navidrome/persistence/criteria_sql.go

Ignoring revisions in .git-blame-ignore-revs. Click here to bypass and see the normal blame view.

714 lines
23 KiB
Go
Raw Normal View History

package persistence
import (
"errors"
"fmt"
fix(smartplaylists): optimize smart playlist performance for role and tag criteria (#5515) * fix(server): optimize smart playlist role queries for large criteria (#5511) Role-based smart playlist criteria (artist, composer, etc.) now query the indexed media_file_artists join table instead of parsing JSON via json_tree() on every row. Multiple conditions for the same role within an OR group are merged into a single EXISTS subquery (batched at 200 to stay under SQLite's expression tree depth limit). A composite index (media_file_id, role) replaces the now-redundant single-column (media_file_id) index on media_file_artists. Benchmark (40k tracks, 500 patterns, 3 artists/track): - Merged join-table: 15ms (9.3x faster) - Merged json_tree: 30ms (4.6x faster) - Unmerged baseline: 137ms * refactor: simplify role condition SQL generation and benchmark Extract shared roleCondSQL/roleExistsSQL helpers to deduplicate the EXISTS template between roleCond and roleCondGroup. Use slices.Chunk for batching per project convention. Extract runBenchQuery helper to eliminate triplicated benchmark execution loop. * chore: raise roleCondBatchSize to 350 The empirical SQLite limit is 496 conditions per merged EXISTS subquery. Raising from 200 to 350 reduces the number of batches (e.g. 500 patterns now splits into 2 batches instead of 3). * fix(server): apply OR-merge optimization to tag conditions too Generalize mergeRoleConds into mergeJsonConds to also collapse multiple tag conditions for the same tag (e.g. genre) within OR groups. This gives the same ~5x speedup for tag-heavy smart playlists as the role optimization gives for artist-heavy ones. * refactor: benchmark uses real criteria pipeline instead of hand-built SQL The "Current" sub-benchmark now builds criteria.Criteria expressions and runs them through the actual newSmartPlaylistCriteria → Where() → ToSql() pipeline, validating the real production code path. The baseline still uses hand-built SQL representing the old json_tree approach. * fix: stabilize merged group ordering and close rows before error check Sort group keys in mergeJsonConds so the merged additions have deterministic order across runs, improving SQLite statement cache reuse. Move rows.Close() before rows.Err() in benchmark helper.
2026-05-22 18:00:13 -03:00
"maps"
"reflect"
fix(smartplaylists): optimize smart playlist performance for role and tag criteria (#5515) * fix(server): optimize smart playlist role queries for large criteria (#5511) Role-based smart playlist criteria (artist, composer, etc.) now query the indexed media_file_artists join table instead of parsing JSON via json_tree() on every row. Multiple conditions for the same role within an OR group are merged into a single EXISTS subquery (batched at 200 to stay under SQLite's expression tree depth limit). A composite index (media_file_id, role) replaces the now-redundant single-column (media_file_id) index on media_file_artists. Benchmark (40k tracks, 500 patterns, 3 artists/track): - Merged join-table: 15ms (9.3x faster) - Merged json_tree: 30ms (4.6x faster) - Unmerged baseline: 137ms * refactor: simplify role condition SQL generation and benchmark Extract shared roleCondSQL/roleExistsSQL helpers to deduplicate the EXISTS template between roleCond and roleCondGroup. Use slices.Chunk for batching per project convention. Extract runBenchQuery helper to eliminate triplicated benchmark execution loop. * chore: raise roleCondBatchSize to 350 The empirical SQLite limit is 496 conditions per merged EXISTS subquery. Raising from 200 to 350 reduces the number of batches (e.g. 500 patterns now splits into 2 batches instead of 3). * fix(server): apply OR-merge optimization to tag conditions too Generalize mergeRoleConds into mergeJsonConds to also collapse multiple tag conditions for the same tag (e.g. genre) within OR groups. This gives the same ~5x speedup for tag-heavy smart playlists as the role optimization gives for artist-heavy ones. * refactor: benchmark uses real criteria pipeline instead of hand-built SQL The "Current" sub-benchmark now builds criteria.Criteria expressions and runs them through the actual newSmartPlaylistCriteria → Where() → ToSql() pipeline, validating the real production code path. The baseline still uses hand-built SQL representing the old json_tree approach. * fix: stabilize merged group ordering and close rows before error check Sort group keys in mergeJsonConds so the merged additions have deterministic order across runs, improving SQLite statement cache reuse. Move rows.Close() before rows.Err() in benchmark helper.
2026-05-22 18:00:13 -03:00
"slices"
"strconv"
"strings"
"time"
feat(smartplaylists): relax playlist visibility in inPlaylist/notInPlaylist rules (#5411) * test(e2e): add end-to-end tests for smart playlists functionality Signed-off-by: Deluan <deluan@navidrome.org> * fix: enforce playlist visibility in smart playlist InPlaylist/NotInPlaylist rules Previously, the InPlaylist/NotInPlaylist smart playlist criteria only allowed referencing public playlists, regardless of who owned the smart playlist. This was too restrictive for owners referencing their own private playlists and for admins who should have unrestricted access. The fix passes the smart playlist owner's identity and admin status into the criteria SQL builder, so that: admins can reference any playlist, regular users can reference public playlists plus their own private ones, and inaccessible referenced playlists produce a warning instead of a hard error. Also prevents recursive refresh of child playlists the owner cannot access. * test(e2e): clarify user roles and fix playlist visibility tests Renamed testUser/otherUser to adminUser/regularUser to make the admin vs regular user distinction explicit in test code. Fixed three playlist visibility tests that were evaluating as admin (bypassing all access checks) instead of as a regular user, so the public playlist path is now actually exercised. All playlist operator tests now use explicit evaluateRuleAs calls with the appropriate user role. * fix: sync rulesSQL criteria after limitPercent resolution The rulesSQL struct captures a copy of rules at creation time. When limitPercent is resolved later, rules.Limit is updated but rulesSQL still holds the stale value. This caused percentage-based smart playlist limits to be silently ignored. Fix by updating rulesSQL.criteria after the resolution. * refactor: convert inList to a method on smartPlaylistCriteria The inList function already receives ownerID and ownerIsAdmin from the smartPlaylistCriteria caller. Making it a method lets it access those fields directly from the receiver, simplifying the signature and staying consistent with exprSQL which was already converted to a method. * refactor: simplify function signatures by removing type parameters in criteria_sql.go Signed-off-by: Deluan <deluan@navidrome.org> --------- Signed-off-by: Deluan <deluan@navidrome.org>
2026-04-25 14:59:06 -04:00
"github.com/Masterminds/squirrel"
refactor: centralize criteria sort parsing and extract smart playlist logic (#5415) * test: add tests for recordingdate alias resolution in smart playlists Signed-off-by: Deluan <deluan@navidrome.org> * refactor: update FieldInfo structure and simplify fieldMap initialization Signed-off-by: Deluan <deluan@navidrome.org> * refactor: move sort parsing logic from persistence to criteria package Extracted sort field parsing, validation, and direction handling from persistence/criteria_sql.go into model/criteria/sort.go. The new OrderByFields method on Criteria parses the Sort/Order strings into validated SortField structs (field name + direction), resolving aliases and handling +/- prefixes and order inversion. The persistence layer now consumes these parsed fields and only handles SQL expression mapping. This centralizes sort parsing to enforce consistent implementations. * refactor: standardize field access in smartPlaylistCriteria structure Signed-off-by: Deluan <deluan@navidrome.org> * refactor: add ResolveLimit method to Criteria Moved the percentage-limit resolution logic from playlist_repository into Criteria.ResolveLimit, replacing the 3-line mutate-after-query pattern with a single method call. The method preserves LimitPercent rather than zeroing it, since IsPercentageLimit already returns false once Limit is set, making the clear redundant and lossy. * refactor: improve child playlist loading and error handling in refresh logic Signed-off-by: Deluan <deluan@navidrome.org> * refactor: extract smart playlist logic to dedicated files Moved refreshSmartPlaylist, addSmartPlaylistAnnotationJoins, and addCriteria methods from playlist_repository.go to a new smart_playlist_repository.go file. Extracted all smart playlist tests to smart_playlist_repository_test.go. Added DeferCleanup to the "valid rules" test to fix ordering flakiness when Ginkgo randomizes test execution across files. * refactor: break refreshSmartPlaylist into smaller focused methods Split the monolithic refreshSmartPlaylist method into discrete helpers for readability: shouldRefreshSmartPlaylist for guard checks, refreshChildPlaylists for recursive dependency refresh, resolvePercentageLimit for count-based limit resolution, buildSmartPlaylistQuery for assembling the SELECT with joins, and addMediaFileAnnotationJoin to DRY up the repeated annotation join clause. * refactor: deduplicate child playlist IDs in Criteria Signed-off-by: Deluan <deluan@navidrome.org> * refactor: simplify withSmartPlaylistOwner to accept model.User Replaced separate ownerID string and ownerIsAdmin bool parameters with a single model.User struct, reducing the field count in smartPlaylistCriteria and making the option function signature clearer. Updated all call sites and tests accordingly. * fix: handle empty sort fields and propagate child playlist load errors OrderByFields now falls back to [{title, asc}] when all user-supplied sort fields are invalid, preventing empty ORDER BY clauses that would produce invalid SQL in row_number() window functions. Also restored the original behavior where a DB error loading child playlists aborts the parent smart playlist refresh, by making refreshChildPlaylists return a bool. * refactor: log warning when no valid sort fields are found Signed-off-by: Deluan <deluan@navidrome.org> --------- Signed-off-by: Deluan <deluan@navidrome.org>
2026-04-26 14:49:59 -04:00
"github.com/navidrome/navidrome/model"
"github.com/navidrome/navidrome/model/criteria"
)
type smartPlaylistJoinType int
const (
smartPlaylistJoinNone smartPlaylistJoinType = 0
smartPlaylistJoinAlbumAnnotation smartPlaylistJoinType = 1 << iota
smartPlaylistJoinArtistAnnotation
)
func (j smartPlaylistJoinType) has(other smartPlaylistJoinType) bool {
return j&other != 0
}
type smartPlaylistField struct {
feat(smartplaylist): extend isMissing/isPresent to bpm, bitDepth and many text fields (#5603) * feat(smartplaylist): support isMissing/isPresent on mbz_* and lyrics fields Mark the six mbz_* MusicBrainz ID columns and the lyrics column as Nullable in the criteria field map, then extend missingExpr to handle string columns where absence is encoded as NULL or empty string (plus '[]' for lyrics). The Numeric/Boolean path (ReplayGain) is preserved via an explicit type check. * refactor(model): make MediaFile BPM and BitDepth nullable pointers Convert BPM and BitDepth fields in model.MediaFile from int to *int so that 'tag absent' is distinguishable from zero. The metadata mapper now uses NullableFloat for BPM (nil when absent or zero/unparseable) and only sets BitDepth when the audio property is non-zero (lossy codecs report 0). All read sites use gg.V() for zero-fallback deref so Subsonic API output and transcoding behaviour are byte-identical to before. The persistence layer bridges the existing NOT NULL DB columns by coercing nil to 0 on write and 0 back to nil on read in PostMapArgs/PostScan; a later migration task will drop those constraints. Hash upgrade safety is verified by a new MediaFile.Hash describe block: nil *int hashes identically to the old int(0) default via ZeroNil+IgnoreZeroValue, so no files will be spuriously re-imported after this change. Extra files touched beyond the plan's list: core/stream/legacy_client_test.go (BitDepth in model.MediaFile literals), persistence/mediafile_repository.go (NOT NULL bridge). * test(model): pin pre-conversion golden hashes for BPM/BitDepth * feat(smartplaylist): support isMissing/isPresent on bpm and bitDepth * feat(db): make bpm and bit_depth columns nullable, backfill 0 to NULL Drop the NOT NULL constraint on media_file.bpm and bit_depth via a lossless migration that converts legacy 0-means-absent values to real NULL. Remove the temporary shim in PostScan/PostMapArgs that was bridging the old NOT NULL columns to the *int model fields. Add round-trip persistence tests asserting NULL storage for nil pointers and correct value round-trip for non-nil pointers. * test(e2e): verify isMissing/isPresent partition for nullable fields Add DescribeTable covering bpm, bitdepth, lyrics, and mbz_recording_id: for each field, isMissing + isPresent song counts must equal the total library count, proving the nullable-column SQL is exhaustive and correct. * test(e2e): seed bpm tag so isMissing/isPresent partition is non-trivial * fix(model): omit bitDepth from JSON when absent instead of emitting null * feat(smartplaylist): support isMissing/isPresent on more string fields Enable isMissing/isPresent operators for album, comment, catalognumber, discsubtitle, albumcomment, sorttitle, sortalbum, sortartist, sortalbumartist, and explicitstatus by marking them Nullable in fieldMap. * refactor(smartplaylist): unify missingExpr column logic into one flow Collapse the numeric/string fork in missingExpr into a single empties-driven loop (numeric/boolean fields simply have no empties), and replace the duplicated IsTag/IsRole guard with a three-way switch that expresses the dispatch model once. No SQL semantics change for string fields; numeric/boolean fields now emit a single-element Or/And which squirrel parenthesizes (e.g. `(col IS NULL)` instead of bare `col IS NULL`) — update the affected test expectations accordingly.
2026-06-13 13:15:20 -04:00
expr string
order string
joinType smartPlaylistJoinType
emptyValues []string // additional values that encode "missing" for string columns (e.g. '[]' for lyrics)
}
type smartPlaylistCriteria struct {
refactor: centralize criteria sort parsing and extract smart playlist logic (#5415) * test: add tests for recordingdate alias resolution in smart playlists Signed-off-by: Deluan <deluan@navidrome.org> * refactor: update FieldInfo structure and simplify fieldMap initialization Signed-off-by: Deluan <deluan@navidrome.org> * refactor: move sort parsing logic from persistence to criteria package Extracted sort field parsing, validation, and direction handling from persistence/criteria_sql.go into model/criteria/sort.go. The new OrderByFields method on Criteria parses the Sort/Order strings into validated SortField structs (field name + direction), resolving aliases and handling +/- prefixes and order inversion. The persistence layer now consumes these parsed fields and only handles SQL expression mapping. This centralizes sort parsing to enforce consistent implementations. * refactor: standardize field access in smartPlaylistCriteria structure Signed-off-by: Deluan <deluan@navidrome.org> * refactor: add ResolveLimit method to Criteria Moved the percentage-limit resolution logic from playlist_repository into Criteria.ResolveLimit, replacing the 3-line mutate-after-query pattern with a single method call. The method preserves LimitPercent rather than zeroing it, since IsPercentageLimit already returns false once Limit is set, making the clear redundant and lossy. * refactor: improve child playlist loading and error handling in refresh logic Signed-off-by: Deluan <deluan@navidrome.org> * refactor: extract smart playlist logic to dedicated files Moved refreshSmartPlaylist, addSmartPlaylistAnnotationJoins, and addCriteria methods from playlist_repository.go to a new smart_playlist_repository.go file. Extracted all smart playlist tests to smart_playlist_repository_test.go. Added DeferCleanup to the "valid rules" test to fix ordering flakiness when Ginkgo randomizes test execution across files. * refactor: break refreshSmartPlaylist into smaller focused methods Split the monolithic refreshSmartPlaylist method into discrete helpers for readability: shouldRefreshSmartPlaylist for guard checks, refreshChildPlaylists for recursive dependency refresh, resolvePercentageLimit for count-based limit resolution, buildSmartPlaylistQuery for assembling the SELECT with joins, and addMediaFileAnnotationJoin to DRY up the repeated annotation join clause. * refactor: deduplicate child playlist IDs in Criteria Signed-off-by: Deluan <deluan@navidrome.org> * refactor: simplify withSmartPlaylistOwner to accept model.User Replaced separate ownerID string and ownerIsAdmin bool parameters with a single model.User struct, reducing the field count in smartPlaylistCriteria and making the option function signature clearer. Updated all call sites and tests accordingly. * fix: handle empty sort fields and propagate child playlist load errors OrderByFields now falls back to [{title, asc}] when all user-supplied sort fields are invalid, preventing empty ORDER BY clauses that would produce invalid SQL in row_number() window functions. Also restored the original behavior where a DB error loading child playlists aborts the parent smart playlist refresh, by making refreshChildPlaylists return a bool. * refactor: log warning when no valid sort fields are found Signed-off-by: Deluan <deluan@navidrome.org> --------- Signed-off-by: Deluan <deluan@navidrome.org>
2026-04-26 14:49:59 -04:00
criteria.Criteria
owner model.User
}
feat(smartplaylists): relax playlist visibility in inPlaylist/notInPlaylist rules (#5411) * test(e2e): add end-to-end tests for smart playlists functionality Signed-off-by: Deluan <deluan@navidrome.org> * fix: enforce playlist visibility in smart playlist InPlaylist/NotInPlaylist rules Previously, the InPlaylist/NotInPlaylist smart playlist criteria only allowed referencing public playlists, regardless of who owned the smart playlist. This was too restrictive for owners referencing their own private playlists and for admins who should have unrestricted access. The fix passes the smart playlist owner's identity and admin status into the criteria SQL builder, so that: admins can reference any playlist, regular users can reference public playlists plus their own private ones, and inaccessible referenced playlists produce a warning instead of a hard error. Also prevents recursive refresh of child playlists the owner cannot access. * test(e2e): clarify user roles and fix playlist visibility tests Renamed testUser/otherUser to adminUser/regularUser to make the admin vs regular user distinction explicit in test code. Fixed three playlist visibility tests that were evaluating as admin (bypassing all access checks) instead of as a regular user, so the public playlist path is now actually exercised. All playlist operator tests now use explicit evaluateRuleAs calls with the appropriate user role. * fix: sync rulesSQL criteria after limitPercent resolution The rulesSQL struct captures a copy of rules at creation time. When limitPercent is resolved later, rules.Limit is updated but rulesSQL still holds the stale value. This caused percentage-based smart playlist limits to be silently ignored. Fix by updating rulesSQL.criteria after the resolution. * refactor: convert inList to a method on smartPlaylistCriteria The inList function already receives ownerID and ownerIsAdmin from the smartPlaylistCriteria caller. Making it a method lets it access those fields directly from the receiver, simplifying the signature and staying consistent with exprSQL which was already converted to a method. * refactor: simplify function signatures by removing type parameters in criteria_sql.go Signed-off-by: Deluan <deluan@navidrome.org> --------- Signed-off-by: Deluan <deluan@navidrome.org>
2026-04-25 14:59:06 -04:00
func newSmartPlaylistCriteria(c criteria.Criteria, opts ...func(*smartPlaylistCriteria)) smartPlaylistCriteria {
refactor: centralize criteria sort parsing and extract smart playlist logic (#5415) * test: add tests for recordingdate alias resolution in smart playlists Signed-off-by: Deluan <deluan@navidrome.org> * refactor: update FieldInfo structure and simplify fieldMap initialization Signed-off-by: Deluan <deluan@navidrome.org> * refactor: move sort parsing logic from persistence to criteria package Extracted sort field parsing, validation, and direction handling from persistence/criteria_sql.go into model/criteria/sort.go. The new OrderByFields method on Criteria parses the Sort/Order strings into validated SortField structs (field name + direction), resolving aliases and handling +/- prefixes and order inversion. The persistence layer now consumes these parsed fields and only handles SQL expression mapping. This centralizes sort parsing to enforce consistent implementations. * refactor: standardize field access in smartPlaylistCriteria structure Signed-off-by: Deluan <deluan@navidrome.org> * refactor: add ResolveLimit method to Criteria Moved the percentage-limit resolution logic from playlist_repository into Criteria.ResolveLimit, replacing the 3-line mutate-after-query pattern with a single method call. The method preserves LimitPercent rather than zeroing it, since IsPercentageLimit already returns false once Limit is set, making the clear redundant and lossy. * refactor: improve child playlist loading and error handling in refresh logic Signed-off-by: Deluan <deluan@navidrome.org> * refactor: extract smart playlist logic to dedicated files Moved refreshSmartPlaylist, addSmartPlaylistAnnotationJoins, and addCriteria methods from playlist_repository.go to a new smart_playlist_repository.go file. Extracted all smart playlist tests to smart_playlist_repository_test.go. Added DeferCleanup to the "valid rules" test to fix ordering flakiness when Ginkgo randomizes test execution across files. * refactor: break refreshSmartPlaylist into smaller focused methods Split the monolithic refreshSmartPlaylist method into discrete helpers for readability: shouldRefreshSmartPlaylist for guard checks, refreshChildPlaylists for recursive dependency refresh, resolvePercentageLimit for count-based limit resolution, buildSmartPlaylistQuery for assembling the SELECT with joins, and addMediaFileAnnotationJoin to DRY up the repeated annotation join clause. * refactor: deduplicate child playlist IDs in Criteria Signed-off-by: Deluan <deluan@navidrome.org> * refactor: simplify withSmartPlaylistOwner to accept model.User Replaced separate ownerID string and ownerIsAdmin bool parameters with a single model.User struct, reducing the field count in smartPlaylistCriteria and making the option function signature clearer. Updated all call sites and tests accordingly. * fix: handle empty sort fields and propagate child playlist load errors OrderByFields now falls back to [{title, asc}] when all user-supplied sort fields are invalid, preventing empty ORDER BY clauses that would produce invalid SQL in row_number() window functions. Also restored the original behavior where a DB error loading child playlists aborts the parent smart playlist refresh, by making refreshChildPlaylists return a bool. * refactor: log warning when no valid sort fields are found Signed-off-by: Deluan <deluan@navidrome.org> --------- Signed-off-by: Deluan <deluan@navidrome.org>
2026-04-26 14:49:59 -04:00
cSQL := smartPlaylistCriteria{Criteria: c}
feat(smartplaylists): relax playlist visibility in inPlaylist/notInPlaylist rules (#5411) * test(e2e): add end-to-end tests for smart playlists functionality Signed-off-by: Deluan <deluan@navidrome.org> * fix: enforce playlist visibility in smart playlist InPlaylist/NotInPlaylist rules Previously, the InPlaylist/NotInPlaylist smart playlist criteria only allowed referencing public playlists, regardless of who owned the smart playlist. This was too restrictive for owners referencing their own private playlists and for admins who should have unrestricted access. The fix passes the smart playlist owner's identity and admin status into the criteria SQL builder, so that: admins can reference any playlist, regular users can reference public playlists plus their own private ones, and inaccessible referenced playlists produce a warning instead of a hard error. Also prevents recursive refresh of child playlists the owner cannot access. * test(e2e): clarify user roles and fix playlist visibility tests Renamed testUser/otherUser to adminUser/regularUser to make the admin vs regular user distinction explicit in test code. Fixed three playlist visibility tests that were evaluating as admin (bypassing all access checks) instead of as a regular user, so the public playlist path is now actually exercised. All playlist operator tests now use explicit evaluateRuleAs calls with the appropriate user role. * fix: sync rulesSQL criteria after limitPercent resolution The rulesSQL struct captures a copy of rules at creation time. When limitPercent is resolved later, rules.Limit is updated but rulesSQL still holds the stale value. This caused percentage-based smart playlist limits to be silently ignored. Fix by updating rulesSQL.criteria after the resolution. * refactor: convert inList to a method on smartPlaylistCriteria The inList function already receives ownerID and ownerIsAdmin from the smartPlaylistCriteria caller. Making it a method lets it access those fields directly from the receiver, simplifying the signature and staying consistent with exprSQL which was already converted to a method. * refactor: simplify function signatures by removing type parameters in criteria_sql.go Signed-off-by: Deluan <deluan@navidrome.org> --------- Signed-off-by: Deluan <deluan@navidrome.org>
2026-04-25 14:59:06 -04:00
for _, opt := range opts {
opt(&cSQL)
}
return cSQL
}
refactor: centralize criteria sort parsing and extract smart playlist logic (#5415) * test: add tests for recordingdate alias resolution in smart playlists Signed-off-by: Deluan <deluan@navidrome.org> * refactor: update FieldInfo structure and simplify fieldMap initialization Signed-off-by: Deluan <deluan@navidrome.org> * refactor: move sort parsing logic from persistence to criteria package Extracted sort field parsing, validation, and direction handling from persistence/criteria_sql.go into model/criteria/sort.go. The new OrderByFields method on Criteria parses the Sort/Order strings into validated SortField structs (field name + direction), resolving aliases and handling +/- prefixes and order inversion. The persistence layer now consumes these parsed fields and only handles SQL expression mapping. This centralizes sort parsing to enforce consistent implementations. * refactor: standardize field access in smartPlaylistCriteria structure Signed-off-by: Deluan <deluan@navidrome.org> * refactor: add ResolveLimit method to Criteria Moved the percentage-limit resolution logic from playlist_repository into Criteria.ResolveLimit, replacing the 3-line mutate-after-query pattern with a single method call. The method preserves LimitPercent rather than zeroing it, since IsPercentageLimit already returns false once Limit is set, making the clear redundant and lossy. * refactor: improve child playlist loading and error handling in refresh logic Signed-off-by: Deluan <deluan@navidrome.org> * refactor: extract smart playlist logic to dedicated files Moved refreshSmartPlaylist, addSmartPlaylistAnnotationJoins, and addCriteria methods from playlist_repository.go to a new smart_playlist_repository.go file. Extracted all smart playlist tests to smart_playlist_repository_test.go. Added DeferCleanup to the "valid rules" test to fix ordering flakiness when Ginkgo randomizes test execution across files. * refactor: break refreshSmartPlaylist into smaller focused methods Split the monolithic refreshSmartPlaylist method into discrete helpers for readability: shouldRefreshSmartPlaylist for guard checks, refreshChildPlaylists for recursive dependency refresh, resolvePercentageLimit for count-based limit resolution, buildSmartPlaylistQuery for assembling the SELECT with joins, and addMediaFileAnnotationJoin to DRY up the repeated annotation join clause. * refactor: deduplicate child playlist IDs in Criteria Signed-off-by: Deluan <deluan@navidrome.org> * refactor: simplify withSmartPlaylistOwner to accept model.User Replaced separate ownerID string and ownerIsAdmin bool parameters with a single model.User struct, reducing the field count in smartPlaylistCriteria and making the option function signature clearer. Updated all call sites and tests accordingly. * fix: handle empty sort fields and propagate child playlist load errors OrderByFields now falls back to [{title, asc}] when all user-supplied sort fields are invalid, preventing empty ORDER BY clauses that would produce invalid SQL in row_number() window functions. Also restored the original behavior where a DB error loading child playlists aborts the parent smart playlist refresh, by making refreshChildPlaylists return a bool. * refactor: log warning when no valid sort fields are found Signed-off-by: Deluan <deluan@navidrome.org> --------- Signed-off-by: Deluan <deluan@navidrome.org>
2026-04-26 14:49:59 -04:00
func withSmartPlaylistOwner(owner model.User) func(*smartPlaylistCriteria) {
feat(smartplaylists): relax playlist visibility in inPlaylist/notInPlaylist rules (#5411) * test(e2e): add end-to-end tests for smart playlists functionality Signed-off-by: Deluan <deluan@navidrome.org> * fix: enforce playlist visibility in smart playlist InPlaylist/NotInPlaylist rules Previously, the InPlaylist/NotInPlaylist smart playlist criteria only allowed referencing public playlists, regardless of who owned the smart playlist. This was too restrictive for owners referencing their own private playlists and for admins who should have unrestricted access. The fix passes the smart playlist owner's identity and admin status into the criteria SQL builder, so that: admins can reference any playlist, regular users can reference public playlists plus their own private ones, and inaccessible referenced playlists produce a warning instead of a hard error. Also prevents recursive refresh of child playlists the owner cannot access. * test(e2e): clarify user roles and fix playlist visibility tests Renamed testUser/otherUser to adminUser/regularUser to make the admin vs regular user distinction explicit in test code. Fixed three playlist visibility tests that were evaluating as admin (bypassing all access checks) instead of as a regular user, so the public playlist path is now actually exercised. All playlist operator tests now use explicit evaluateRuleAs calls with the appropriate user role. * fix: sync rulesSQL criteria after limitPercent resolution The rulesSQL struct captures a copy of rules at creation time. When limitPercent is resolved later, rules.Limit is updated but rulesSQL still holds the stale value. This caused percentage-based smart playlist limits to be silently ignored. Fix by updating rulesSQL.criteria after the resolution. * refactor: convert inList to a method on smartPlaylistCriteria The inList function already receives ownerID and ownerIsAdmin from the smartPlaylistCriteria caller. Making it a method lets it access those fields directly from the receiver, simplifying the signature and staying consistent with exprSQL which was already converted to a method. * refactor: simplify function signatures by removing type parameters in criteria_sql.go Signed-off-by: Deluan <deluan@navidrome.org> --------- Signed-off-by: Deluan <deluan@navidrome.org>
