schema

package
v0.0.0-...-77bcb31 Latest Latest
Warning

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

 Go to latest Published: Jun 14, 2019 License: MIT Imports: 15 Imported by: 0

Documentation

Overview

Package schema implements core Relational Algrebra schema objects such as Table, Schema, DataSource, Fields, Headers, Index.

Package schema is a generated protocol buffer package.

It is generated from these files: 

schema.proto

It has these top-level messages: 

TablePartition
Partition
TablePb
FieldPb
Index

Index

Constants

View Source 
const (
	// NoNulls defines if we allow nulls
	NoNulls = false
	// AllowNulls ?
	AllowNulls = true
)

Variables

View Source 
var (
	// ErrNotFound is error expressing sought item was not found.
	ErrNotFound = fmt.Errorf("Not Found")
	// ErrNotImplemented this feature is not implemented for this source.
	ErrNotImplemented = fmt.Errorf("Not Implemented")
)
View Source 
var (
	// SchemaRefreshInterval default schema Refresh Interval
	SchemaRefreshInterval = -time.Minute * 5

	// Static list of common field names for describe header on Show, Describe
	EngineFullCols       = []string{"Engine", "Support", "Comment", "Transactions", "XA", "Savepoints"}
	ProdedureFullCols    = []string{"Db", "Name", "Type", "Definer", "Modified", "Created", "Security_type", "Comment", "character_set_client ", "collation_connection", "Database Collation"}
	DescribeFullCols     = []string{"Field", "Type", "Collation", "Null", "Key", "Default", "Extra", "Privileges", "Comment"}
	DescribeFullColMap   = map[string]int{"Field": 0, "Type": 1, "Collation": 2, "Null": 3, "Key": 4, "Default": 5, "Extra": 6, "Privileges": 7, "Comment": 8}
	DescribeCols         = []string{"Field", "Type", "Null", "Key", "Default", "Extra"}
	DescribeColMap       = map[string]int{"Field": 0, "Type": 1, "Null": 2, "Key": 3, "Default": 4, "Extra": 5}
	ShowTableColumns     = []string{"Table", "Table_Type"}
	ShowVariablesColumns = []string{"Variable_name", "Value"}
	ShowDatabasesColumns = []string{"Database"}
	ShowTableColumnMap   = map[string]int{"Table": 0}
	ShowIndexCols        = []string{"Table", "Non_unique", "Key_name", "Seq_in_index", "Column_name", "Collation", "Cardinality", "Sub_part", "Packed", "Null", "Index_type", "Index_comment"}
	DescribeFullHeaders  = NewDescribeFullHeaders()
	DescribeHeaders      = NewDescribeHeaders()
)
View Source 
var (

	// DisableRecover If true, we will not capture/suppress panics.
	// Test only feature hopefully
	DisableRecover bool
)

Functions

func CreateDefaultRegistry 

func CreateDefaultRegistry(applyer Applyer)

CreateDefaultRegistry create the default registry.

func RegisterSchema 

func RegisterSchema(schema *Schema) error

RegisterSchema makes a named schema available by the provided @name If Register is called twice with the same name or if source is nil, it panics.

Sources are specific schemas of type csv, elasticsearch, etc containing multiple tables.

func RegisterSourceAsSchema 

func RegisterSourceAsSchema(name string, source Source) error

RegisterSourceAsSchema means you have a datasource, that is going to act as a named schema. ie, this will not be a nested schema with sub-schemas and the source will not be re-useable as a source-type.

func RegisterSourceType 

func RegisterSourceType(sourceType string, source Source)

RegisterSourceType makes a datasource type available by the provided @sourceType If Register is called twice with the same name or if source is nil, it panics.

Sources are specific schemas of type csv, elasticsearch, etc containing multiple tables.

