dataset

package
v2.0.2 Latest Latest
Warning

This package is not in the latest version of its module.

 Go to latest Published: Jun 24, 2026 License: Apache-2.0 Imports: 21 Imported by: 0

Documentation

Overview

Package dataset defines the interface for managing time seres objects and provides time range manipulation capabilities

Index

Constants

This section is empty.

Variables

This section is empty.

Functions

func MarshalDataSet 

func MarshalDataSet(ts timeseries.Timeseries, _ *timeseries.RequestOptions,
	_ int,
) ([]byte, error)

MarshalDataSet marshals the dataset into a msgpack-formatted byte slice

func PointsAreEqual 

func PointsAreEqual(p1, p2 Point) bool

Equal returns true if p and p2 are exactly equal

func UnmarshalDataSet 

func UnmarshalDataSet(b []byte, trq *timeseries.TimeRangeQuery) (timeseries.Timeseries, error)

UnmarshalDataSet unmarshals the dataset from a msgpack-formatted byte slice

Types

type DataSet 

type DataSet struct {
	// Status is the optional status indicator for the DataSet
	Status string `msg:"status"`
	// ExtentList is the list of Extents (time ranges) represented in the Results
	ExtentList timeseries.ExtentList `msg:"extent_list"`
	// Results is the list of type Result. Each Result represents information about a
	// different statement in the source query for this DataSet
	Results Results `msg:"results"`
	// UpdateLock is used to synchronize updates to the DataSet
	UpdateLock sync.Mutex `msg:"-"`
	// Error is a container for any DataSet-level Errors
	Error string `msg:"error"`
	// ErrorType describes the type for any DataSet-level Errors
	ErrorType string `msg:"errorType"`
	// Warnings is a container for any DataSet-level Warnings
	Warnings []string `msg:"warnings"`
	// TimeRangeQuery is the trq associated with the Timeseries
	TimeRangeQuery *timeseries.TimeRangeQuery `msg:"trq"`
	// VolatileExtents is the list extents in the dataset that should be refreshed
	// on the next request to the Origin
	VolatileExtentList timeseries.ExtentList `msg:"volatile_extents"`
	// Sorter is the DataSet's Sort function, which defaults to DefaultSort
	Sorter func() `msg:"-"`
	// Merger is the DataSet's Merge function, which defaults to DefaultMerge
	Merger func(sortPoints bool, ts ...timeseries.Timeseries) `msg:"-"`
	// SizeCropper is the DataSet's CropToSize function, which defaults to DefaultSizeCropper
	SizeCropper func(int, time.Time, timeseries.Extent) `msg:"-"`
	// RangeCropper is the DataSet's CropToRange function, which defaults to DefaultRangeCropper
	RangeCropper func(timeseries.Extent) `msg:"-"`
}

DataSet is the Common Time Series Format that Trickster uses to accelerate most of its supported TSDB backends DataSet conforms to the Timeseries interface

func (*DataSet) Clone 

func (ds *DataSet) Clone() timeseries.Timeseries

Clone returns a new, perfect copy of the DataSet

func (*DataSet) CropToRange 

func (ds *DataSet) CropToRange(e timeseries.Extent)

CropToRange reduces the DataSet down to timestamps contained within the provided Extents (inclusive). CropToRange assumes the base DataSet is already sorted, and will corrupt an unsorted DataSet

func (*DataSet) CropToSize 

func (ds *DataSet) CropToSize(sz int, t time.Time, lur timeseries.Extent)

CropToSize reduces the number of elements in the Timeseries to the provided count, by evicting elements using a least-recently-used methodology. The time parameter limits the upper extent to the provided time, in order to support backfill tolerance

func (*DataSet) CroppedClone 

func (ds *DataSet) CroppedClone(e timeseries.Extent) timeseries.Timeseries

CroppedClone returns a new, perfect copy of the DataSet, efficiently cropped to the provided Extent. CroppedClone assumes the DataSet is sorted.

func (*DataSet) DecodeMsg 

func (z *DataSet) DecodeMsg(dc *msgp.Reader) (err error)

DecodeMsg implements msgp.Decodable

func (*DataSet) DefaultMerger 

func (ds *DataSet) DefaultMerger(sortPoints bool, collection ...timeseries.Timeseries)