2026-04-25 14:59:06 -04:00
return func(c *smartPlaylistCriteria) {
refactor: centralize criteria sort parsing and extract smart playlist logic (#5415) * test: add tests for recordingdate alias resolution in smart playlists Signed-off-by: Deluan <deluan@navidrome.org> * refactor: update FieldInfo structure and simplify fieldMap initialization Signed-off-by: Deluan <deluan@navidrome.org> * refactor: move sort parsing logic from persistence to criteria package Extracted sort field parsing, validation, and direction handling from persistence/criteria_sql.go into model/criteria/sort.go. The new OrderByFields method on Criteria parses the Sort/Order strings into validated SortField structs (field name + direction), resolving aliases and handling +/- prefixes and order inversion. The persistence layer now consumes these parsed fields and only handles SQL expression mapping. This centralizes sort parsing to enforce consistent implementations. * refactor: standardize field access in smartPlaylistCriteria structure Signed-off-by: Deluan <deluan@navidrome.org> * refactor: add ResolveLimit method to Criteria Moved the percentage-limit resolution logic from playlist_repository into Criteria.ResolveLimit, replacing the 3-line mutate-after-query pattern with a single method call. The method preserves LimitPercent rather than zeroing it, since IsPercentageLimit already returns false once Limit is set, making the clear redundant and lossy. * refactor: improve child playlist loading and error handling in refresh logic Signed-off-by: Deluan <deluan@navidrome.org> * refactor: extract smart playlist logic to dedicated files Moved refreshSmartPlaylist, addSmartPlaylistAnnotationJoins, and addCriteria methods from playlist_repository.go to a new smart_playlist_repository.go file. Extracted all smart playlist tests to smart_playlist_repository_test.go. Added DeferCleanup to the "valid rules" test to fix ordering flakiness when Ginkgo randomizes test execution across files. * refactor: break refreshSmartPlaylist into smaller focused methods Split the monolithic refreshSmartPlaylist method into discrete helpers for readability: shouldRefreshSmartPlaylist for guard checks, refreshChildPlaylists for recursive dependency refresh, resolvePercentageLimit for count-based limit resolution, buildSmartPlaylistQuery for assembling the SELECT with joins, and addMediaFileAnnotationJoin to DRY up the repeated annotation join clause. * refactor: deduplicate child playlist IDs in Criteria Signed-off-by: Deluan <deluan@navidrome.org> * refactor: simplify withSmartPlaylistOwner to accept model.User Replaced separate ownerID string and ownerIsAdmin bool parameters with a single model.User struct, reducing the field count in smartPlaylistCriteria and making the option function signature clearer. Updated all call sites and tests accordingly. * fix: handle empty sort fields and propagate child playlist load errors OrderByFields now falls back to [{title, asc}] when all user-supplied sort fields are invalid, preventing empty ORDER BY clauses that would produce invalid SQL in row_number() window functions. Also restored the original behavior where a DB error loading child playlists aborts the parent smart playlist refresh, by making refreshChildPlaylists return a bool. * refactor: log warning when no valid sort fields are found Signed-off-by: Deluan <deluan@navidrome.org> --------- Signed-off-by: Deluan <deluan@navidrome.org>
2026-04-26 14:49:59 -04:00
c.owner = owner
feat(smartplaylists): relax playlist visibility in inPlaylist/notInPlaylist rules (#5411) * test(e2e): add end-to-end tests for smart playlists functionality Signed-off-by: Deluan <deluan@navidrome.org> * fix: enforce playlist visibility in smart playlist InPlaylist/NotInPlaylist rules Previously, the InPlaylist/NotInPlaylist smart playlist criteria only allowed referencing public playlists, regardless of who owned the smart playlist. This was too restrictive for owners referencing their own private playlists and for admins who should have unrestricted access. The fix passes the smart playlist owner's identity and admin status into the criteria SQL builder, so that: admins can reference any playlist, regular users can reference public playlists plus their own private ones, and inaccessible referenced playlists produce a warning instead of a hard error. Also prevents recursive refresh of child playlists the owner cannot access. * test(e2e): clarify user roles and fix playlist visibility tests Renamed testUser/otherUser to adminUser/regularUser to make the admin vs regular user distinction explicit in test code. Fixed three playlist visibility tests that were evaluating as admin (bypassing all access checks) instead of as a regular user, so the public playlist path is now actually exercised. All playlist operator tests now use explicit evaluateRuleAs calls with the appropriate user role. * fix: sync rulesSQL criteria after limitPercent resolution The rulesSQL struct captures a copy of rules at creation time. When limitPercent is resolved later, rules.Limit is updated but rulesSQL still holds the stale value. This caused percentage-based smart playlist limits to be silently ignored. Fix by updating rulesSQL.criteria after the resolution. * refactor: convert inList to a method on smartPlaylistCriteria The inList function already receives ownerID and ownerIsAdmin from the smartPlaylistCriteria caller. Making it a method lets it access those fields directly from the receiver, simplifying the signature and staying consistent with exprSQL which was already converted to a method. * refactor: simplify function signatures by removing type parameters in criteria_sql.go Signed-off-by: Deluan <deluan@navidrome.org> --------- Signed-off-by: Deluan <deluan@navidrome.org>
2026-04-25 14:59:06 -04:00
}
}
var smartPlaylistFields = map[string]smartPlaylistField{
"title": {expr: "media_file.title"},
"album": {expr: "media_file.album"},
"hascoverart": {expr: "media_file.has_cover_art"},
"tracknumber": {expr: "media_file.track_number"},
"discnumber": {expr: "media_file.disc_number"},
"year": {expr: "media_file.year"},
"date": {expr: "media_file.date"},
"originalyear": {expr: "media_file.original_year"},
"originaldate": {expr: "media_file.original_date"},
"releaseyear": {expr: "media_file.release_year"},
"releasedate": {expr: "media_file.release_date"},
"size": {expr: "media_file.size"},
"compilation": {expr: "media_file.compilation"},
"missing": {expr: "media_file.missing"},
"explicitstatus": {expr: "media_file.explicit_status"},
"dateadded": {expr: "media_file.created_at"},
"datemodified": {expr: "media_file.updated_at"},
"discsubtitle": {expr: "media_file.disc_subtitle"},
"comment": {expr: "media_file.comment"},
feat(smartplaylist): extend isMissing/isPresent to bpm, bitDepth and many text fields (#5603) * feat(smartplaylist): support isMissing/isPresent on mbz_* and lyrics fields Mark the six mbz_* MusicBrainz ID columns and the lyrics column as Nullable in the criteria field map, then extend missingExpr to handle string columns where absence is encoded as NULL or empty string (plus '[]' for lyrics). The Numeric/Boolean path (ReplayGain) is preserved via an explicit type check. * refactor(model): make MediaFile BPM and BitDepth nullable pointers Convert BPM and BitDepth fields in model.MediaFile from int to *int so that 'tag absent' is distinguishable from zero. The metadata mapper now uses NullableFloat for BPM (nil when absent or zero/unparseable) and only sets BitDepth when the audio property is non-zero (lossy codecs report 0). All read sites use gg.V() for zero-fallback deref so Subsonic API output and transcoding behaviour are byte-identical to before. The persistence layer bridges the existing NOT NULL DB columns by coercing nil to 0 on write and 0 back to nil on read in PostMapArgs/PostScan; a later migration task will drop those constraints. Hash upgrade safety is verified by a new MediaFile.Hash describe block: nil *int hashes identically to the old int(0) default via ZeroNil+IgnoreZeroValue, so no files will be spuriously re-imported after this change. Extra files touched beyond the plan's list: core/stream/legacy_client_test.go (BitDepth in model.MediaFile literals), persistence/mediafile_repository.go (NOT NULL bridge). * test(model): pin pre-conversion golden hashes for BPM/BitDepth * feat(smartplaylist): support isMissing/isPresent on bpm and bitDepth * feat(db): make bpm and bit_depth columns nullable, backfill 0 to NULL Drop the NOT NULL constraint on media_file.bpm and bit_depth via a lossless migration that converts legacy 0-means-absent values to real NULL. Remove the temporary shim in PostScan/PostMapArgs that was bridging the old NOT NULL columns to the *int model fields. Add round-trip persistence tests asserting NULL storage for nil pointers and correct value round-trip for non-nil pointers. * test(e2e): verify isMissing/isPresent partition for nullable fields Add DescribeTable covering bpm, bitdepth, lyrics, and mbz_recording_id: for each field, isMissing + isPresent song counts must equal the total library count, proving the nullable-column SQL is exhaustive and correct. * test(e2e): seed bpm tag so isMissing/isPresent partition is non-trivial * fix(model): omit bitDepth from JSON when absent instead of emitting null * feat(smartplaylist): support isMissing/isPresent on more string fields Enable isMissing/isPresent operators for album, comment, catalognumber, discsubtitle, albumcomment, sorttitle, sortalbum, sortartist, sortalbumartist, and explicitstatus by marking them Nullable in fieldMap. * refactor(smartplaylist): unify missingExpr column logic into one flow Collapse the numeric/string fork in missingExpr into a single empties-driven loop (numeric/boolean fields simply have no empties), and replace the duplicated IsTag/IsRole guard with a three-way switch that expresses the dispatch model once. No SQL semantics change for string fields; numeric/boolean fields now emit a single-element Or/And which squirrel parenthesizes (e.g. `(col IS NULL)` instead of bare `col IS NULL`) — update the affected test expectations accordingly.
2026-06-13 13:15:20 -04:00
"lyrics": {expr: "media_file.lyrics", emptyValues: []string{"[]"}},
"sorttitle": {expr: "media_file.sort_title"},
"sortalbum": {expr: "media_file.sort_album_name"},
"sortartist": {expr: "media_file.sort_artist_name"},
"sortalbumartist": {expr: "media_file.sort_album_artist_name"},
"albumcomment": {expr: "media_file.mbz_album_comment"},
"catalognumber": {expr: "media_file.catalog_num"},
"filepath": {expr: "media_file.path"},
"filetype": {expr: "media_file.suffix"},
"codec": {expr: "media_file.codec"},
"duration": {expr: "media_file.duration"},
"bitrate": {expr: "media_file.bit_rate"},
"bitdepth": {expr: "media_file.bit_depth"},
"samplerate": {expr: "media_file.sample_rate"},
"bpm": {expr: "media_file.bpm"},
"channels": {expr: "media_file.channels"},
"loved": {expr: "COALESCE(annotation.starred, false)"},
"dateloved": {expr: "annotation.starred_at"},
"lastplayed": {expr: "annotation.play_date"},
"daterated": {expr: "annotation.rated_at"},
"playcount": {expr: "COALESCE(annotation.play_count, 0)"},
"rating": {expr: "COALESCE(annotation.rating, 0)"},
"averagerating": {expr: "media_file.average_rating"},
"albumrating": {expr: "COALESCE(album_annotation.rating, 0)", joinType: smartPlaylistJoinAlbumAnnotation},
"albumloved": {expr: "COALESCE(album_annotation.starred, false)", joinType: smartPlaylistJoinAlbumAnnotation},
"albumplaycount": {expr: "COALESCE(album_annotation.play_count, 0)", joinType: smartPlaylistJoinAlbumAnnotation},
"albumlastplayed": {expr: "album_annotation.play_date", joinType: smartPlaylistJoinAlbumAnnotation},
"albumdateloved": {expr: "album_annotation.starred_at", joinType: smartPlaylistJoinAlbumAnnotation},
"albumdaterated": {expr: "album_annotation.rated_at", joinType: smartPlaylistJoinAlbumAnnotation},
"artistrating": {expr: "COALESCE(artist_annotation.rating, 0)", joinType: smartPlaylistJoinArtistAnnotation},
"artistloved": {expr: "COALESCE(artist_annotation.starred, false)", joinType: smartPlaylistJoinArtistAnnotation},
"artistplaycount": {expr: "COALESCE(artist_annotation.play_count, 0)", joinType: smartPlaylistJoinArtistAnnotation},
"artistlastplayed": {expr: "artist_annotation.play_date", joinType: smartPlaylistJoinArtistAnnotation},
"artistdateloved": {expr: "artist_annotation.starred_at", joinType: smartPlaylistJoinArtistAnnotation},
"artistdaterated": {expr: "artist_annotation.rated_at", joinType: smartPlaylistJoinArtistAnnotation},
"mbz_album_id": {expr: "media_file.mbz_album_id"},
"mbz_album_artist_id": {expr: "media_file.mbz_album_artist_id"},
"mbz_artist_id": {expr: "media_file.mbz_artist_id"},
"mbz_recording_id": {expr: "media_file.mbz_recording_id"},
"mbz_release_track_id": {expr: "media_file.mbz_release_track_id"},
"mbz_release_group_id": {expr: "media_file.mbz_release_group_id"},
"rgalbumgain": {expr: "media_file.rg_album_gain"},
"rgalbumpeak": {expr: "media_file.rg_album_peak"},
"rgtrackgain": {expr: "media_file.rg_track_gain"},
"rgtrackpeak": {expr: "media_file.rg_track_peak"},
"library_id": {expr: "media_file.library_id"},
"random": {order: "random()"},
}
func (c smartPlaylistCriteria) Where() (squirrel.Sqlizer, error) {
refactor: centralize criteria sort parsing and extract smart playlist logic (#5415) * test: add tests for recordingdate alias resolution in smart playlists Signed-off-by: Deluan <deluan@navidrome.org> * refactor: update FieldInfo structure and simplify fieldMap initialization Signed-off-by: Deluan <deluan@navidrome.org> * refactor: move sort parsing logic from persistence to criteria package Extracted sort field parsing, validation, and direction handling from persistence/criteria_sql.go into model/criteria/sort.go. The new OrderByFields method on Criteria parses the Sort/Order strings into validated SortField structs (field name + direction), resolving aliases and handling +/- prefixes and order inversion. The persistence layer now consumes these parsed fields and only handles SQL expression mapping. This centralizes sort parsing to enforce consistent implementations. * refactor: standardize field access in smartPlaylistCriteria structure Signed-off-by: Deluan <deluan@navidrome.org> * refactor: add ResolveLimit method to Criteria Moved the percentage-limit resolution logic from playlist_repository into Criteria.ResolveLimit, replacing the 3-line mutate-after-query pattern with a single method call. The method preserves LimitPercent rather than zeroing it, since IsPercentageLimit already returns false once Limit is set, making the clear redundant and lossy. * refactor: improve child playlist loading and error handling in refresh logic Signed-off-by: Deluan <deluan@navidrome.org> * refactor: extract smart playlist logic to dedicated files Moved refreshSmartPlaylist, addSmartPlaylistAnnotationJoins, and addCriteria methods from playlist_repository.go to a new smart_playlist_repository.go file. Extracted all smart playlist tests to smart_playlist_repository_test.go. Added DeferCleanup to the "valid rules" test to fix ordering flakiness when Ginkgo randomizes test execution across files. * refactor: break refreshSmartPlaylist into smaller focused methods Split the monolithic refreshSmartPlaylist method into discrete helpers for readability: shouldRefreshSmartPlaylist for guard checks, refreshChildPlaylists for recursive dependency refresh, resolvePercentageLimit for count-based limit resolution, buildSmartPlaylistQuery for assembling the SELECT with joins, and addMediaFileAnnotationJoin to DRY up the repeated annotation join clause. * refactor: deduplicate child playlist IDs in Criteria Signed-off-by: Deluan <deluan@navidrome.org> * refactor: simplify withSmartPlaylistOwner to accept model.User Replaced separate ownerID string and ownerIsAdmin bool parameters with a single model.User struct, reducing the field count in smartPlaylistCriteria and making the option function signature clearer. Updated all call sites and tests accordingly. * fix: handle empty sort fields and propagate child playlist load errors OrderByFields now falls back to [{title, asc}] when all user-supplied sort fields are invalid, preventing empty ORDER BY clauses that would produce invalid SQL in row_number() window functions. Also restored the original behavior where a DB error loading child playlists aborts the parent smart playlist refresh, by making refreshChildPlaylists return a bool. * refactor: log warning when no valid sort fields are found Signed-off-by: Deluan <deluan@navidrome.org> --------- Signed-off-by: Deluan <deluan@navidrome.org>
2026-04-26 14:49:59 -04:00
if c.Criteria.Expression == nil {
return squirrel.Expr("1 = 1"), nil
}
refactor: centralize criteria sort parsing and extract smart playlist logic (#5415) * test: add tests for recordingdate alias resolution in smart playlists Signed-off-by: Deluan <deluan@navidrome.org> * refactor: update FieldInfo structure and simplify fieldMap initialization Signed-off-by: Deluan <deluan@navidrome.org> * refactor: move sort parsing logic from persistence to criteria package Extracted sort field parsing, validation, and direction handling from persistence/criteria_sql.go into model/criteria/sort.go. The new OrderByFields method on Criteria parses the Sort/Order strings into validated SortField structs (field name + direction), resolving aliases and handling +/- prefixes and order inversion. The persistence layer now consumes these parsed fields and only handles SQL expression mapping. This centralizes sort parsing to enforce consistent implementations. * refactor: standardize field access in smartPlaylistCriteria structure Signed-off-by: Deluan <deluan@navidrome.org> * refactor: add ResolveLimit method to Criteria Moved the percentage-limit resolution logic from playlist_repository into Criteria.ResolveLimit, replacing the 3-line mutate-after-query pattern with a single method call. The method preserves LimitPercent rather than zeroing it, since IsPercentageLimit already returns false once Limit is set, making the clear redundant and lossy. * refactor: improve child playlist loading and error handling in refresh logic Signed-off-by: Deluan <deluan@navidrome.org> * refactor: extract smart playlist logic to dedicated files Moved refreshSmartPlaylist, addSmartPlaylistAnnotationJoins, and addCriteria methods from playlist_repository.go to a new smart_playlist_repository.go file. Extracted all smart playlist tests to smart_playlist_repository_test.go. Added DeferCleanup to the "valid rules" test to fix ordering flakiness when Ginkgo randomizes test execution across files. * refactor: break refreshSmartPlaylist into smaller focused methods Split the monolithic refreshSmartPlaylist method into discrete helpers for readability: shouldRefreshSmartPlaylist for guard checks, refreshChildPlaylists for recursive dependency refresh, resolvePercentageLimit for count-based limit resolution, buildSmartPlaylistQuery for assembling the SELECT with joins, and addMediaFileAnnotationJoin to DRY up the repeated annotation join clause. * refactor: deduplicate child playlist IDs in Criteria Signed-off-by: Deluan <deluan@navidrome.org> * refactor: simplify withSmartPlaylistOwner to accept model.User Replaced separate ownerID string and ownerIsAdmin bool parameters with a single model.User struct, reducing the field count in smartPlaylistCriteria and making the option function signature clearer. Updated all call sites and tests accordingly. * fix: handle empty sort fields and propagate child playlist load errors OrderByFields now falls back to [{title, asc}] when all user-supplied sort fields are invalid, preventing empty ORDER BY clauses that would produce invalid SQL in row_number() window functions. Also restored the original behavior where a DB error loading child playlists aborts the parent smart playlist refresh, by making refreshChildPlaylists return a bool. * refactor: log warning when no valid sort fields are found Signed-off-by: Deluan <deluan@navidrome.org> --------- Signed-off-by: Deluan <deluan@navidrome.org>
2026-04-26 14:49:59 -04:00
return c.exprSQL(c.Criteria.Expression)
}
feat(smartplaylists): relax playlist visibility in inPlaylist/notInPlaylist rules (#5411) * test(e2e): add end-to-end tests for smart playlists functionality Signed-off-by: Deluan <deluan@navidrome.org> * fix: enforce playlist visibility in smart playlist InPlaylist/NotInPlaylist rules Previously, the InPlaylist/NotInPlaylist smart playlist criteria only allowed referencing public playlists, regardless of who owned the smart playlist. This was too restrictive for owners referencing their own private playlists and for admins who should have unrestricted access. The fix passes the smart playlist owner's identity and admin status into the criteria SQL builder, so that: admins can reference any playlist, regular users can reference public playlists plus their own private ones, and inaccessible referenced playlists produce a warning instead of a hard error. Also prevents recursive refresh of child playlists the owner cannot access. * test(e2e): clarify user roles and fix playlist visibility tests Renamed testUser/otherUser to adminUser/regularUser to make the admin vs regular user distinction explicit in test code. Fixed three playlist visibility tests that were evaluating as admin (bypassing all access checks) instead of as a regular user, so the public playlist path is now actually exercised. All playlist operator tests now use explicit evaluateRuleAs calls with the appropriate user role. * fix: sync rulesSQL criteria after limitPercent resolution The rulesSQL struct captures a copy of rules at creation time. When limitPercent is resolved later, rules.Limit is updated but rulesSQL still holds the stale value. This caused percentage-based smart playlist limits to be silently ignored. Fix by updating rulesSQL.criteria after the resolution. * refactor: convert inList to a method on smartPlaylistCriteria The inList function already receives ownerID and ownerIsAdmin from the smartPlaylistCriteria caller. Making it a method lets it access those fields directly from the receiver, simplifying the signature and staying consistent with exprSQL which was already converted to a method. * refactor: simplify function signatures by removing type parameters in criteria_sql.go Signed-off-by: Deluan <deluan@navidrome.org> --------- Signed-off-by: Deluan <deluan@navidrome.org>
2026-04-25 14:59:06 -04:00
func (c smartPlaylistCriteria) exprSQL(expr criteria.Expression) (squirrel.Sqlizer, error) {
switch e := expr.(type) {
case criteria.All:
and := squirrel.And{}
for _, child := range e {
feat(smartplaylists): relax playlist visibility in inPlaylist/notInPlaylist rules (#5411) * test(e2e): add end-to-end tests for smart playlists functionality Signed-off-by: Deluan <deluan@navidrome.org> * fix: enforce playlist visibility in smart playlist InPlaylist/NotInPlaylist rules Previously, the InPlaylist/NotInPlaylist smart playlist criteria only allowed referencing public playlists, regardless of who owned the smart playlist. This was too restrictive for owners referencing their own private playlists and for admins who should have unrestricted access. The fix passes the smart playlist owner's identity and admin status into the criteria SQL builder, so that: admins can reference any playlist, regular users can reference public playlists plus their own private ones, and inaccessible referenced playlists produce a warning instead of a hard error. Also prevents recursive refresh of child playlists the owner cannot access. * test(e2e): clarify user roles and fix playlist visibility tests Renamed testUser/otherUser to adminUser/regularUser to make the admin vs regular user distinction explicit in test code. Fixed three playlist visibility tests that were evaluating as admin (bypassing all access checks) instead of as a regular user, so the public playlist path is now actually exercised. All playlist operator tests now use explicit evaluateRuleAs calls with the appropriate user role. * fix: sync rulesSQL criteria after limitPercent resolution The rulesSQL struct captures a copy of rules at creation time. When limitPercent is resolved later, rules.Limit is updated but rulesSQL still holds the stale value. This caused percentage-based smart playlist limits to be silently ignored. Fix by updating rulesSQL.criteria after the resolution. * refactor: convert inList to a method on smartPlaylistCriteria The inList function already receives ownerID and ownerIsAdmin from the smartPlaylistCriteria caller. Making it a method lets it access those fields directly from the receiver, simplifying the signature and staying consistent with exprSQL which was already converted to a method. * refactor: simplify function signatures by removing type parameters in criteria_sql.go Signed-off-by: Deluan <deluan@navidrome.org> --------- Signed-off-by: Deluan <deluan@navidrome.org>
2026-04-25 14:59:06 -04:00
cond, err := c.exprSQL(child)
if err != nil {
return nil, err
}
and = append(and, cond)
}
return mergeNegatedJsonConds(and), nil
case criteria.Any:
or := squirrel.Or{}
for _, child := range e {
feat(smartplaylists): relax playlist visibility in inPlaylist/notInPlaylist rules (#5411) * test(e2e): add end-to-end tests for smart playlists functionality Signed-off-by: Deluan <deluan@navidrome.org> * fix: enforce playlist visibility in smart playlist InPlaylist/NotInPlaylist rules Previously, the InPlaylist/NotInPlaylist smart playlist criteria only allowed referencing public playlists, regardless of who owned the smart playlist. This was too restrictive for owners referencing their own private playlists and for admins who should have unrestricted access. The fix passes the smart playlist owner's identity and admin status into the criteria SQL builder, so that: admins can reference any playlist, regular users can reference public playlists plus their own private ones, and inaccessible referenced playlists produce a warning instead of a hard error. Also prevents recursive refresh of child playlists the owner cannot access. * test(e2e): clarify user roles and fix playlist visibility tests Renamed testUser/otherUser to adminUser/regularUser to make the admin vs regular user distinction explicit in test code. Fixed three playlist visibility tests that were evaluating as admin (bypassing all access checks) instead of as a regular user, so the public playlist path is now actually exercised. All playlist operator tests now use explicit evaluateRuleAs calls with the appropriate user role. * fix: sync rulesSQL criteria after limitPercent resolution The rulesSQL struct captures a copy of rules at creation time. When limitPercent is resolved later, rules.Limit is updated but rulesSQL still holds the stale value. This caused percentage-based smart playlist limits to be silently ignored. Fix by updating rulesSQL.criteria after the resolution. * refactor: convert inList to a method on smartPlaylistCriteria The inList function already receives ownerID and ownerIsAdmin from the smartPlaylistCriteria caller. Making it a method lets it access those fields directly from the receiver, simplifying the signature and staying consistent with exprSQL which was already converted to a method. * refactor: simplify function signatures by removing type parameters in criteria_sql.go Signed-off-by: Deluan <deluan@navidrome.org> --------- Signed-off-by: Deluan <deluan@navidrome.org>
2026-04-25 14:59:06 -04:00
cond, err := c.exprSQL(child)
if err != nil {
return nil, err
}
or = append(or, cond)
}
fix(smartplaylists): optimize smart playlist performance for role and tag criteria (#5515) * fix(server): optimize smart playlist role queries for large criteria (#5511) Role-based smart playlist criteria (artist, composer, etc.) now query the indexed media_file_artists join table instead of parsing JSON via json_tree() on every row. Multiple conditions for the same role within an OR group are merged into a single EXISTS subquery (batched at 200 to stay under SQLite's expression tree depth limit). A composite index (media_file_id, role) replaces the now-redundant single-column (media_file_id) index on media_file_artists. Benchmark (40k tracks, 500 patterns, 3 artists/track): - Merged join-table: 15ms (9.3x faster) - Merged json_tree: 30ms (4.6x faster) - Unmerged baseline: 137ms * refactor: simplify role condition SQL generation and benchmark Extract shared roleCondSQL/roleExistsSQL helpers to deduplicate the EXISTS template between roleCond and roleCondGroup. Use slices.Chunk for batching per project convention. Extract runBenchQuery helper to eliminate triplicated benchmark execution loop. * chore: raise roleCondBatchSize to 350 The empirical SQLite limit is 496 conditions per merged EXISTS subquery. Raising from 200 to 350 reduces the number of batches (e.g. 500 patterns now splits into 2 batches instead of 3). * fix(server): apply OR-merge optimization to tag conditions too Generalize mergeRoleConds into mergeJsonConds to also collapse multiple tag conditions for the same tag (e.g. genre) within OR groups. This gives the same ~5x speedup for tag-heavy smart playlists as the role optimization gives for artist-heavy ones. * refactor: benchmark uses real criteria pipeline instead of hand-built SQL The "Current" sub-benchmark now builds criteria.Criteria expressions and runs them through the actual newSmartPlaylistCriteria → Where() → ToSql() pipeline, validating the real production code path. The baseline still uses hand-built SQL representing the old json_tree approach. * fix: stabilize merged group ordering and close rows before error check Sort group keys in mergeJsonConds so the merged additions have deterministic order across runs, improving SQLite statement cache reuse. Move rows.Close() before rows.Err() in benchmark helper.
