consul/vendor/github.com/hashicorp/go-memdb/index.go

884 lines
24 KiB
Go
Raw Normal View History

package memdb
import (
"encoding/binary"
"encoding/hex"
"errors"
"fmt"
"reflect"
"strings"
)
New ACLs (#4791) This PR is almost a complete rewrite of the ACL system within Consul. It brings the features more in line with other HashiCorp products. Obviously there is quite a bit left to do here but most of it is related docs, testing and finishing the last few commands in the CLI. I will update the PR description and check off the todos as I finish them over the next few days/week. Description At a high level this PR is mainly to split ACL tokens from Policies and to split the concepts of Authorization from Identities. A lot of this PR is mostly just to support CRUD operations on ACLTokens and ACLPolicies. These in and of themselves are not particularly interesting. The bigger conceptual changes are in how tokens get resolved, how backwards compatibility is handled and the separation of policy from identity which could lead the way to allowing for alternative identity providers. On the surface and with a new cluster the ACL system will look very similar to that of Nomads. Both have tokens and policies. Both have local tokens. The ACL management APIs for both are very similar. I even ripped off Nomad's ACL bootstrap resetting procedure. There are a few key differences though. Nomad requires token and policy replication where Consul only requires policy replication with token replication being opt-in. In Consul local tokens only work with token replication being enabled though. All policies in Nomad are globally applicable. In Consul all policies are stored and replicated globally but can be scoped to a subset of the datacenters. This allows for more granular access management. Unlike Nomad, Consul has legacy baggage in the form of the original ACL system. The ramifications of this are: A server running the new system must still support other clients using the legacy system. A client running the new system must be able to use the legacy RPCs when the servers in its datacenter are running the legacy system. The primary ACL DC's servers running in legacy mode needs to be a gate that keeps everything else in the entire multi-DC cluster running in legacy mode. So not only does this PR implement the new ACL system but has a legacy mode built in for when the cluster isn't ready for new ACLs. Also detecting that new ACLs can be used is automatic and requires no configuration on the part of administrators. This process is detailed more in the "Transitioning from Legacy to New ACL Mode" section below.
2018-10-19 16:04:07 +00:00
// Indexer is an interface used for defining indexes. Indexes are used
// for efficient lookup of objects in a MemDB table. An Indexer must also
// implement one of SingleIndexer or MultiIndexer.
//
// Indexers are primarily responsible for returning the lookup key as
// a byte slice. The byte slice is the key data in the underlying data storage.
type Indexer interface {
New ACLs (#4791) This PR is almost a complete rewrite of the ACL system within Consul. It brings the features more in line with other HashiCorp products. Obviously there is quite a bit left to do here but most of it is related docs, testing and finishing the last few commands in the CLI. I will update the PR description and check off the todos as I finish them over the next few days/week. Description At a high level this PR is mainly to split ACL tokens from Policies and to split the concepts of Authorization from Identities. A lot of this PR is mostly just to support CRUD operations on ACLTokens and ACLPolicies. These in and of themselves are not particularly interesting. The bigger conceptual changes are in how tokens get resolved, how backwards compatibility is handled and the separation of policy from identity which could lead the way to allowing for alternative identity providers. On the surface and with a new cluster the ACL system will look very similar to that of Nomads. Both have tokens and policies. Both have local tokens. The ACL management APIs for both are very similar. I even ripped off Nomad's ACL bootstrap resetting procedure. There are a few key differences though. Nomad requires token and policy replication where Consul only requires policy replication with token replication being opt-in. In Consul local tokens only work with token replication being enabled though. All policies in Nomad are globally applicable. In Consul all policies are stored and replicated globally but can be scoped to a subset of the datacenters. This allows for more granular access management. Unlike Nomad, Consul has legacy baggage in the form of the original ACL system. The ramifications of this are: A server running the new system must still support other clients using the legacy system. A client running the new system must be able to use the legacy RPCs when the servers in its datacenter are running the legacy system. The primary ACL DC's servers running in legacy mode needs to be a gate that keeps everything else in the entire multi-DC cluster running in legacy mode. So not only does this PR implement the new ACL system but has a legacy mode built in for when the cluster isn't ready for new ACLs. Also detecting that new ACLs can be used is automatic and requires no configuration on the part of administrators. This process is detailed more in the "Transitioning from Legacy to New ACL Mode" section below.
2018-10-19 16:04:07 +00:00
// FromArgs is called to build the exact index key from a list of arguments.