Types

type Alter 

type Alter interface {
	// DropTable drop given table
	DropTable(table string) error
}

Alter interface for schema storage sources

type Applyer 

type Applyer interface {
	// Init initialize the applyer with registry.
	Init(r *Registry)
	// AddOrUpdateOnSchema Add or Update object (Table, Index)
	AddOrUpdateOnSchema(s *Schema, obj interface{}) error
	// Drop an object from schema
	Drop(s *Schema, obj interface{}) error
}

Applyer takes schema writes and applies them. This is used both as a database is being loaded, and schema is loaded by store as well as responsible for applying schema changes such as Alters. In distributed db's this is very, very huge part of work so is a very important interface that is under flux.

func NewApplyer 

func NewApplyer(sp SchemaSourceProvider) Applyer

NewApplyer new in memory applyer. For distributed db's we would need a different applyer (Raft).

type ConfigNode 

type ConfigNode struct {
	Name     string       `json:"name"`     // Name of this Node optional
	Source   string       `json:"source"`   // Name of source this node belongs to
	Address  string       `json:"address"`  // host/ip
	Settings u.JsonHelper `json:"settings"` // Arbitrary settings
}

ConfigNode are Servers/Services, ie a running instance of said Source - each must represent a single source type - normal use is a server, describing partitions of servers - may have arbitrary config info in Settings.

type ConfigSchema 

type ConfigSchema struct {
	Name       string   `json:"name"`    // Virtual Schema Name, must be unique
	Sources    []string `json:"sources"` // List of sources , the names of the "Db" in source
	ConfigNode []string `json:"-"`       // List of backend Servers
}

ConfigSchema is the json/config block for Schema, the data-sources that make up this Virtual Schema. Must have a name and list of sources to include.

type ConfigSource 

type ConfigSource struct {
	Name         string            `json:"name"`            // Name
	Schema       string            `json:"schema"`          // Schema Name if different than Name, will join existing schema
	SourceType   string            `json:"type"`            // [mysql,elasticsearch,csv,etc] Name in DataSource Registry
	TablesToLoad []string          `json:"tables_to_load"`  // if non empty, only load these tables
	TableAliases map[string]string `json:"table_aliases"`   // if non empty, only load these tables
	Nodes        []*ConfigNode     `json:"nodes"`           // List of nodes
	Hosts        []string          `json:"hosts"`           // List of hosts, replaces older "nodes"
	Settings     u.JsonHelper      `json:"settings"`        // Arbitrary settings specific to each source type
	Partitions   []*TablePartition `json:"partitions"`      // List of partitions per table (optional)
	PartitionCt  uint32            `json:"partition_count"` // Instead of array of per table partitions, raw partition count
}

ConfigSource are backend datasources ie : storage/database/csvfiles Each represents a single source type/config. May belong to more than one schema.

func NewSourceConfig 

func NewSourceConfig(name, sourceType string) *ConfigSource

func (*ConfigSource) String 

func (m *ConfigSource) String() string

type Conn 

type Conn interface {
	Close() error
}

Conn A Connection/Session to a file, api, backend database. Depending on the features of the backing source, it may optionally implement different portions of this interface.

Minimum Read Features to provide Sql Select

  • Scanning: iterate through messages/rows
  • Schema Tables: at a minium list of tables available, the column level data can be introspected so is optional

Planning:

  • CreateMutator(ctx *plan.Context) : execute a mutation task insert, delete, update

Non Select based Sql DML Operations for Mutator:

  • Deletion: (sql delete) Delete() DeleteExpression()
  • Upsert Interface (sql Update, Upsert, Insert) Put() PutMulti()

DataSource Connection Session that is Stateful. this is really a generic interface, will actually implement features below: SchemaColumns, Scanner, Seeker, Mutator

func OpenConn 

func OpenConn(schemaName, table string) (Conn, error)

OpenConn a schema-source Connection, Global open connection function using default schema registry.

type ConnAll 

type ConnAll interface {
	Close() error
	ConnColumns
	Iterator
	ConnSeeker
	ConnUpsert
	ConnDeletion
}

ConnAll interface describes the FULL set of features a connection can implement.