2026-05-22 18:00:13 -03:00
return mergeJsonConds(or), nil
case criteria.Is:
return mapExpr(e, func(fields map[string]any) squirrel.Sqlizer {
return squirrel.Eq(fields)
}, false)
case criteria.IsNot:
return isNotExpr(e)
case criteria.Gt:
return mapExpr(e, func(fields map[string]any) squirrel.Sqlizer {
return squirrel.Gt(fields)
}, false)
case criteria.Lt:
return mapExpr(e, func(fields map[string]any) squirrel.Sqlizer {
return squirrel.Lt(fields)
}, false)
case criteria.Before:
return mapExpr(e, func(fields map[string]any) squirrel.Sqlizer {
return squirrel.Lt(fields)
}, false)
case criteria.After:
return mapExpr(e, func(fields map[string]any) squirrel.Sqlizer {
return squirrel.Gt(fields)
}, false)
case criteria.Contains:
return likeExpr(e, "%%%v%%", false)
case criteria.NotContains:
return likeExpr(e, "%%%v%%", true)
case criteria.StartsWith:
return likeExpr(e, "%v%%", false)
case criteria.EndsWith:
return likeExpr(e, "%%%v", false)
case criteria.InTheRange:
return rangeExpr(e)
case criteria.InTheLast:
return periodExpr(e, false)
case criteria.NotInTheLast:
return periodExpr(e, true)
case criteria.InPlaylist:
feat(smartplaylists): relax playlist visibility in inPlaylist/notInPlaylist rules (#5411) * test(e2e): add end-to-end tests for smart playlists functionality Signed-off-by: Deluan <deluan@navidrome.org> * fix: enforce playlist visibility in smart playlist InPlaylist/NotInPlaylist rules Previously, the InPlaylist/NotInPlaylist smart playlist criteria only allowed referencing public playlists, regardless of who owned the smart playlist. This was too restrictive for owners referencing their own private playlists and for admins who should have unrestricted access. The fix passes the smart playlist owner's identity and admin status into the criteria SQL builder, so that: admins can reference any playlist, regular users can reference public playlists plus their own private ones, and inaccessible referenced playlists produce a warning instead of a hard error. Also prevents recursive refresh of child playlists the owner cannot access. * test(e2e): clarify user roles and fix playlist visibility tests Renamed testUser/otherUser to adminUser/regularUser to make the admin vs regular user distinction explicit in test code. Fixed three playlist visibility tests that were evaluating as admin (bypassing all access checks) instead of as a regular user, so the public playlist path is now actually exercised. All playlist operator tests now use explicit evaluateRuleAs calls with the appropriate user role. * fix: sync rulesSQL criteria after limitPercent resolution The rulesSQL struct captures a copy of rules at creation time. When limitPercent is resolved later, rules.Limit is updated but rulesSQL still holds the stale value. This caused percentage-based smart playlist limits to be silently ignored. Fix by updating rulesSQL.criteria after the resolution. * refactor: convert inList to a method on smartPlaylistCriteria The inList function already receives ownerID and ownerIsAdmin from the smartPlaylistCriteria caller. Making it a method lets it access those fields directly from the receiver, simplifying the signature and staying consistent with exprSQL which was already converted to a method. * refactor: simplify function signatures by removing type parameters in criteria_sql.go Signed-off-by: Deluan <deluan@navidrome.org> --------- Signed-off-by: Deluan <deluan@navidrome.org>
2026-04-25 14:59:06 -04:00
return c.inList(e, false)
case criteria.NotInPlaylist:
feat(smartplaylists): relax playlist visibility in inPlaylist/notInPlaylist rules (#5411) * test(e2e): add end-to-end tests for smart playlists functionality Signed-off-by: Deluan <deluan@navidrome.org> * fix: enforce playlist visibility in smart playlist InPlaylist/NotInPlaylist rules Previously, the InPlaylist/NotInPlaylist smart playlist criteria only allowed referencing public playlists, regardless of who owned the smart playlist. This was too restrictive for owners referencing their own private playlists and for admins who should have unrestricted access. The fix passes the smart playlist owner's identity and admin status into the criteria SQL builder, so that: admins can reference any playlist, regular users can reference public playlists plus their own private ones, and inaccessible referenced playlists produce a warning instead of a hard error. Also prevents recursive refresh of child playlists the owner cannot access. * test(e2e): clarify user roles and fix playlist visibility tests Renamed testUser/otherUser to adminUser/regularUser to make the admin vs regular user distinction explicit in test code. Fixed three playlist visibility tests that were evaluating as admin (bypassing all access checks) instead of as a regular user, so the public playlist path is now actually exercised. All playlist operator tests now use explicit evaluateRuleAs calls with the appropriate user role. * fix: sync rulesSQL criteria after limitPercent resolution The rulesSQL struct captures a copy of rules at creation time. When limitPercent is resolved later, rules.Limit is updated but rulesSQL still holds the stale value. This caused percentage-based smart playlist limits to be silently ignored. Fix by updating rulesSQL.criteria after the resolution. * refactor: convert inList to a method on smartPlaylistCriteria The inList function already receives ownerID and ownerIsAdmin from the smartPlaylistCriteria caller. Making it a method lets it access those fields directly from the receiver, simplifying the signature and staying consistent with exprSQL which was already converted to a method. * refactor: simplify function signatures by removing type parameters in criteria_sql.go Signed-off-by: Deluan <deluan@navidrome.org> --------- Signed-off-by: Deluan <deluan@navidrome.org>
2026-04-25 14:59:06 -04:00
return c.inList(e, true)
feat(smartplaylist): add isMissing and isPresent operators (#5436) * feat(smartplaylist): add IsMissing and IsPresent operator types Add two new Expression types for detecting absent/present tags and roles in smart playlist criteria. Includes JSON marshal/unmarshal support and Walk visitor registration. * test(smartplaylist): add JSON marshal/unmarshal tests for isMissing/isPresent * feat(smartplaylist): add SQL generation for isMissing/isPresent operators Tags check json_tree(media_file.tags) for key existence. Roles check json_tree(media_file.participants) for key existence. Regular DB column fields are rejected with an error. * test(smartplaylist): add e2e tests for isMissing/isPresent operators Tests cover tag presence/absence with selective matching (grouping), universal absence (lyricist role), universal presence (composer role), and combined operator usage. * refactor(smartplaylist): use strconv.ParseBool in IsTruthy Replace hand-rolled string truthiness check with strconv.ParseBool, which correctly handles standard boolean strings and rejects unrecognized values as false. * refactor(smartplaylist): clarify missingExpr parameter naming Rename defaultNegate to checkAbsence and extract truthy local for readability. The XNOR logic (checkAbsence == truthy) is now easier to follow: isMissing passes true, isPresent passes false. * refactor(smartplaylist): reuse jsonExpr in missingExpr, improve errors - tagCond/roleCond now handle nil cond (existence-only check) - missingExpr delegates to jsonExpr(info, nil, negate) instead of building SQL manually - Better error messages: unknown fields now report the field name
2026-04-28 19:40:08 -04:00
case criteria.IsMissing:
return missingExpr(e, true)
case criteria.IsPresent:
return missingExpr(e, false)
default:
return nil, fmt.Errorf("unknown criteria expression type %T", expr)
}
}
feat(smartplaylists): relax playlist visibility in inPlaylist/notInPlaylist rules (#5411) * test(e2e): add end-to-end tests for smart playlists functionality Signed-off-by: Deluan <deluan@navidrome.org> * fix: enforce playlist visibility in smart playlist InPlaylist/NotInPlaylist rules Previously, the InPlaylist/NotInPlaylist smart playlist criteria only allowed referencing public playlists, regardless of who owned the smart playlist. This was too restrictive for owners referencing their own private playlists and for admins who should have unrestricted access. The fix passes the smart playlist owner's identity and admin status into the criteria SQL builder, so that: admins can reference any playlist, regular users can reference public playlists plus their own private ones, and inaccessible referenced playlists produce a warning instead of a hard error. Also prevents recursive refresh of child playlists the owner cannot access. * test(e2e): clarify user roles and fix playlist visibility tests Renamed testUser/otherUser to adminUser/regularUser to make the admin vs regular user distinction explicit in test code. Fixed three playlist visibility tests that were evaluating as admin (bypassing all access checks) instead of as a regular user, so the public playlist path is now actually exercised. All playlist operator tests now use explicit evaluateRuleAs calls with the appropriate user role. * fix: sync rulesSQL criteria after limitPercent resolution The rulesSQL struct captures a copy of rules at creation time. When limitPercent is resolved later, rules.Limit is updated but rulesSQL still holds the stale value. This caused percentage-based smart playlist limits to be silently ignored. Fix by updating rulesSQL.criteria after the resolution. * refactor: convert inList to a method on smartPlaylistCriteria The inList function already receives ownerID and ownerIsAdmin from the smartPlaylistCriteria caller. Making it a method lets it access those fields directly from the receiver, simplifying the signature and staying consistent with exprSQL which was already converted to a method. * refactor: simplify function signatures by removing type parameters in criteria_sql.go Signed-off-by: Deluan <deluan@navidrome.org> --------- Signed-off-by: Deluan <deluan@navidrome.org>
2026-04-25 14:59:06 -04:00
func isNotExpr(values map[string]any) (squirrel.Sqlizer, error) {
if _, value, info, ok := singleField(values); ok && (info.IsTag || info.IsRole) {
return jsonExpr(info, squirrel.Eq{"value": value}, true), nil
}
fields, err := sqlFields(values)
if err != nil {
return nil, err
}
return squirrel.NotEq(fields), nil
}
feat(smartplaylist): add isMissing and isPresent operators (#5436) * feat(smartplaylist): add IsMissing and IsPresent operator types Add two new Expression types for detecting absent/present tags and roles in smart playlist criteria. Includes JSON marshal/unmarshal support and Walk visitor registration. * test(smartplaylist): add JSON marshal/unmarshal tests for isMissing/isPresent * feat(smartplaylist): add SQL generation for isMissing/isPresent operators Tags check json_tree(media_file.tags) for key existence. Roles check json_tree(media_file.participants) for key existence. Regular DB column fields are rejected with an error. * test(smartplaylist): add e2e tests for isMissing/isPresent operators Tests cover tag presence/absence with selective matching (grouping), universal absence (lyricist role), universal presence (composer role), and combined operator usage. * refactor(smartplaylist): use strconv.ParseBool in IsTruthy Replace hand-rolled string truthiness check with strconv.ParseBool, which correctly handles standard boolean strings and rejects unrecognized values as false. * refactor(smartplaylist): clarify missingExpr parameter naming Rename defaultNegate to checkAbsence and extract truthy local for readability. The XNOR logic (checkAbsence == truthy) is now easier to follow: isMissing passes true, isPresent passes false. * refactor(smartplaylist): reuse jsonExpr in missingExpr, improve errors - tagCond/roleCond now handle nil cond (existence-only check) - missingExpr delegates to jsonExpr(info, nil, negate) instead of building SQL manually - Better error messages: unknown fields now report the field name
2026-04-28 19:40:08 -04:00
func missingExpr(values map[string]any, checkAbsence bool) (squirrel.Sqlizer, error) {
field, value, info, ok := singleField(values)
if !ok {
if len(values) != 1 {
return nil, fmt.Errorf("invalid field in criteria: isMissing/isPresent requires exactly one field")
}
return nil, fmt.Errorf("invalid field in criteria: %s", field)
}
b, ok := value.(bool)
if !ok {
return nil, fmt.Errorf("invalid boolean value for 'missing' expression: %s: %v", field, value)
}
negate := checkAbsence == b
fix(smartplaylist): support isMissing/isPresent operators on ReplayGain fields (#5585) * fix(smartplaylist): support isMissing/isPresent operators on ReplayGain fields ReplayGain values are stored in dedicated nullable columns (rg_album_gain, rg_album_peak, rg_track_gain, rg_track_peak) rather than in the media_file.tags JSON blob. The isMissing/isPresent operators previously only supported tag and role fields, causing two failure modes: 1. Using the documented alias names (replaygain_album_gain etc.) from PR #5256: these got registered as JSON tags from mappings.yaml, so isMissing queried json_tree(media_file.tags, '$.replaygain_album_gain') which is always empty -> the playlist matched ALL songs. 2. Using the canonical field names (rgalbumgain etc.): not a tag/role, so SQL generation returned an error. Because refreshSmartPlaylist deletes old tracks before regenerating, the abort left the playlist empty. Fix: add Nullable bool to FieldInfo and mark the four ReplayGain fields. Add static alias entries (replaygain_album_gain -> rgalbumgain etc.) with Numeric+Nullable set; because AddTagNames skips names already in the field map, these static entries take precedence over the mappings.yaml tag registration. missingExpr now emits IS NULL / IS NOT NULL for nullable column fields instead of the json_tree lookup. Fixes #5584 * chore(smartplaylist): address code review feedback - Simplify the isMissing/isPresent unsupported-field error message, removing the internal "nullable fields" jargon - Standardize comments on mappings.yaml (the actual filename) - Clarify the alias precedence comment in the LookupField test
2026-06-10 21:12:31 -04:00
feat(smartplaylist): extend isMissing/isPresent to bpm, bitDepth and many text fields (#5603) * feat(smartplaylist): support isMissing/isPresent on mbz_* and lyrics fields Mark the six mbz_* MusicBrainz ID columns and the lyrics column as Nullable in the criteria field map, then extend missingExpr to handle string columns where absence is encoded as NULL or empty string (plus '[]' for lyrics). The Numeric/Boolean path (ReplayGain) is preserved via an explicit type check. * refactor(model): make MediaFile BPM and BitDepth nullable pointers Convert BPM and BitDepth fields in model.MediaFile from int to *int so that 'tag absent' is distinguishable from zero. The metadata mapper now uses NullableFloat for BPM (nil when absent or zero/unparseable) and only sets BitDepth when the audio property is non-zero (lossy codecs report 0). All read sites use gg.V() for zero-fallback deref so Subsonic API output and transcoding behaviour are byte-identical to before. The persistence layer bridges the existing NOT NULL DB columns by coercing nil to 0 on write and 0 back to nil on read in PostMapArgs/PostScan; a later migration task will drop those constraints. Hash upgrade safety is verified by a new MediaFile.Hash describe block: nil *int hashes identically to the old int(0) default via ZeroNil+IgnoreZeroValue, so no files will be spuriously re-imported after this change. Extra files touched beyond the plan's list: core/stream/legacy_client_test.go (BitDepth in model.MediaFile literals), persistence/mediafile_repository.go (NOT NULL bridge). * test(model): pin pre-conversion golden hashes for BPM/BitDepth * feat(smartplaylist): support isMissing/isPresent on bpm and bitDepth * feat(db): make bpm and bit_depth columns nullable, backfill 0 to NULL Drop the NOT NULL constraint on media_file.bpm and bit_depth via a lossless migration that converts legacy 0-means-absent values to real NULL. Remove the temporary shim in PostScan/PostMapArgs that was bridging the old NOT NULL columns to the *int model fields. Add round-trip persistence tests asserting NULL storage for nil pointers and correct value round-trip for non-nil pointers. * test(e2e): verify isMissing/isPresent partition for nullable fields Add DescribeTable covering bpm, bitdepth, lyrics, and mbz_recording_id: for each field, isMissing + isPresent song counts must equal the total library count, proving the nullable-column SQL is exhaustive and correct. * test(e2e): seed bpm tag so isMissing/isPresent partition is non-trivial * fix(model): omit bitDepth from JSON when absent instead of emitting null * feat(smartplaylist): support isMissing/isPresent on more string fields Enable isMissing/isPresent operators for album, comment, catalognumber, discsubtitle, albumcomment, sorttitle, sortalbum, sortartist, sortalbumartist, and explicitstatus by marking them Nullable in fieldMap. * refactor(smartplaylist): unify missingExpr column logic into one flow Collapse the numeric/string fork in missingExpr into a single empties-driven loop (numeric/boolean fields simply have no empties), and replace the duplicated IsTag/IsRole guard with a three-way switch that expresses the dispatch model once. No SQL semantics change for string fields; numeric/boolean fields now emit a single-element Or/And which squirrel parenthesizes (e.g. `(col IS NULL)` instead of bare `col IS NULL`) — update the affected test expectations accordingly.
2026-06-13 13:15:20 -04:00
switch {
case info.IsTag || info.IsRole:
return jsonExpr(info, nil, negate), nil
case info.Nullable:
// Nullable column fields are stored in dedicated columns, not in the tags JSON, so
// "missing" maps to a column check rather than a json_tree lookup. Numeric/boolean
// columns (e.g. ReplayGain, BPM) encode absence as NULL only; string columns (e.g.