DefaultMerger is the default Merger function

func (*DataSet) DefaultRangeCropper 

func (ds *DataSet) DefaultRangeCropper(e timeseries.Extent)

DefaultRangeCropper is the default RangeCropper Function

func (*DataSet) DefaultSizeCropper 

func (ds *DataSet) DefaultSizeCropper(sz int, t time.Time, lur timeseries.Extent)

DefaultSizeCropper is the default SizeCropper Function

func (*DataSet) EncodeMsg 

func (z *DataSet) EncodeMsg(en *msgp.Writer) (err error)

EncodeMsg implements msgp.Encodable

func (*DataSet) Extents 

func (ds *DataSet) Extents() timeseries.ExtentList

Extents returns the DataSet's ExentList

func (*DataSet) FieldDefinitions 

func (ds *DataSet) FieldDefinitions() (timeseries.FieldDefinitions,
	timeseries.FieldDefinitions, timeseries.FieldDefinitions,
	timeseries.FieldDefinition,
)

FieldDefinitions returns a de-duped slice of Field Definitions (FDs) from all series in the DataSet. The return values are: All FDs, Tag FDs, Value FDs, Timestamp FD

func (*DataSet) FinalizeAvg added in v2.0.2 

func (ds *DataSet) FinalizeAvg(count int)

FinalizeAvg divides all point values by count, converting accumulated sums into averages.

func (*DataSet) FinalizeWeightedAvg added in v2.0.2 

func (ds *DataSet) FinalizeWeightedAvg(countDS *DataSet, pairingQueryStatement string)

FinalizeWeightedAvg divides each point value in ds (accumulated sums from a sum-rewritten query) by the corresponding count value from countDS (from a count-rewritten query), producing a true weighted arithmetic mean per series per timestamp. Both datasets must have been accumulated using MergeStrategySum across the same fanout backends.

pairingQueryStatement, when non-empty, is used only for series pairing: the same value must be passed for both sum and count DataSets so their rows align even when each SeriesHeader.QueryStatement differs (e.g. PromQL sum(...) vs count(...) rewrites of one logical avg(...)). It should be the original user query string (or another canonical statement shared by both sides). When empty, pairing uses SeriesHeader.CalculateHash() so distinct statements keep distinct identities (e.g. multiple Influx series that share tags but differ by query).

global_avg = sum_of_all_values / count_of_all_values

func (*DataSet) InjectTags 

func (ds *DataSet) InjectTags(tags Tags)

InjectTags injects the provided tags into all series in all results in the DataSet in an insert-or-update fashion

func (*DataSet) MarshalMsg 

func (z *DataSet) MarshalMsg(b []byte) (o []byte, err error)

MarshalMsg implements msgp.Marshaler

func (*DataSet) Merge 

func (ds *DataSet) Merge(sortPoints bool, collection ...timeseries.Timeseries)

Merge merges the provided Timeseries list into the base DataSet (in the order provided) and optionally sorts the merged DataSet This implementation ignores any Timeseries that are not of type *DataSet

func (*DataSet) MergeWithOpts added in v2.0.2 

func (ds *DataSet) MergeWithOpts(opts MergeOpts, collection ...timeseries.Timeseries)

MergeWithOpts merges the provided Timeseries list into the base DataSet honoring every field of opts (including ToleranceNanos for sub-step dedup).

func (*DataSet) MergeWithStrategy added in v2.0.2 

func (ds *DataSet) MergeWithStrategy(sortPoints bool, strategy int, collection ...timeseries.Timeseries)

MergeWithStrategy merges the provided Timeseries list into the base DataSet using the specified MergeStrategy for combining values from matching series.

func (*DataSet) MergeWithStrategyTolerant added in v2.0.2 

func (ds *DataSet) MergeWithStrategyTolerant(sortPoints bool, strategy int, toleranceNanos int64,
	collection ...timeseries.Timeseries,
)

MergeWithStrategyTolerant is the primitive-typed entrypoint used by the response/merge package to plumb tolerance through without importing the dataset package (which would form an import cycle).

func (*DataSet) Msgsize 

func (z *DataSet) Msgsize() (s int)