type ConnColumns 

type ConnColumns interface {
	Columns() []string
}

ConnColumns Interface for a data source connection exposing column positions for []driver.Value iteration

type ConnDeletion 

type ConnDeletion interface {
	// Delete using this key
	Delete(driver.Value) (int, error)
	// Delete with given expression
	DeleteExpression(p interface{}, n expr.Node) (int, error)
}

ConnDeletion deletion interface for data sources

type ConnMutation 

type ConnMutation interface {
	CreateMutator(pc interface{}) (ConnMutator, error)
}

ConnMutation creates a Mutator connection similar to Open() connection for select - accepts the plan context used in this upsert/insert/update - returns a connection which must be closed

type ConnMutator 

type ConnMutator interface {
	ConnUpsert
	ConnDeletion
}

ConnMutator Mutator Connection

type ConnPatchWhere 

type ConnPatchWhere interface {
	PatchWhere(ctx context.Context, where expr.Node, patch interface{}) (int64, error)
}

ConnPatchWhere pass through where expression to underlying datasource Used for update statements WHERE x = y

type ConnScanner 

type ConnScanner interface {
	Conn
	Iterator
}

ConnScanner is the primary basis for reading data sources. It exposes an interface to scan through rows. If the Source supports Predicate Push Down (ie, push the where/sql down to underlying store) this is just the resulting rows. Otherwise, Qlbridge engine must polyfill.

type ConnSeeker 

type ConnSeeker interface {
	Get(key driver.Value) (Message, error)
}

ConnSeeker is a conn that is Key-Value store, allows relational implementation to be faster for Seeking row values instead of scanning

type ConnUpsert 

type ConnUpsert interface {
	Put(ctx context.Context, key Key, value interface{}) (Key, error)
	PutMulti(ctx context.Context, keys []Key, src interface{}) ([]Key, error)
}

ConnUpsert Mutation interface for Put

  • assumes datasource understands key(s?)

type DialectWriter 

type DialectWriter interface {
	// Dialect ie "mysql", "postgres", "cassandra", "bigquery"
	Dialect() string
	Table(tbl *Table) string
	FieldType(t value.ValueType) string
}

DialectWriter knows how to format the schema output specific to a dialect such as postgres, mysql, bigquery all have different identity, value escape characters.

type Field 

type Field struct {
	FieldPb
	Context map[string]interface{} // During schema discovery of underlying source, may need to store additional info
	// contains filtered or unexported fields
}

Field Describes the column info, name, data type, defaults, index, null

  • dialects (mysql, mongo, cassandra) have their own descriptors for these, so this is generic meant to be converted to Frontend at runtime

func NewDescribeFullHeaders 

func NewDescribeFullHeaders() []*Field

func NewDescribeHeaders 

func NewDescribeHeaders() []*Field

func NewField 

func NewField(name string, valType value.ValueType, size int, allowNulls bool, defaultVal driver.Value, key, collation, description string) *Field

func NewFieldBase 

func NewFieldBase(name string, valType value.ValueType, size int, desc string) *Field

func (*Field) AddContext 

func (m *Field) AddContext(key string, value interface{})

func (*Field) AsRow 

func (m *Field) AsRow() []driver.Value

func (*Field) Body 

func (m *Field) Body() interface{}

func (*Field) Id 

func (m *Field) Id() uint64

func (*Field) String 

func (m *Field) String() string

func (*Field) ValueType 

func (m *Field) ValueType() value.ValueType

type FieldData 

type FieldData []byte

FieldData is the byte value of a "Described" field ready to write to the wire so we don't have to continually re-serialize it.