// mbz_* IDs, lyrics) additionally treat empty string — and any field-specific empty
// encodings (e.g. '[]' for lyrics) — as missing. The unified flow below handles both:
// numeric/boolean fields simply have no empties, so the loops are no-ops.
f, ok := smartPlaylistFields[info.Name()]
if !ok || f.expr == "" {
fix(smartplaylist): support isMissing/isPresent operators on ReplayGain fields (#5585) * fix(smartplaylist): support isMissing/isPresent operators on ReplayGain fields ReplayGain values are stored in dedicated nullable columns (rg_album_gain, rg_album_peak, rg_track_gain, rg_track_peak) rather than in the media_file.tags JSON blob. The isMissing/isPresent operators previously only supported tag and role fields, causing two failure modes: 1. Using the documented alias names (replaygain_album_gain etc.) from PR #5256: these got registered as JSON tags from mappings.yaml, so isMissing queried json_tree(media_file.tags, '$.replaygain_album_gain') which is always empty -> the playlist matched ALL songs. 2. Using the canonical field names (rgalbumgain etc.): not a tag/role, so SQL generation returned an error. Because refreshSmartPlaylist deletes old tracks before regenerating, the abort left the playlist empty. Fix: add Nullable bool to FieldInfo and mark the four ReplayGain fields. Add static alias entries (replaygain_album_gain -> rgalbumgain etc.) with Numeric+Nullable set; because AddTagNames skips names already in the field map, these static entries take precedence over the mappings.yaml tag registration. missingExpr now emits IS NULL / IS NOT NULL for nullable column fields instead of the json_tree lookup. Fixes #5584 * chore(smartplaylist): address code review feedback - Simplify the isMissing/isPresent unsupported-field error message, removing the internal "nullable fields" jargon - Standardize comments on mappings.yaml (the actual filename) - Clarify the alias precedence comment in the LookupField test
2026-06-10 21:12:31 -04:00
return nil, fmt.Errorf("invalid field in criteria: %s", field)
}
feat(smartplaylist): extend isMissing/isPresent to bpm, bitDepth and many text fields (#5603) * feat(smartplaylist): support isMissing/isPresent on mbz_* and lyrics fields Mark the six mbz_* MusicBrainz ID columns and the lyrics column as Nullable in the criteria field map, then extend missingExpr to handle string columns where absence is encoded as NULL or empty string (plus '[]' for lyrics). The Numeric/Boolean path (ReplayGain) is preserved via an explicit type check. * refactor(model): make MediaFile BPM and BitDepth nullable pointers Convert BPM and BitDepth fields in model.MediaFile from int to *int so that 'tag absent' is distinguishable from zero. The metadata mapper now uses NullableFloat for BPM (nil when absent or zero/unparseable) and only sets BitDepth when the audio property is non-zero (lossy codecs report 0). All read sites use gg.V() for zero-fallback deref so Subsonic API output and transcoding behaviour are byte-identical to before. The persistence layer bridges the existing NOT NULL DB columns by coercing nil to 0 on write and 0 back to nil on read in PostMapArgs/PostScan; a later migration task will drop those constraints. Hash upgrade safety is verified by a new MediaFile.Hash describe block: nil *int hashes identically to the old int(0) default via ZeroNil+IgnoreZeroValue, so no files will be spuriously re-imported after this change. Extra files touched beyond the plan's list: core/stream/legacy_client_test.go (BitDepth in model.MediaFile literals), persistence/mediafile_repository.go (NOT NULL bridge). * test(model): pin pre-conversion golden hashes for BPM/BitDepth * feat(smartplaylist): support isMissing/isPresent on bpm and bitDepth * feat(db): make bpm and bit_depth columns nullable, backfill 0 to NULL Drop the NOT NULL constraint on media_file.bpm and bit_depth via a lossless migration that converts legacy 0-means-absent values to real NULL. Remove the temporary shim in PostScan/PostMapArgs that was bridging the old NOT NULL columns to the *int model fields. Add round-trip persistence tests asserting NULL storage for nil pointers and correct value round-trip for non-nil pointers. * test(e2e): verify isMissing/isPresent partition for nullable fields Add DescribeTable covering bpm, bitdepth, lyrics, and mbz_recording_id: for each field, isMissing + isPresent song counts must equal the total library count, proving the nullable-column SQL is exhaustive and correct. * test(e2e): seed bpm tag so isMissing/isPresent partition is non-trivial * fix(model): omit bitDepth from JSON when absent instead of emitting null * feat(smartplaylist): support isMissing/isPresent on more string fields Enable isMissing/isPresent operators for album, comment, catalognumber, discsubtitle, albumcomment, sorttitle, sortalbum, sortartist, sortalbumartist, and explicitstatus by marking them Nullable in fieldMap. * refactor(smartplaylist): unify missingExpr column logic into one flow Collapse the numeric/string fork in missingExpr into a single empties-driven loop (numeric/boolean fields simply have no empties), and replace the duplicated IsTag/IsRole guard with a three-way switch that expresses the dispatch model once. No SQL semantics change for string fields; numeric/boolean fields now emit a single-element Or/And which squirrel parenthesizes (e.g. `(col IS NULL)` instead of bare `col IS NULL`) — update the affected test expectations accordingly.
2026-06-13 13:15:20 -04:00
col := f.expr
var empties []string
if !info.Numeric && !info.Boolean {
empties = append([]string{""}, f.emptyValues...)
}
missing := squirrel.Or{squirrel.Eq{col: nil}}
present := squirrel.And{squirrel.NotEq{col: nil}}
for _, e := range empties {
missing = append(missing, squirrel.Eq{col: e})
present = append(present, squirrel.NotEq{col: e})
}
fix(smartplaylist): support isMissing/isPresent operators on ReplayGain fields (#5585) * fix(smartplaylist): support isMissing/isPresent operators on ReplayGain fields ReplayGain values are stored in dedicated nullable columns (rg_album_gain, rg_album_peak, rg_track_gain, rg_track_peak) rather than in the media_file.tags JSON blob. The isMissing/isPresent operators previously only supported tag and role fields, causing two failure modes: 1. Using the documented alias names (replaygain_album_gain etc.) from PR #5256: these got registered as JSON tags from mappings.yaml, so isMissing queried json_tree(media_file.tags, '$.replaygain_album_gain') which is always empty -> the playlist matched ALL songs. 2. Using the canonical field names (rgalbumgain etc.): not a tag/role, so SQL generation returned an error. Because refreshSmartPlaylist deletes old tracks before regenerating, the abort left the playlist empty. Fix: add Nullable bool to FieldInfo and mark the four ReplayGain fields. Add static alias entries (replaygain_album_gain -> rgalbumgain etc.) with Numeric+Nullable set; because AddTagNames skips names already in the field map, these static entries take precedence over the mappings.yaml tag registration. missingExpr now emits IS NULL / IS NOT NULL for nullable column fields instead of the json_tree lookup. Fixes #5584 * chore(smartplaylist): address code review feedback - Simplify the isMissing/isPresent unsupported-field error message, removing the internal "nullable fields" jargon - Standardize comments on mappings.yaml (the actual filename) - Clarify the alias precedence comment in the LookupField test
2026-06-10 21:12:31 -04:00
if negate {
feat(smartplaylist): extend isMissing/isPresent to bpm, bitDepth and many text fields (#5603) * feat(smartplaylist): support isMissing/isPresent on mbz_* and lyrics fields Mark the six mbz_* MusicBrainz ID columns and the lyrics column as Nullable in the criteria field map, then extend missingExpr to handle string columns where absence is encoded as NULL or empty string (plus '[]' for lyrics). The Numeric/Boolean path (ReplayGain) is preserved via an explicit type check. * refactor(model): make MediaFile BPM and BitDepth nullable pointers Convert BPM and BitDepth fields in model.MediaFile from int to *int so that 'tag absent' is distinguishable from zero. The metadata mapper now uses NullableFloat for BPM (nil when absent or zero/unparseable) and only sets BitDepth when the audio property is non-zero (lossy codecs report 0). All read sites use gg.V() for zero-fallback deref so Subsonic API output and transcoding behaviour are byte-identical to before. The persistence layer bridges the existing NOT NULL DB columns by coercing nil to 0 on write and 0 back to nil on read in PostMapArgs/PostScan; a later migration task will drop those constraints. Hash upgrade safety is verified by a new MediaFile.Hash describe block: nil *int hashes identically to the old int(0) default via ZeroNil+IgnoreZeroValue, so no files will be spuriously re-imported after this change. Extra files touched beyond the plan's list: core/stream/legacy_client_test.go (BitDepth in model.MediaFile literals), persistence/mediafile_repository.go (NOT NULL bridge). * test(model): pin pre-conversion golden hashes for BPM/BitDepth * feat(smartplaylist): support isMissing/isPresent on bpm and bitDepth * feat(db): make bpm and bit_depth columns nullable, backfill 0 to NULL Drop the NOT NULL constraint on media_file.bpm and bit_depth via a lossless migration that converts legacy 0-means-absent values to real NULL. Remove the temporary shim in PostScan/PostMapArgs that was bridging the old NOT NULL columns to the *int model fields. Add round-trip persistence tests asserting NULL storage for nil pointers and correct value round-trip for non-nil pointers. * test(e2e): verify isMissing/isPresent partition for nullable fields Add DescribeTable covering bpm, bitdepth, lyrics, and mbz_recording_id: for each field, isMissing + isPresent song counts must equal the total library count, proving the nullable-column SQL is exhaustive and correct. * test(e2e): seed bpm tag so isMissing/isPresent partition is non-trivial * fix(model): omit bitDepth from JSON when absent instead of emitting null * feat(smartplaylist): support isMissing/isPresent on more string fields Enable isMissing/isPresent operators for album, comment, catalognumber, discsubtitle, albumcomment, sorttitle, sortalbum, sortartist, sortalbumartist, and explicitstatus by marking them Nullable in fieldMap. * refactor(smartplaylist): unify missingExpr column logic into one flow Collapse the numeric/string fork in missingExpr into a single empties-driven loop (numeric/boolean fields simply have no empties), and replace the duplicated IsTag/IsRole guard with a three-way switch that expresses the dispatch model once. No SQL semantics change for string fields; numeric/boolean fields now emit a single-element Or/And which squirrel parenthesizes (e.g. `(col IS NULL)` instead of bare `col IS NULL`) — update the affected test expectations accordingly.
2026-06-13 13:15:20 -04:00
return missing, nil
fix(smartplaylist): support isMissing/isPresent operators on ReplayGain fields (#5585) * fix(smartplaylist): support isMissing/isPresent operators on ReplayGain fields ReplayGain values are stored in dedicated nullable columns (rg_album_gain, rg_album_peak, rg_track_gain, rg_track_peak) rather than in the media_file.tags JSON blob. The isMissing/isPresent operators previously only supported tag and role fields, causing two failure modes: 1. Using the documented alias names (replaygain_album_gain etc.) from PR #5256: these got registered as JSON tags from mappings.yaml, so isMissing queried json_tree(media_file.tags, '$.replaygain_album_gain') which is always empty -> the playlist matched ALL songs. 2. Using the canonical field names (rgalbumgain etc.): not a tag/role, so SQL generation returned an error. Because refreshSmartPlaylist deletes old tracks before regenerating, the abort left the playlist empty. Fix: add Nullable bool to FieldInfo and mark the four ReplayGain fields. Add static alias entries (replaygain_album_gain -> rgalbumgain etc.) with Numeric+Nullable set; because AddTagNames skips names already in the field map, these static entries take precedence over the mappings.yaml tag registration. missingExpr now emits IS NULL / IS NOT NULL for nullable column fields instead of the json_tree lookup. Fixes #5584 * chore(smartplaylist): address code review feedback - Simplify the isMissing/isPresent unsupported-field error message, removing the internal "nullable fields" jargon - Standardize comments on mappings.yaml (the actual filename) - Clarify the alias precedence comment in the LookupField test
2026-06-10 21:12:31 -04:00
}
feat(smartplaylist): extend isMissing/isPresent to bpm, bitDepth and many text fields (#5603) * feat(smartplaylist): support isMissing/isPresent on mbz_* and lyrics fields Mark the six mbz_* MusicBrainz ID columns and the lyrics column as Nullable in the criteria field map, then extend missingExpr to handle string columns where absence is encoded as NULL or empty string (plus '[]' for lyrics). The Numeric/Boolean path (ReplayGain) is preserved via an explicit type check. * refactor(model): make MediaFile BPM and BitDepth nullable pointers Convert BPM and BitDepth fields in model.MediaFile from int to *int so that 'tag absent' is distinguishable from zero. The metadata mapper now uses NullableFloat for BPM (nil when absent or zero/unparseable) and only sets BitDepth when the audio property is non-zero (lossy codecs report 0). All read sites use gg.V() for zero-fallback deref so Subsonic API output and transcoding behaviour are byte-identical to before. The persistence layer bridges the existing NOT NULL DB columns by coercing nil to 0 on write and 0 back to nil on read in PostMapArgs/PostScan; a later migration task will drop those constraints. Hash upgrade safety is verified by a new MediaFile.Hash describe block: nil *int hashes identically to the old int(0) default via ZeroNil+IgnoreZeroValue, so no files will be spuriously re-imported after this change. Extra files touched beyond the plan's list: core/stream/legacy_client_test.go (BitDepth in model.MediaFile literals), persistence/mediafile_repository.go (NOT NULL bridge). * test(model): pin pre-conversion golden hashes for BPM/BitDepth * feat(smartplaylist): support isMissing/isPresent on bpm and bitDepth * feat(db): make bpm and bit_depth columns nullable, backfill 0 to NULL Drop the NOT NULL constraint on media_file.bpm and bit_depth via a lossless migration that converts legacy 0-means-absent values to real NULL. Remove the temporary shim in PostScan/PostMapArgs that was bridging the old NOT NULL columns to the *int model fields. Add round-trip persistence tests asserting NULL storage for nil pointers and correct value round-trip for non-nil pointers. * test(e2e): verify isMissing/isPresent partition for nullable fields Add DescribeTable covering bpm, bitdepth, lyrics, and mbz_recording_id: for each field, isMissing + isPresent song counts must equal the total library count, proving the nullable-column SQL is exhaustive and correct. * test(e2e): seed bpm tag so isMissing/isPresent partition is non-trivial * fix(model): omit bitDepth from JSON when absent instead of emitting null * feat(smartplaylist): support isMissing/isPresent on more string fields Enable isMissing/isPresent operators for album, comment, catalognumber, discsubtitle, albumcomment, sorttitle, sortalbum, sortartist, sortalbumartist, and explicitstatus by marking them Nullable in fieldMap. * refactor(smartplaylist): unify missingExpr column logic into one flow Collapse the numeric/string fork in missingExpr into a single empties-driven loop (numeric/boolean fields simply have no empties), and replace the duplicated IsTag/IsRole guard with a three-way switch that expresses the dispatch model once. No SQL semantics change for string fields; numeric/boolean fields now emit a single-element Or/And which squirrel parenthesizes (e.g. `(col IS NULL)` instead of bare `col IS NULL`) — update the affected test expectations accordingly.
2026-06-13 13:15:20 -04:00
return present, nil
default:
return nil, fmt.Errorf("isMissing/isPresent operator is not supported for field: %s", field)
fix(smartplaylist): support isMissing/isPresent operators on ReplayGain fields (#5585) * fix(smartplaylist): support isMissing/isPresent operators on ReplayGain fields ReplayGain values are stored in dedicated nullable columns (rg_album_gain, rg_album_peak, rg_track_gain, rg_track_peak) rather than in the media_file.tags JSON blob. The isMissing/isPresent operators previously only supported tag and role fields, causing two failure modes: 1. Using the documented alias names (replaygain_album_gain etc.) from PR #5256: these got registered as JSON tags from mappings.yaml, so isMissing queried json_tree(media_file.tags, '$.replaygain_album_gain') which is always empty -> the playlist matched ALL songs. 2. Using the canonical field names (rgalbumgain etc.): not a tag/role, so SQL generation returned an error. Because refreshSmartPlaylist deletes old tracks before regenerating, the abort left the playlist empty. Fix: add Nullable bool to FieldInfo and mark the four ReplayGain fields. Add static alias entries (replaygain_album_gain -> rgalbumgain etc.) with Numeric+Nullable set; because AddTagNames skips names already in the field map, these static entries take precedence over the mappings.yaml tag registration. missingExpr now emits IS NULL / IS NOT NULL for nullable column fields instead of the json_tree lookup. Fixes #5584 * chore(smartplaylist): address code review feedback - Simplify the isMissing/isPresent unsupported-field error message, removing the internal "nullable fields" jargon - Standardize comments on mappings.yaml (the actual filename) - Clarify the alias precedence comment in the LookupField test
2026-06-10 21:12:31 -04:00
}
feat(smartplaylist): add isMissing and isPresent operators (#5436) * feat(smartplaylist): add IsMissing and IsPresent operator types Add two new Expression types for detecting absent/present tags and roles in smart playlist criteria. Includes JSON marshal/unmarshal support and Walk visitor registration. * test(smartplaylist): add JSON marshal/unmarshal tests for isMissing/isPresent * feat(smartplaylist): add SQL generation for isMissing/isPresent operators Tags check json_tree(media_file.tags) for key existence. Roles check json_tree(media_file.participants) for key existence. Regular DB column fields are rejected with an error. * test(smartplaylist): add e2e tests for isMissing/isPresent operators Tests cover tag presence/absence with selective matching (grouping), universal absence (lyricist role), universal presence (composer role), and combined operator usage. * refactor(smartplaylist): use strconv.ParseBool in IsTruthy Replace hand-rolled string truthiness check with strconv.ParseBool, which correctly handles standard boolean strings and rejects unrecognized values as false. * refactor(smartplaylist): clarify missingExpr parameter naming Rename defaultNegate to checkAbsence and extract truthy local for readability. The XNOR logic (checkAbsence == truthy) is now easier to follow: isMissing passes true, isPresent passes false. * refactor(smartplaylist): reuse jsonExpr in missingExpr, improve errors - tagCond/roleCond now handle nil cond (existence-only check) - missingExpr delegates to jsonExpr(info, nil, negate) instead of building SQL manually - Better error messages: unknown fields now report the field name
2026-04-28 19:40:08 -04:00
}
feat(smartplaylists): relax playlist visibility in inPlaylist/notInPlaylist rules (#5411) * test(e2e): add end-to-end tests for smart playlists functionality Signed-off-by: Deluan <deluan@navidrome.org> * fix: enforce playlist visibility in smart playlist InPlaylist/NotInPlaylist rules Previously, the InPlaylist/NotInPlaylist smart playlist criteria only allowed referencing public playlists, regardless of who owned the smart playlist. This was too restrictive for owners referencing their own private playlists and for admins who should have unrestricted access. The fix passes the smart playlist owner's identity and admin status into the criteria SQL builder, so that: admins can reference any playlist, regular users can reference public playlists plus their own private ones, and inaccessible referenced playlists produce a warning instead of a hard error. Also prevents recursive refresh of child playlists the owner cannot access. * test(e2e): clarify user roles and fix playlist visibility tests Renamed testUser/otherUser to adminUser/regularUser to make the admin vs regular user distinction explicit in test code. Fixed three playlist visibility tests that were evaluating as admin (bypassing all access checks) instead of as a regular user, so the public playlist path is now actually exercised. All playlist operator tests now use explicit evaluateRuleAs calls with the appropriate user role. * fix: sync rulesSQL criteria after limitPercent resolution The rulesSQL struct captures a copy of rules at creation time. When limitPercent is resolved later, rules.Limit is updated but rulesSQL still holds the stale value. This caused percentage-based smart playlist limits to be silently ignored. Fix by updating rulesSQL.criteria after the resolution. * refactor: convert inList to a method on smartPlaylistCriteria The inList function already receives ownerID and ownerIsAdmin from the smartPlaylistCriteria caller. Making it a method lets it access those fields directly from the receiver, simplifying the signature and staying consistent with exprSQL which was already converted to a method. * refactor: simplify function signatures by removing type parameters in criteria_sql.go Signed-off-by: Deluan <deluan@navidrome.org> --------- Signed-off-by: Deluan <deluan@navidrome.org>
2026-04-25 14:59:06 -04:00
func mapExpr(values map[string]any, makeCond func(map[string]any) squirrel.Sqlizer, negateJSON bool) (squirrel.Sqlizer, error) {
if _, value, info, ok := singleField(values); ok && (info.IsTag || info.IsRole) {
return jsonExpr(info, makeCond(map[string]any{"value": value}), negateJSON), nil
}
fields, err := sqlFields(values)
if err != nil {
return nil, err
}
return makeCond(fields), nil
}
feat(smartplaylists): relax playlist visibility in inPlaylist/notInPlaylist rules (#5411) * test(e2e): add end-to-end tests for smart playlists functionality Signed-off-by: Deluan <deluan@navidrome.org> * fix: enforce playlist visibility in smart playlist InPlaylist/NotInPlaylist rules Previously, the InPlaylist/NotInPlaylist smart playlist criteria only allowed referencing public playlists, regardless of who owned the smart playlist. This was too restrictive for owners referencing their own private playlists and for admins who should have unrestricted access. The fix passes the smart playlist owner's identity and admin status into the criteria SQL builder, so that: admins can reference any playlist, regular users can reference public playlists plus their own private ones, and inaccessible referenced playlists produce a warning instead of a hard error. Also prevents recursive refresh of child playlists the owner cannot access. * test(e2e): clarify user roles and fix playlist visibility tests Renamed testUser/otherUser to adminUser/regularUser to make the admin vs regular user distinction explicit in test code. Fixed three playlist visibility tests that were evaluating as admin (bypassing all access checks) instead of as a regular user, so the public playlist path is now actually exercised. All playlist operator tests now use explicit evaluateRuleAs calls with the appropriate user role. * fix: sync rulesSQL criteria after limitPercent resolution The rulesSQL struct captures a copy of rules at creation time. When limitPercent is resolved later, rules.Limit is updated but rulesSQL still holds the stale value. This caused percentage-based smart playlist limits to be silently ignored. Fix by updating rulesSQL.criteria after the resolution. * refactor: convert inList to a method on smartPlaylistCriteria The inList function already receives ownerID and ownerIsAdmin from the smartPlaylistCriteria caller. Making it a method lets it access those fields directly from the receiver, simplifying the signature and staying consistent with exprSQL which was already converted to a method. * refactor: simplify function signatures by removing type parameters in criteria_sql.go Signed-off-by: Deluan <deluan@navidrome.org> --------- Signed-off-by: Deluan <deluan@navidrome.org>
2026-04-25 14:59:06 -04:00
func likeExpr(values map[string]any, pattern string, negate bool) (squirrel.Sqlizer, error) {
if _, value, info, ok := singleField(values); ok && (info.IsTag || info.IsRole) {
return jsonExpr(info, squirrel.Like{"value": fmt.Sprintf(pattern, value)}, negate), nil
}
fields, err := sqlFields(values)
if err != nil {
return nil, err
}
if negate {
lk := squirrel.NotLike{}
for field, value := range fields {
lk[field] = fmt.Sprintf(pattern, value)
}
return lk, nil
}
lk := squirrel.Like{}
for field, value := range fields {
lk[field] = fmt.Sprintf(pattern, value)
}
return lk, nil
}
feat(smartplaylists): relax playlist visibility in inPlaylist/notInPlaylist rules (#5411) * test(e2e): add end-to-end tests for smart playlists functionality Signed-off-by: Deluan <deluan@navidrome.org> * fix: enforce playlist visibility in smart playlist InPlaylist/NotInPlaylist rules Previously, the InPlaylist/NotInPlaylist smart playlist criteria only allowed referencing public playlists, regardless of who owned the smart playlist. This was too restrictive for owners referencing their own private playlists and for admins who should have unrestricted access. The fix passes the smart playlist owner's identity and admin status into the criteria SQL builder, so that: admins can reference any playlist, regular users can reference public playlists plus their own private ones, and inaccessible referenced playlists produce a warning instead of a hard error. Also prevents recursive refresh of child playlists the owner cannot access. * test(e2e): clarify user roles and fix playlist visibility tests Renamed testUser/otherUser to adminUser/regularUser to make the admin vs regular user distinction explicit in test code. Fixed three playlist visibility tests that were evaluating as admin (bypassing all access checks) instead of as a regular user, so the public playlist path is now actually exercised. All playlist operator tests now use explicit evaluateRuleAs calls with the appropriate user role. * fix: sync rulesSQL criteria after limitPercent resolution The rulesSQL struct captures a copy of rules at creation time. When limitPercent is resolved later, rules.Limit is updated but rulesSQL still holds the stale value. This caused percentage-based smart playlist limits to be silently ignored. Fix by updating rulesSQL.criteria after the resolution. * refactor: convert inList to a method on smartPlaylistCriteria The inList function already receives ownerID and ownerIsAdmin from the smartPlaylistCriteria caller. Making it a method lets it access those fields directly from the receiver, simplifying the signature and staying consistent with exprSQL which was already converted to a method. * refactor: simplify function signatures by removing type parameters in criteria_sql.go Signed-off-by: Deluan <deluan@navidrome.org> --------- Signed-off-by: Deluan <deluan@navidrome.org>
2026-04-25 14:59:06 -04:00
func rangeExpr(values map[string]any) (squirrel.Sqlizer, error) {
fields, err := sqlFields(values)
if err != nil {
return nil, err
}
and := squirrel.And{}
for field, value := range fields {
s := reflect.ValueOf(value)
if s.Kind() != reflect.Slice || s.Len() != 2 {
return nil, fmt.Errorf("invalid range for 'in' operator: %s", value)
}
and = append(and,
squirrel.GtOrEq{field: s.Index(0).Interface()},
squirrel.LtOrEq{field: s.Index(1).Interface()},
)
}
return and, nil
}