2017-01-09 19:23:09 +00:00
FromArgs(args ...interface{}) ([]byte, error)
}
New ACLs (#4791) This PR is almost a complete rewrite of the ACL system within Consul. It brings the features more in line with other HashiCorp products. Obviously there is quite a bit left to do here but most of it is related docs, testing and finishing the last few commands in the CLI. I will update the PR description and check off the todos as I finish them over the next few days/week. Description At a high level this PR is mainly to split ACL tokens from Policies and to split the concepts of Authorization from Identities. A lot of this PR is mostly just to support CRUD operations on ACLTokens and ACLPolicies. These in and of themselves are not particularly interesting. The bigger conceptual changes are in how tokens get resolved, how backwards compatibility is handled and the separation of policy from identity which could lead the way to allowing for alternative identity providers. On the surface and with a new cluster the ACL system will look very similar to that of Nomads. Both have tokens and policies. Both have local tokens. The ACL management APIs for both are very similar. I even ripped off Nomad's ACL bootstrap resetting procedure. There are a few key differences though. Nomad requires token and policy replication where Consul only requires policy replication with token replication being opt-in. In Consul local tokens only work with token replication being enabled though. All policies in Nomad are globally applicable. In Consul all policies are stored and replicated globally but can be scoped to a subset of the datacenters. This allows for more granular access management. Unlike Nomad, Consul has legacy baggage in the form of the original ACL system. The ramifications of this are: A server running the new system must still support other clients using the legacy system. A client running the new system must be able to use the legacy RPCs when the servers in its datacenter are running the legacy system. The primary ACL DC's servers running in legacy mode needs to be a gate that keeps everything else in the entire multi-DC cluster running in legacy mode. So not only does this PR implement the new ACL system but has a legacy mode built in for when the cluster isn't ready for new ACLs. Also detecting that new ACLs can be used is automatic and requires no configuration on the part of administrators. This process is detailed more in the "Transitioning from Legacy to New ACL Mode" section below.
2018-10-19 16:04:07 +00:00
// SingleIndexer is an interface used for defining indexes that generate a
// single value per object
2017-01-09 19:23:09 +00:00
type SingleIndexer interface {
New ACLs (#4791) This PR is almost a complete rewrite of the ACL system within Consul. It brings the features more in line with other HashiCorp products. Obviously there is quite a bit left to do here but most of it is related docs, testing and finishing the last few commands in the CLI. I will update the PR description and check off the todos as I finish them over the next few days/week. Description At a high level this PR is mainly to split ACL tokens from Policies and to split the concepts of Authorization from Identities. A lot of this PR is mostly just to support CRUD operations on ACLTokens and ACLPolicies. These in and of themselves are not particularly interesting. The bigger conceptual changes are in how tokens get resolved, how backwards compatibility is handled and the separation of policy from identity which could lead the way to allowing for alternative identity providers. On the surface and with a new cluster the ACL system will look very similar to that of Nomads. Both have tokens and policies. Both have local tokens. The ACL management APIs for both are very similar. I even ripped off Nomad's ACL bootstrap resetting procedure. There are a few key differences though. Nomad requires token and policy replication where Consul only requires policy replication with token replication being opt-in. In Consul local tokens only work with token replication being enabled though. All policies in Nomad are globally applicable. In Consul all policies are stored and replicated globally but can be scoped to a subset of the datacenters. This allows for more granular access management. Unlike Nomad, Consul has legacy baggage in the form of the original ACL system. The ramifications of this are: A server running the new system must still support other clients using the legacy system. A client running the new system must be able to use the legacy RPCs when the servers in its datacenter are running the legacy system. The primary ACL DC's servers running in legacy mode needs to be a gate that keeps everything else in the entire multi-DC cluster running in legacy mode. So not only does this PR implement the new ACL system but has a legacy mode built in for when the cluster isn't ready for new ACLs. Also detecting that new ACLs can be used is automatic and requires no configuration on the part of administrators. This process is detailed more in the "Transitioning from Legacy to New ACL Mode" section below.
2018-10-19 16:04:07 +00:00
// FromObject extracts the index value from an object. The return values
// are whether the index value was found, the index value, and any error
// while extracting the index value, respectively.
FromObject(raw interface{}) (bool, []byte, error)
2017-01-09 19:23:09 +00:00
}
New ACLs (#4791) This PR is almost a complete rewrite of the ACL system within Consul. It brings the features more in line with other HashiCorp products. Obviously there is quite a bit left to do here but most of it is related docs, testing and finishing the last few commands in the CLI. I will update the PR description and check off the todos as I finish them over the next few days/week. Description At a high level this PR is mainly to split ACL tokens from Policies and to split the concepts of Authorization from Identities. A lot of this PR is mostly just to support CRUD operations on ACLTokens and ACLPolicies. These in and of themselves are not particularly interesting. The bigger conceptual changes are in how tokens get resolved, how backwards compatibility is handled and the separation of policy from identity which could lead the way to allowing for alternative identity providers. On the surface and with a new cluster the ACL system will look very similar to that of Nomads. Both have tokens and policies. Both have local tokens. The ACL management APIs for both are very similar. I even ripped off Nomad's ACL bootstrap resetting procedure. There are a few key differences though. Nomad requires token and policy replication where Consul only requires policy replication with token replication being opt-in. In Consul local tokens only work with token replication being enabled though. All policies in Nomad are globally applicable. In Consul all policies are stored and replicated globally but can be scoped to a subset of the datacenters. This allows for more granular access management. Unlike Nomad, Consul has legacy baggage in the form of the original ACL system. The ramifications of this are: A server running the new system must still support other clients using the legacy system. A client running the new system must be able to use the legacy RPCs when the servers in its datacenter are running the legacy system. The primary ACL DC's servers running in legacy mode needs to be a gate that keeps everything else in the entire multi-DC cluster running in legacy mode. So not only does this PR implement the new ACL system but has a legacy mode built in for when the cluster isn't ready for new ACLs. Also detecting that new ACLs can be used is automatic and requires no configuration on the part of administrators. This process is detailed more in the "Transitioning from Legacy to New ACL Mode" section below.
2018-10-19 16:04:07 +00:00
// MultiIndexer is an interface used for defining indexes that generate
// multiple values per object. Each value is stored as a seperate index
// pointing to the same object.