type FieldPb 

type FieldPb struct {
	Name        string   `protobuf:"bytes,1,opt,name=name" json:"name,omitempty"`
	Description string   `protobuf:"bytes,2,opt,name=description" json:"description,omitempty"`
	Key         string   `protobuf:"bytes,3,opt,name=key" json:"key,omitempty"`
	Extra       string   `protobuf:"bytes,4,opt,name=extra" json:"extra,omitempty"`
	Data        string   `protobuf:"bytes,5,opt,name=data" json:"data,omitempty"`
	Length      uint32   `protobuf:"varint,6,opt,name=length" json:"length,omitempty"`
	Type        uint32   `protobuf:"varint,7,opt,name=type" json:"type,omitempty"`
	NativeType  uint32   `protobuf:"varint,8,opt,name=nativeType" json:"nativeType,omitempty"`
	DefLength   uint64   `protobuf:"varint,9,opt,name=defLength" json:"defLength,omitempty"`
	DefVal      []byte   `protobuf:"bytes,11,opt,name=defVal,proto3" json:"defVal,omitempty"`
	Indexed     bool     `protobuf:"varint,13,opt,name=indexed" json:"indexed,omitempty"`
	NoNulls     bool     `protobuf:"varint,14,opt,name=noNulls" json:"noNulls,omitempty"`
	Collation   string   `protobuf:"bytes,15,opt,name=collation" json:"collation,omitempty"`
	Roles       []string `protobuf:"bytes,16,rep,name=roles" json:"roles,omitempty"`
	Indexes     []*Index `protobuf:"bytes,17,rep,name=indexes" json:"indexes,omitempty"`
	ContextJson []byte   `protobuf:"bytes,18,opt,name=contextJson,proto3" json:"contextJson,omitempty"`
}

func (*FieldPb) Descriptor 

func (*FieldPb) Descriptor() ([]byte, []int)

func (*FieldPb) GetCollation 

func (m *FieldPb) GetCollation() string

func (*FieldPb) GetContextJson 

func (m *FieldPb) GetContextJson() []byte

func (*FieldPb) GetData 

func (m *FieldPb) GetData() string

func (*FieldPb) GetDefLength 

func (m *FieldPb) GetDefLength() uint64

func (*FieldPb) GetDefVal 

func (m *FieldPb) GetDefVal() []byte

func (*FieldPb) GetDescription 

func (m *FieldPb) GetDescription() string

func (*FieldPb) GetExtra 

func (m *FieldPb) GetExtra() string

func (*FieldPb) GetIndexed 

func (m *FieldPb) GetIndexed() bool

func (*FieldPb) GetIndexes 

func (m *FieldPb) GetIndexes() []*Index

func (*FieldPb) GetKey 

func (m *FieldPb) GetKey() string

func (*FieldPb) GetLength 

func (m *FieldPb) GetLength() uint32

func (*FieldPb) GetName 

func (m *FieldPb) GetName() string

func (*FieldPb) GetNativeType 

func (m *FieldPb) GetNativeType() uint32

func (*FieldPb) GetNoNulls 

func (m *FieldPb) GetNoNulls() bool

func (*FieldPb) GetRoles 

func (m *FieldPb) GetRoles() []string

func (*FieldPb) GetType 

func (m *FieldPb) GetType() uint32

func (*FieldPb) ProtoMessage 

func (*FieldPb) ProtoMessage()

func (*FieldPb) Reset 

func (m *FieldPb) Reset()

func (*FieldPb) String 

func (m *FieldPb) String() string

type InMemApplyer 

type InMemApplyer struct {
	// contains filtered or unexported fields
}

InMemApplyer applies schema changes in memory. As changes to schema come in (such as ALTER statements, new tables, new databases) we need to apply them to the underlying schema.

func (*InMemApplyer) AddOrUpdateOnSchema 

func (m *InMemApplyer) AddOrUpdateOnSchema(s *Schema, v interface{}) error

AddOrUpdateOnSchema we have a schema change to apply. A schema change is a new table, index, or whole new schema being registered. We provide the first argument which is which schema it is being applied to (ie, add table x to schema y).

func (*InMemApplyer) Drop 

func (m *InMemApplyer) Drop(s *Schema, v interface{}) error

Drop we have a schema change to apply.

func (*InMemApplyer) Init 

func (m *InMemApplyer) Init(r *Registry)

Init store the registry as part of in-mem applyer which needs it.