feat(smartplaylists): relax playlist visibility in inPlaylist/notInPlaylist rules (#5411) * test(e2e): add end-to-end tests for smart playlists functionality Signed-off-by: Deluan <deluan@navidrome.org> * fix: enforce playlist visibility in smart playlist InPlaylist/NotInPlaylist rules Previously, the InPlaylist/NotInPlaylist smart playlist criteria only allowed referencing public playlists, regardless of who owned the smart playlist. This was too restrictive for owners referencing their own private playlists and for admins who should have unrestricted access. The fix passes the smart playlist owner's identity and admin status into the criteria SQL builder, so that: admins can reference any playlist, regular users can reference public playlists plus their own private ones, and inaccessible referenced playlists produce a warning instead of a hard error. Also prevents recursive refresh of child playlists the owner cannot access. * test(e2e): clarify user roles and fix playlist visibility tests Renamed testUser/otherUser to adminUser/regularUser to make the admin vs regular user distinction explicit in test code. Fixed three playlist visibility tests that were evaluating as admin (bypassing all access checks) instead of as a regular user, so the public playlist path is now actually exercised. All playlist operator tests now use explicit evaluateRuleAs calls with the appropriate user role. * fix: sync rulesSQL criteria after limitPercent resolution The rulesSQL struct captures a copy of rules at creation time. When limitPercent is resolved later, rules.Limit is updated but rulesSQL still holds the stale value. This caused percentage-based smart playlist limits to be silently ignored. Fix by updating rulesSQL.criteria after the resolution. * refactor: convert inList to a method on smartPlaylistCriteria The inList function already receives ownerID and ownerIsAdmin from the smartPlaylistCriteria caller. Making it a method lets it access those fields directly from the receiver, simplifying the signature and staying consistent with exprSQL which was already converted to a method. * refactor: simplify function signatures by removing type parameters in criteria_sql.go Signed-off-by: Deluan <deluan@navidrome.org> --------- Signed-off-by: Deluan <deluan@navidrome.org>
2026-04-25 14:59:06 -04:00
func periodExpr(values map[string]any, negate bool) (squirrel.Sqlizer, error) {
fields, err := sqlFields(values)
if err != nil {
return nil, err
}
var field string
var value any
for f, v := range fields {
field, value = f, v
break
}
days, err := strconv.ParseInt(fmt.Sprintf("%v", value), 10, 64)
if err != nil {
return nil, err
}
firstDate := startOfPeriod(days, time.Now())
if negate {
return squirrel.Or{
squirrel.Lt{field: firstDate},
squirrel.Eq{field: nil},
}, nil
}
return squirrel.Gt{field: firstDate}, nil
}
func startOfPeriod(numDays int64, from time.Time) string {
return from.Add(time.Duration(-24*numDays) * time.Hour).Format("2006-01-02")
}
feat(smartplaylists): relax playlist visibility in inPlaylist/notInPlaylist rules (#5411) * test(e2e): add end-to-end tests for smart playlists functionality Signed-off-by: Deluan <deluan@navidrome.org> * fix: enforce playlist visibility in smart playlist InPlaylist/NotInPlaylist rules Previously, the InPlaylist/NotInPlaylist smart playlist criteria only allowed referencing public playlists, regardless of who owned the smart playlist. This was too restrictive for owners referencing their own private playlists and for admins who should have unrestricted access. The fix passes the smart playlist owner's identity and admin status into the criteria SQL builder, so that: admins can reference any playlist, regular users can reference public playlists plus their own private ones, and inaccessible referenced playlists produce a warning instead of a hard error. Also prevents recursive refresh of child playlists the owner cannot access. * test(e2e): clarify user roles and fix playlist visibility tests Renamed testUser/otherUser to adminUser/regularUser to make the admin vs regular user distinction explicit in test code. Fixed three playlist visibility tests that were evaluating as admin (bypassing all access checks) instead of as a regular user, so the public playlist path is now actually exercised. All playlist operator tests now use explicit evaluateRuleAs calls with the appropriate user role. * fix: sync rulesSQL criteria after limitPercent resolution The rulesSQL struct captures a copy of rules at creation time. When limitPercent is resolved later, rules.Limit is updated but rulesSQL still holds the stale value. This caused percentage-based smart playlist limits to be silently ignored. Fix by updating rulesSQL.criteria after the resolution. * refactor: convert inList to a method on smartPlaylistCriteria The inList function already receives ownerID and ownerIsAdmin from the smartPlaylistCriteria caller. Making it a method lets it access those fields directly from the receiver, simplifying the signature and staying consistent with exprSQL which was already converted to a method. * refactor: simplify function signatures by removing type parameters in criteria_sql.go Signed-off-by: Deluan <deluan@navidrome.org> --------- Signed-off-by: Deluan <deluan@navidrome.org>
2026-04-25 14:59:06 -04:00
func (c smartPlaylistCriteria) inList(values map[string]any, negate bool) (squirrel.Sqlizer, error) {
playlistID, ok := values["id"].(string)
if !ok {
return nil, errors.New("playlist id not given")
}
feat(smartplaylists): relax playlist visibility in inPlaylist/notInPlaylist rules (#5411) * test(e2e): add end-to-end tests for smart playlists functionality Signed-off-by: Deluan <deluan@navidrome.org> * fix: enforce playlist visibility in smart playlist InPlaylist/NotInPlaylist rules Previously, the InPlaylist/NotInPlaylist smart playlist criteria only allowed referencing public playlists, regardless of who owned the smart playlist. This was too restrictive for owners referencing their own private playlists and for admins who should have unrestricted access. The fix passes the smart playlist owner's identity and admin status into the criteria SQL builder, so that: admins can reference any playlist, regular users can reference public playlists plus their own private ones, and inaccessible referenced playlists produce a warning instead of a hard error. Also prevents recursive refresh of child playlists the owner cannot access. * test(e2e): clarify user roles and fix playlist visibility tests Renamed testUser/otherUser to adminUser/regularUser to make the admin vs regular user distinction explicit in test code. Fixed three playlist visibility tests that were evaluating as admin (bypassing all access checks) instead of as a regular user, so the public playlist path is now actually exercised. All playlist operator tests now use explicit evaluateRuleAs calls with the appropriate user role. * fix: sync rulesSQL criteria after limitPercent resolution The rulesSQL struct captures a copy of rules at creation time. When limitPercent is resolved later, rules.Limit is updated but rulesSQL still holds the stale value. This caused percentage-based smart playlist limits to be silently ignored. Fix by updating rulesSQL.criteria after the resolution. * refactor: convert inList to a method on smartPlaylistCriteria The inList function already receives ownerID and ownerIsAdmin from the smartPlaylistCriteria caller. Making it a method lets it access those fields directly from the receiver, simplifying the signature and staying consistent with exprSQL which was already converted to a method. * refactor: simplify function signatures by removing type parameters in criteria_sql.go Signed-off-by: Deluan <deluan@navidrome.org> --------- Signed-off-by: Deluan <deluan@navidrome.org>
2026-04-25 14:59:06 -04:00
filters := squirrel.And{squirrel.Eq{"pl.playlist_id": playlistID}}
refactor: centralize criteria sort parsing and extract smart playlist logic (#5415) * test: add tests for recordingdate alias resolution in smart playlists Signed-off-by: Deluan <deluan@navidrome.org> * refactor: update FieldInfo structure and simplify fieldMap initialization Signed-off-by: Deluan <deluan@navidrome.org> * refactor: move sort parsing logic from persistence to criteria package Extracted sort field parsing, validation, and direction handling from persistence/criteria_sql.go into model/criteria/sort.go. The new OrderByFields method on Criteria parses the Sort/Order strings into validated SortField structs (field name + direction), resolving aliases and handling +/- prefixes and order inversion. The persistence layer now consumes these parsed fields and only handles SQL expression mapping. This centralizes sort parsing to enforce consistent implementations. * refactor: standardize field access in smartPlaylistCriteria structure Signed-off-by: Deluan <deluan@navidrome.org> * refactor: add ResolveLimit method to Criteria Moved the percentage-limit resolution logic from playlist_repository into Criteria.ResolveLimit, replacing the 3-line mutate-after-query pattern with a single method call. The method preserves LimitPercent rather than zeroing it, since IsPercentageLimit already returns false once Limit is set, making the clear redundant and lossy. * refactor: improve child playlist loading and error handling in refresh logic Signed-off-by: Deluan <deluan@navidrome.org> * refactor: extract smart playlist logic to dedicated files Moved refreshSmartPlaylist, addSmartPlaylistAnnotationJoins, and addCriteria methods from playlist_repository.go to a new smart_playlist_repository.go file. Extracted all smart playlist tests to smart_playlist_repository_test.go. Added DeferCleanup to the "valid rules" test to fix ordering flakiness when Ginkgo randomizes test execution across files. * refactor: break refreshSmartPlaylist into smaller focused methods Split the monolithic refreshSmartPlaylist method into discrete helpers for readability: shouldRefreshSmartPlaylist for guard checks, refreshChildPlaylists for recursive dependency refresh, resolvePercentageLimit for count-based limit resolution, buildSmartPlaylistQuery for assembling the SELECT with joins, and addMediaFileAnnotationJoin to DRY up the repeated annotation join clause. * refactor: deduplicate child playlist IDs in Criteria Signed-off-by: Deluan <deluan@navidrome.org> * refactor: simplify withSmartPlaylistOwner to accept model.User Replaced separate ownerID string and ownerIsAdmin bool parameters with a single model.User struct, reducing the field count in smartPlaylistCriteria and making the option function signature clearer. Updated all call sites and tests accordingly. * fix: handle empty sort fields and propagate child playlist load errors OrderByFields now falls back to [{title, asc}] when all user-supplied sort fields are invalid, preventing empty ORDER BY clauses that would produce invalid SQL in row_number() window functions. Also restored the original behavior where a DB error loading child playlists aborts the parent smart playlist refresh, by making refreshChildPlaylists return a bool. * refactor: log warning when no valid sort fields are found Signed-off-by: Deluan <deluan@navidrome.org> --------- Signed-off-by: Deluan <deluan@navidrome.org>
2026-04-26 14:49:59 -04:00
if !c.owner.IsAdmin {
if c.owner.ID == "" {
feat(smartplaylists): relax playlist visibility in inPlaylist/notInPlaylist rules (#5411) * test(e2e): add end-to-end tests for smart playlists functionality Signed-off-by: Deluan <deluan@navidrome.org> * fix: enforce playlist visibility in smart playlist InPlaylist/NotInPlaylist rules Previously, the InPlaylist/NotInPlaylist smart playlist criteria only allowed referencing public playlists, regardless of who owned the smart playlist. This was too restrictive for owners referencing their own private playlists and for admins who should have unrestricted access. The fix passes the smart playlist owner's identity and admin status into the criteria SQL builder, so that: admins can reference any playlist, regular users can reference public playlists plus their own private ones, and inaccessible referenced playlists produce a warning instead of a hard error. Also prevents recursive refresh of child playlists the owner cannot access. * test(e2e): clarify user roles and fix playlist visibility tests Renamed testUser/otherUser to adminUser/regularUser to make the admin vs regular user distinction explicit in test code. Fixed three playlist visibility tests that were evaluating as admin (bypassing all access checks) instead of as a regular user, so the public playlist path is now actually exercised. All playlist operator tests now use explicit evaluateRuleAs calls with the appropriate user role. * fix: sync rulesSQL criteria after limitPercent resolution The rulesSQL struct captures a copy of rules at creation time. When limitPercent is resolved later, rules.Limit is updated but rulesSQL still holds the stale value. This caused percentage-based smart playlist limits to be silently ignored. Fix by updating rulesSQL.criteria after the resolution. * refactor: convert inList to a method on smartPlaylistCriteria The inList function already receives ownerID and ownerIsAdmin from the smartPlaylistCriteria caller. Making it a method lets it access those fields directly from the receiver, simplifying the signature and staying consistent with exprSQL which was already converted to a method. * refactor: simplify function signatures by removing type parameters in criteria_sql.go Signed-off-by: Deluan <deluan@navidrome.org> --------- Signed-off-by: Deluan <deluan@navidrome.org>
2026-04-25 14:59:06 -04:00
filters = append(filters, squirrel.Eq{"playlist.public": 1})
} else {
filters = append(filters, squirrel.Or{
squirrel.Eq{"playlist.public": 1},
refactor: centralize criteria sort parsing and extract smart playlist logic (#5415) * test: add tests for recordingdate alias resolution in smart playlists Signed-off-by: Deluan <deluan@navidrome.org> * refactor: update FieldInfo structure and simplify fieldMap initialization Signed-off-by: Deluan <deluan@navidrome.org> * refactor: move sort parsing logic from persistence to criteria package Extracted sort field parsing, validation, and direction handling from persistence/criteria_sql.go into model/criteria/sort.go. The new OrderByFields method on Criteria parses the Sort/Order strings into validated SortField structs (field name + direction), resolving aliases and handling +/- prefixes and order inversion. The persistence layer now consumes these parsed fields and only handles SQL expression mapping. This centralizes sort parsing to enforce consistent implementations. * refactor: standardize field access in smartPlaylistCriteria structure Signed-off-by: Deluan <deluan@navidrome.org> * refactor: add ResolveLimit method to Criteria Moved the percentage-limit resolution logic from playlist_repository into Criteria.ResolveLimit, replacing the 3-line mutate-after-query pattern with a single method call. The method preserves LimitPercent rather than zeroing it, since IsPercentageLimit already returns false once Limit is set, making the clear redundant and lossy. * refactor: improve child playlist loading and error handling in refresh logic Signed-off-by: Deluan <deluan@navidrome.org> * refactor: extract smart playlist logic to dedicated files Moved refreshSmartPlaylist, addSmartPlaylistAnnotationJoins, and addCriteria methods from playlist_repository.go to a new smart_playlist_repository.go file. Extracted all smart playlist tests to smart_playlist_repository_test.go. Added DeferCleanup to the "valid rules" test to fix ordering flakiness when Ginkgo randomizes test execution across files. * refactor: break refreshSmartPlaylist into smaller focused methods Split the monolithic refreshSmartPlaylist method into discrete helpers for readability: shouldRefreshSmartPlaylist for guard checks, refreshChildPlaylists for recursive dependency refresh, resolvePercentageLimit for count-based limit resolution, buildSmartPlaylistQuery for assembling the SELECT with joins, and addMediaFileAnnotationJoin to DRY up the repeated annotation join clause. * refactor: deduplicate child playlist IDs in Criteria Signed-off-by: Deluan <deluan@navidrome.org> * refactor: simplify withSmartPlaylistOwner to accept model.User Replaced separate ownerID string and ownerIsAdmin bool parameters with a single model.User struct, reducing the field count in smartPlaylistCriteria and making the option function signature clearer. Updated all call sites and tests accordingly. * fix: handle empty sort fields and propagate child playlist load errors OrderByFields now falls back to [{title, asc}] when all user-supplied sort fields are invalid, preventing empty ORDER BY clauses that would produce invalid SQL in row_number() window functions. Also restored the original behavior where a DB error loading child playlists aborts the parent smart playlist refresh, by making refreshChildPlaylists return a bool. * refactor: log warning when no valid sort fields are found Signed-off-by: Deluan <deluan@navidrome.org> --------- Signed-off-by: Deluan <deluan@navidrome.org>
2026-04-26 14:49:59 -04:00
squirrel.Eq{"playlist.owner_id": c.owner.ID},
feat(smartplaylists): relax playlist visibility in inPlaylist/notInPlaylist rules (#5411) * test(e2e): add end-to-end tests for smart playlists functionality Signed-off-by: Deluan <deluan@navidrome.org> * fix: enforce playlist visibility in smart playlist InPlaylist/NotInPlaylist rules Previously, the InPlaylist/NotInPlaylist smart playlist criteria only allowed referencing public playlists, regardless of who owned the smart playlist. This was too restrictive for owners referencing their own private playlists and for admins who should have unrestricted access. The fix passes the smart playlist owner's identity and admin status into the criteria SQL builder, so that: admins can reference any playlist, regular users can reference public playlists plus their own private ones, and inaccessible referenced playlists produce a warning instead of a hard error. Also prevents recursive refresh of child playlists the owner cannot access. * test(e2e): clarify user roles and fix playlist visibility tests Renamed testUser/otherUser to adminUser/regularUser to make the admin vs regular user distinction explicit in test code. Fixed three playlist visibility tests that were evaluating as admin (bypassing all access checks) instead of as a regular user, so the public playlist path is now actually exercised. All playlist operator tests now use explicit evaluateRuleAs calls with the appropriate user role. * fix: sync rulesSQL criteria after limitPercent resolution The rulesSQL struct captures a copy of rules at creation time. When limitPercent is resolved later, rules.Limit is updated but rulesSQL still holds the stale value. This caused percentage-based smart playlist limits to be silently ignored. Fix by updating rulesSQL.criteria after the resolution. * refactor: convert inList to a method on smartPlaylistCriteria The inList function already receives ownerID and ownerIsAdmin from the smartPlaylistCriteria caller. Making it a method lets it access those fields directly from the receiver, simplifying the signature and staying consistent with exprSQL which was already converted to a method. * refactor: simplify function signatures by removing type parameters in criteria_sql.go Signed-off-by: Deluan <deluan@navidrome.org> --------- Signed-off-by: Deluan <deluan@navidrome.org>
2026-04-25 14:59:06 -04:00
})
}
}
subQuery := squirrel.Select("media_file_id").
From("playlist_tracks pl").
LeftJoin("playlist on pl.playlist_id = playlist.id").
feat(smartplaylists): relax playlist visibility in inPlaylist/notInPlaylist rules (#5411) * test(e2e): add end-to-end tests for smart playlists functionality Signed-off-by: Deluan <deluan@navidrome.org> * fix: enforce playlist visibility in smart playlist InPlaylist/NotInPlaylist rules Previously, the InPlaylist/NotInPlaylist smart playlist criteria only allowed referencing public playlists, regardless of who owned the smart playlist. This was too restrictive for owners referencing their own private playlists and for admins who should have unrestricted access. The fix passes the smart playlist owner's identity and admin status into the criteria SQL builder, so that: admins can reference any playlist, regular users can reference public playlists plus their own private ones, and inaccessible referenced playlists produce a warning instead of a hard error. Also prevents recursive refresh of child playlists the owner cannot access. * test(e2e): clarify user roles and fix playlist visibility tests Renamed testUser/otherUser to adminUser/regularUser to make the admin vs regular user distinction explicit in test code. Fixed three playlist visibility tests that were evaluating as admin (bypassing all access checks) instead of as a regular user, so the public playlist path is now actually exercised. All playlist operator tests now use explicit evaluateRuleAs calls with the appropriate user role. * fix: sync rulesSQL criteria after limitPercent resolution The rulesSQL struct captures a copy of rules at creation time. When limitPercent is resolved later, rules.Limit is updated but rulesSQL still holds the stale value. This caused percentage-based smart playlist limits to be silently ignored. Fix by updating rulesSQL.criteria after the resolution. * refactor: convert inList to a method on smartPlaylistCriteria The inList function already receives ownerID and ownerIsAdmin from the smartPlaylistCriteria caller. Making it a method lets it access those fields directly from the receiver, simplifying the signature and staying consistent with exprSQL which was already converted to a method. * refactor: simplify function signatures by removing type parameters in criteria_sql.go Signed-off-by: Deluan <deluan@navidrome.org> --------- Signed-off-by: Deluan <deluan@navidrome.org>
2026-04-25 14:59:06 -04:00
Where(filters)
subSQL, subArgs, err := subQuery.PlaceholderFormat(squirrel.Question).ToSql()
if err != nil {
return nil, err
}
if negate {
return squirrel.Expr("media_file.id NOT IN ("+subSQL+")", subArgs...), nil
}
return squirrel.Expr("media_file.id IN ("+subSQL+")", subArgs...), nil
}
func jsonExpr(info criteria.FieldInfo, cond squirrel.Sqlizer, negate bool) squirrel.Sqlizer {
if info.IsRole {
return roleCond{role: info.Name(), cond: cond, not: negate}
}
return tagCond{tag: info.Name(), numeric: info.Numeric, cond: cond, not: negate}
}
type tagCond struct {
tag string
numeric bool
cond squirrel.Sqlizer
not bool
}
func (e tagCond) ToSql() (string, []any, error) {
feat(smartplaylist): add isMissing and isPresent operators (#5436) * feat(smartplaylist): add IsMissing and IsPresent operator types Add two new Expression types for detecting absent/present tags and roles in smart playlist criteria. Includes JSON marshal/unmarshal support and Walk visitor registration. * test(smartplaylist): add JSON marshal/unmarshal tests for isMissing/isPresent * feat(smartplaylist): add SQL generation for isMissing/isPresent operators Tags check json_tree(media_file.tags) for key existence. Roles check json_tree(media_file.participants) for key existence. Regular DB column fields are rejected with an error. * test(smartplaylist): add e2e tests for isMissing/isPresent operators Tests cover tag presence/absence with selective matching (grouping), universal absence (lyricist role), universal presence (composer role), and combined operator usage. * refactor(smartplaylist): use strconv.ParseBool in IsTruthy Replace hand-rolled string truthiness check with strconv.ParseBool, which correctly handles standard boolean strings and rejects unrecognized values as false. * refactor(smartplaylist): clarify missingExpr parameter naming Rename defaultNegate to checkAbsence and extract truthy local for readability. The XNOR logic (checkAbsence == truthy) is now easier to follow: isMissing passes true, isPresent passes false. * refactor(smartplaylist): reuse jsonExpr in missingExpr, improve errors - tagCond/roleCond now handle nil cond (existence-only check) - missingExpr delegates to jsonExpr(info, nil, negate) instead of building SQL manually - Better error messages: unknown fields now report the field name
2026-04-28 19:40:08 -04:00
var cond string
var args []any
var err error
if e.cond != nil {
cond, args, err = e.cond.ToSql()
if e.numeric {
cond = strings.ReplaceAll(cond, "value", "CAST(value AS REAL)")
}
cond = fmt.Sprintf("exists (select 1 from json_tree(media_file.tags, '$.%s') where key='value' and %s)", e.tag, cond)
} else {
cond = fmt.Sprintf("exists (select 1 from json_tree(media_file.tags, '$.%s') where key='value')", e.tag)
}
if e.not {
cond = "not " + cond
}
return cond, args, err
}
type roleCond struct {
role string
cond squirrel.Sqlizer
not bool
}
func (e roleCond) ToSql() (string, []any, error) {
feat(smartplaylist): add isMissing and isPresent operators (#5436) * feat(smartplaylist): add IsMissing and IsPresent operator types Add two new Expression types for detecting absent/present tags and roles in smart playlist criteria. Includes JSON marshal/unmarshal support and Walk visitor registration. * test(smartplaylist): add JSON marshal/unmarshal tests for isMissing/isPresent * feat(smartplaylist): add SQL generation for isMissing/isPresent operators Tags check json_tree(media_file.tags) for key existence. Roles check json_tree(media_file.participants) for key existence. Regular DB column fields are rejected with an error. * test(smartplaylist): add e2e tests for isMissing/isPresent operators Tests cover tag presence/absence with selective matching (grouping), universal absence (lyricist role), universal presence (composer role), and combined operator usage. * refactor(smartplaylist): use strconv.ParseBool in IsTruthy Replace hand-rolled string truthiness check with strconv.ParseBool, which correctly handles standard boolean strings and rejects unrecognized values as false. * refactor(smartplaylist): clarify missingExpr parameter naming Rename defaultNegate to checkAbsence and extract truthy local for readability. The XNOR logic (checkAbsence == truthy) is now easier to follow: isMissing passes true, isPresent passes false. * refactor(smartplaylist): reuse jsonExpr in missingExpr, improve errors - tagCond/roleCond now handle nil cond (existence-only check) - missingExpr delegates to jsonExpr(info, nil, negate) instead of building SQL manually - Better error messages: unknown fields now report the field name
2026-04-28 19:40:08 -04:00
var cond string
var args []any
if e.cond != nil {
fix(smartplaylists): optimize smart playlist performance for role and tag criteria (#5515) * fix(server): optimize smart playlist role queries for large criteria (#5511) Role-based smart playlist criteria (artist, composer, etc.) now query the indexed media_file_artists join table instead of parsing JSON via json_tree() on every row. Multiple conditions for the same role within an OR group are merged into a single EXISTS subquery (batched at 200 to stay under SQLite's expression tree depth limit). A composite index (media_file_id, role) replaces the now-redundant single-column (media_file_id) index on media_file_artists. Benchmark (40k tracks, 500 patterns, 3 artists/track): - Merged join-table: 15ms (9.3x faster) - Merged json_tree: 30ms (4.6x faster) - Unmerged baseline: 137ms * refactor: simplify role condition SQL generation and benchmark Extract shared roleCondSQL/roleExistsSQL helpers to deduplicate the EXISTS template between roleCond and roleCondGroup. Use slices.Chunk for batching per project convention. Extract runBenchQuery helper to eliminate triplicated benchmark execution loop. * chore: raise roleCondBatchSize to 350 The empirical SQLite limit is 496 conditions per merged EXISTS subquery. Raising from 200 to 350 reduces the number of batches (e.g. 500 patterns now splits into 2 batches instead of 3). * fix(server): apply OR-merge optimization to tag conditions too Generalize mergeRoleConds into mergeJsonConds to also collapse multiple tag conditions for the same tag (e.g. genre) within OR groups. This gives the same ~5x speedup for tag-heavy smart playlists as the role optimization gives for artist-heavy ones. * refactor: benchmark uses real criteria pipeline instead of hand-built SQL The "Current" sub-benchmark now builds criteria.Criteria expressions and runs them through the actual newSmartPlaylistCriteria → Where() → ToSql() pipeline, validating the real production code path. The baseline still uses hand-built SQL representing the old json_tree approach. * fix: stabilize merged group ordering and close rows before error check Sort group keys in mergeJsonConds so the merged additions have deterministic order across runs, improving SQLite statement cache reuse. Move rows.Close() before rows.Err() in benchmark helper.
