[performance] remove hard reliance on .Cached field to indicate whether media / emoji is cached (#4545)

This removes our hard reliance on the `.Cached` field of media and emojis to determine whether it exists in storage. We still make use of it as a useful flag to know whether to even bother checking storage, but we ultimately rely on the `ErrNotFound` response of storage to determine whether the media exists and needs recaching. This now removes our hard reliance on performing the `FixCacheStatus()` cleanup operations for media and emojis, which should reduce a whole bunch of S3 storage driver calls (thus, reducing cost for metered S3 buckets).

Reviewed-on: https://codeberg.org/superseriousbusiness/gotosocial/pulls/4545
Co-authored-by: kim <grufwub@gmail.com>
Co-committed-by: kim <grufwub@gmail.com>

3 files changed, 16 insertions, 4 deletions

diff --git a/internal/cleaner/cleaner.go b/internal/cleaner/cleaner.go
index 7e32566ee..ab9a1867e 100644
--- a/internal/cleaner/cleaner.go
+++ b/internal/cleaner/cleaner.go
@@ -141,8 +141,8 @@ func (c *Cleaner) ScheduleJobs() error {
 		)
 	}
 
-	// Time travel from
-	// year zero, groovy.
+	// Time travel from the
+	// year zero, groovy baby.
 	firstCleanupAt := time.Date(
 		now.Year(),
 		now.Month(),
diff --git a/internal/cleaner/emoji.go b/internal/cleaner/emoji.go
index 245dfac27..e72ee9665 100644
--- a/internal/cleaner/emoji.go
+++ b/internal/cleaner/emoji.go
@@ -41,10 +41,16 @@ func (e *Emoji) All(ctx context.Context, maxRemoteDays int) {
 	e.LogUncacheRemote(ctx, t)
 	e.LogFixBroken(ctx)
 	e.LogPruneUnused(ctx)
-	e.LogFixCacheStates(ctx)
 	_ = e.state.Storage.Storage.Clean(ctx)
 }
 
+// AllAndFix calls LogFixCacheStates(), followed by All(), it
+// is done this way round so Storage.Clean() is performed last.
+func (e *Emoji) AllAndFix(ctx context.Context, maxRemoteDays int) {
+	e.LogFixCacheStates(ctx)
+	e.All(ctx, maxRemoteDays)
+}
+
 // LogUncacheRemote performs Emoji.UncacheRemote(...), logging the start and outcome.
 func (e *Emoji) LogUncacheRemote(ctx context.Context, olderThan time.Time) {
 	log.Infof(ctx, "start older than: %s", olderThan.Format(time.Stamp))
diff --git a/internal/cleaner/media.go b/internal/cleaner/media.go
index 99fd5779f..bd02cb2a3 100644
--- a/internal/cleaner/media.go
+++ b/internal/cleaner/media.go
@@ -44,10 +44,16 @@ func (m *Media) All(ctx context.Context, maxRemoteDays int) {
 	m.LogUncacheRemote(ctx, t)
 	m.LogPruneOrphaned(ctx)
 	m.LogPruneUnused(ctx)
-	m.LogFixCacheStates(ctx)
 	_ = m.state.Storage.Storage.Clean(ctx)
 }
 
+// AllAndFix calls LogFixCacheStates(), followed by All(), it
+// is done this way round so Storage.Clean() is performed last.
+func (m *Media) AllAndFix(ctx context.Context, maxRemoteDays int) {
+	m.LogFixCacheStates(ctx)
+	m.All(ctx, maxRemoteDays)
+}
+
 // LogUncacheRemote performs Media.UncacheRemote(...), logging the start and outcome.
 func (m *Media) LogUncacheRemote(ctx context.Context, olderThan time.Time) {
 	log.Infof(ctx, "start older than: %s", olderThan.Format(time.Stamp))