type Index 

type Index struct {
	Name          string   `protobuf:"bytes,1,opt,name=name" json:"name,omitempty"`
	Fields        []string `protobuf:"bytes,2,rep,name=fields" json:"fields,omitempty"`
	PrimaryKey    bool     `protobuf:"varint,3,opt,name=primaryKey" json:"primaryKey,omitempty"`
	HashPartition []string `protobuf:"bytes,4,rep,name=hashPartition" json:"hashPartition,omitempty"`
	PartitionSize int32    `protobuf:"varint,5,opt,name=partitionSize" json:"partitionSize,omitempty"`
}

Index a description of how field(s) should be indexed for a table.

func (*Index) Descriptor 

func (*Index) Descriptor() ([]byte, []int)

func (*Index) GetFields 

func (m *Index) GetFields() []string

func (*Index) GetHashPartition 

func (m *Index) GetHashPartition() []string

func (*Index) GetName 

func (m *Index) GetName() string

func (*Index) GetPartitionSize 

func (m *Index) GetPartitionSize() int32

func (*Index) GetPrimaryKey 

func (m *Index) GetPrimaryKey() bool

func (*Index) ProtoMessage 

func (*Index) ProtoMessage()

func (*Index) Reset 

func (m *Index) Reset()

func (*Index) String 

func (m *Index) String() string

type Iterator 

type Iterator interface {
	// Next returns the next message.  If none remain, returns nil.
	Next() Message
}

Iterator is simple iterator for paging through a datastore Message(rows) to be used for scanning. Building block for Tasks that process part of a DAG of tasks to process data.

type Key 

type Key interface {
	Key() driver.Value
}

Key interface is the Unique Key identifying a row.

type KeyUint 

type KeyUint struct {
	ID uint64
}

KeyUint implements Key interface and is simple uint64 key

func NewKeyUint 

func NewKeyUint(key uint64) *KeyUint

NewKeyUint simple new uint64 key

func (*KeyUint) Key 

func (m *KeyUint) Key() driver.Value

Key is key interface

type Message 

type Message interface {
	Id() uint64
	Body() interface{}
}

Message is an interface to describe a Row being processed by query engine/vm or it is a message between distributed parts of the system. It provides a Id() method which can be used by consistent-hash algorithms for routing a message consistently to different processes/servers.

Body() returns interface allowing this to be generic structure for routing

see http://github.com/lytics/grid

type MessageValues 

type MessageValues interface {
	Values() []driver.Value
}

MessageValues describes a message with array of driver.Value.

type Partition 

type Partition struct {
	Id    string `protobuf:"bytes,1,opt,name=id" json:"id,omitempty"`
	Left  string `protobuf:"bytes,2,opt,name=left" json:"left,omitempty"`
	Right string `protobuf:"bytes,3,opt,name=right" json:"right,omitempty"`
}

Partition describes a range of data the left-key is contained in this partition the right key is not contained in this partition, in the next one

func (*Partition) Descriptor 

func (*Partition) Descriptor() ([]byte, []int)

func (*Partition) GetId 

func (m *Partition) GetId() string

func (*Partition) GetLeft 

func (m *Partition) GetLeft() string

func (*Partition) GetRight 

func (m *Partition) GetRight() string

func (*Partition) ProtoMessage 

func (*Partition) ProtoMessage()

func (*Partition) Reset 

func (m *Partition) Reset()

func (*Partition) String 

func (m *Partition) String() string

type Registry 

type Registry struct {
	// contains filtered or unexported fields
}

Registry is a global or namespace registry of datasources and schema. Datasources have a "sourcetype" and define somewhat the driver. Schemas are made up of one or more underlying source-types and have normal schema info about tables etc.

func DefaultRegistry 

func DefaultRegistry() *Registry

DefaultRegistry get access to the shared/global registry of all datasource implementations

func NewRegistry 

func NewRegistry(applyer Applyer) *Registry

NewRegistry create schema registry.

func (*Registry) GetSource 

func (m *Registry) GetSource(sourceType string) (Source, error)