2026-05-22 18:00:13 -03:00
innerSQL, innerArgs, err := roleCondSQL(e.cond)
if err != nil {
return "", nil, err
}
cond = roleExistsSQL(innerSQL)
args = append([]any{e.role}, innerArgs...)
feat(smartplaylist): add isMissing and isPresent operators (#5436) * feat(smartplaylist): add IsMissing and IsPresent operator types Add two new Expression types for detecting absent/present tags and roles in smart playlist criteria. Includes JSON marshal/unmarshal support and Walk visitor registration. * test(smartplaylist): add JSON marshal/unmarshal tests for isMissing/isPresent * feat(smartplaylist): add SQL generation for isMissing/isPresent operators Tags check json_tree(media_file.tags) for key existence. Roles check json_tree(media_file.participants) for key existence. Regular DB column fields are rejected with an error. * test(smartplaylist): add e2e tests for isMissing/isPresent operators Tests cover tag presence/absence with selective matching (grouping), universal absence (lyricist role), universal presence (composer role), and combined operator usage. * refactor(smartplaylist): use strconv.ParseBool in IsTruthy Replace hand-rolled string truthiness check with strconv.ParseBool, which correctly handles standard boolean strings and rejects unrecognized values as false. * refactor(smartplaylist): clarify missingExpr parameter naming Rename defaultNegate to checkAbsence and extract truthy local for readability. The XNOR logic (checkAbsence == truthy) is now easier to follow: isMissing passes true, isPresent passes false. * refactor(smartplaylist): reuse jsonExpr in missingExpr, improve errors - tagCond/roleCond now handle nil cond (existence-only check) - missingExpr delegates to jsonExpr(info, nil, negate) instead of building SQL manually - Better error messages: unknown fields now report the field name
2026-04-28 19:40:08 -04:00
} else {
fix(smartplaylists): optimize smart playlist performance for role and tag criteria (#5515) * fix(server): optimize smart playlist role queries for large criteria (#5511) Role-based smart playlist criteria (artist, composer, etc.) now query the indexed media_file_artists join table instead of parsing JSON via json_tree() on every row. Multiple conditions for the same role within an OR group are merged into a single EXISTS subquery (batched at 200 to stay under SQLite's expression tree depth limit). A composite index (media_file_id, role) replaces the now-redundant single-column (media_file_id) index on media_file_artists. Benchmark (40k tracks, 500 patterns, 3 artists/track): - Merged join-table: 15ms (9.3x faster) - Merged json_tree: 30ms (4.6x faster) - Unmerged baseline: 137ms * refactor: simplify role condition SQL generation and benchmark Extract shared roleCondSQL/roleExistsSQL helpers to deduplicate the EXISTS template between roleCond and roleCondGroup. Use slices.Chunk for batching per project convention. Extract runBenchQuery helper to eliminate triplicated benchmark execution loop. * chore: raise roleCondBatchSize to 350 The empirical SQLite limit is 496 conditions per merged EXISTS subquery. Raising from 200 to 350 reduces the number of batches (e.g. 500 patterns now splits into 2 batches instead of 3). * fix(server): apply OR-merge optimization to tag conditions too Generalize mergeRoleConds into mergeJsonConds to also collapse multiple tag conditions for the same tag (e.g. genre) within OR groups. This gives the same ~5x speedup for tag-heavy smart playlists as the role optimization gives for artist-heavy ones. * refactor: benchmark uses real criteria pipeline instead of hand-built SQL The "Current" sub-benchmark now builds criteria.Criteria expressions and runs them through the actual newSmartPlaylistCriteria → Where() → ToSql() pipeline, validating the real production code path. The baseline still uses hand-built SQL representing the old json_tree approach. * fix: stabilize merged group ordering and close rows before error check Sort group keys in mergeJsonConds so the merged additions have deterministic order across runs, improving SQLite statement cache reuse. Move rows.Close() before rows.Err() in benchmark helper.
2026-05-22 18:00:13 -03:00
cond = "exists (select 1 from media_file_artists mfa where mfa.media_file_id = media_file.id and mfa.role = ?)"
args = []any{e.role}
feat(smartplaylist): add isMissing and isPresent operators (#5436) * feat(smartplaylist): add IsMissing and IsPresent operator types Add two new Expression types for detecting absent/present tags and roles in smart playlist criteria. Includes JSON marshal/unmarshal support and Walk visitor registration. * test(smartplaylist): add JSON marshal/unmarshal tests for isMissing/isPresent * feat(smartplaylist): add SQL generation for isMissing/isPresent operators Tags check json_tree(media_file.tags) for key existence. Roles check json_tree(media_file.participants) for key existence. Regular DB column fields are rejected with an error. * test(smartplaylist): add e2e tests for isMissing/isPresent operators Tests cover tag presence/absence with selective matching (grouping), universal absence (lyricist role), universal presence (composer role), and combined operator usage. * refactor(smartplaylist): use strconv.ParseBool in IsTruthy Replace hand-rolled string truthiness check with strconv.ParseBool, which correctly handles standard boolean strings and rejects unrecognized values as false. * refactor(smartplaylist): clarify missingExpr parameter naming Rename defaultNegate to checkAbsence and extract truthy local for readability. The XNOR logic (checkAbsence == truthy) is now easier to follow: isMissing passes true, isPresent passes false. * refactor(smartplaylist): reuse jsonExpr in missingExpr, improve errors - tagCond/roleCond now handle nil cond (existence-only check) - missingExpr delegates to jsonExpr(info, nil, negate) instead of building SQL manually - Better error messages: unknown fields now report the field name
2026-04-28 19:40:08 -04:00
}
if e.not {
cond = "not " + cond
}
fix(smartplaylists): optimize smart playlist performance for role and tag criteria (#5515) * fix(server): optimize smart playlist role queries for large criteria (#5511) Role-based smart playlist criteria (artist, composer, etc.) now query the indexed media_file_artists join table instead of parsing JSON via json_tree() on every row. Multiple conditions for the same role within an OR group are merged into a single EXISTS subquery (batched at 200 to stay under SQLite's expression tree depth limit). A composite index (media_file_id, role) replaces the now-redundant single-column (media_file_id) index on media_file_artists. Benchmark (40k tracks, 500 patterns, 3 artists/track): - Merged join-table: 15ms (9.3x faster) - Merged json_tree: 30ms (4.6x faster) - Unmerged baseline: 137ms * refactor: simplify role condition SQL generation and benchmark Extract shared roleCondSQL/roleExistsSQL helpers to deduplicate the EXISTS template between roleCond and roleCondGroup. Use slices.Chunk for batching per project convention. Extract runBenchQuery helper to eliminate triplicated benchmark execution loop. * chore: raise roleCondBatchSize to 350 The empirical SQLite limit is 496 conditions per merged EXISTS subquery. Raising from 200 to 350 reduces the number of batches (e.g. 500 patterns now splits into 2 batches instead of 3). * fix(server): apply OR-merge optimization to tag conditions too Generalize mergeRoleConds into mergeJsonConds to also collapse multiple tag conditions for the same tag (e.g. genre) within OR groups. This gives the same ~5x speedup for tag-heavy smart playlists as the role optimization gives for artist-heavy ones. * refactor: benchmark uses real criteria pipeline instead of hand-built SQL The "Current" sub-benchmark now builds criteria.Criteria expressions and runs them through the actual newSmartPlaylistCriteria → Where() → ToSql() pipeline, validating the real production code path. The baseline still uses hand-built SQL representing the old json_tree approach. * fix: stabilize merged group ordering and close rows before error check Sort group keys in mergeJsonConds so the merged additions have deterministic order across runs, improving SQLite statement cache reuse. Move rows.Close() before rows.Err() in benchmark helper.
2026-05-22 18:00:13 -03:00
return cond, args, nil
}
// roleCondSQL extracts SQL from a squirrel condition and rewrites the placeholder column name.
func roleCondSQL(cond squirrel.Sqlizer) (string, []any, error) {
sql, args, err := cond.ToSql()
if err != nil {
return "", nil, err
}
return strings.ReplaceAll(sql, "value", "artist.name"), args, nil
}
// roleExistsSQL wraps a condition fragment in the standard role EXISTS subquery.
func roleExistsSQL(innerCond string) string {
return fmt.Sprintf("exists (select 1 from media_file_artists mfa join artist on artist.id = mfa.artist_id "+
"where mfa.media_file_id = media_file.id and mfa.role = ? and %s)", innerCond)
}
// jsonCondBatchSize limits how many conditions are ORed inside a single EXISTS subquery
// to stay within SQLite's expression tree depth limit (max 1000). The EXISTS wrapper
// consumes ~4 levels; each ORed condition adds 1 level. Empirically, 496 is the maximum.
const jsonCondBatchSize = 350
// mergeJsonConds collapses multiple non-negated roleCond or tagCond entries for the same
// field within an OR group into batched EXISTS subqueries with the conditions ORed inside.
// This turns N separate correlated subqueries into ceil(N/batchSize), dramatically
// improving performance for smart playlists with many patterns.
func mergeJsonConds(or squirrel.Or) squirrel.Sqlizer {
if merged, ok := mergeSameFieldConds(or, false); ok {
return squirrel.Or(merged)
}
return or
}
// mergeNegatedJsonConds is the AND-group counterpart to mergeJsonConds, merging negated
// conditions. By De Morgan, "NOT EXISTS(X) AND NOT EXISTS(Y)" == "NOT EXISTS(X OR Y)".
func mergeNegatedJsonConds(and squirrel.And) squirrel.Sqlizer {
if merged, ok := mergeSameFieldConds(and, true); ok {
return squirrel.And(merged)
}
return and
}
// mergeSameFieldConds groups roleCond/tagCond entries that share a field and the requested
// polarity, replacing each group of 2+ with batched roleCondGroup/tagCondGroup subqueries.
// Returns the rewritten conditions and whether any merge happened.
func mergeSameFieldConds(conds []squirrel.Sqlizer, negated bool) ([]squirrel.Sqlizer, bool) {
fix(smartplaylists): optimize smart playlist performance for role and tag criteria (#5515) * fix(server): optimize smart playlist role queries for large criteria (#5511) Role-based smart playlist criteria (artist, composer, etc.) now query the indexed media_file_artists join table instead of parsing JSON via json_tree() on every row. Multiple conditions for the same role within an OR group are merged into a single EXISTS subquery (batched at 200 to stay under SQLite's expression tree depth limit). A composite index (media_file_id, role) replaces the now-redundant single-column (media_file_id) index on media_file_artists. Benchmark (40k tracks, 500 patterns, 3 artists/track): - Merged join-table: 15ms (9.3x faster) - Merged json_tree: 30ms (4.6x faster) - Unmerged baseline: 137ms * refactor: simplify role condition SQL generation and benchmark Extract shared roleCondSQL/roleExistsSQL helpers to deduplicate the EXISTS template between roleCond and roleCondGroup. Use slices.Chunk for batching per project convention. Extract runBenchQuery helper to eliminate triplicated benchmark execution loop. * chore: raise roleCondBatchSize to 350 The empirical SQLite limit is 496 conditions per merged EXISTS subquery. Raising from 200 to 350 reduces the number of batches (e.g. 500 patterns now splits into 2 batches instead of 3). * fix(server): apply OR-merge optimization to tag conditions too Generalize mergeRoleConds into mergeJsonConds to also collapse multiple tag conditions for the same tag (e.g. genre) within OR groups. This gives the same ~5x speedup for tag-heavy smart playlists as the role optimization gives for artist-heavy ones. * refactor: benchmark uses real criteria pipeline instead of hand-built SQL The "Current" sub-benchmark now builds criteria.Criteria expressions and runs them through the actual newSmartPlaylistCriteria → Where() → ToSql() pipeline, validating the real production code path. The baseline still uses hand-built SQL representing the old json_tree approach. * fix: stabilize merged group ordering and close rows before error check Sort group keys in mergeJsonConds so the merged additions have deterministic order across runs, improving SQLite statement cache reuse. Move rows.Close() before rows.Err() in benchmark helper.
2026-05-22 18:00:13 -03:00
type condEntry struct {
index int
cond squirrel.Sqlizer
}
type group struct {
entries []condEntry
isRole bool
numeric bool
tag string
}
groups := make(map[string]*group)
for i, s := range conds {
fix(smartplaylists): optimize smart playlist performance for role and tag criteria (#5515) * fix(server): optimize smart playlist role queries for large criteria (#5511) Role-based smart playlist criteria (artist, composer, etc.) now query the indexed media_file_artists join table instead of parsing JSON via json_tree() on every row. Multiple conditions for the same role within an OR group are merged into a single EXISTS subquery (batched at 200 to stay under SQLite's expression tree depth limit). A composite index (media_file_id, role) replaces the now-redundant single-column (media_file_id) index on media_file_artists. Benchmark (40k tracks, 500 patterns, 3 artists/track): - Merged join-table: 15ms (9.3x faster) - Merged json_tree: 30ms (4.6x faster) - Unmerged baseline: 137ms * refactor: simplify role condition SQL generation and benchmark Extract shared roleCondSQL/roleExistsSQL helpers to deduplicate the EXISTS template between roleCond and roleCondGroup. Use slices.Chunk for batching per project convention. Extract runBenchQuery helper to eliminate triplicated benchmark execution loop. * chore: raise roleCondBatchSize to 350 The empirical SQLite limit is 496 conditions per merged EXISTS subquery. Raising from 200 to 350 reduces the number of batches (e.g. 500 patterns now splits into 2 batches instead of 3). * fix(server): apply OR-merge optimization to tag conditions too Generalize mergeRoleConds into mergeJsonConds to also collapse multiple tag conditions for the same tag (e.g. genre) within OR groups. This gives the same ~5x speedup for tag-heavy smart playlists as the role optimization gives for artist-heavy ones. * refactor: benchmark uses real criteria pipeline instead of hand-built SQL The "Current" sub-benchmark now builds criteria.Criteria expressions and runs them through the actual newSmartPlaylistCriteria → Where() → ToSql() pipeline, validating the real production code path. The baseline still uses hand-built SQL representing the old json_tree approach. * fix: stabilize merged group ordering and close rows before error check Sort group keys in mergeJsonConds so the merged additions have deterministic order across runs, improving SQLite statement cache reuse. Move rows.Close() before rows.Err() in benchmark helper.
2026-05-22 18:00:13 -03:00
switch c := s.(type) {
case roleCond:
if c.not != negated || c.cond == nil {
fix(smartplaylists): optimize smart playlist performance for role and tag criteria (#5515) * fix(server): optimize smart playlist role queries for large criteria (#5511) Role-based smart playlist criteria (artist, composer, etc.) now query the indexed media_file_artists join table instead of parsing JSON via json_tree() on every row. Multiple conditions for the same role within an OR group are merged into a single EXISTS subquery (batched at 200 to stay under SQLite's expression tree depth limit). A composite index (media_file_id, role) replaces the now-redundant single-column (media_file_id) index on media_file_artists. Benchmark (40k tracks, 500 patterns, 3 artists/track): - Merged join-table: 15ms (9.3x faster) - Merged json_tree: 30ms (4.6x faster) - Unmerged baseline: 137ms * refactor: simplify role condition SQL generation and benchmark Extract shared roleCondSQL/roleExistsSQL helpers to deduplicate the EXISTS template between roleCond and roleCondGroup. Use slices.Chunk for batching per project convention. Extract runBenchQuery helper to eliminate triplicated benchmark execution loop. * chore: raise roleCondBatchSize to 350 The empirical SQLite limit is 496 conditions per merged EXISTS subquery. Raising from 200 to 350 reduces the number of batches (e.g. 500 patterns now splits into 2 batches instead of 3). * fix(server): apply OR-merge optimization to tag conditions too Generalize mergeRoleConds into mergeJsonConds to also collapse multiple tag conditions for the same tag (e.g. genre) within OR groups. This gives the same ~5x speedup for tag-heavy smart playlists as the role optimization gives for artist-heavy ones. * refactor: benchmark uses real criteria pipeline instead of hand-built SQL The "Current" sub-benchmark now builds criteria.Criteria expressions and runs them through the actual newSmartPlaylistCriteria → Where() → ToSql() pipeline, validating the real production code path. The baseline still uses hand-built SQL representing the old json_tree approach. * fix: stabilize merged group ordering and close rows before error check Sort group keys in mergeJsonConds so the merged additions have deterministic order across runs, improving SQLite statement cache reuse. Move rows.Close() before rows.Err() in benchmark helper.
2026-05-22 18:00:13 -03:00
continue
}
g, exists := groups["role:"+c.role]
if !exists {
g = &group{isRole: true}
groups["role:"+c.role] = g
}
g.entries = append(g.entries, condEntry{index: i, cond: c.cond})
case tagCond:
if c.not != negated || c.cond == nil {
fix(smartplaylists): optimize smart playlist performance for role and tag criteria (#5515) * fix(server): optimize smart playlist role queries for large criteria (#5511) Role-based smart playlist criteria (artist, composer, etc.) now query the indexed media_file_artists join table instead of parsing JSON via json_tree() on every row. Multiple conditions for the same role within an OR group are merged into a single EXISTS subquery (batched at 200 to stay under SQLite's expression tree depth limit). A composite index (media_file_id, role) replaces the now-redundant single-column (media_file_id) index on media_file_artists. Benchmark (40k tracks, 500 patterns, 3 artists/track): - Merged join-table: 15ms (9.3x faster) - Merged json_tree: 30ms (4.6x faster) - Unmerged baseline: 137ms * refactor: simplify role condition SQL generation and benchmark Extract shared roleCondSQL/roleExistsSQL helpers to deduplicate the EXISTS template between roleCond and roleCondGroup. Use slices.Chunk for batching per project convention. Extract runBenchQuery helper to eliminate triplicated benchmark execution loop. * chore: raise roleCondBatchSize to 350 The empirical SQLite limit is 496 conditions per merged EXISTS subquery. Raising from 200 to 350 reduces the number of batches (e.g. 500 patterns now splits into 2 batches instead of 3). * fix(server): apply OR-merge optimization to tag conditions too Generalize mergeRoleConds into mergeJsonConds to also collapse multiple tag conditions for the same tag (e.g. genre) within OR groups. This gives the same ~5x speedup for tag-heavy smart playlists as the role optimization gives for artist-heavy ones. * refactor: benchmark uses real criteria pipeline instead of hand-built SQL The "Current" sub-benchmark now builds criteria.Criteria expressions and runs them through the actual newSmartPlaylistCriteria → Where() → ToSql() pipeline, validating the real production code path. The baseline still uses hand-built SQL representing the old json_tree approach. * fix: stabilize merged group ordering and close rows before error check Sort group keys in mergeJsonConds so the merged additions have deterministic order across runs, improving SQLite statement cache reuse. Move rows.Close() before rows.Err() in benchmark helper.
2026-05-22 18:00:13 -03:00
continue
}
g, exists := groups["tag:"+c.tag]
if !exists {
g = &group{tag: c.tag, numeric: c.numeric}
groups["tag:"+c.tag] = g
}
g.entries = append(g.entries, condEntry{index: i, cond: c.cond})
}
}
remove := make(map[int]bool)
var additions []squirrel.Sqlizer
for _, key := range slices.Sorted(maps.Keys(groups)) {
g := groups[key]
if len(g.entries) < 2 {
continue
}
batchConds := make([]squirrel.Sqlizer, len(g.entries))
fix(smartplaylists): optimize smart playlist performance for role and tag criteria (#5515) * fix(server): optimize smart playlist role queries for large criteria (#5511) Role-based smart playlist criteria (artist, composer, etc.) now query the indexed media_file_artists join table instead of parsing JSON via json_tree() on every row. Multiple conditions for the same role within an OR group are merged into a single EXISTS subquery (batched at 200 to stay under SQLite's expression tree depth limit). A composite index (media_file_id, role) replaces the now-redundant single-column (media_file_id) index on media_file_artists. Benchmark (40k tracks, 500 patterns, 3 artists/track): - Merged join-table: 15ms (9.3x faster) - Merged json_tree: 30ms (4.6x faster) - Unmerged baseline: 137ms * refactor: simplify role condition SQL generation and benchmark Extract shared roleCondSQL/roleExistsSQL helpers to deduplicate the EXISTS template between roleCond and roleCondGroup. Use slices.Chunk for batching per project convention. Extract runBenchQuery helper to eliminate triplicated benchmark execution loop. * chore: raise roleCondBatchSize to 350 The empirical SQLite limit is 496 conditions per merged EXISTS subquery. Raising from 200 to 350 reduces the number of batches (e.g. 500 patterns now splits into 2 batches instead of 3). * fix(server): apply OR-merge optimization to tag conditions too Generalize mergeRoleConds into mergeJsonConds to also collapse multiple tag conditions for the same tag (e.g. genre) within OR groups. This gives the same ~5x speedup for tag-heavy smart playlists as the role optimization gives for artist-heavy ones. * refactor: benchmark uses real criteria pipeline instead of hand-built SQL The "Current" sub-benchmark now builds criteria.Criteria expressions and runs them through the actual newSmartPlaylistCriteria → Where() → ToSql() pipeline, validating the real production code path. The baseline still uses hand-built SQL representing the old json_tree approach. * fix: stabilize merged group ordering and close rows before error check Sort group keys in mergeJsonConds so the merged additions have deterministic order across runs, improving SQLite statement cache reuse. Move rows.Close() before rows.Err() in benchmark helper.
2026-05-22 18:00:13 -03:00
for i, e := range g.entries {
remove[e.index] = true
batchConds[i] = e.cond
fix(smartplaylists): optimize smart playlist performance for role and tag criteria (#5515) * fix(server): optimize smart playlist role queries for large criteria (#5511) Role-based smart playlist criteria (artist, composer, etc.) now query the indexed media_file_artists join table instead of parsing JSON via json_tree() on every row. Multiple conditions for the same role within an OR group are merged into a single EXISTS subquery (batched at 200 to stay under SQLite's expression tree depth limit). A composite index (media_file_id, role) replaces the now-redundant single-column (media_file_id) index on media_file_artists. Benchmark (40k tracks, 500 patterns, 3 artists/track): - Merged join-table: 15ms (9.3x faster) - Merged json_tree: 30ms (4.6x faster) - Unmerged baseline: 137ms * refactor: simplify role condition SQL generation and benchmark Extract shared roleCondSQL/roleExistsSQL helpers to deduplicate the EXISTS template between roleCond and roleCondGroup. Use slices.Chunk for batching per project convention. Extract runBenchQuery helper to eliminate triplicated benchmark execution loop. * chore: raise roleCondBatchSize to 350 The empirical SQLite limit is 496 conditions per merged EXISTS subquery. Raising from 200 to 350 reduces the number of batches (e.g. 500 patterns now splits into 2 batches instead of 3). * fix(server): apply OR-merge optimization to tag conditions too Generalize mergeRoleConds into mergeJsonConds to also collapse multiple tag conditions for the same tag (e.g. genre) within OR groups. This gives the same ~5x speedup for tag-heavy smart playlists as the role optimization gives for artist-heavy ones. * refactor: benchmark uses real criteria pipeline instead of hand-built SQL The "Current" sub-benchmark now builds criteria.Criteria expressions and runs them through the actual newSmartPlaylistCriteria → Where() → ToSql() pipeline, validating the real production code path. The baseline still uses hand-built SQL representing the old json_tree approach. * fix: stabilize merged group ordering and close rows before error check Sort group keys in mergeJsonConds so the merged additions have deterministic order across runs, improving SQLite statement cache reuse. Move rows.Close() before rows.Err() in benchmark helper.