//
// For example, an index that extracts the first and last name of a person
// and allows lookup based on eitherd would be a MultiIndexer. The FromObject
// of this example would split the first and last name and return both as
// values.
2017-01-09 19:23:09 +00:00
type MultiIndexer interface {
New ACLs (#4791) This PR is almost a complete rewrite of the ACL system within Consul. It brings the features more in line with other HashiCorp products. Obviously there is quite a bit left to do here but most of it is related docs, testing and finishing the last few commands in the CLI. I will update the PR description and check off the todos as I finish them over the next few days/week. Description At a high level this PR is mainly to split ACL tokens from Policies and to split the concepts of Authorization from Identities. A lot of this PR is mostly just to support CRUD operations on ACLTokens and ACLPolicies. These in and of themselves are not particularly interesting. The bigger conceptual changes are in how tokens get resolved, how backwards compatibility is handled and the separation of policy from identity which could lead the way to allowing for alternative identity providers. On the surface and with a new cluster the ACL system will look very similar to that of Nomads. Both have tokens and policies. Both have local tokens. The ACL management APIs for both are very similar. I even ripped off Nomad's ACL bootstrap resetting procedure. There are a few key differences though. Nomad requires token and policy replication where Consul only requires policy replication with token replication being opt-in. In Consul local tokens only work with token replication being enabled though. All policies in Nomad are globally applicable. In Consul all policies are stored and replicated globally but can be scoped to a subset of the datacenters. This allows for more granular access management. Unlike Nomad, Consul has legacy baggage in the form of the original ACL system. The ramifications of this are: A server running the new system must still support other clients using the legacy system. A client running the new system must be able to use the legacy RPCs when the servers in its datacenter are running the legacy system. The primary ACL DC's servers running in legacy mode needs to be a gate that keeps everything else in the entire multi-DC cluster running in legacy mode. So not only does this PR implement the new ACL system but has a legacy mode built in for when the cluster isn't ready for new ACLs. Also detecting that new ACLs can be used is automatic and requires no configuration on the part of administrators. This process is detailed more in the "Transitioning from Legacy to New ACL Mode" section below.
2018-10-19 16:04:07 +00:00
// FromObject extracts index values from an object. The return values
// are the same as a SingleIndexer except there can be multiple index
// values.
2017-01-09 19:23:09 +00:00
FromObject(raw interface{}) (bool, [][]byte, error)
}
New ACLs (#4791) This PR is almost a complete rewrite of the ACL system within Consul. It brings the features more in line with other HashiCorp products. Obviously there is quite a bit left to do here but most of it is related docs, testing and finishing the last few commands in the CLI. I will update the PR description and check off the todos as I finish them over the next few days/week. Description At a high level this PR is mainly to split ACL tokens from Policies and to split the concepts of Authorization from Identities. A lot of this PR is mostly just to support CRUD operations on ACLTokens and ACLPolicies. These in and of themselves are not particularly interesting. The bigger conceptual changes are in how tokens get resolved, how backwards compatibility is handled and the separation of policy from identity which could lead the way to allowing for alternative identity providers. On the surface and with a new cluster the ACL system will look very similar to that of Nomads. Both have tokens and policies. Both have local tokens. The ACL management APIs for both are very similar. I even ripped off Nomad's ACL bootstrap resetting procedure. There are a few key differences though. Nomad requires token and policy replication where Consul only requires policy replication with token replication being opt-in. In Consul local tokens only work with token replication being enabled though. All policies in Nomad are globally applicable. In Consul all policies are stored and replicated globally but can be scoped to a subset of the datacenters. This allows for more granular access management. Unlike Nomad, Consul has legacy baggage in the form of the original ACL system. The ramifications of this are: A server running the new system must still support other clients using the legacy system. A client running the new system must be able to use the legacy RPCs when the servers in its datacenter are running the legacy system. The primary ACL DC's servers running in legacy mode needs to be a gate that keeps everything else in the entire multi-DC cluster running in legacy mode. So not only does this PR implement the new ACL system but has a legacy mode built in for when the cluster isn't ready for new ACLs. Also detecting that new ACLs can be used is automatic and requires no configuration on the part of administrators. This process is detailed more in the "Transitioning from Legacy to New ACL Mode" section below.
2018-10-19 16:04:07 +00:00
// PrefixIndexer is an optional interface on top of an Indexer that allows
// indexes to support prefix-based iteration.
type PrefixIndexer interface {
New ACLs (#4791) This PR is almost a complete rewrite of the ACL system within Consul. It brings the features more in line with other HashiCorp products. Obviously there is quite a bit left to do here but most of it is related docs, testing and finishing the last few commands in the CLI. I will update the PR description and check off the todos as I finish them over the next few days/week. Description At a high level this PR is mainly to split ACL tokens from Policies and to split the concepts of Authorization from Identities. A lot of this PR is mostly just to support CRUD operations on ACLTokens and ACLPolicies. These in and of themselves are not particularly interesting. The bigger conceptual changes are in how tokens get resolved, how backwards compatibility is handled and the separation of policy from identity which could lead the way to allowing for alternative identity providers. On the surface and with a new cluster the ACL system will look very similar to that of Nomads. Both have tokens and policies. Both have local tokens. The ACL management APIs for both are very similar. I even ripped off Nomad's ACL bootstrap resetting procedure. There are a few key differences though. Nomad requires token and policy replication where Consul only requires policy replication with token replication being opt-in. In Consul local tokens only work with token replication being enabled though. All policies in Nomad are globally applicable. In Consul all policies are stored and replicated globally but can be scoped to a subset of the datacenters. This allows for more granular access management. Unlike Nomad, Consul has legacy baggage in the form of the original ACL system. The ramifications of this are: A server running the new system must still support other clients using the legacy system. A client running the new system must be able to use the legacy RPCs when the servers in its datacenter are running the legacy system. The primary ACL DC's servers running in legacy mode needs to be a gate that keeps everything else in the entire multi-DC cluster running in legacy mode. So not only does this PR implement the new ACL system but has a legacy mode built in for when the cluster isn't ready for new ACLs. Also detecting that new ACLs can be used is automatic and requires no configuration on the part of administrators. This process is detailed more in the "Transitioning from Legacy to New ACL Mode" section below.