Msgsize returns an upper bound estimate of the number of bytes occupied by the serialized message

func (*DataSet) PointCount 

func (ds *DataSet) PointCount() int

PointCount returns the total number of Points across all Results and Series

func (*DataSet) SeriesCount 

func (ds *DataSet) SeriesCount() int

SeriesCount returns the count of all Series across all Results in the DataSet

func (*DataSet) SetExtents 

func (ds *DataSet) SetExtents(el timeseries.ExtentList)

SetExtents overwrites a DataSet's known extents with the provided extent list

func (*DataSet) SetTimeRangeQuery 

func (ds *DataSet) SetTimeRangeQuery(trq *timeseries.TimeRangeQuery)

SetTimeRangeQuery sets the TimeRangeQuery for the DataSet

func (*DataSet) SetVolatileExtents 

func (ds *DataSet) SetVolatileExtents(e timeseries.ExtentList)

SetVolatileExtents sets the list of time Extents in the dataset that should be re-fetched

func (*DataSet) Size 

func (ds *DataSet) Size() int64

Size returns the memory utilization in bytes of the DataSet

func (*DataSet) Sort 

func (ds *DataSet) Sort()

Sort sorts all Values in each Series chronologically by their timestamp Sorting is efficiently baked into DataSet.Merge(), therefore this interface function is unused unless overridden

func (*DataSet) Step 

func (ds *DataSet) Step() time.Duration

Step returns the step for the DataSet

func (*DataSet) StripTags added in v2.0.2 

func (ds *DataSet) StripTags(keys []string)

StripTags removes the specified tag keys from all series in all results in the DataSet and forces a hash recalculation so that series identity reflects the updated tags.

func (*DataSet) TimestampCount 

func (ds *DataSet) TimestampCount() int64

TimestampCount returns the count of unique timestamps across all series in the DataSet

func (*DataSet) UnmarshalMsg 

func (z *DataSet) UnmarshalMsg(bts []byte) (o []byte, err error)

UnmarshalMsg implements msgp.Unmarshaler

func (*DataSet) ValueCount 

func (ds *DataSet) ValueCount() int64

ValueCount returns the count of all values across all Series in the DataSet

func (*DataSet) VolatileExtents 

func (ds *DataSet) VolatileExtents() timeseries.ExtentList

VolatileExtents returns the list of time Extents in the dataset that should be re-fetched

type DataSets 

type DataSets []*DataSet

func (*DataSets) DecodeMsg 

func (z *DataSets) DecodeMsg(dc *msgp.Reader) (err error)

DecodeMsg implements msgp.Decodable

func (DataSets) EncodeMsg 

func (z DataSets) EncodeMsg(en *msgp.Writer) (err error)

EncodeMsg implements msgp.Encodable

func (DataSets) MarshalMsg 

func (z DataSets) MarshalMsg(b []byte) (o []byte, err error)

MarshalMsg implements msgp.Marshaler

func (DataSets) Msgsize 

func (z DataSets) Msgsize() (s int)

Msgsize returns an upper bound estimate of the number of bytes occupied by the serialized message

func (*DataSets) UnmarshalMsg 

func (z *DataSets) UnmarshalMsg(bts []byte) (o []byte, err error)

UnmarshalMsg implements msgp.Unmarshaler

type Hash 

type Hash uint64

Hash is a numeric value representing a calculated hash

func (*Hash) DecodeMsg 

func (z *Hash) DecodeMsg(dc *msgp.Reader) (err error)

DecodeMsg implements msgp.Decodable

func (Hash) EncodeMsg 

func (z Hash) EncodeMsg(en *msgp.Writer) (err error)

EncodeMsg implements msgp.Encodable

func (Hash) MarshalMsg 

func (z Hash) MarshalMsg(b []byte) (o []byte, err error)

MarshalMsg implements msgp.Marshaler

func (Hash) Msgsize 

func (z Hash) Msgsize() (s int)

Msgsize returns an upper bound estimate of the number of bytes occupied by the serialized message

func (*Hash) UnmarshalMsg 

func (z *Hash) UnmarshalMsg(bts []byte) (o []byte, err error)