2026-05-22 18:00:13 -03:00
}
if g.isRole {
role := key[len("role:"):]
for batch := range slices.Chunk(batchConds, jsonCondBatchSize) {
additions = append(additions, roleCondGroup{role: role, conds: batch, not: negated})
fix(smartplaylists): optimize smart playlist performance for role and tag criteria (#5515) * fix(server): optimize smart playlist role queries for large criteria (#5511) Role-based smart playlist criteria (artist, composer, etc.) now query the indexed media_file_artists join table instead of parsing JSON via json_tree() on every row. Multiple conditions for the same role within an OR group are merged into a single EXISTS subquery (batched at 200 to stay under SQLite's expression tree depth limit). A composite index (media_file_id, role) replaces the now-redundant single-column (media_file_id) index on media_file_artists. Benchmark (40k tracks, 500 patterns, 3 artists/track): - Merged join-table: 15ms (9.3x faster) - Merged json_tree: 30ms (4.6x faster) - Unmerged baseline: 137ms * refactor: simplify role condition SQL generation and benchmark Extract shared roleCondSQL/roleExistsSQL helpers to deduplicate the EXISTS template between roleCond and roleCondGroup. Use slices.Chunk for batching per project convention. Extract runBenchQuery helper to eliminate triplicated benchmark execution loop. * chore: raise roleCondBatchSize to 350 The empirical SQLite limit is 496 conditions per merged EXISTS subquery. Raising from 200 to 350 reduces the number of batches (e.g. 500 patterns now splits into 2 batches instead of 3). * fix(server): apply OR-merge optimization to tag conditions too Generalize mergeRoleConds into mergeJsonConds to also collapse multiple tag conditions for the same tag (e.g. genre) within OR groups. This gives the same ~5x speedup for tag-heavy smart playlists as the role optimization gives for artist-heavy ones. * refactor: benchmark uses real criteria pipeline instead of hand-built SQL The "Current" sub-benchmark now builds criteria.Criteria expressions and runs them through the actual newSmartPlaylistCriteria → Where() → ToSql() pipeline, validating the real production code path. The baseline still uses hand-built SQL representing the old json_tree approach. * fix: stabilize merged group ordering and close rows before error check Sort group keys in mergeJsonConds so the merged additions have deterministic order across runs, improving SQLite statement cache reuse. Move rows.Close() before rows.Err() in benchmark helper.
2026-05-22 18:00:13 -03:00
}
} else {
for batch := range slices.Chunk(batchConds, jsonCondBatchSize) {
additions = append(additions, tagCondGroup{tag: g.tag, numeric: g.numeric, conds: batch, not: negated})
fix(smartplaylists): optimize smart playlist performance for role and tag criteria (#5515) * fix(server): optimize smart playlist role queries for large criteria (#5511) Role-based smart playlist criteria (artist, composer, etc.) now query the indexed media_file_artists join table instead of parsing JSON via json_tree() on every row. Multiple conditions for the same role within an OR group are merged into a single EXISTS subquery (batched at 200 to stay under SQLite's expression tree depth limit). A composite index (media_file_id, role) replaces the now-redundant single-column (media_file_id) index on media_file_artists. Benchmark (40k tracks, 500 patterns, 3 artists/track): - Merged join-table: 15ms (9.3x faster) - Merged json_tree: 30ms (4.6x faster) - Unmerged baseline: 137ms * refactor: simplify role condition SQL generation and benchmark Extract shared roleCondSQL/roleExistsSQL helpers to deduplicate the EXISTS template between roleCond and roleCondGroup. Use slices.Chunk for batching per project convention. Extract runBenchQuery helper to eliminate triplicated benchmark execution loop. * chore: raise roleCondBatchSize to 350 The empirical SQLite limit is 496 conditions per merged EXISTS subquery. Raising from 200 to 350 reduces the number of batches (e.g. 500 patterns now splits into 2 batches instead of 3). * fix(server): apply OR-merge optimization to tag conditions too Generalize mergeRoleConds into mergeJsonConds to also collapse multiple tag conditions for the same tag (e.g. genre) within OR groups. This gives the same ~5x speedup for tag-heavy smart playlists as the role optimization gives for artist-heavy ones. * refactor: benchmark uses real criteria pipeline instead of hand-built SQL The "Current" sub-benchmark now builds criteria.Criteria expressions and runs them through the actual newSmartPlaylistCriteria → Where() → ToSql() pipeline, validating the real production code path. The baseline still uses hand-built SQL representing the old json_tree approach. * fix: stabilize merged group ordering and close rows before error check Sort group keys in mergeJsonConds so the merged additions have deterministic order across runs, improving SQLite statement cache reuse. Move rows.Close() before rows.Err() in benchmark helper.
2026-05-22 18:00:13 -03:00
}
}
}
if len(remove) == 0 {
return conds, false
fix(smartplaylists): optimize smart playlist performance for role and tag criteria (#5515) * fix(server): optimize smart playlist role queries for large criteria (#5511) Role-based smart playlist criteria (artist, composer, etc.) now query the indexed media_file_artists join table instead of parsing JSON via json_tree() on every row. Multiple conditions for the same role within an OR group are merged into a single EXISTS subquery (batched at 200 to stay under SQLite's expression tree depth limit). A composite index (media_file_id, role) replaces the now-redundant single-column (media_file_id) index on media_file_artists. Benchmark (40k tracks, 500 patterns, 3 artists/track): - Merged join-table: 15ms (9.3x faster) - Merged json_tree: 30ms (4.6x faster) - Unmerged baseline: 137ms * refactor: simplify role condition SQL generation and benchmark Extract shared roleCondSQL/roleExistsSQL helpers to deduplicate the EXISTS template between roleCond and roleCondGroup. Use slices.Chunk for batching per project convention. Extract runBenchQuery helper to eliminate triplicated benchmark execution loop. * chore: raise roleCondBatchSize to 350 The empirical SQLite limit is 496 conditions per merged EXISTS subquery. Raising from 200 to 350 reduces the number of batches (e.g. 500 patterns now splits into 2 batches instead of 3). * fix(server): apply OR-merge optimization to tag conditions too Generalize mergeRoleConds into mergeJsonConds to also collapse multiple tag conditions for the same tag (e.g. genre) within OR groups. This gives the same ~5x speedup for tag-heavy smart playlists as the role optimization gives for artist-heavy ones. * refactor: benchmark uses real criteria pipeline instead of hand-built SQL The "Current" sub-benchmark now builds criteria.Criteria expressions and runs them through the actual newSmartPlaylistCriteria → Where() → ToSql() pipeline, validating the real production code path. The baseline still uses hand-built SQL representing the old json_tree approach. * fix: stabilize merged group ordering and close rows before error check Sort group keys in mergeJsonConds so the merged additions have deterministic order across runs, improving SQLite statement cache reuse. Move rows.Close() before rows.Err() in benchmark helper.
2026-05-22 18:00:13 -03:00
}
result := make([]squirrel.Sqlizer, 0, len(conds)-len(remove)+len(additions))
for i, s := range conds {
fix(smartplaylists): optimize smart playlist performance for role and tag criteria (#5515) * fix(server): optimize smart playlist role queries for large criteria (#5511) Role-based smart playlist criteria (artist, composer, etc.) now query the indexed media_file_artists join table instead of parsing JSON via json_tree() on every row. Multiple conditions for the same role within an OR group are merged into a single EXISTS subquery (batched at 200 to stay under SQLite's expression tree depth limit). A composite index (media_file_id, role) replaces the now-redundant single-column (media_file_id) index on media_file_artists. Benchmark (40k tracks, 500 patterns, 3 artists/track): - Merged join-table: 15ms (9.3x faster) - Merged json_tree: 30ms (4.6x faster) - Unmerged baseline: 137ms * refactor: simplify role condition SQL generation and benchmark Extract shared roleCondSQL/roleExistsSQL helpers to deduplicate the EXISTS template between roleCond and roleCondGroup. Use slices.Chunk for batching per project convention. Extract runBenchQuery helper to eliminate triplicated benchmark execution loop. * chore: raise roleCondBatchSize to 350 The empirical SQLite limit is 496 conditions per merged EXISTS subquery. Raising from 200 to 350 reduces the number of batches (e.g. 500 patterns now splits into 2 batches instead of 3). * fix(server): apply OR-merge optimization to tag conditions too Generalize mergeRoleConds into mergeJsonConds to also collapse multiple tag conditions for the same tag (e.g. genre) within OR groups. This gives the same ~5x speedup for tag-heavy smart playlists as the role optimization gives for artist-heavy ones. * refactor: benchmark uses real criteria pipeline instead of hand-built SQL The "Current" sub-benchmark now builds criteria.Criteria expressions and runs them through the actual newSmartPlaylistCriteria → Where() → ToSql() pipeline, validating the real production code path. The baseline still uses hand-built SQL representing the old json_tree approach. * fix: stabilize merged group ordering and close rows before error check Sort group keys in mergeJsonConds so the merged additions have deterministic order across runs, improving SQLite statement cache reuse. Move rows.Close() before rows.Err() in benchmark helper.
2026-05-22 18:00:13 -03:00
if !remove[i] {
result = append(result, s)
}
}
return append(result, additions...), true
fix(smartplaylists): optimize smart playlist performance for role and tag criteria (#5515) * fix(server): optimize smart playlist role queries for large criteria (#5511) Role-based smart playlist criteria (artist, composer, etc.) now query the indexed media_file_artists join table instead of parsing JSON via json_tree() on every row. Multiple conditions for the same role within an OR group are merged into a single EXISTS subquery (batched at 200 to stay under SQLite's expression tree depth limit). A composite index (media_file_id, role) replaces the now-redundant single-column (media_file_id) index on media_file_artists. Benchmark (40k tracks, 500 patterns, 3 artists/track): - Merged join-table: 15ms (9.3x faster) - Merged json_tree: 30ms (4.6x faster) - Unmerged baseline: 137ms * refactor: simplify role condition SQL generation and benchmark Extract shared roleCondSQL/roleExistsSQL helpers to deduplicate the EXISTS template between roleCond and roleCondGroup. Use slices.Chunk for batching per project convention. Extract runBenchQuery helper to eliminate triplicated benchmark execution loop. * chore: raise roleCondBatchSize to 350 The empirical SQLite limit is 496 conditions per merged EXISTS subquery. Raising from 200 to 350 reduces the number of batches (e.g. 500 patterns now splits into 2 batches instead of 3). * fix(server): apply OR-merge optimization to tag conditions too Generalize mergeRoleConds into mergeJsonConds to also collapse multiple tag conditions for the same tag (e.g. genre) within OR groups. This gives the same ~5x speedup for tag-heavy smart playlists as the role optimization gives for artist-heavy ones. * refactor: benchmark uses real criteria pipeline instead of hand-built SQL The "Current" sub-benchmark now builds criteria.Criteria expressions and runs them through the actual newSmartPlaylistCriteria → Where() → ToSql() pipeline, validating the real production code path. The baseline still uses hand-built SQL representing the old json_tree approach. * fix: stabilize merged group ordering and close rows before error check Sort group keys in mergeJsonConds so the merged additions have deterministic order across runs, improving SQLite statement cache reuse. Move rows.Close() before rows.Err() in benchmark helper.
2026-05-22 18:00:13 -03:00
}
// roleCondGroup represents multiple role conditions for the same role, merged into a single
// (optionally negated) EXISTS subquery for performance.
fix(smartplaylists): optimize smart playlist performance for role and tag criteria (#5515) * fix(server): optimize smart playlist role queries for large criteria (#5511) Role-based smart playlist criteria (artist, composer, etc.) now query the indexed media_file_artists join table instead of parsing JSON via json_tree() on every row. Multiple conditions for the same role within an OR group are merged into a single EXISTS subquery (batched at 200 to stay under SQLite's expression tree depth limit). A composite index (media_file_id, role) replaces the now-redundant single-column (media_file_id) index on media_file_artists. Benchmark (40k tracks, 500 patterns, 3 artists/track): - Merged join-table: 15ms (9.3x faster) - Merged json_tree: 30ms (4.6x faster) - Unmerged baseline: 137ms * refactor: simplify role condition SQL generation and benchmark Extract shared roleCondSQL/roleExistsSQL helpers to deduplicate the EXISTS template between roleCond and roleCondGroup. Use slices.Chunk for batching per project convention. Extract runBenchQuery helper to eliminate triplicated benchmark execution loop. * chore: raise roleCondBatchSize to 350 The empirical SQLite limit is 496 conditions per merged EXISTS subquery. Raising from 200 to 350 reduces the number of batches (e.g. 500 patterns now splits into 2 batches instead of 3). * fix(server): apply OR-merge optimization to tag conditions too Generalize mergeRoleConds into mergeJsonConds to also collapse multiple tag conditions for the same tag (e.g. genre) within OR groups. This gives the same ~5x speedup for tag-heavy smart playlists as the role optimization gives for artist-heavy ones. * refactor: benchmark uses real criteria pipeline instead of hand-built SQL The "Current" sub-benchmark now builds criteria.Criteria expressions and runs them through the actual newSmartPlaylistCriteria → Where() → ToSql() pipeline, validating the real production code path. The baseline still uses hand-built SQL representing the old json_tree approach. * fix: stabilize merged group ordering and close rows before error check Sort group keys in mergeJsonConds so the merged additions have deterministic order across runs, improving SQLite statement cache reuse. Move rows.Close() before rows.Err() in benchmark helper.
2026-05-22 18:00:13 -03:00
type roleCondGroup struct {
role string
conds []squirrel.Sqlizer
not bool
fix(smartplaylists): optimize smart playlist performance for role and tag criteria (#5515) * fix(server): optimize smart playlist role queries for large criteria (#5511) Role-based smart playlist criteria (artist, composer, etc.) now query the indexed media_file_artists join table instead of parsing JSON via json_tree() on every row. Multiple conditions for the same role within an OR group are merged into a single EXISTS subquery (batched at 200 to stay under SQLite's expression tree depth limit). A composite index (media_file_id, role) replaces the now-redundant single-column (media_file_id) index on media_file_artists. Benchmark (40k tracks, 500 patterns, 3 artists/track): - Merged join-table: 15ms (9.3x faster) - Merged json_tree: 30ms (4.6x faster) - Unmerged baseline: 137ms * refactor: simplify role condition SQL generation and benchmark Extract shared roleCondSQL/roleExistsSQL helpers to deduplicate the EXISTS template between roleCond and roleCondGroup. Use slices.Chunk for batching per project convention. Extract runBenchQuery helper to eliminate triplicated benchmark execution loop. * chore: raise roleCondBatchSize to 350 The empirical SQLite limit is 496 conditions per merged EXISTS subquery. Raising from 200 to 350 reduces the number of batches (e.g. 500 patterns now splits into 2 batches instead of 3). * fix(server): apply OR-merge optimization to tag conditions too Generalize mergeRoleConds into mergeJsonConds to also collapse multiple tag conditions for the same tag (e.g. genre) within OR groups. This gives the same ~5x speedup for tag-heavy smart playlists as the role optimization gives for artist-heavy ones. * refactor: benchmark uses real criteria pipeline instead of hand-built SQL The "Current" sub-benchmark now builds criteria.Criteria expressions and runs them through the actual newSmartPlaylistCriteria → Where() → ToSql() pipeline, validating the real production code path. The baseline still uses hand-built SQL representing the old json_tree approach. * fix: stabilize merged group ordering and close rows before error check Sort group keys in mergeJsonConds so the merged additions have deterministic order across runs, improving SQLite statement cache reuse. Move rows.Close() before rows.Err() in benchmark helper.
2026-05-22 18:00:13 -03:00
}
func (g roleCondGroup) ToSql() (string, []any, error) {
innerParts := make([]string, 0, len(g.conds))
allArgs := []any{g.role}
for _, c := range g.conds {
part, args, err := roleCondSQL(c)
if err != nil {
return "", nil, err
}
innerParts = append(innerParts, part)
allArgs = append(allArgs, args...)
}
cond := roleExistsSQL("(" + strings.Join(innerParts, " OR ") + ")")
if g.not {
cond = "not " + cond
}
fix(smartplaylists): optimize smart playlist performance for role and tag criteria (#5515) * fix(server): optimize smart playlist role queries for large criteria (#5511) Role-based smart playlist criteria (artist, composer, etc.) now query the indexed media_file_artists join table instead of parsing JSON via json_tree() on every row. Multiple conditions for the same role within an OR group are merged into a single EXISTS subquery (batched at 200 to stay under SQLite's expression tree depth limit). A composite index (media_file_id, role) replaces the now-redundant single-column (media_file_id) index on media_file_artists. Benchmark (40k tracks, 500 patterns, 3 artists/track): - Merged join-table: 15ms (9.3x faster) - Merged json_tree: 30ms (4.6x faster) - Unmerged baseline: 137ms * refactor: simplify role condition SQL generation and benchmark Extract shared roleCondSQL/roleExistsSQL helpers to deduplicate the EXISTS template between roleCond and roleCondGroup. Use slices.Chunk for batching per project convention. Extract runBenchQuery helper to eliminate triplicated benchmark execution loop. * chore: raise roleCondBatchSize to 350 The empirical SQLite limit is 496 conditions per merged EXISTS subquery. Raising from 200 to 350 reduces the number of batches (e.g. 500 patterns now splits into 2 batches instead of 3). * fix(server): apply OR-merge optimization to tag conditions too Generalize mergeRoleConds into mergeJsonConds to also collapse multiple tag conditions for the same tag (e.g. genre) within OR groups. This gives the same ~5x speedup for tag-heavy smart playlists as the role optimization gives for artist-heavy ones. * refactor: benchmark uses real criteria pipeline instead of hand-built SQL The "Current" sub-benchmark now builds criteria.Criteria expressions and runs them through the actual newSmartPlaylistCriteria → Where() → ToSql() pipeline, validating the real production code path. The baseline still uses hand-built SQL representing the old json_tree approach. * fix: stabilize merged group ordering and close rows before error check Sort group keys in mergeJsonConds so the merged additions have deterministic order across runs, improving SQLite statement cache reuse. Move rows.Close() before rows.Err() in benchmark helper.
2026-05-22 18:00:13 -03:00
return cond, allArgs, nil
}
// tagCondGroup represents multiple tag conditions for the same tag, merged into a single
// (optionally negated) EXISTS subquery for performance.
fix(smartplaylists): optimize smart playlist performance for role and tag criteria (#5515) * fix(server): optimize smart playlist role queries for large criteria (#5511) Role-based smart playlist criteria (artist, composer, etc.) now query the indexed media_file_artists join table instead of parsing JSON via json_tree() on every row. Multiple conditions for the same role within an OR group are merged into a single EXISTS subquery (batched at 200 to stay under SQLite's expression tree depth limit). A composite index (media_file_id, role) replaces the now-redundant single-column (media_file_id) index on media_file_artists. Benchmark (40k tracks, 500 patterns, 3 artists/track): - Merged join-table: 15ms (9.3x faster) - Merged json_tree: 30ms (4.6x faster) - Unmerged baseline: 137ms * refactor: simplify role condition SQL generation and benchmark Extract shared roleCondSQL/roleExistsSQL helpers to deduplicate the EXISTS template between roleCond and roleCondGroup. Use slices.Chunk for batching per project convention. Extract runBenchQuery helper to eliminate triplicated benchmark execution loop. * chore: raise roleCondBatchSize to 350 The empirical SQLite limit is 496 conditions per merged EXISTS subquery. Raising from 200 to 350 reduces the number of batches (e.g. 500 patterns now splits into 2 batches instead of 3). * fix(server): apply OR-merge optimization to tag conditions too Generalize mergeRoleConds into mergeJsonConds to also collapse multiple tag conditions for the same tag (e.g. genre) within OR groups. This gives the same ~5x speedup for tag-heavy smart playlists as the role optimization gives for artist-heavy ones. * refactor: benchmark uses real criteria pipeline instead of hand-built SQL The "Current" sub-benchmark now builds criteria.Criteria expressions and runs them through the actual newSmartPlaylistCriteria → Where() → ToSql() pipeline, validating the real production code path. The baseline still uses hand-built SQL representing the old json_tree approach. * fix: stabilize merged group ordering and close rows before error check Sort group keys in mergeJsonConds so the merged additions have deterministic order across runs, improving SQLite statement cache reuse. Move rows.Close() before rows.Err() in benchmark helper.
2026-05-22 18:00:13 -03:00
type tagCondGroup struct {
tag string
numeric bool
conds []squirrel.Sqlizer
not bool
fix(smartplaylists): optimize smart playlist performance for role and tag criteria (#5515) * fix(server): optimize smart playlist role queries for large criteria (#5511) Role-based smart playlist criteria (artist, composer, etc.) now query the indexed media_file_artists join table instead of parsing JSON via json_tree() on every row. Multiple conditions for the same role within an OR group are merged into a single EXISTS subquery (batched at 200 to stay under SQLite's expression tree depth limit). A composite index (media_file_id, role) replaces the now-redundant single-column (media_file_id) index on media_file_artists. Benchmark (40k tracks, 500 patterns, 3 artists/track): - Merged join-table: 15ms (9.3x faster) - Merged json_tree: 30ms (4.6x faster) - Unmerged baseline: 137ms * refactor: simplify role condition SQL generation and benchmark Extract shared roleCondSQL/roleExistsSQL helpers to deduplicate the EXISTS template between roleCond and roleCondGroup. Use slices.Chunk for batching per project convention. Extract runBenchQuery helper to eliminate triplicated benchmark execution loop. * chore: raise roleCondBatchSize to 350 The empirical SQLite limit is 496 conditions per merged EXISTS subquery. Raising from 200 to 350 reduces the number of batches (e.g. 500 patterns now splits into 2 batches instead of 3). * fix(server): apply OR-merge optimization to tag conditions too Generalize mergeRoleConds into mergeJsonConds to also collapse multiple tag conditions for the same tag (e.g. genre) within OR groups. This gives the same ~5x speedup for tag-heavy smart playlists as the role optimization gives for artist-heavy ones. * refactor: benchmark uses real criteria pipeline instead of hand-built SQL The "Current" sub-benchmark now builds criteria.Criteria expressions and runs them through the actual newSmartPlaylistCriteria → Where() → ToSql() pipeline, validating the real production code path. The baseline still uses hand-built SQL representing the old json_tree approach. * fix: stabilize merged group ordering and close rows before error check Sort group keys in mergeJsonConds so the merged additions have deterministic order across runs, improving SQLite statement cache reuse. Move rows.Close() before rows.Err() in benchmark helper.
2026-05-22 18:00:13 -03:00
}
func (g tagCondGroup) ToSql() (string, []any, error) {
innerParts := make([]string, 0, len(g.conds))
var allArgs []any
for _, c := range g.conds {
part, args, err := c.ToSql()
if err != nil {
return "", nil, err
}
if g.numeric {
part = strings.ReplaceAll(part, "value", "CAST(value AS REAL)")
}
innerParts = append(innerParts, part)
allArgs = append(allArgs, args...)
}
cond := fmt.Sprintf("exists (select 1 from json_tree(media_file.tags, '$.%s') where key='value' and (%s))",
g.tag, strings.Join(innerParts, " OR "))
if g.not {
cond = "not " + cond
}
fix(smartplaylists): optimize smart playlist performance for role and tag criteria (#5515) * fix(server): optimize smart playlist role queries for large criteria (#5511) Role-based smart playlist criteria (artist, composer, etc.) now query the indexed media_file_artists join table instead of parsing JSON via json_tree() on every row. Multiple conditions for the same role within an OR group are merged into a single EXISTS subquery (batched at 200 to stay under SQLite's expression tree depth limit). A composite index (media_file_id, role) replaces the now-redundant single-column (media_file_id) index on media_file_artists. Benchmark (40k tracks, 500 patterns, 3 artists/track): - Merged join-table: 15ms (9.3x faster) - Merged json_tree: 30ms (4.6x faster) - Unmerged baseline: 137ms * refactor: simplify role condition SQL generation and benchmark Extract shared roleCondSQL/roleExistsSQL helpers to deduplicate the EXISTS template between roleCond and roleCondGroup. Use slices.Chunk for batching per project convention. Extract runBenchQuery helper to eliminate triplicated benchmark execution loop. * chore: raise roleCondBatchSize to 350 The empirical SQLite limit is 496 conditions per merged EXISTS subquery. Raising from 200 to 350 reduces the number of batches (e.g. 500 patterns now splits into 2 batches instead of 3). * fix(server): apply OR-merge optimization to tag conditions too Generalize mergeRoleConds into mergeJsonConds to also collapse multiple tag conditions for the same tag (e.g. genre) within OR groups. This gives the same ~5x speedup for tag-heavy smart playlists as the role optimization gives for artist-heavy ones. * refactor: benchmark uses real criteria pipeline instead of hand-built SQL The "Current" sub-benchmark now builds criteria.Criteria expressions and runs them through the actual newSmartPlaylistCriteria → Where() → ToSql() pipeline, validating the real production code path. The baseline still uses hand-built SQL representing the old json_tree approach. * fix: stabilize merged group ordering and close rows before error check Sort group keys in mergeJsonConds so the merged additions have deterministic order across runs, improving SQLite statement cache reuse. Move rows.Close() before rows.Err() in benchmark helper.