2018-10-19 16:04:07 +00:00
// PrefixFromArgs is the same as FromArgs for an Indexer except that
// the index value returned should return all prefix-matched values.
PrefixFromArgs(args ...interface{}) ([]byte, error)
}
// StringFieldIndex is used to extract a field from an object
// using reflection and builds an index on that field.
type StringFieldIndex struct {
Field string
Lowercase bool
}
func (s *StringFieldIndex) FromObject(obj interface{}) (bool, []byte, error) {
v := reflect.ValueOf(obj)
v = reflect.Indirect(v) // Dereference the pointer if any
fv := v.FieldByName(s.Field)
isPtr := fv.Kind() == reflect.Ptr
fv = reflect.Indirect(fv)
if !isPtr && !fv.IsValid() {
return false, nil,
fmt.Errorf("field '%s' for %#v is invalid %v ", s.Field, obj, isPtr)
}
if isPtr && !fv.IsValid() {
val := ""
return false, []byte(val), nil
}
val := fv.String()
if val == "" {
return false, nil, nil
}
if s.Lowercase {
val = strings.ToLower(val)
}
// Add the null character as a terminator
val += "\x00"
return true, []byte(val), nil
}
func (s *StringFieldIndex) FromArgs(args ...interface{}) ([]byte, error) {
if len(args) != 1 {
return nil, fmt.Errorf("must provide only a single argument")
}
arg, ok := args[0].(string)
if !ok {
return nil, fmt.Errorf("argument must be a string: %#v", args[0])
}
if s.Lowercase {
arg = strings.ToLower(arg)
}
// Add the null character as a terminator
arg += "\x00"
return []byte(arg), nil
}
func (s *StringFieldIndex) PrefixFromArgs(args ...interface{}) ([]byte, error) {
val, err := s.FromArgs(args...)
if err != nil {
return nil, err
}
// Strip the null terminator, the rest is a prefix
n := len(val)
if n > 0 {
return val[:n-1], nil
}
return val, nil
}
New ACLs (#4791) This PR is almost a complete rewrite of the ACL system within Consul. It brings the features more in line with other HashiCorp products. Obviously there is quite a bit left to do here but most of it is related docs, testing and finishing the last few commands in the CLI. I will update the PR description and check off the todos as I finish them over the next few days/week. Description At a high level this PR is mainly to split ACL tokens from Policies and to split the concepts of Authorization from Identities. A lot of this PR is mostly just to support CRUD operations on ACLTokens and ACLPolicies. These in and of themselves are not particularly interesting. The bigger conceptual changes are in how tokens get resolved, how backwards compatibility is handled and the separation of policy from identity which could lead the way to allowing for alternative identity providers. On the surface and with a new cluster the ACL system will look very similar to that of Nomads. Both have tokens and policies. Both have local tokens. The ACL management APIs for both are very similar. I even ripped off Nomad's ACL bootstrap resetting procedure. There are a few key differences though. Nomad requires token and policy replication where Consul only requires policy replication with token replication being opt-in. In Consul local tokens only work with token replication being enabled though. All policies in Nomad are globally applicable. In Consul all policies are stored and replicated globally but can be scoped to a subset of the datacenters. This allows for more granular access management. Unlike Nomad, Consul has legacy baggage in the form of the original ACL system. The ramifications of this are: A server running the new system must still support other clients using the legacy system. A client running the new system must be able to use the legacy RPCs when the servers in its datacenter are running the legacy system. The primary ACL DC's servers running in legacy mode needs to be a gate that keeps everything else in the entire multi-DC cluster running in legacy mode. So not only does this PR implement the new ACL system but has a legacy mode built in for when the cluster isn't ready for new ACLs. Also detecting that new ACLs can be used is automatic and requires no configuration on the part of administrators. This process is detailed more in the "Transitioning from Legacy to New ACL Mode" section below.
2018-10-19 16:04:07 +00:00
// StringSliceFieldIndex builds an index from a field on an object that is a
// string slice ([]string). Each value within the string slice can be used for
// lookup.