UnmarshalMsg implements msgp.Unmarshaler

type Hashes 

type Hashes []Hash

Hashes is a slice of type Hash

func (*Hashes) DecodeMsg 

func (z *Hashes) DecodeMsg(dc *msgp.Reader) (err error)

DecodeMsg implements msgp.Decodable

func (Hashes) EncodeMsg 

func (z Hashes) EncodeMsg(en *msgp.Writer) (err error)

EncodeMsg implements msgp.Encodable

func (Hashes) MarshalMsg 

func (z Hashes) MarshalMsg(b []byte) (o []byte, err error)

MarshalMsg implements msgp.Marshaler

func (Hashes) Msgsize 

func (z Hashes) Msgsize() (s int)

Msgsize returns an upper bound estimate of the number of bytes occupied by the serialized message

func (*Hashes) UnmarshalMsg 

func (z *Hashes) UnmarshalMsg(bts []byte) (o []byte, err error)

UnmarshalMsg implements msgp.Unmarshaler

type MarshalWriter 

type MarshalWriter func(io.Writer, *DataSet, *timeseries.RequestOptions, int) error

MarshalWriter is a function that serializes the DataSet via an io.Writer

type Marshaler 

type Marshaler func(*DataSet, *timeseries.RequestOptions, int) ([]byte, error)

Marshaler is a function that serializes the provided DataSet into a byte slice

type MergeOpts added in v2.0.2 

type MergeOpts struct {
	SortPoints     bool
	Strategy       MergeStrategy
	ToleranceNanos int64
}

MergeOpts controls a single merge invocation across the dataset package. All fields have zero-value defaults that preserve historical behavior: SortPoints=false leaves the output unsorted, Strategy=MergeStrategyDedup applies last-value-wins, and ToleranceNanos=0 requires exact epoch matches for deduplication. ToleranceNanos is a nanosecond window for clustering near-duplicate samples from independent shards (Thanos-style); when >0 and Strategy is MergeStrategyDedup, adjacent (sorted) points whose epoch difference is <= ToleranceNanos collapse to a single survivor.

type MergeStrategy added in v2.0.2 

type MergeStrategy int

MergeStrategy defines how values from matching series (identical labels) are combined when merging time-series data from multiple backends. 

const (
	// MergeStrategyDedup is the default: for matching epochs, the last value wins.
	MergeStrategyDedup MergeStrategy = iota
	// MergeStrategySum sums values at matching epochs.
	MergeStrategySum
	// MergeStrategyAvg averages values at matching epochs.
	MergeStrategyAvg
	// MergeStrategyMin takes the minimum value at matching epochs.
	MergeStrategyMin
	// MergeStrategyMax takes the maximum value at matching epochs.
	MergeStrategyMax
	// MergeStrategyCount counts the number of values at matching epochs.
	MergeStrategyCount
)

func ParseMergeStrategy added in v2.0.2 

func ParseMergeStrategy(s string) (MergeStrategy, error)

ParseMergeStrategy converts a string to a MergeStrategy.

func (MergeStrategy) String added in v2.0.2 

func (ms MergeStrategy) String() string

String returns the string representation of a MergeStrategy.

type Point 

type Point struct {
	Epoch  epoch.Epoch `msg:"epoch"`
	Size   int         `msg:"size"`
	Values []any       `msg:"values"`
}

Point represents a timeseries data point

func (*Point) Clone 

func (p *Point) Clone() Point

Clone returns a perfect copy of the Point

func (*Point) DecodeMsg 

func (z *Point) DecodeMsg(dc *msgp.Reader) (err error)

DecodeMsg implements msgp.Decodable

func (*Point) EncodeMsg 

func (z *Point) EncodeMsg(en *msgp.Writer) (err error)

EncodeMsg implements msgp.Encodable

func (*Point) MarshalMsg 

func (z *Point) MarshalMsg(b []byte) (o []byte, err error)

MarshalMsg implements msgp.Marshaler

func (*Point) Msgsize 

func (z *Point) Msgsize() (s int)

Msgsize returns an upper bound estimate of the number of bytes occupied by the serialized message

func (*Point) UnmarshalMsg 

func (z *Point) UnmarshalMsg(bts []byte) (o []byte, err error)

