Global/wakapi
1
0
Fork 0
mirror of https://github.com/muety/wakapi.git synced 2025-12-05 22:20:24 -08:00
Code Issues Packages Projects Releases Wiki Activity

fix: include most prominent entity for each duration

Browse Source
This commit is contained in:
Ferdinand Mütsch
2025-02-21 09:52:02 +01:00
parent 71b18f1aaf
commit 7b7f5e94bd
1 changed files with 29 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files

31
services/duration.go
View File

@@ -6,6 +6,7 @@ import (
"github.com/duke-git/lancet/v2/datetime"
"github.com/duke-git/lancet/v2/maputil"
"github.com/duke-git/lancet/v2/slice"
"github.com/duke-git/lancet/v2/tuple"
"github.com/leandro-lugaresi/hub"
"github.com/muety/artifex/v2"
"github.com/muety/wakapi/config"
@@ -62,7 +63,7 @@ func (srv *DurationService) Get(from, to time.Time, user *models.User, filters *
// note about "multi-level" durations at different intervals:
// while durations themselves store the interval (aka. heartbeats timeout) they were computed for, we currently don't support actually storing durations at different intervals
// if an interval different from the user's preference is requested, recompute durations live from heartbeats and skip cache
effectiveInterval := srv.getEffectiveInterval(user, customInterval)
effectiveInterval := getEffectiveInterval(user, customInterval)
skipCache = skipCache || effectiveInterval != user.HeartbeatsTimeout()
// recompute live
@@ -174,6 +175,7 @@ func (srv *DurationService) getLive(from, to time.Time, user *models.User, inter
var latest *models.Duration
mapping := make(map[string][]*models.Duration)
entityDurations := make(map[tuple.Tuple2[string, string]]time.Duration)
for h := range heartbeats {
if h.User == nil {
@@ -210,6 +212,14 @@ func (srv *DurationService) getLive(from, to time.Time, user *models.User, inter
latest = d1
} else {
latest.NumHeartbeats++
// TODO: think about how to fix this properly
// Problem: we don't consider entities (aka. file names) when distinguishing between durations, that is, in other words,
// durations essentially aggregate heartbeats by (a) by time (squash all heartbeats within <heartbeatsTimeout> and (b) by entity.
// This conflicts with file-level summaries (as shown on the project details page), because fine-grained file information is lost.
// As a heuristic, for each duration we keep the file (entity) that was "most prominent" (i.e. was coded on most) with the duration.
// However, when you code for 10 minutes straight, where you work 6 minutes on file A and 4 minutes on file B, the latter still won't show up in the summary at all.
// This is a tricky trade-off, because if we changed durations to respect entities as well, we'd end up with much higher cardinality, thus much less "compression" in comparison to raw heartbeats and thus higher compute- and storage load.
latest = updateDurationEntity(latest, h, entityDurations) // dirty hack
}
count++
@@ -301,9 +311,26 @@ func (srv *DurationService) filtersToColumnMap(filters *models.Filters) map[stri
return columnMap
}
func (r *DurationService) getEffectiveInterval(user *models.User, overrideInterval *time.Duration) time.Duration {
func getEffectiveInterval(user *models.User, overrideInterval *time.Duration) time.Duration {
if overrideInterval == nil {
return user.HeartbeatsTimeout()
}
return *overrideInterval
}
func updateDurationEntity(d *models.Duration, h *models.Heartbeat, entityDurations map[tuple.Tuple2[string, string]]time.Duration) *models.Duration {
// check if total time for the entity of the given heartbeat exceeds the duration's entity's total time
// if yes, update duration entity so that it always reflect the "most prominent" entity of this group
key1 := tuple.NewTuple2(d.GroupHash, h.Entity)
key2 := tuple.NewTuple2(d.GroupHash, d.Entity)
if _, ok := entityDurations[key1]; !ok {
entityDurations[key1] = time.Duration(0)
}
entityDurations[key1] += d.Duration
dur1 := entityDurations[key1]
dur2 := entityDurations[key2]
if dur1 > dur2 {
d.Entity = h.Entity
}
return d
}