2017-01-09 19:23:09 +00:00
type StringSliceFieldIndex struct {
Field string
Lowercase bool
}
func (s *StringSliceFieldIndex) FromObject(obj interface{}) (bool, [][]byte, error) {
v := reflect.ValueOf(obj)
v = reflect.Indirect(v) // Dereference the pointer if any
fv := v.FieldByName(s.Field)
if !fv.IsValid() {
return false, nil,
fmt.Errorf("field '%s' for %#v is invalid", s.Field, obj)
}
if fv.Kind() != reflect.Slice || fv.Type().Elem().Kind() != reflect.String {
return false, nil, fmt.Errorf("field '%s' is not a string slice", s.Field)
}
length := fv.Len()
vals := make([][]byte, 0, length)
for i := 0; i < fv.Len(); i++ {
val := fv.Index(i).String()
if val == "" {
continue
}
if s.Lowercase {
val = strings.ToLower(val)
}
// Add the null character as a terminator
val += "\x00"
vals = append(vals, []byte(val))
}
if len(vals) == 0 {
return false, nil, nil
}
return true, vals, nil
}
func (s *StringSliceFieldIndex) FromArgs(args ...interface{}) ([]byte, error) {
if len(args) != 1 {
return nil, fmt.Errorf("must provide only a single argument")
}
arg, ok := args[0].(string)
if !ok {
return nil, fmt.Errorf("argument must be a string: %#v", args[0])
}
if s.Lowercase {
arg = strings.ToLower(arg)
}
// Add the null character as a terminator
arg += "\x00"
return []byte(arg), nil
}
func (s *StringSliceFieldIndex) PrefixFromArgs(args ...interface{}) ([]byte, error) {
val, err := s.FromArgs(args...)
if err != nil {
return nil, err
}
// Strip the null terminator, the rest is a prefix
n := len(val)
if n > 0 {
return val[:n-1], nil
}
return val, nil
}
// StringMapFieldIndex is used to extract a field of type map[string]string
// from an object using reflection and builds an index on that field.
//
// Note that although FromArgs in theory supports using either one or
// two arguments, there is a bug: FromObject only creates an index
// using key/value, and does not also create an index using key. This
// means a lookup using one argument will never actually work.
//
// It is currently left as-is to prevent backwards compatibility
// issues.
//
// TODO: Fix this in the next major bump.
2017-01-09 19:23:09 +00:00
type StringMapFieldIndex struct {
Field string
Lowercase bool
}
var MapType = reflect.MapOf(reflect.TypeOf(""), reflect.TypeOf("")).Kind()
func (s *StringMapFieldIndex) FromObject(obj interface{}) (bool, [][]byte, error) {
v := reflect.ValueOf(obj)
v = reflect.Indirect(v) // Dereference the pointer if any
fv := v.FieldByName(s.Field)
if !fv.IsValid() {
return false, nil, fmt.Errorf("field '%s' for %#v is invalid", s.Field, obj)
}
if fv.Kind() != MapType {
return false, nil, fmt.Errorf("field '%s' is not a map[string]string", s.Field)
}
length := fv.Len()
vals := make([][]byte, 0, length)
for _, key := range fv.MapKeys() {
k := key.String()
if k == "" {
continue
}
val := fv.MapIndex(key).String()
if s.Lowercase {
k = strings.ToLower(k)
val = strings.ToLower(val)
}
// Add the null character as a terminator
k += "\x00" + val + "\x00"
vals = append(vals, []byte(k))
}
if len(vals) == 0 {
return false, nil, nil
}
return true, vals, nil
}
// WARNING: Because of a bug in FromObject, this function will never return
// a value when using the single-argument version.
2017-01-09 19:23:09 +00:00
func (s *StringMapFieldIndex) FromArgs(args ...interface{}) ([]byte, error) {
if len(args) > 2 || len(args) == 0 {
return nil, fmt.Errorf("must provide one or two arguments")
}
key, ok := args[0].(string)
if !ok {
return nil, fmt.Errorf("argument must be a string: %#v", args[0])
}
if s.Lowercase {
key = strings.ToLower(key)
}
// Add the null character as a terminator
key += "\x00"
if len(args) == 2 {
val, ok := args[1].(string)
if !ok {
return nil, fmt.Errorf("argument must be a string: %#v", args[1])
}
if s.Lowercase {
val = strings.ToLower(val)
}
// Add the null character as a terminator
key += val + "\x00"
}
return []byte(key), nil
}
// IntFieldIndex is used to extract an int field from an object using
// reflection and builds an index on that field.
type IntFieldIndex struct {
Field string
}
func (i *IntFieldIndex) FromObject(obj interface{}) (bool, []byte, error) {
v := reflect.ValueOf(obj)
v = reflect.Indirect(v) // Dereference the pointer if any
fv := v.FieldByName(i.Field)
if !fv.IsValid() {
return false, nil,
fmt.Errorf("field '%s' for %#v is invalid", i.Field, obj)
}
// Check the type
k := fv.Kind()
size, ok := IsIntType(k)
if !ok {
return false, nil, fmt.Errorf("field %q is of type %v; want an int", i.Field, k)
}
// Get the value and encode it
val := fv.Int()
buf := make([]byte, size)
binary.PutVarint(buf, val)
return true, buf, nil
}
func (i *IntFieldIndex) FromArgs(args ...interface{}) ([]byte, error) {
if len(args) != 1 {
return nil, fmt.Errorf("must provide only a single argument")
}
v := reflect.ValueOf(args[0])
if !v.IsValid() {
return nil, fmt.Errorf("%#v is invalid", args[0])
}
k := v.Kind()
size, ok := IsIntType(k)
if !ok {
return nil, fmt.Errorf("arg is of type %v; want a int", k)
}
val := v.Int()
buf := make([]byte, size)
binary.PutVarint(buf, val)
return buf, nil
}
// IsIntType returns whether the passed type is a type of int and the number
// of bytes needed to encode the type.