UnmarshalMsg implements msgp.Unmarshaler

type Points 

type Points []Point

Points is a slice of type *Point

func MergePoints 

func MergePoints(p, p2 Points, sortPoints bool) Points

Merge returns a new Points slice of p and p2 merged together. If sort is true the new slice is sorted and dupe-killed before being returned

func MergePointsWithOpts added in v2.0.2 

func MergePointsWithOpts(p, p2 Points, opts MergeOpts) Points

MergePointsWithOpts is the MergeOpts-aware merge primitive. Tolerance > 0 is only honored when opts.Strategy == MergeStrategyDedup (see sortAndAggregateTolerant for the rationale).

func MergePointsWithStrategy added in v2.0.2 

func MergePointsWithStrategy(p, p2 Points, sortPoints bool, strategy MergeStrategy) Points

MergePointsWithStrategy merges two Points slices using the specified strategy. For MergeStrategyDedup, this behaves identically to MergePoints.

func (Points) Clone 

func (p Points) Clone() Points

Clone returns a perfect copy of the Points

func (Points) CloneRange 

func (p Points) CloneRange(start, end int) Points

CloneRange returns a perfect copy of the Points, cloning only the points in the provided index range (upper-bound exclusive)

func (*Points) DecodeMsg 

func (z *Points) DecodeMsg(dc *msgp.Reader) (err error)

DecodeMsg implements msgp.Decodable

func (Points) EncodeMsg 

func (z Points) EncodeMsg(en *msgp.Writer) (err error)

EncodeMsg implements msgp.Encodable

func (Points) Equal 

func (p Points) Equal(p2 Points) bool

Equal returns true if both slices are exactly equal

func (Points) Len 

func (p Points) Len() int

Len returns the length of a slice of time series data points

func (Points) Less 

func (p Points) Less(i, j int) bool

Less returns true if i comes before j

func (Points) MarshalMsg 

func (z Points) MarshalMsg(b []byte) (o []byte, err error)

MarshalMsg implements msgp.Marshaler

func (Points) Msgsize 

func (z Points) Msgsize() (s int)

Msgsize returns an upper bound estimate of the number of bytes occupied by the serialized message

func (Points) Size 

func (p Points) Size() int64

Size returns the memory utilization of the Points in bytes

func (Points) Swap 

func (p Points) Swap(i, j int)

Swap modifies a slice of time series data points by swapping the values in indexes i and j

func (*Points) UnmarshalMsg 

func (z *Points) UnmarshalMsg(bts []byte) (o []byte, err error)

UnmarshalMsg implements msgp.Unmarshaler

type Result 

type Result struct {
	// StatementID represents the ID of the statement for this result. This field may not be
	// used by all tsdb implementations
	StatementID int `msg:"statement_id"`
	// Error represents a statement-level error
	Error string `msg:"error"`
	// SeriesList is an ordered list of the Series in this result
	SeriesList SeriesList `msg:"series"`
	// Name is the name of the Result, used by some Providers that include
	// multiple results in the same response, in lieu of a StatementID
	Name string `msg:"name"`
}

Result represents the results of a single query statement in the DataSet

func (*Result) Clone 

func (r *Result) Clone() *Result

Clone returns an exact copy of the Result

func (*Result) DecodeMsg 

func (z *Result) DecodeMsg(dc *msgp.Reader) (err error)

DecodeMsg implements msgp.Decodable

func (*Result) EncodeMsg 

func (z *Result) EncodeMsg(en *msgp.Writer) (err error)

EncodeMsg implements msgp.Encodable

func (*Result) Hashes 

func (r *Result) Hashes() Hashes

Hashes returns the ordered list of Hashes for the SeriesList in the Result

func (*Result) MarshalMsg 

func (z *Result) MarshalMsg(b []byte) (o []byte, err error)

MarshalMsg implements msgp.Marshaler

func (*Result) Msgsize 

func (z *Result) Msgsize() (s int)

Msgsize returns an upper bound estimate of the number of bytes occupied by the serialized message

func (*Result) Size 

func (r *Result) Size() int64

Size returns the size of the Result in bytes

func (*Result) String 

func (r *Result) String() string

func (*Result) UnmarshalMsg 

func (z *Result) UnmarshalMsg(bts []byte) (o []byte, err error)

