lookups through references requires appropriate access permissions.

Arguments

  • ...props (String) Property paths to return in results.

Returns

this

QueryCursor.reduce(fn, memo)

Exhausts the cursor, calling fn with memo for each document, and returning memo. Because reduce() iterates through the cursor, the caller should set a sane limit().

Arguments

  • fn (Function) a function to call for each result in the form of (memo, document) => { return memo; }.

  • memo (Object) an initial value for fn.

Returns

Object

QueryCursor.skip(count)

Skips count number of results. Please not that large offsets can be expensive and slow down your query. It's more efficient to use range queries.

Arguments

  • count (Number) The number of results to skip.

Returns

this

QueryCursor.skipAcl(bool = true)

Skips initial lookup access control checks. Use in conjunction with grant() to set property level access.

Arguments

  • bool (Boolean)

Returns

this

QueryCursor.sort(object)

Sets the sort options (e.g. {c_num_prescriptions: -1, c_fullname: 1});

Arguments

  • sort (Object) an object containing indexed, sortable properties to sort on and their direction (1, -1)

Returns

this

QueryCursor.stream(fnEach, fnMap)

Exhausts the cursor, setting the content type to application/json and writing directly to the Response as a list object. Values returned from the script after stream is called are ignored.

Arguments

  • fnEach (Function) optional function - (result, total) => { return false; } - called after each item is written. Returning false ends the stream early.

  • fnMap (Function) optional function - (result) => { return result; } - called to transform each result before output, the result of which is JSON.stringify()'d.

Returns

undefined

QueryCursor.where(filter)

Overwrites the initial filter passed in the constructor.

Arguments

  • filter (Object)

Returns

thisJavaScript

const skip = require('request').params.skip;

const cursor = org.objects.c_nums
  .find()
  .paths('_id')
  .skipAcl()
  .skip(skip)
  .grant(1)
  .limit(10000)

return cursor.passthru();

QueryCursor.toArray()

Exhausts the cursor and returns an array of results (not a list! see toList() ). Because toArray() iterates through the cursor, the caller should set a sane limit().

Returns

Object[]

QueryCursor.toList()

Returns a list object based on limit() (or the default limit of 10).

Returns

List ({object: 'list', data: [...], hasMore: true})

AggregationCursor

Cursor defaults are access(1), batchSize(100), grant(null), maxTimeMs(2000), skipAcl(false).

access(level) batchSize(sz) close() forEach(fn) grant(level = null) group(object) hasNext() isClosed() limit(count) map(fn) match(object) maxTimeMS(ms) next() passthru() pathPrefix(path = null) project(object) reduce(memo, fn) skip(count) skipAcl(bool = true) sort(object) stream(fnEach, fnMap) toArray() toList() unwind(prop)

Exhausting a db.AggregationCursor

const cursor = org.objects.account
  .aggregate()
  .skipAcl()
  .grant(consts.accessLevels.read)
  .sort({
    'name.first': 1
  })
  .project({
    _id: 1,
    'name.first': 1
  })
  .limit(50);

const first_names = []; 
for (let account of cursor) {
  first_names.push( account.name.first );
}
return first_names;

// OR return cursor.map(v => v.name.first)

Reading accounts using db.AggregationCursor and streaming

const cursor = org.objects.account
  .aggregate()
  .skipAcl() 
  .grant(consts.accessLevels.read)
  .sort({
    'name.first': 1
  })
  .project({
    _id: 1,
    first_name: 'name.first'
  })
  .limit(50)
  .stream(); // stream json directly to the response object.

AggregationCursor.access(level)

Sets the minimum context access level the caller needs to return results. To return only results to which the caller has update access, for example, call cursor.access(consts.accessLevels.update).

Arguments

  • level (Number)

Returns

this

AggregationCursor.batchSize(sz)

Sets the internal cursor batch size for fetching results. Normally, the default is optimal but in cases where a large number of very small documents is being fetched, a larger batch size may be slightly more performant.

Arguments

  • sz (Number) The internal cursor batch size (between 1 and 1000).

Returns

this

AggregationCursor.close()

Releases the cursor resources. It's not necessary to close cursors as this is done automatically when a script ends or is exhausted, but in some cases a script may need to free up a cursor mid-stream in order to open up another one.

Returns

this

AggregationCursor.forEach(fn)

Executes the cursor, calling fn with each document.

Arguments

  • fn (Function) a function to call for each element in the form of (document) => {}.