func IsIntType(k reflect.Kind) (size int, okay bool) {
switch k {
case reflect.Int:
return binary.MaxVarintLen64, true
case reflect.Int8:
return 2, true
case reflect.Int16:
return binary.MaxVarintLen16, true
case reflect.Int32:
return binary.MaxVarintLen32, true
case reflect.Int64:
return binary.MaxVarintLen64, true
default:
return 0, false
}
}
// UintFieldIndex is used to extract a uint field from an object using
// reflection and builds an index on that field.
type UintFieldIndex struct {
Field string
}
func (u *UintFieldIndex) FromObject(obj interface{}) (bool, []byte, error) {
v := reflect.ValueOf(obj)
v = reflect.Indirect(v) // Dereference the pointer if any
fv := v.FieldByName(u.Field)
if !fv.IsValid() {
return false, nil,
fmt.Errorf("field '%s' for %#v is invalid", u.Field, obj)
}
// Check the type
k := fv.Kind()
size, ok := IsUintType(k)
if !ok {
return false, nil, fmt.Errorf("field %q is of type %v; want a uint", u.Field, k)
}
// Get the value and encode it
val := fv.Uint()
buf := make([]byte, size)
binary.PutUvarint(buf, val)
return true, buf, nil
}
func (u *UintFieldIndex) FromArgs(args ...interface{}) ([]byte, error) {
if len(args) != 1 {
return nil, fmt.Errorf("must provide only a single argument")
}
v := reflect.ValueOf(args[0])
if !v.IsValid() {
return nil, fmt.Errorf("%#v is invalid", args[0])
}
k := v.Kind()
size, ok := IsUintType(k)
if !ok {
return nil, fmt.Errorf("arg is of type %v; want a uint", k)
}
val := v.Uint()
buf := make([]byte, size)
binary.PutUvarint(buf, val)
return buf, nil
}
// IsUintType returns whether the passed type is a type of uint and the number
// of bytes needed to encode the type.
func IsUintType(k reflect.Kind) (size int, okay bool) {
switch k {
case reflect.Uint:
return binary.MaxVarintLen64, true
case reflect.Uint8:
return 2, true
case reflect.Uint16:
return binary.MaxVarintLen16, true
case reflect.Uint32:
return binary.MaxVarintLen32, true
case reflect.Uint64:
return binary.MaxVarintLen64, true
default:
return 0, false
}
}
// BoolFieldIndex is used to extract an boolean field from an object using
// reflection and builds an index on that field.
type BoolFieldIndex struct {
Field string
}
func (i *BoolFieldIndex) FromObject(obj interface{}) (bool, []byte, error) {
v := reflect.ValueOf(obj)
v = reflect.Indirect(v) // Dereference the pointer if any
fv := v.FieldByName(i.Field)
if !fv.IsValid() {
return false, nil,
fmt.Errorf("field '%s' for %#v is invalid", i.Field, obj)
}
// Check the type
k := fv.Kind()
if k != reflect.Bool {
return false, nil, fmt.Errorf("field %q is of type %v; want a bool", i.Field, k)
}
// Get the value and encode it
buf := make([]byte, 1)
if fv.Bool() {
buf[0] = 1
}
return true, buf, nil
}
func (i *BoolFieldIndex) FromArgs(args ...interface{}) ([]byte, error) {
return fromBoolArgs(args)
}
// UUIDFieldIndex is used to extract a field from an object
// using reflection and builds an index on that field by treating
// it as a UUID. This is an optimization to using a StringFieldIndex
// as the UUID can be more compactly represented in byte form.
type UUIDFieldIndex struct {
Field string
}
func (u *UUIDFieldIndex) FromObject(obj interface{}) (bool, []byte, error) {
v := reflect.ValueOf(obj)
v = reflect.Indirect(v) // Dereference the pointer if any
fv := v.FieldByName(u.Field)
if !fv.IsValid() {
return false, nil,
fmt.Errorf("field '%s' for %#v is invalid", u.Field, obj)
}
val := fv.String()
if val == "" {
return false, nil, nil
}
buf, err := u.parseString(val, true)
return true, buf, err
}
func (u *UUIDFieldIndex) FromArgs(args ...interface{}) ([]byte, error) {
if len(args) != 1 {
return nil, fmt.Errorf("must provide only a single argument")
}
switch arg := args[0].(type) {
case string:
return u.parseString(arg, true)
case []byte:
if len(arg) != 16 {
return nil, fmt.Errorf("byte slice must be 16 characters")
}
return arg, nil
default:
return nil,
fmt.Errorf("argument must be a string or byte slice: %#v", args[0])
}
}
func (u *UUIDFieldIndex) PrefixFromArgs(args ...interface{}) ([]byte, error) {
if len(args) != 1 {
return nil, fmt.Errorf("must provide only a single argument")
}
switch arg := args[0].(type) {
case string:
return u.parseString(arg, false)
case []byte:
return arg, nil
default:
return nil,
fmt.Errorf("argument must be a string or byte slice: %#v", args[0])
}
}
// parseString parses a UUID from the string. If enforceLength is false, it will
// parse a partial UUID. An error is returned if the input, stripped of hyphens,
// is not even length.
func (u *UUIDFieldIndex) parseString(s string, enforceLength bool) ([]byte, error) {
// Verify the length
l := len(s)
if enforceLength && l != 36 {
return nil, fmt.Errorf("UUID must be 36 characters")
} else if l > 36 {
return nil, fmt.Errorf("Invalid UUID length. UUID have 36 characters; got %d", l)
}
hyphens := strings.Count(s, "-")
if hyphens > 4 {
return nil, fmt.Errorf(`UUID should have maximum of 4 "-"; got %d`, hyphens)
}
// The sanitized length is the length of the original string without the "-".