UnmarshalMsg implements msgp.Unmarshaler

type Results 

type Results []*Result

func (*Results) DecodeMsg 

func (z *Results) DecodeMsg(dc *msgp.Reader) (err error)

DecodeMsg implements msgp.Decodable

func (Results) EncodeMsg 

func (z Results) EncodeMsg(en *msgp.Writer) (err error)

EncodeMsg implements msgp.Encodable

func (Results) MarshalMsg 

func (z Results) MarshalMsg(b []byte) (o []byte, err error)

MarshalMsg implements msgp.Marshaler

func (Results) Msgsize 

func (z Results) Msgsize() (s int)

Msgsize returns an upper bound estimate of the number of bytes occupied by the serialized message

func (*Results) UnmarshalMsg 

func (z *Results) UnmarshalMsg(bts []byte) (o []byte, err error)

UnmarshalMsg implements msgp.Unmarshaler

type ResultsLookup 

type ResultsLookup map[int]*Result

ResultsLookup is a map of Results searchable by the Statement ID

type Series 

type Series struct {
	// Header is the Series Header describing the Series
	Header SeriesHeader `msg:"header"`
	// Points is the list of Points in the Series
	Points Points `msg:"points"`
	// PointSize is the memory utilization of the Points in bytes
	PointSize int64 `msg:"ps"`
}

Series represents a single timeseries in a Result

func (*Series) Clone 

func (s *Series) Clone() *Series

Clone returns a perfect, new copy of the Series

func (*Series) DecodeMsg 

func (z *Series) DecodeMsg(dc *msgp.Reader) (err error)

DecodeMsg implements msgp.Decodable

func (*Series) EncodeMsg 

func (z *Series) EncodeMsg(en *msgp.Writer) (err error)

EncodeMsg implements msgp.Encodable

func (*Series) MarshalMsg 

func (z *Series) MarshalMsg(b []byte) (o []byte, err error)

MarshalMsg implements msgp.Marshaler

func (*Series) Msgsize 

func (z *Series) Msgsize() (s int)

Msgsize returns an upper bound estimate of the number of bytes occupied by the serialized message

func (*Series) Size 

func (s *Series) Size() int64

Size returns the memory utilization of the Series in bytes

func (*Series) String 

func (s *Series) String() string

func (*Series) UnmarshalMsg 

func (z *Series) UnmarshalMsg(bts []byte) (o []byte, err error)

UnmarshalMsg implements msgp.Unmarshaler

type SeriesHeader 

type SeriesHeader struct {
	// Name is the name of the Series
	Name string `msg:"name"`
	// Tags is the map of tags associated with the Series. Each key will map to
	// a fd.Name in TagFieldsList, with values representing the specific tag
	// values for this Series.
	Tags Tags `msg:"tags"`
	// TimestampField is the Field Definitions for the timestamp field.
	// Optional and used by some providers.
	TimestampField timeseries.FieldDefinition `msg:"timestampField"`
	// TagFieldsList is the ordered list of tag-based Field Definitions in the
	// Series. Optional and used by some providers.
	TagFieldsList timeseries.FieldDefinitions `msg:"tagFields"`
	// ValueFieldsList is the ordered list of value-based Field Definitions in the Series.
	ValueFieldsList timeseries.FieldDefinitions `msg:"valueFields"`
	// UntrackedFieldsList is alist of Field Definitions in the Series whose row values are ignored.
	UntrackedFieldsList timeseries.FieldDefinitions `msg:"untrackedFields"`
	// QueryStatement is the original query to which this DataSet is associated
	QueryStatement string `msg:"query"`
	// Size is the memory utilization of the Header in bytes
	Size int `msg:"size"`
	// contains filtered or unexported fields
}

SeriesHeader is the header section of a series, and describes its shape, size, and attributes

func (*SeriesHeader) CalculateHash 

func (sh *SeriesHeader) CalculateHash(rehash ...bool) Hash

CalculateHash sums the FNV64a hash for the Header and stores it to the Hash member

func (*SeriesHeader) CalculateHashWithQueryStatement added in v2.0.2 

func (sh *SeriesHeader) CalculateHashWithQueryStatement(queryStatement string) Hash

