AddOne add one to last byte in string, on empty string return empty it helps with key iteration upper bound.

After get substring after sub string.

AfterN Split slices After(str) into all substrings separated by sep.

AppendFieldData appends fields data of specified index from src to dst.

BigEndianBytesToUint64 converts a byte slice (big endian) to uint64.

BytesToFloat32 converts a byte slice to float32.

BytesToInt64 converts a byte slice to uint64.

BytesToUint64 converts a byte slice to uint64.

Generic Clone for proto message.

CompareIndexParams compares indexParam1 with indexParam2.

ComparePK returns if i-th PK of dataA > j-th PK of dataB.

ComparePKInSlice returns if i-th PK < j-th PK.

CreateSchemaHelper returns a new SchemaHelper object.

accepted format: - {"indices": [1, 2, 3], "values": [0.1, 0.2, 0.3]} # format1 - {"1": 0.1, "2": 0.2, "3": 0.3} # format2 we don't require the indices to be sorted from user input, but the returned byte representation must have indices sorted.

DeleteFieldData delete fields data appended last time.

EstimateSizePerRecord returns the estimate size of a record in a collection.

placeholderGroup is a serialized PlaceholderGroup, return estimated total number of non-zero elements of all the sparse vectors in the placeholderGroup This is a rough estimate, and should be used only for statistics.

Float32ArrayToBFloat16Bytes converts float32 vector `fv` to bfloat16 vector.

Float32ArrayToBytes serialize vector into byte slice, used in search placeholder LittleEndian is used for convention.

Float32ArrayToFloat16Bytes converts float32 vector `fv` to float16 vector.

Float32ToBytes converts a float to byte slice.

GetDim get dimension of field.

GetDynamicField returns the dynamic field if it exists.

GetMapKeys return keys of a map.

GetPartitionKeyFieldSchema get partition field schema from collection schema.

GetPrimaryFieldData get primary field data from all field data inserted from sdk.

GetPrimaryFieldSchema get primary field schema from collection schema.

GetVectorFieldSchema get vector field schema from collection schema.

GetVectorFieldSchemas get vector fields schema from collection schema.

HasClusterKey check if a collection schema has ClusterKey field.

Hash32Bytes hashing a byte array to uint32.

Hash32Int64 hashing an int64 number to uint32.

Hash32String hashing a string to int64.

Hash32Uint64 hashing an uint64 nubmer to uint32.

HashKey2Partitions hash partition keys to partitions.

HashPK2Channels hash primary keys to channels.

HashString2LessUint32 hashing a string to uint32 but less than uint32 max.

HashString2Uint32 hashing a string to uint32.

HasPartitionKey check if a collection schema has PartitionKey field.

Int64ToBytes converts uint64 to a byte slice.

IsArithmetic returns true if input is of arithmetic type, otherwise false.

IsBoolType returns true if input is a bool type, otherwise false.

IsChanClosed returns whether input signal channel is closed or not.

IsFloatingType returns true if input is a floating type, otherwise false.

IsIntegerType returns true if input is an integer type, otherwise false.

IsStringType returns true if input is a varChar type, otherwise false.

IsVectorType returns true if input is a vector type, otherwise false.

MapEqual returns true if the two map contain the same keys and values.

MergeFieldData appends fields data to dst.

MergeMap merge one map to another.

NewArrayBasedMaximumHeap returns a new maximum heap.

NewArrayBasedMinimumHeap returns a new minimum heap.

NewBackoffTimer creates a new balanceTimer.

NewBackoffWithInstant creates a new backoff with instant.

NewGenericLifetime returns a new instance of Lifetime with init state and isHealthy logic.

NewHeap returns a new heap from a underlying representation.

NewIDAllocator creates a new IDAllocator.

NewLifetime returns a new instance of Lifetime with default state logic.

NewMultipartQueue create a new multi-part queue.

NewSharedReference creates a new SharedReference with the inner object.

NewSkipList creates a new SkipList with provided option.

ParseTimestamp returns a timestamp for a given byte slice.

PrepareResultFieldData construct this slice fo FieldData for final result reduce this shall preallocate the space for field data internal slice prevent slice growing cost.

this method returns a static sequence for partitions for partiton key mode.

SelectMinPK select the index of the minPK in results T of the cursors.

SliceRemoveDuplicate is used to dedup a Slice.

dim of a sparse float vector is the maximum/last index + 1.

SparseFloatRowUtils.

does not check for out-of-range access.

does not check for out-of-range access.

SubTimeByWallClock returns the duration between two different timestamps.

SwapPK swaps i-th PK with j-th PK.

Uint64ToBytes converts uint64 to a byte slice.

Uint64ToBytesBigEndian converts uint64 to a byte slice(big endian).

#nosec G103 */.

#nosec G103 */.

return VectorTypeSize for each dim (byte).

WithMaxLevel returns SkipListOption to setup max level.

WithSkip returns SkipListOption to setup skip.

IndexNodeRole is a constant represent IndexNode.

DataNodeRole is a constant represent DataNode.

EmbeddedRole is for embedded Milvus.

IndexCoordRole is a constant represent IndexCoord.

IndexNodeRole is a constant represent IndexNode.

MaxTimestamp is the max timestamp.

MixtureRole is a constant represents Mixture running modtoe.

ProxyRole is a constant represent Proxy.

QueryCoordRole is a constant represent QueryCoord.

QueryNodeRole is a constant represent QueryNode.

RootCoordRole is a constant represent RootCoord.

StandaloneRole is a constant represent Standalone.

StreamingCoord is a constant represent StreamingCoord.

StreamingNode is a constant represent StreamingNode.

ZeroTime is a zero time.

ZeroTimestamp is a zero timestamp.

BackoffConfig is the configuration for backoff.

BackoffTimer is a timer for balance operation.

BackoffTimerConfig is the configuration for backoff timer It's also used to be const config fetcher.

BackoffWithInstant is a backoff with instant.

GenericLifetime is a common component lifetime control logic.

IDAllocator is a thread-safe ID allocator.

MultipartQueue is a multi-part queue.

SchemaHelper provides methods to get the schema of fields.

SharedReference is a reference type that can be shared with only one close operation.

SkipList generic skip list.

VersionInt64Pair is a pair of int64 type version.

WeakReference is a weak reference type that can be shared with only one close operation.

BackoffTimerConfigFetcher is the interface to fetch backoff timer configuration.

Closable is an interface that can be closed or release the resource.

Heap is a heap of E.

HeapInterface is the interface that a heap must implement.

Version is a interface for version comparison.

IntPrimaryKey is an alias of int64.

SkipListOption setup function alias.

Timestamp is an alias of uint64.

UniqueID is an alias of int64.

VersionInt64 is a int64 type version.