GetSource Find a DataSource by SourceType

func (*Registry) Init 

func (m *Registry) Init()

Init pre-schema load call any sources that need pre-schema init

func (*Registry) Schema 

func (m *Registry) Schema(schemaName string) (*Schema, bool)

Schema Get schema for given name.

func (*Registry) SchemaAdd 

func (m *Registry) SchemaAdd(s *Schema) error

SchemaAdd Add a new Schema

func (*Registry) SchemaAddChild 

func (m *Registry) SchemaAddChild(name string, child *Schema) error

SchemaAddChild Add a new Child Schema

func (*Registry) SchemaAddFromConfig 

func (m *Registry) SchemaAddFromConfig(conf *ConfigSource) error

SchemaAddFromConfig means you have a Schema-Source you want to add

func (*Registry) SchemaDrop 

func (m *Registry) SchemaDrop(schema, name string, objectType lex.TokenType) error

SchemaDrop removes a schema

func (*Registry) SchemaRefresh 

func (m *Registry) SchemaRefresh(name string) error

SchemaRefresh means reload the schema from underlying store. Possibly requires introspection.

func (*Registry) Schemas 

func (m *Registry) Schemas() []string

Schemas returns a list of schema names

func (*Registry) String 

func (m *Registry) String() string

String describe contents of registry.

type Schema 

type Schema struct {
	Name       string        // Name of schema
	Conf       *ConfigSource // source configuration
	DS         Source        // This datasource Interface
	InfoSchema *Schema       // represent this Schema as sql schema like "information_schema"
	SchemaRef  *Schema       // IF this is infoschema, the schema it refers to
	// contains filtered or unexported fields
}

Schema is a "Virtual" Schema and may have multiple different backing sources. - Multiple DataSource(s) (each may be discrete source type such as mysql, elasticsearch, etc) - each schema supplies tables to the virtual table pool - each table name across schemas must be unique (or aliased)

func NewInfoSchema 

func NewInfoSchema(schemaName string, s *Schema) *Schema

NewInfoSchema create a new empty schema with given name.

func NewSchema 

func NewSchema(schemaName string) *Schema

NewSchema create a new empty schema with given name.

func NewSchemaSource 

func NewSchemaSource(schemaName string, ds Source) *Schema

NewSchemaSource create a new empty schema with given name and source.

func (*Schema) Current 

func (m *Schema) Current() bool

Current Is this schema up to date?

func (*Schema) OpenConn 

func (m *Schema) OpenConn(tableName string) (Conn, error)

OpenConn get a connection from this schema by table name.

func (*Schema) Schema 

func (m *Schema) Schema(schemaName string) (*Schema, error)

Schema Find a child Schema for given schema name,

func (*Schema) SchemaForTable 

func (m *Schema) SchemaForTable(tableName string) (*Schema, error)

SchemaForTable Find a Schema for given Table

func (*Schema) Since 

func (m *Schema) Since(dur time.Duration) bool

Since Is this schema object been refreshed within time window described by @dur time ago ?

func (*Schema) Table 

func (m *Schema) Table(tableIn string) (*Table, error)

Table gets Table definition for given table name

func (*Schema) Tables 

func (m *Schema) Tables() []string

Tables gets list of all tables for this schema.

type SchemaSourceProvider 

type SchemaSourceProvider func(s *Schema) Source

SchemaSourceProvider is factory for creating schema storage

type Source 

type Source interface {
	// Init provides opportunity for those sources that require/ no configuration and
	// introspect schema from their environment time to load pre-schema discovery
	Init()
	// Setup optional interface for getting the Schema injected during creation/starup.
	// Since the Source is a singleton, stateful manager,  it has a startup/shutdown process.
	Setup(*Schema) error
	// Close this source, ensure connections, underlying resources are closed.
	Close() error
	// Open create a connection (not thread safe) to this source.
	Open(source string) (Conn, error)
	// Tables is a list of table names provided by this source.
	Tables() []string
	// Table get table schema for given table name.
	Table(table string) (*Table, error)
}

Source is an interface describing a datasource such as a database, file, api, in-mem data etc. It is thread-safe, singleton, responsible for creating connections and exposing schema. It also exposes partition information optionally if a distributed source.