sanitized := strings.Replace(s, "-", "", -1)
sanitizedLength := len(sanitized)
if sanitizedLength%2 != 0 {
return nil, fmt.Errorf("Input (without hyphens) must be even length")
}
dec, err := hex.DecodeString(sanitized)
if err != nil {
return nil, fmt.Errorf("Invalid UUID: %v", err)
}
return dec, nil
}
// FieldSetIndex is used to extract a field from an object using reflection and
// builds an index on whether the field is set by comparing it against its
// type's nil value.
type FieldSetIndex struct {
Field string
}
func (f *FieldSetIndex) FromObject(obj interface{}) (bool, []byte, error) {
v := reflect.ValueOf(obj)
v = reflect.Indirect(v) // Dereference the pointer if any
fv := v.FieldByName(f.Field)
if !fv.IsValid() {
return false, nil,
fmt.Errorf("field '%s' for %#v is invalid", f.Field, obj)
}
if fv.Interface() == reflect.Zero(fv.Type()).Interface() {
return true, []byte{0}, nil
}
return true, []byte{1}, nil
}
func (f *FieldSetIndex) FromArgs(args ...interface{}) ([]byte, error) {
return fromBoolArgs(args)
}
// ConditionalIndex builds an index based on a condition specified by a passed
// user function. This function may examine the passed object and return a
// boolean to encapsulate an arbitrarily complex conditional.
type ConditionalIndex struct {
Conditional ConditionalIndexFunc
}
// ConditionalIndexFunc is the required function interface for a
// ConditionalIndex.
type ConditionalIndexFunc func(obj interface{}) (bool, error)
func (c *ConditionalIndex) FromObject(obj interface{}) (bool, []byte, error) {
// Call the user's function
res, err := c.Conditional(obj)
if err != nil {
return false, nil, fmt.Errorf("ConditionalIndexFunc(%#v) failed: %v", obj, err)
}
if res {
return true, []byte{1}, nil
}
return true, []byte{0}, nil
}
func (c *ConditionalIndex) FromArgs(args ...interface{}) ([]byte, error) {
return fromBoolArgs(args)
}
// fromBoolArgs is a helper that expects only a single boolean argument and
// returns a single length byte array containing either a one or zero depending
// on whether the passed input is true or false respectively.
func fromBoolArgs(args []interface{}) ([]byte, error) {
if len(args) != 1 {
return nil, fmt.Errorf("must provide only a single argument")
}
if val, ok := args[0].(bool); !ok {
return nil, fmt.Errorf("argument must be a boolean type: %#v", args[0])
} else if val {
return []byte{1}, nil
}
return []byte{0}, nil
}
// CompoundIndex is used to build an index using multiple sub-indexes
// Prefix based iteration is supported as long as the appropriate prefix
// of indexers support it. All sub-indexers are only assumed to expect
// a single argument.
type CompoundIndex struct {
Indexes []Indexer
// AllowMissing results in an index based on only the indexers
// that return data. If true, you may end up with 2/3 columns
// indexed which might be useful for an index scan. Otherwise,
// the CompoundIndex requires all indexers to be satisfied.
AllowMissing bool
}
func (c *CompoundIndex) FromObject(raw interface{}) (bool, []byte, error) {
var out []byte
2017-01-09 19:23:09 +00:00
for i, idxRaw := range c.Indexes {
idx, ok := idxRaw.(SingleIndexer)
if !ok {
return false, nil, fmt.Errorf("sub-index %d error: %s", i, "sub-index must be a SingleIndexer")
}
ok, val, err := idx.FromObject(raw)
if err != nil {
return false, nil, fmt.Errorf("sub-index %d error: %v", i, err)
}
if !ok {
if c.AllowMissing {
break
} else {
return false, nil, nil
}
}
out = append(out, val...)
}
return true, out, nil
}
func (c *CompoundIndex) FromArgs(args ...interface{}) ([]byte, error) {
if len(args) != len(c.Indexes) {
return nil, fmt.Errorf("non-equivalent argument count and index fields")
}
var out []byte
for i, arg := range args {
val, err := c.Indexes[i].FromArgs(arg)
if err != nil {
return nil, fmt.Errorf("sub-index %d error: %v", i, err)
}
out = append(out, val...)
}
return out, nil
}
func (c *CompoundIndex) PrefixFromArgs(args ...interface{}) ([]byte, error) {
if len(args) > len(c.Indexes) {
return nil, fmt.Errorf("more arguments than index fields")
}
var out []byte
for i, arg := range args {
if i+1 < len(args) {
val, err := c.Indexes[i].FromArgs(arg)
if err != nil {
return nil, fmt.Errorf("sub-index %d error: %v", i, err)
}
out = append(out, val...)
} else {
prefixIndexer, ok := c.Indexes[i].(PrefixIndexer)
if !ok {
return nil, fmt.Errorf("sub-index %d does not support prefix scanning", i)
}
val, err := prefixIndexer.PrefixFromArgs(arg)
if err != nil {
return nil, fmt.Errorf("sub-index %d error: %v", i, err)
}
out = append(out, val...)
