gosora/common/topic_cache.go
Azareal 2296008655 Basic search now works for the Search & Filter Widget. ElasticSearch has been temporarily delayed, so I can push through this update.
Added the three month time range to the analytics panes.
Began work on adding new graphs to the analytics panes.
Began work on the ElasticSearch adapter for the search system.
Added the currently limited AddKey method to the database adapters.
Expanded upon the column parsing logic in the database adapters to ease the use of InsertSelects.

Added the BulkGet method to TopicCache.
Added the BulkGetMap method to TopicStore.
TopicStore methods should now properly retrieve lastReplyBy.
Added the panel_analytics_script template to de-dupe part of the analytics logic. We plan to tidy this up further, but for now, it'll suffice.
Added plugin_sendmail and plugin_hyperdrive to the continuous integration test list.
Tweaked the width and heights of the textareas for the Widget Editor.
Added the AddKey method to *qgen.builder
Fixed a bug where using the inline forum editor would crash Gosora and wouldn't set the preset permissions for that forum properly.
Added DotBot to the user agent analytics.
Invisibles should be better handled when they're encountered now in user agent strings.
Unknown language ISO Codes in headers now have the requests fully logged for debugging purposes.
Shortened some of the pointer receiver names.
Shortened some variable names.

Added the dotbot phrase.
Added the panel_statistics_time_range_three_months phrase.

Added gopkg.in/olivere/elastic.v6 as a dependency.

You will need to run the patcher or updater for this commit.
2019-02-23 16:29:19 +10:00

162 lines
4.4 KiB
Go

package common
import (
"sync"
"sync/atomic"
)
// TopicCache is an interface which spits out topics from a fast cache rather than the database, whether from memory or from an application like Redis. Topics may not be present in the cache but may be in the database
type TopicCache interface {
Get(id int) (*Topic, error)
GetUnsafe(id int) (*Topic, error)
BulkGet(ids []int) (list []*Topic)
Set(item *Topic) error
Add(item *Topic) error
AddUnsafe(item *Topic) error
Remove(id int) error
RemoveUnsafe(id int) error
Flush()
Length() int
SetCapacity(capacity int)
GetCapacity() int
}
// MemoryTopicCache stores and pulls topics out of the current process' memory
type MemoryTopicCache struct {
items map[int]*Topic
length int64 // sync/atomic only lets us operate on int32s and int64s
capacity int
sync.RWMutex
}
// NewMemoryTopicCache gives you a new instance of MemoryTopicCache
func NewMemoryTopicCache(capacity int) *MemoryTopicCache {
return &MemoryTopicCache{
items: make(map[int]*Topic),
capacity: capacity,
}
}
// Get fetches a topic by ID. Returns ErrNoRows if not present.
func (mts *MemoryTopicCache) Get(id int) (*Topic, error) {
mts.RLock()
item, ok := mts.items[id]
mts.RUnlock()
if ok {
return item, nil
}
return item, ErrNoRows
}
// GetUnsafe fetches a topic by ID. Returns ErrNoRows if not present. THIS METHOD IS NOT THREAD-SAFE.
func (mts *MemoryTopicCache) GetUnsafe(id int) (*Topic, error) {
item, ok := mts.items[id]
if ok {
return item, nil
}
return item, ErrNoRows
}
// BulkGet fetches multiple topics by their IDs. Indices without topics will be set to nil, so make sure you check for those, we might want to change this behaviour to make it less confusing.
func (c *MemoryTopicCache) BulkGet(ids []int) (list []*Topic) {
list = make([]*Topic, len(ids))
c.RLock()
for i, id := range ids {
list[i] = c.items[id]
}
c.RUnlock()
return list
}
// Set overwrites the value of a topic in the cache, whether it's present or not. May return a capacity overflow error.
func (mts *MemoryTopicCache) Set(item *Topic) error {
mts.Lock()
_, ok := mts.items[item.ID]
if ok {
mts.items[item.ID] = item
} else if int(mts.length) >= mts.capacity {
mts.Unlock()
return ErrStoreCapacityOverflow
} else {
mts.items[item.ID] = item
atomic.AddInt64(&mts.length, 1)
}
mts.Unlock()
return nil
}
// Add adds a topic to the cache, similar to Set, but it's only intended for new items. This method might be deprecated in the near future, use Set. May return a capacity overflow error.
// ? Is this redundant if we have Set? Are the efficiency wins worth this? Is this even used?
func (mts *MemoryTopicCache) Add(item *Topic) error {
mts.Lock()
if int(mts.length) >= mts.capacity {
mts.Unlock()
return ErrStoreCapacityOverflow
}
mts.items[item.ID] = item
mts.Unlock()
atomic.AddInt64(&mts.length, 1)
return nil
}
// AddUnsafe is the unsafe version of Add. May return a capacity overflow error. THIS METHOD IS NOT THREAD-SAFE.
func (mts *MemoryTopicCache) AddUnsafe(item *Topic) error {
if int(mts.length) >= mts.capacity {
return ErrStoreCapacityOverflow
}
mts.items[item.ID] = item
mts.length = int64(len(mts.items))
return nil
}
// Remove removes a topic from the cache by ID, if they exist. Returns ErrNoRows if no items exist.
func (mts *MemoryTopicCache) Remove(id int) error {
mts.Lock()
_, ok := mts.items[id]
if !ok {
mts.Unlock()
return ErrNoRows
}
delete(mts.items, id)
mts.Unlock()
atomic.AddInt64(&mts.length, -1)
return nil
}
// RemoveUnsafe is the unsafe version of Remove. THIS METHOD IS NOT THREAD-SAFE.
func (mts *MemoryTopicCache) RemoveUnsafe(id int) error {
_, ok := mts.items[id]
if !ok {
return ErrNoRows
}
delete(mts.items, id)
atomic.AddInt64(&mts.length, -1)
return nil
}
// Flush removes all the topics from the cache, useful for tests.
func (mts *MemoryTopicCache) Flush() {
mts.Lock()
mts.items = make(map[int]*Topic)
mts.length = 0
mts.Unlock()
}
// ! Is this concurrent?
// Length returns the number of topics in the memory cache
func (mts *MemoryTopicCache) Length() int {
return int(mts.length)
}
// SetCapacity sets the maximum number of topics which this cache can hold
func (mts *MemoryTopicCache) SetCapacity(capacity int) {
// Ints are moved in a single instruction, so this should be thread-safe
mts.capacity = capacity
}
// GetCapacity returns the maximum number of topics this cache can hold
func (mts *MemoryTopicCache) GetCapacity() int {
return mts.capacity
}