CalculateHashWithQueryStatement returns the same hash CalculateHash would yield if SeriesHeader.QueryStatement were queryStatement. It does not read or update the header's cached hash, so callers can compare series across DataSets that intentionally use different stored statements (e.g. TSM sum/count rewrites of one logical avg query) without affecting normal CalculateHash behavior elsewhere.

func (*SeriesHeader) CalculateSize 

func (sh *SeriesHeader) CalculateSize() int

CalculateSize sets and returns the header size

func (*SeriesHeader) Clone 

func (sh *SeriesHeader) Clone() SeriesHeader

Clone returns a perfect, new copy of the SeriesHeader

func (*SeriesHeader) DecodeMsg 

func (z *SeriesHeader) DecodeMsg(dc *msgp.Reader) (err error)

DecodeMsg implements msgp.Decodable

func (*SeriesHeader) EncodeMsg 

func (z *SeriesHeader) EncodeMsg(en *msgp.Writer) (err error)

EncodeMsg implements msgp.Encodable

func (*SeriesHeader) FieldDefinitions 

func (sh *SeriesHeader) FieldDefinitions() timeseries.FieldDefinitions

FieldDefinitions returns all FieldDefinitions in the series ordered by OutputPosition

func (*SeriesHeader) MarshalMsg 

func (z *SeriesHeader) MarshalMsg(b []byte) (o []byte, err error)

MarshalMsg implements msgp.Marshaler

func (*SeriesHeader) Msgsize 

func (z *SeriesHeader) Msgsize() (s int)

Msgsize returns an upper bound estimate of the number of bytes occupied by the serialized message

func (*SeriesHeader) String 

func (sh *SeriesHeader) String() string

func (*SeriesHeader) UnmarshalMsg 

func (z *SeriesHeader) UnmarshalMsg(bts []byte) (o []byte, err error)

UnmarshalMsg implements msgp.Unmarshaler

type SeriesList 

type SeriesList []*Series

SeriesList is an ordered list of Series

func (SeriesList) Clone 

func (sl SeriesList) Clone() SeriesList

func (*SeriesList) DecodeMsg 

func (z *SeriesList) DecodeMsg(dc *msgp.Reader) (err error)

DecodeMsg implements msgp.Decodable

func (SeriesList) EncodeMsg 

func (z SeriesList) EncodeMsg(en *msgp.Writer) (err error)

EncodeMsg implements msgp.Encodable

func (SeriesList) EqualHeader 

func (sl SeriesList) EqualHeader(sl2 SeriesList) bool

EqualHeader returns true if the slice elements contain identical header values in the identical order.

func (SeriesList) MarshalMsg 

func (z SeriesList) MarshalMsg(b []byte) (o []byte, err error)

MarshalMsg implements msgp.Marshaler

func (SeriesList) Merge 

func (sl SeriesList) Merge(sl2 SeriesList, sortPoints bool) SeriesList

Merge merges sl2 into the subject SeriesList, using sl2's authoritative order to adaptively reorder the existing+merged list such that it best emulates the fully constituted series order as it would be served by the origin. Merge assumes that a *Series in both lists, having the identical header hash, are the same series and will merge sl2[i].Points into sl.Points

func (SeriesList) MergeWithOpts added in v2.0.2 

func (sl SeriesList) MergeWithOpts(sl2 SeriesList, opts MergeOpts) SeriesList

MergeWithOpts is the MergeOpts-aware variant of MergeWithStrategy. The dedup path with opts.ToleranceNanos > 0 collapses sub-step duplicates produced by independent shards sampling the same metric at slightly different timestamps.

func (SeriesList) MergeWithStrategy added in v2.0.2 

func (sl SeriesList) MergeWithStrategy(sl2 SeriesList, sortPoints bool, strategy MergeStrategy) SeriesList

MergeWithStrategy merges sl2 into the subject SeriesList using the specified MergeStrategy to combine values from series with identical headers. For MergeStrategyDedup, this behaves identically to Merge.

func (SeriesList) Msgsize 

func (z SeriesList) Msgsize() (s int)