}
}
return out, nil
}
// CompoundMultiIndex is used to build an index using multiple
// sub-indexes.
//
// Unlike CompoundIndex, CompoundMultiIndex can have both
// SingleIndexer and MultiIndexer sub-indexers. However, each
// MultiIndexer adds considerable overhead/complexity in terms of
// the number of indexes created under-the-hood. It is not suggested
// to use more than one or two, if possible.
//
// Another change from CompoundIndexer is that if AllowMissing is
// set, not only is it valid to have empty index fields, but it will
// still create index values up to the first empty index. This means
// that if you have a value with an empty field, rather than using a
// prefix for lookup, you can simply pass in less arguments. As an
// example, if {Foo, Bar} is indexed but Bar is missing for a value
// and AllowMissing is set, an index will still be created for {Foo}
// and it is valid to do a lookup passing in only Foo as an argument.
// Note that the ordering isn't guaranteed -- it's last-insert wins,
// but this is true if you have two objects that have the same
// indexes not using AllowMissing anyways.
//
// Because StringMapFieldIndexers can take a varying number of args,
// it is currently a requirement that whenever it is used, two
// arguments must _always_ be provided for it. In theory we only
// need one, except a bug in that indexer means the single-argument
// version will never work. You can leave the second argument nil,
// but it will never produce a value. We support this for whenever
// that bug is fixed, likely in a next major version bump.
//
// Prefix-based indexing is not currently supported.
type CompoundMultiIndex struct {
Indexes []Indexer
// AllowMissing results in an index based on only the indexers
// that return data. If true, you may end up with 2/3 columns
// indexed which might be useful for an index scan. Otherwise,
// CompoundMultiIndex requires all indexers to be satisfied.
AllowMissing bool
}
func (c *CompoundMultiIndex) FromObject(raw interface{}) (bool, [][]byte, error) {
// At each entry, builder is storing the results from the next index
builder := make([][][]byte, 0, len(c.Indexes))
// Start with something higher to avoid resizing if possible
out := make([][]byte, 0, len(c.Indexes)^3)
forloop:
// This loop goes through each indexer and adds the value(s) provided to the next
// entry in the slice. We can then later walk it like a tree to construct the indices.
for i, idxRaw := range c.Indexes {
switch idx := idxRaw.(type) {
case SingleIndexer:
ok, val, err := idx.FromObject(raw)
if err != nil {
return false, nil, fmt.Errorf("single sub-index %d error: %v", i, err)
}
if !ok {
if c.AllowMissing {
break forloop
} else {
return false, nil, nil
}
}
builder = append(builder, [][]byte{val})
case MultiIndexer:
ok, vals, err := idx.FromObject(raw)
if err != nil {
return false, nil, fmt.Errorf("multi sub-index %d error: %v", i, err)
}
if !ok {
if c.AllowMissing {
break forloop
} else {
return false, nil, nil
}
}
// Add each of the new values to each of the old values
builder = append(builder, vals)
default:
return false, nil, fmt.Errorf("sub-index %d does not satisfy either SingleIndexer or MultiIndexer", i)
}
}
// We are walking through the builder slice essentially in a depth-first fashion,
// building the prefix and leaves as we go. If AllowMissing is false, we only insert
// these full paths to leaves. Otherwise, we also insert each prefix along the way.
// This allows for lookup in FromArgs when AllowMissing is true that does not contain
// the full set of arguments. e.g. for {Foo, Bar} where an object has only the Foo
// field specified as "abc", it is valid to call FromArgs with just "abc".
var walkVals func([]byte, int)
walkVals = func(currPrefix []byte, depth int) {
if depth == len(builder)-1 {
// These are the "leaves", so append directly
for _, v := range builder[depth] {
out = append(out, append(currPrefix, v...))
}
return
}
for _, v := range builder[depth] {
nextPrefix := append(currPrefix, v...)
if c.AllowMissing {
out = append(out, nextPrefix)
}
walkVals(nextPrefix, depth+1)
}
}
walkVals(nil, 0)
return true, out, nil
}
func (c *CompoundMultiIndex) FromArgs(args ...interface{}) ([]byte, error) {
var stringMapCount int
var argCount int
for _, index := range c.Indexes {
if argCount >= len(args) {
break
}
if _, ok := index.(*StringMapFieldIndex); ok {
// We require pairs for StringMapFieldIndex, but only got one
if argCount+1 >= len(args) {
return nil, errors.New("invalid number of arguments")
}
stringMapCount++
argCount += 2
} else {
argCount++
}
}
argCount = 0
switch c.AllowMissing {
case true:
if len(args) > len(c.Indexes)+stringMapCount {
return nil, errors.New("too many arguments")
}
default:
if len(args) != len(c.Indexes)+stringMapCount {
return nil, errors.New("number of arguments does not equal number of indexers")
}
}
var out []byte
var val []byte
var err error
for i, idx := range c.Indexes {
if argCount >= len(args) {
// We're done; should only hit this if AllowMissing
break
}
if _, ok := idx.(*StringMapFieldIndex); ok {
if args[argCount+1] == nil {
val, err = idx.FromArgs(args[argCount])
} else {
val, err = idx.FromArgs(args[argCount : argCount+2]...)
}
argCount += 2
} else {
val, err = idx.FromArgs(args[argCount])
argCount++
}
if err != nil {
return nil, fmt.Errorf("sub-index %d error: %v", i, err)
}
out = append(out, val...)
}
return out, nil
}