2026-05-22 18:00:13 -03:00
return cond, allArgs, nil
}
feat(smartplaylists): relax playlist visibility in inPlaylist/notInPlaylist rules (#5411) * test(e2e): add end-to-end tests for smart playlists functionality Signed-off-by: Deluan <deluan@navidrome.org> * fix: enforce playlist visibility in smart playlist InPlaylist/NotInPlaylist rules Previously, the InPlaylist/NotInPlaylist smart playlist criteria only allowed referencing public playlists, regardless of who owned the smart playlist. This was too restrictive for owners referencing their own private playlists and for admins who should have unrestricted access. The fix passes the smart playlist owner's identity and admin status into the criteria SQL builder, so that: admins can reference any playlist, regular users can reference public playlists plus their own private ones, and inaccessible referenced playlists produce a warning instead of a hard error. Also prevents recursive refresh of child playlists the owner cannot access. * test(e2e): clarify user roles and fix playlist visibility tests Renamed testUser/otherUser to adminUser/regularUser to make the admin vs regular user distinction explicit in test code. Fixed three playlist visibility tests that were evaluating as admin (bypassing all access checks) instead of as a regular user, so the public playlist path is now actually exercised. All playlist operator tests now use explicit evaluateRuleAs calls with the appropriate user role. * fix: sync rulesSQL criteria after limitPercent resolution The rulesSQL struct captures a copy of rules at creation time. When limitPercent is resolved later, rules.Limit is updated but rulesSQL still holds the stale value. This caused percentage-based smart playlist limits to be silently ignored. Fix by updating rulesSQL.criteria after the resolution. * refactor: convert inList to a method on smartPlaylistCriteria The inList function already receives ownerID and ownerIsAdmin from the smartPlaylistCriteria caller. Making it a method lets it access those fields directly from the receiver, simplifying the signature and staying consistent with exprSQL which was already converted to a method. * refactor: simplify function signatures by removing type parameters in criteria_sql.go Signed-off-by: Deluan <deluan@navidrome.org> --------- Signed-off-by: Deluan <deluan@navidrome.org>
2026-04-25 14:59:06 -04:00
func singleField(values map[string]any) (string, any, criteria.FieldInfo, bool) {
if len(values) != 1 {
return "", nil, criteria.FieldInfo{}, false
}
for field, value := range values {
info, ok := criteria.LookupField(field)
return field, value, info, ok
}
return "", nil, criteria.FieldInfo{}, false
}
feat(smartplaylists): relax playlist visibility in inPlaylist/notInPlaylist rules (#5411) * test(e2e): add end-to-end tests for smart playlists functionality Signed-off-by: Deluan <deluan@navidrome.org> * fix: enforce playlist visibility in smart playlist InPlaylist/NotInPlaylist rules Previously, the InPlaylist/NotInPlaylist smart playlist criteria only allowed referencing public playlists, regardless of who owned the smart playlist. This was too restrictive for owners referencing their own private playlists and for admins who should have unrestricted access. The fix passes the smart playlist owner's identity and admin status into the criteria SQL builder, so that: admins can reference any playlist, regular users can reference public playlists plus their own private ones, and inaccessible referenced playlists produce a warning instead of a hard error. Also prevents recursive refresh of child playlists the owner cannot access. * test(e2e): clarify user roles and fix playlist visibility tests Renamed testUser/otherUser to adminUser/regularUser to make the admin vs regular user distinction explicit in test code. Fixed three playlist visibility tests that were evaluating as admin (bypassing all access checks) instead of as a regular user, so the public playlist path is now actually exercised. All playlist operator tests now use explicit evaluateRuleAs calls with the appropriate user role. * fix: sync rulesSQL criteria after limitPercent resolution The rulesSQL struct captures a copy of rules at creation time. When limitPercent is resolved later, rules.Limit is updated but rulesSQL still holds the stale value. This caused percentage-based smart playlist limits to be silently ignored. Fix by updating rulesSQL.criteria after the resolution. * refactor: convert inList to a method on smartPlaylistCriteria The inList function already receives ownerID and ownerIsAdmin from the smartPlaylistCriteria caller. Making it a method lets it access those fields directly from the receiver, simplifying the signature and staying consistent with exprSQL which was already converted to a method. * refactor: simplify function signatures by removing type parameters in criteria_sql.go Signed-off-by: Deluan <deluan@navidrome.org> --------- Signed-off-by: Deluan <deluan@navidrome.org>
2026-04-25 14:59:06 -04:00
func sqlFields(values map[string]any) (map[string]any, error) {
fields := make(map[string]any, len(values))
for field, value := range values {
info, ok := criteria.LookupField(field)
if !ok {
return nil, fmt.Errorf("invalid field in criteria: %s", field)
}
if info.IsTag || info.IsRole {
return nil, fmt.Errorf("tag and role criteria must contain exactly one field: %s", field)
}
sqlField, ok := fieldExpr(info.Name())
if !ok || sqlField == "" {
return nil, fmt.Errorf("invalid field in criteria: %s", field)
}
fields[sqlField] = value
}
return fields, nil
}
func fieldExpr(name string) (string, bool) {
field, ok := smartPlaylistFields[strings.ToLower(name)]
return field.expr, ok
}
func fieldJoinType(name string) smartPlaylistJoinType {
info, ok := criteria.LookupField(name)
if !ok {
return smartPlaylistJoinNone
}
field, ok := smartPlaylistFields[info.Name()]
if !ok {
return smartPlaylistJoinNone
}
return field.joinType
}
func (c smartPlaylistCriteria) ExpressionJoins() smartPlaylistJoinType {
var joins smartPlaylistJoinType
refactor: centralize criteria sort parsing and extract smart playlist logic (#5415) * test: add tests for recordingdate alias resolution in smart playlists Signed-off-by: Deluan <deluan@navidrome.org> * refactor: update FieldInfo structure and simplify fieldMap initialization Signed-off-by: Deluan <deluan@navidrome.org> * refactor: move sort parsing logic from persistence to criteria package Extracted sort field parsing, validation, and direction handling from persistence/criteria_sql.go into model/criteria/sort.go. The new OrderByFields method on Criteria parses the Sort/Order strings into validated SortField structs (field name + direction), resolving aliases and handling +/- prefixes and order inversion. The persistence layer now consumes these parsed fields and only handles SQL expression mapping. This centralizes sort parsing to enforce consistent implementations. * refactor: standardize field access in smartPlaylistCriteria structure Signed-off-by: Deluan <deluan@navidrome.org> * refactor: add ResolveLimit method to Criteria Moved the percentage-limit resolution logic from playlist_repository into Criteria.ResolveLimit, replacing the 3-line mutate-after-query pattern with a single method call. The method preserves LimitPercent rather than zeroing it, since IsPercentageLimit already returns false once Limit is set, making the clear redundant and lossy. * refactor: improve child playlist loading and error handling in refresh logic Signed-off-by: Deluan <deluan@navidrome.org> * refactor: extract smart playlist logic to dedicated files Moved refreshSmartPlaylist, addSmartPlaylistAnnotationJoins, and addCriteria methods from playlist_repository.go to a new smart_playlist_repository.go file. Extracted all smart playlist tests to smart_playlist_repository_test.go. Added DeferCleanup to the "valid rules" test to fix ordering flakiness when Ginkgo randomizes test execution across files. * refactor: break refreshSmartPlaylist into smaller focused methods Split the monolithic refreshSmartPlaylist method into discrete helpers for readability: shouldRefreshSmartPlaylist for guard checks, refreshChildPlaylists for recursive dependency refresh, resolvePercentageLimit for count-based limit resolution, buildSmartPlaylistQuery for assembling the SELECT with joins, and addMediaFileAnnotationJoin to DRY up the repeated annotation join clause. * refactor: deduplicate child playlist IDs in Criteria Signed-off-by: Deluan <deluan@navidrome.org> * refactor: simplify withSmartPlaylistOwner to accept model.User Replaced separate ownerID string and ownerIsAdmin bool parameters with a single model.User struct, reducing the field count in smartPlaylistCriteria and making the option function signature clearer. Updated all call sites and tests accordingly. * fix: handle empty sort fields and propagate child playlist load errors OrderByFields now falls back to [{title, asc}] when all user-supplied sort fields are invalid, preventing empty ORDER BY clauses that would produce invalid SQL in row_number() window functions. Also restored the original behavior where a DB error loading child playlists aborts the parent smart playlist refresh, by making refreshChildPlaylists return a bool. * refactor: log warning when no valid sort fields are found Signed-off-by: Deluan <deluan@navidrome.org> --------- Signed-off-by: Deluan <deluan@navidrome.org>
2026-04-26 14:49:59 -04:00
_ = criteria.Walk(c.Criteria.Expression, func(expr criteria.Expression) error {
for field := range criteria.Fields(expr) {
joins |= fieldJoinType(field)
}
return nil
})
return joins
}
func (c smartPlaylistCriteria) RequiredJoins() smartPlaylistJoinType {
joins := c.ExpressionJoins()
refactor: centralize criteria sort parsing and extract smart playlist logic (#5415) * test: add tests for recordingdate alias resolution in smart playlists Signed-off-by: Deluan <deluan@navidrome.org> * refactor: update FieldInfo structure and simplify fieldMap initialization Signed-off-by: Deluan <deluan@navidrome.org> * refactor: move sort parsing logic from persistence to criteria package Extracted sort field parsing, validation, and direction handling from persistence/criteria_sql.go into model/criteria/sort.go. The new OrderByFields method on Criteria parses the Sort/Order strings into validated SortField structs (field name + direction), resolving aliases and handling +/- prefixes and order inversion. The persistence layer now consumes these parsed fields and only handles SQL expression mapping. This centralizes sort parsing to enforce consistent implementations. * refactor: standardize field access in smartPlaylistCriteria structure Signed-off-by: Deluan <deluan@navidrome.org> * refactor: add ResolveLimit method to Criteria Moved the percentage-limit resolution logic from playlist_repository into Criteria.ResolveLimit, replacing the 3-line mutate-after-query pattern with a single method call. The method preserves LimitPercent rather than zeroing it, since IsPercentageLimit already returns false once Limit is set, making the clear redundant and lossy. * refactor: improve child playlist loading and error handling in refresh logic Signed-off-by: Deluan <deluan@navidrome.org> * refactor: extract smart playlist logic to dedicated files Moved refreshSmartPlaylist, addSmartPlaylistAnnotationJoins, and addCriteria methods from playlist_repository.go to a new smart_playlist_repository.go file. Extracted all smart playlist tests to smart_playlist_repository_test.go. Added DeferCleanup to the "valid rules" test to fix ordering flakiness when Ginkgo randomizes test execution across files. * refactor: break refreshSmartPlaylist into smaller focused methods Split the monolithic refreshSmartPlaylist method into discrete helpers for readability: shouldRefreshSmartPlaylist for guard checks, refreshChildPlaylists for recursive dependency refresh, resolvePercentageLimit for count-based limit resolution, buildSmartPlaylistQuery for assembling the SELECT with joins, and addMediaFileAnnotationJoin to DRY up the repeated annotation join clause. * refactor: deduplicate child playlist IDs in Criteria Signed-off-by: Deluan <deluan@navidrome.org> * refactor: simplify withSmartPlaylistOwner to accept model.User Replaced separate ownerID string and ownerIsAdmin bool parameters with a single model.User struct, reducing the field count in smartPlaylistCriteria and making the option function signature clearer. Updated all call sites and tests accordingly. * fix: handle empty sort fields and propagate child playlist load errors OrderByFields now falls back to [{title, asc}] when all user-supplied sort fields are invalid, preventing empty ORDER BY clauses that would produce invalid SQL in row_number() window functions. Also restored the original behavior where a DB error loading child playlists aborts the parent smart playlist refresh, by making refreshChildPlaylists return a bool. * refactor: log warning when no valid sort fields are found Signed-off-by: Deluan <deluan@navidrome.org> --------- Signed-off-by: Deluan <deluan@navidrome.org>
2026-04-26 14:49:59 -04:00
for _, name := range c.Criteria.SortFieldNames() {
joins |= fieldJoinType(name)
}
return joins
}
func (c smartPlaylistCriteria) OrderBy() string {
refactor: centralize criteria sort parsing and extract smart playlist logic (#5415) * test: add tests for recordingdate alias resolution in smart playlists Signed-off-by: Deluan <deluan@navidrome.org> * refactor: update FieldInfo structure and simplify fieldMap initialization Signed-off-by: Deluan <deluan@navidrome.org> * refactor: move sort parsing logic from persistence to criteria package Extracted sort field parsing, validation, and direction handling from persistence/criteria_sql.go into model/criteria/sort.go. The new OrderByFields method on Criteria parses the Sort/Order strings into validated SortField structs (field name + direction), resolving aliases and handling +/- prefixes and order inversion. The persistence layer now consumes these parsed fields and only handles SQL expression mapping. This centralizes sort parsing to enforce consistent implementations. * refactor: standardize field access in smartPlaylistCriteria structure Signed-off-by: Deluan <deluan@navidrome.org> * refactor: add ResolveLimit method to Criteria Moved the percentage-limit resolution logic from playlist_repository into Criteria.ResolveLimit, replacing the 3-line mutate-after-query pattern with a single method call. The method preserves LimitPercent rather than zeroing it, since IsPercentageLimit already returns false once Limit is set, making the clear redundant and lossy. * refactor: improve child playlist loading and error handling in refresh logic Signed-off-by: Deluan <deluan@navidrome.org> * refactor: extract smart playlist logic to dedicated files Moved refreshSmartPlaylist, addSmartPlaylistAnnotationJoins, and addCriteria methods from playlist_repository.go to a new smart_playlist_repository.go file. Extracted all smart playlist tests to smart_playlist_repository_test.go. Added DeferCleanup to the "valid rules" test to fix ordering flakiness when Ginkgo randomizes test execution across files. * refactor: break refreshSmartPlaylist into smaller focused methods Split the monolithic refreshSmartPlaylist method into discrete helpers for readability: shouldRefreshSmartPlaylist for guard checks, refreshChildPlaylists for recursive dependency refresh, resolvePercentageLimit for count-based limit resolution, buildSmartPlaylistQuery for assembling the SELECT with joins, and addMediaFileAnnotationJoin to DRY up the repeated annotation join clause. * refactor: deduplicate child playlist IDs in Criteria Signed-off-by: Deluan <deluan@navidrome.org> * refactor: simplify withSmartPlaylistOwner to accept model.User Replaced separate ownerID string and ownerIsAdmin bool parameters with a single model.User struct, reducing the field count in smartPlaylistCriteria and making the option function signature clearer. Updated all call sites and tests accordingly. * fix: handle empty sort fields and propagate child playlist load errors OrderByFields now falls back to [{title, asc}] when all user-supplied sort fields are invalid, preventing empty ORDER BY clauses that would produce invalid SQL in row_number() window functions. Also restored the original behavior where a DB error loading child playlists aborts the parent smart playlist refresh, by making refreshChildPlaylists return a bool. * refactor: log warning when no valid sort fields are found Signed-off-by: Deluan <deluan@navidrome.org> --------- Signed-off-by: Deluan <deluan@navidrome.org>
2026-04-26 14:49:59 -04:00
sortFields := c.Criteria.OrderByFields()
parts := make([]string, 0, len(sortFields))
for _, sf := range sortFields {
mapped, ok := sortExpr(sf.Field)
if !ok {
continue
}
refactor: centralize criteria sort parsing and extract smart playlist logic (#5415) * test: add tests for recordingdate alias resolution in smart playlists Signed-off-by: Deluan <deluan@navidrome.org> * refactor: update FieldInfo structure and simplify fieldMap initialization Signed-off-by: Deluan <deluan@navidrome.org> * refactor: move sort parsing logic from persistence to criteria package Extracted sort field parsing, validation, and direction handling from persistence/criteria_sql.go into model/criteria/sort.go. The new OrderByFields method on Criteria parses the Sort/Order strings into validated SortField structs (field name + direction), resolving aliases and handling +/- prefixes and order inversion. The persistence layer now consumes these parsed fields and only handles SQL expression mapping. This centralizes sort parsing to enforce consistent implementations. * refactor: standardize field access in smartPlaylistCriteria structure Signed-off-by: Deluan <deluan@navidrome.org> * refactor: add ResolveLimit method to Criteria Moved the percentage-limit resolution logic from playlist_repository into Criteria.ResolveLimit, replacing the 3-line mutate-after-query pattern with a single method call. The method preserves LimitPercent rather than zeroing it, since IsPercentageLimit already returns false once Limit is set, making the clear redundant and lossy. * refactor: improve child playlist loading and error handling in refresh logic Signed-off-by: Deluan <deluan@navidrome.org> * refactor: extract smart playlist logic to dedicated files Moved refreshSmartPlaylist, addSmartPlaylistAnnotationJoins, and addCriteria methods from playlist_repository.go to a new smart_playlist_repository.go file. Extracted all smart playlist tests to smart_playlist_repository_test.go. Added DeferCleanup to the "valid rules" test to fix ordering flakiness when Ginkgo randomizes test execution across files. * refactor: break refreshSmartPlaylist into smaller focused methods Split the monolithic refreshSmartPlaylist method into discrete helpers for readability: shouldRefreshSmartPlaylist for guard checks, refreshChildPlaylists for recursive dependency refresh, resolvePercentageLimit for count-based limit resolution, buildSmartPlaylistQuery for assembling the SELECT with joins, and addMediaFileAnnotationJoin to DRY up the repeated annotation join clause. * refactor: deduplicate child playlist IDs in Criteria Signed-off-by: Deluan <deluan@navidrome.org> * refactor: simplify withSmartPlaylistOwner to accept model.User Replaced separate ownerID string and ownerIsAdmin bool parameters with a single model.User struct, reducing the field count in smartPlaylistCriteria and making the option function signature clearer. Updated all call sites and tests accordingly. * fix: handle empty sort fields and propagate child playlist load errors OrderByFields now falls back to [{title, asc}] when all user-supplied sort fields are invalid, preventing empty ORDER BY clauses that would produce invalid SQL in row_number() window functions. Also restored the original behavior where a DB error loading child playlists aborts the parent smart playlist refresh, by making refreshChildPlaylists return a bool. * refactor: log warning when no valid sort fields are found Signed-off-by: Deluan <deluan@navidrome.org> --------- Signed-off-by: Deluan <deluan@navidrome.org>
2026-04-26 14:49:59 -04:00
dir := "asc"
if sf.Desc {
dir = "desc"
}
refactor: centralize criteria sort parsing and extract smart playlist logic (#5415) * test: add tests for recordingdate alias resolution in smart playlists Signed-off-by: Deluan <deluan@navidrome.org> * refactor: update FieldInfo structure and simplify fieldMap initialization Signed-off-by: Deluan <deluan@navidrome.org> * refactor: move sort parsing logic from persistence to criteria package Extracted sort field parsing, validation, and direction handling from persistence/criteria_sql.go into model/criteria/sort.go. The new OrderByFields method on Criteria parses the Sort/Order strings into validated SortField structs (field name + direction), resolving aliases and handling +/- prefixes and order inversion. The persistence layer now consumes these parsed fields and only handles SQL expression mapping. This centralizes sort parsing to enforce consistent implementations. * refactor: standardize field access in smartPlaylistCriteria structure Signed-off-by: Deluan <deluan@navidrome.org> * refactor: add ResolveLimit method to Criteria Moved the percentage-limit resolution logic from playlist_repository into Criteria.ResolveLimit, replacing the 3-line mutate-after-query pattern with a single method call. The method preserves LimitPercent rather than zeroing it, since IsPercentageLimit already returns false once Limit is set, making the clear redundant and lossy. * refactor: improve child playlist loading and error handling in refresh logic Signed-off-by: Deluan <deluan@navidrome.org> * refactor: extract smart playlist logic to dedicated files Moved refreshSmartPlaylist, addSmartPlaylistAnnotationJoins, and addCriteria methods from playlist_repository.go to a new smart_playlist_repository.go file. Extracted all smart playlist tests to smart_playlist_repository_test.go. Added DeferCleanup to the "valid rules" test to fix ordering flakiness when Ginkgo randomizes test execution across files. * refactor: break refreshSmartPlaylist into smaller focused methods Split the monolithic refreshSmartPlaylist method into discrete helpers for readability: shouldRefreshSmartPlaylist for guard checks, refreshChildPlaylists for recursive dependency refresh, resolvePercentageLimit for count-based limit resolution, buildSmartPlaylistQuery for assembling the SELECT with joins, and addMediaFileAnnotationJoin to DRY up the repeated annotation join clause. * refactor: deduplicate child playlist IDs in Criteria Signed-off-by: Deluan <deluan@navidrome.org> * refactor: simplify withSmartPlaylistOwner to accept model.User Replaced separate ownerID string and ownerIsAdmin bool parameters with a single model.User struct, reducing the field count in smartPlaylistCriteria and making the option function signature clearer. Updated all call sites and tests accordingly. * fix: handle empty sort fields and propagate child playlist load errors OrderByFields now falls back to [{title, asc}] when all user-supplied sort fields are invalid, preventing empty ORDER BY clauses that would produce invalid SQL in row_number() window functions. Also restored the original behavior where a DB error loading child playlists aborts the parent smart playlist refresh, by making refreshChildPlaylists return a bool. * refactor: log warning when no valid sort fields are found Signed-off-by: Deluan <deluan@navidrome.org> --------- Signed-off-by: Deluan <deluan@navidrome.org>
2026-04-26 14:49:59 -04:00
parts = append(parts, mapped+" "+dir)
}
refactor: centralize criteria sort parsing and extract smart playlist logic (#5415) * test: add tests for recordingdate alias resolution in smart playlists Signed-off-by: Deluan <deluan@navidrome.org> * refactor: update FieldInfo structure and simplify fieldMap initialization Signed-off-by: Deluan <deluan@navidrome.org> * refactor: move sort parsing logic from persistence to criteria package Extracted sort field parsing, validation, and direction handling from persistence/criteria_sql.go into model/criteria/sort.go. The new OrderByFields method on Criteria parses the Sort/Order strings into validated SortField structs (field name + direction), resolving aliases and handling +/- prefixes and order inversion. The persistence layer now consumes these parsed fields and only handles SQL expression mapping. This centralizes sort parsing to enforce consistent implementations. * refactor: standardize field access in smartPlaylistCriteria structure Signed-off-by: Deluan <deluan@navidrome.org> * refactor: add ResolveLimit method to Criteria Moved the percentage-limit resolution logic from playlist_repository into Criteria.ResolveLimit, replacing the 3-line mutate-after-query pattern with a single method call. The method preserves LimitPercent rather than zeroing it, since IsPercentageLimit already returns false once Limit is set, making the clear redundant and lossy. * refactor: improve child playlist loading and error handling in refresh logic Signed-off-by: Deluan <deluan@navidrome.org> * refactor: extract smart playlist logic to dedicated files Moved refreshSmartPlaylist, addSmartPlaylistAnnotationJoins, and addCriteria methods from playlist_repository.go to a new smart_playlist_repository.go file. Extracted all smart playlist tests to smart_playlist_repository_test.go. Added DeferCleanup to the "valid rules" test to fix ordering flakiness when Ginkgo randomizes test execution across files. * refactor: break refreshSmartPlaylist into smaller focused methods Split the monolithic refreshSmartPlaylist method into discrete helpers for readability: shouldRefreshSmartPlaylist for guard checks, refreshChildPlaylists for recursive dependency refresh, resolvePercentageLimit for count-based limit resolution, buildSmartPlaylistQuery for assembling the SELECT with joins, and addMediaFileAnnotationJoin to DRY up the repeated annotation join clause. * refactor: deduplicate child playlist IDs in Criteria Signed-off-by: Deluan <deluan@navidrome.org> * refactor: simplify withSmartPlaylistOwner to accept model.User Replaced separate ownerID string and ownerIsAdmin bool parameters with a single model.User struct, reducing the field count in smartPlaylistCriteria and making the option function signature clearer. Updated all call sites and tests accordingly. * fix: handle empty sort fields and propagate child playlist load errors OrderByFields now falls back to [{title, asc}] when all user-supplied sort fields are invalid, preventing empty ORDER BY clauses that would produce invalid SQL in row_number() window functions. Also restored the original behavior where a DB error loading child playlists aborts the parent smart playlist refresh, by making refreshChildPlaylists return a bool. * refactor: log warning when no valid sort fields are found Signed-off-by: Deluan <deluan@navidrome.org> --------- Signed-off-by: Deluan <deluan@navidrome.org>
2026-04-26 14:49:59 -04:00
return strings.Join(parts, ", ")
}
func sortExpr(sortField string) (string, bool) {
info, ok := criteria.LookupField(sortField)
if !ok {
return "", false
}
if field, ok := smartPlaylistFields[info.Name()]; ok && field.order != "" {
return field.order, true
}
var mapped string
switch {
case info.IsTag:
mapped = "COALESCE(json_extract(media_file.tags, '$." + info.Name() + "[0].value'), '')"
case info.IsRole:
mapped = "COALESCE(json_extract(media_file.participants, '$." + info.Name() + "[0].name'), '')"
default:
field, ok := smartPlaylistFields[info.Name()]
if !ok || field.expr == "" {
return "", false
}
mapped = field.expr
}
if info.Numeric {
mapped = fmt.Sprintf("CAST(%s AS REAL)", mapped)
}
return mapped, true
}