Sources are registered in a registry, to be dynamically created as schema demands.

Lifecycle: 

Init()
Setup()
// running ....  Open() , Table(name)  etc .....
Close()

type SourcePartitionable 

type SourcePartitionable interface {
	// Partitions list of partitions.
	Partitions() []*Partition
	PartitionSource(p *Partition) (Conn, error)
}

SourcePartitionable is an optional interface a source may implement that announces it (source) as partitionable into ranges for splitting reads, writes onto different nodes of a cluster.

Many databases's already have internal Partition schemas this allow's those to be exposed for use in our partitioning, so the query-planner can distributed work across nodes.

type SourceTableColumn 

type SourceTableColumn interface {
	// Underlying data type of column
	Column(col string) (value.ValueType, bool)
}

SourceTableColumn is a partial source that just provides access to Column schema info, used in Generators.

type SourceTableSchema 

type SourceTableSchema interface {
	Table(table string) (*Table, error)
}

SourceTableSchema Partial interface from Source to define just Table()

type Table 

type Table struct {
	TablePb
	Fields         []*Field               // List of Fields, in order
	Context        map[string]interface{} // During schema discovery of underlying source, may need to store additional info
	FieldPositions map[string]int         // Maps name of column to ordinal position in array of []driver.Value's
	FieldMap       map[string]*Field      // Map of Field-name -> Field
	Schema         *Schema                // The schema this is member of
	Source         Source                 // The source
	// contains filtered or unexported fields
}

Table represents traditional definition of Database Table. It belongs to a Schema and can be used to create a Datasource used to read this table.

func NewTable 

func NewTable(table string) *Table

NewTable create a new table for a schema.

func (*Table) AddContext 

func (m *Table) AddContext(key string, value interface{})

AddContext add key/value pairs to context (settings, metatadata).

func (*Table) AddField 

func (m *Table) AddField(fld *Field)

AddField register a new field

func (*Table) AddFieldType 

func (m *Table) AddFieldType(name string, valType value.ValueType)

AddFieldType describe and register a new column

func (*Table) AsRows 

func (m *Table) AsRows() [][]driver.Value

AsRows return all fields suiteable as list of values for Describe/Show statements.

func (*Table) Body 

func (m *Table) Body() interface{}

Body satisifies Message Interface

func (*Table) Column 

func (m *Table) Column(col string) (value.ValueType, bool)

Column get the Underlying data type.

func (*Table) Columns 

func (m *Table) Columns() []string

Columns list of all column names.

func (*Table) Current 

func (m *Table) Current() bool

Current Is this schema object current? ie, have we refreshed it from source since refresh interval.

func (*Table) FieldNamesPositions 

func (m *Table) FieldNamesPositions() map[string]int

FieldNamesPositions List of Field Names and ordinal position in Column list

func (*Table) FieldsAsMessages 

func (m *Table) FieldsAsMessages() []Message

FieldsAsMessages get list of all fields as interface Message used in schema as sql "describe table"

func (*Table) HasField 

func (m *Table) HasField(name string) bool

HasField does this table have given field/column?

func (*Table) Id 

func (m *Table) Id() uint64

Id satisifieds Message Interface

func (*Table) Marshal 

func (m *Table) Marshal() ([]byte, error)

func (*Table) SetColumns 

func (m *Table) SetColumns(cols []string)

SetColumns Explicityly set column names.

func (*Table) SetColumnsFromFields 

func (m *Table) SetColumnsFromFields()

SetColumnsFromFields Explicityly set column names from fields.

func (*Table) SetRefreshed 

func (m *Table) SetRefreshed()

SetRefreshed update the refreshed date to now.

func (*Table) SetRows 

func (m *Table) SetRows(rows [][]driver.Value)