Returns

undefined

AggregationCursor.grant(level = null)

Sets the property grant level for the operation.

Arguments

  • level (Number)

Returns

this

AggregationCursor.group(object)

Adds a $group stage to the aggregation.

Arguments

  • object (Object)

Returns

this

AggregationCursor.hasNext()

Executes the cursor, returning true if next() can be safely called to fetch the next result.

Returns

bool

AggregationCursor.isClosed()

Returns true if a cursor has been opened and has been subsequently closed or exhausted.

Returns

Boolean

AggregationCursor.limit(count)

Adds a $limit stage to the aggregation. By default, the script toList() limit is 10, though cursors have no default limit.

Arguments

  • count (Number)

Returns

this

AggregationCursor.map(fn)

Exhausts the cursor, calling fn with each document, and returning an array of results. Because map() iterates through the cursor, the caller should set a sane limit().

Arguments

  • fn (Function) a function to call for each result in the form of (document) => { return document; }.

Returns

Object[]

AggregationCursor.match(object)

Adds a $match stage to the aggregation.

Arguments

  • object (Object)

Returns

this

AggregationCursor.maxTimeMs(ms)

Set the maximum time a query is allowed to run.

Arguments

  • ms (Number) A number between 10 and 10000.

Returns

this

AggregationCursor.next()

Returns the next result. Throws a RangeError is there are no more results. Use while ( cursor.hasNext() ) { cursor.next() } or for result of cursor loops to iterate cursors.

Returns

Object

AggregationCursor.passthru()

Opens a cursor (without fetching any results) which can be returned from a route script and streamed directly through the api. When returned, the script is detached and the cursor remains open until it's exhausted, incurring no further script ops or serialization costs. This is ideal for returning query results that do not require any post-query in-script processing. - Without passthru: return org.objects.c_example.aggregate().toList() (script->api->script->api->response) - With passthru: return org.objects.c_example.aggregate().limit(10).passthru() (script->api->response)

Returns

Object

AggregationCursor.pathPrefix(path = null)

Sets the pipeline prefix for retrieving a nested List property. All cursor options are passed through the path down to the list (including grant and skipAcl). Compatible with next(), toArray(), and with passthru() (to retrieve a streamable cursor). For example, org.objects.c_grandparent.aggregate().pathPrefix( '5a2c44bc8664e438e45387d5.c_parents.5a2c44f88664e438e45387d6.c_children' ).match( {c_prop: value} ) passes the match() option down through 2 list properties to the c_parents.c_children.

Arguments

  • path (String) Path to a list property.

Returns

Object

AggregationCursor.project(object)

Adds a $project stage to the aggregation.

Arguments

  • object (Object)

Returns

this

AggregationCursor.reduce(fn, memo)

Exhausts the cursor, calling fn with memo for each document, and returning memo. Because reduce() iterates through the cursor, the caller should set a sane limit().

Arguments

  • fn (Function) a function to call for each result in the form of (memo, document) => { return memo; }.

  • memo (Object) an initial value for fn.

Returns

Object

AggregationCursor.skip(count)

Skips count number of results. Please not that large offsets can be expensive and slow down your query. It's more efficient to use range queries.

Arguments

  • count (Number) The number of results to skip.

Returns

this

AggregationCursor.skipAcl(bool = true)

Skips initial lookup access control checks. Use in conjunction with grant() to set property level access.

Arguments

  • bool (Boolean)

Returns

this

AggregationCursor.sort(object)

Adds a $sort stage to the aggregation.

Arguments

  • object (Object)

Returns

this

AggregationCursor.stream(fnEach, fnMap)

Exhausts the cursor, setting the content type to application/json and writing directly to the Response as a list object. Values returned from the script after stream is called are ignored.

Arguments

  • fnEach (Function) optional function - (result, total) => { return false; } - called after each item is written. Returning false ends the stream early.

  • fnMap (Function) optional function - (result) => { return result; } - called to transform each result before output, the result of which is JSON.stringify()'d.

Returns

undefined

AggregationCursor.toArray()

Exhausts the cursor and returns an array of results (not a list! see toList() ). Because toArray() iterates through the cursor, the caller should set a sane limit().

Returns

Object[]

AggregationCursor.toList()

Returns a list object based on limit() (or the default limit of 10).

Returns

List ({object: 'list', data: [...], hasMore: true})

AggregationCursor.unwind(prop)

Adds an $unwind stage to the aggregation.

Arguments

  • prop (String) the property to unwind.

Returns

this

Last updated