Msgsize returns an upper bound estimate of the number of bytes occupied by the serialized message

func (SeriesList) SortByTags 

func (sl SeriesList) SortByTags()

func (SeriesList) SortPoints 

func (sl SeriesList) SortPoints()

func (SeriesList) String 

func (sl SeriesList) String() string

func (*SeriesList) UnmarshalMsg 

func (z *SeriesList) UnmarshalMsg(bts []byte) (o []byte, err error)

UnmarshalMsg implements msgp.Unmarshaler

type SeriesLookup 

type SeriesLookup map[SeriesLookupKey]*Series

SeriesLookup is a map of Series searchable by Series Header Hash

type SeriesLookupKey 

type SeriesLookupKey struct {
	StatementID int
	Hash        Hash
}

SeriesLookupKey is the key for a SeriesLookup, consisting of a Result.StatementID and a Series.Hash

func (*SeriesLookupKey) DecodeMsg 

func (z *SeriesLookupKey) DecodeMsg(dc *msgp.Reader) (err error)

DecodeMsg implements msgp.Decodable

func (SeriesLookupKey) EncodeMsg 

func (z SeriesLookupKey) EncodeMsg(en *msgp.Writer) (err error)

EncodeMsg implements msgp.Encodable

func (SeriesLookupKey) MarshalMsg 

func (z SeriesLookupKey) MarshalMsg(b []byte) (o []byte, err error)

MarshalMsg implements msgp.Marshaler

func (SeriesLookupKey) Msgsize 

func (z SeriesLookupKey) Msgsize() (s int)

Msgsize returns an upper bound estimate of the number of bytes occupied by the serialized message

func (*SeriesLookupKey) UnmarshalMsg 

func (z *SeriesLookupKey) UnmarshalMsg(bts []byte) (o []byte, err error)

UnmarshalMsg implements msgp.Unmarshaler

type Tags 

type Tags map[string]string

Tags is a key/value pair associated with a Series to scope the cardinality of the DataSet

func (Tags) Clone 

func (t Tags) Clone() Tags

Clone returns an exact copy of the Tags

func (*Tags) DecodeMsg 

func (z *Tags) DecodeMsg(dc *msgp.Reader) (err error)

DecodeMsg implements msgp.Decodable

func (Tags) EncodeMsg 

func (z Tags) EncodeMsg(en *msgp.Writer) (err error)

EncodeMsg implements msgp.Encodable

func (Tags) JSON 

func (t Tags) JSON() string

JSON returns a string representation of the Tags as a JSON object. Keys are emitted in sorted order. Keys and values are escaped via encoding/json since Prometheus label values are arbitrary UTF-8 (see data model spec) and the older string-concat form produced invalid JSON for any value containing `"` or `\`.

func (Tags) KVP 

func (t Tags) KVP() string

KVP returns a string representation of the Tags as "key"="value","key2"="value2"

func (Tags) Keys 

func (t Tags) Keys() []string

Keys returns a string-sorted list of the Tags's keys

func (Tags) MarshalMsg 

func (z Tags) MarshalMsg(b []byte) (o []byte, err error)

MarshalMsg implements msgp.Marshaler

func (Tags) Merge 

func (t Tags) Merge(t2 Tags)

Merge merges the provided tags into the subject tags, replacing any duplicate tag names

func (Tags) Msgsize 

func (z Tags) Msgsize() (s int)

Msgsize returns an upper bound estimate of the number of bytes occupied by the serialized message

func (Tags) Size 

func (t Tags) Size() int

Size returns the byte size of the Tags

func (Tags) String 

func (t Tags) String() string

String returns a string representation of the Tags

func (Tags) StringsWithSep 

func (t Tags) StringsWithSep(sep1, sep2 string) string

StringsWithSep returns a string representation of the Tags with the provided key/value separator

func (*Tags) UnmarshalMsg 

func (z *Tags) UnmarshalMsg(bts []byte) (o []byte, err error)

UnmarshalMsg implements msgp.Unmarshaler

Source Files

View all Source files

Directories

Path Synopsis

Jump to

Keyboard shortcuts

? : This menu
/ : Search site
f or F : Jump to
y or Y : Canonical URL
go.dev uses cookies from Google to deliver and enhance the quality of its services and to analyze traffic. Learn more.