SetRows set rows aka values for this table. Used for schema/testing.

func (*Table) Since 

func (m *Table) Since(dur time.Duration) bool

Since Is this schema object within time window described by @dur time ago ?

type TablePartition 

type TablePartition struct {
	Table      string       `protobuf:"bytes,1,opt,name=table" json:"table,omitempty"`
	Keys       []string     `protobuf:"bytes,2,rep,name=keys" json:"keys,omitempty"`
	Partitions []*Partition `protobuf:"bytes,3,rep,name=partitions" json:"partitions,omitempty"`
}

Partition describes a range of data (in a Table). left-key is contained in this partition right key is not contained in this partition, in the next partition. So any value >= left-key, and < right-key is contained herein.

func (*TablePartition) Descriptor 

func (*TablePartition) Descriptor() ([]byte, []int)

func (*TablePartition) GetKeys 

func (m *TablePartition) GetKeys() []string

func (*TablePartition) GetPartitions 

func (m *TablePartition) GetPartitions() []*Partition

func (*TablePartition) GetTable 

func (m *TablePartition) GetTable() string

func (*TablePartition) ProtoMessage 

func (*TablePartition) ProtoMessage()

func (*TablePartition) Reset 

func (m *TablePartition) Reset()

func (*TablePartition) String 

func (m *TablePartition) String() string

type TablePb 

type TablePb struct {
	// Name of table lowercased
	Name string `protobuf:"bytes,1,opt,name=name" json:"name,omitempty"`
	// Name of table (not lowercased)
	NameOriginal string `protobuf:"bytes,2,opt,name=nameOriginal" json:"nameOriginal,omitempty"`
	// some dbs are more hiearchical (table-column-family)
	Parent string `protobuf:"bytes,3,opt,name=parent" json:"parent,omitempty"`
	// Character set, default = utf8
	Charset uint32 `protobuf:"varint,4,opt,name=Charset" json:"Charset,omitempty"`
	// Partitions in this table, optional may be empty
	Partition *TablePartition `protobuf:"bytes,5,opt,name=partition" json:"partition,omitempty"`
	// Partition Count
	PartitionCt uint32 `protobuf:"varint,6,opt,name=PartitionCt" json:"PartitionCt,omitempty"`
	// List of indexes for this table
	Indexes []*Index `protobuf:"bytes,7,rep,name=indexes" json:"indexes,omitempty"`
	// context json bytes
	ContextJson []byte `protobuf:"bytes,8,opt,name=contextJson,proto3" json:"contextJson,omitempty"`
	// List of Fields, in order
	Fieldpbs []*FieldPb `protobuf:"bytes,9,rep,name=fieldpbs" json:"fieldpbs,omitempty"`
}

func (*TablePb) Descriptor 

func (*TablePb) Descriptor() ([]byte, []int)

func (*TablePb) GetCharset 

func (m *TablePb) GetCharset() uint32

func (*TablePb) GetContextJson 

func (m *TablePb) GetContextJson() []byte

func (*TablePb) GetFieldpbs 

func (m *TablePb) GetFieldpbs() []*FieldPb

func (*TablePb) GetIndexes 

func (m *TablePb) GetIndexes() []*Index

func (*TablePb) GetName 

func (m *TablePb) GetName() string

func (*TablePb) GetNameOriginal 

func (m *TablePb) GetNameOriginal() string

func (*TablePb) GetParent 

func (m *TablePb) GetParent() string

func (*TablePb) GetPartition 

func (m *TablePb) GetPartition() *TablePartition

func (*TablePb) GetPartitionCt 

func (m *TablePb) GetPartitionCt() uint32

func (*TablePb) ProtoMessage 

func (*TablePb) ProtoMessage()

func (*TablePb) Reset 

func (m *TablePb) Reset()

func (*TablePb) String 

func (m *TablePb) String() string

type TimeMessage 

type TimeMessage interface {
	Ts() time.Time
}

TimeMessage describes a message with a timestamp.

Source Files

View all Source files

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.