Skip to content

Speakr is a personal, self-hosted web application designed for transcribing audio recordings

Python

Add needs filters for recordings

Fran Fitzpatrick60a4cac8ab79
diff --git a/src/api/recordings.py b/src/api/recordings.py
index 9476a965..dbdef882 100644
--- a/src/api/recordings.py
+++ b/src/api/recordings.py
@@ -93,6 +93,21 @@ from src.utils.file_hash import compute_file_sha256
# Create blueprint
recordings_bp = Blueprint('recordings', __name__)
+ACTIVE_RECORDING_STATUSES = ['QUEUED', 'PENDING', 'PROCESSING', 'SUMMARIZING']
+TRANSCRIPTION_ERROR_PREFIXES = [
+ 'ERROR_JSON:',
+ 'Transcription failed:',
+ 'Processing failed:',
+ 'ASR processing failed:',
+ 'Audio extraction failed:',
+ 'Upload/Processing failed:',
+]
+SUMMARY_NEEDS_PREFIXES = [
+ '[Summary generation failed:',
+ '[Summary skipped:',
+ '[Summary not generated]',
+]
+
# Configuration from environment
ENABLE_INQUIRE_MODE = os.environ.get('ENABLE_INQUIRE_MODE', 'false').lower() == 'true'
ENABLE_AUTO_DELETION = os.environ.get('ENABLE_AUTO_DELETION', 'false').lower() == 'true'
@@ -104,6 +119,62 @@ VIDEO_PASSTHROUGH_ASR = os.environ.get('VIDEO_PASSTHROUGH_ASR', 'false').lower()
USE_ASR_ENDPOINT = os.environ.get('USE_ASR_ENDPOINT', 'false').lower() == 'true'
ENABLE_CHUNKING = os.environ.get('ENABLE_CHUNKING', 'true').lower() == 'true'
+
+def _idle_recording_condition():
+ """Recordings that are not currently queued or running work."""
+ return db.or_(
+ Recording.status.is_(None),
+ ~Recording.status.in_(ACTIVE_RECORDING_STATUSES),
+ )
+
+
+def _blank_column_condition(column):
+ return db.or_(column.is_(None), db.func.trim(column) == '')
+
+
+def _prefixed_column_condition(column, prefixes):
+ return db.or_(*[column.like(f'{prefix}%') for prefix in prefixes])
+
+
+def _transcription_error_condition():
+ return _prefixed_column_condition(Recording.transcription, TRANSCRIPTION_ERROR_PREFIXES)
+
+
+def _valid_transcription_condition():
+ return db.and_(
+ ~_blank_column_condition(Recording.transcription),
+ ~_transcription_error_condition(),
+ )
+
+
+def _needs_transcription_condition():
+ return db.and_(
+ _idle_recording_condition(),
+ db.or_(
+ _blank_column_condition(Recording.transcription),
+ _transcription_error_condition(),
+ ),
+ )
+
+
+def _needs_summary_condition():
+ return db.and_(
+ _idle_recording_condition(),
+ _valid_transcription_condition(),
+ db.or_(
+ _blank_column_condition(Recording.summary),
+ _prefixed_column_condition(Recording.summary, SUMMARY_NEEDS_PREFIXES),
+ ),
+ )
+
+
+def _needs_speakers_condition():
+ return db.and_(
+ _idle_recording_condition(),
+ _valid_transcription_condition(),
+ Recording.transcription.like('%SPEAKER\\_%', escape='\\'),
+ )
+
def reindex_recording_chunks_async(recording_id):
"""Rebuild a recording's semantic-search (Inquire) chunks in the background.
@@ -1522,6 +1593,9 @@ def get_recordings_paginated():
show_shared = request.args.get('shared', '').lower() == 'true'
show_starred = request.args.get('starred', '').lower() == 'true'
show_inbox = request.args.get('inbox', '').lower() == 'true'
+ needs_transcription = request.args.get('needs_transcription', '').lower() == 'true'
+ needs_summary = request.args.get('needs_summary', '').lower() == 'true'
+ needs_speakers = request.args.get('needs_speakers', '').lower() == 'true'
sort_by = request.args.get('sort_by', 'created_at') # 'created_at' or 'meeting_date'
folder_filter = request.args.get('folder', '').strip() # folder_id or 'none' for no folder
@@ -1604,6 +1678,15 @@ def get_recordings_paginated():
except ValueError:
pass # Invalid folder_id, ignore filter
+ if needs_transcription:
+ stmt = stmt.where(_needs_transcription_condition())
+
+ if needs_summary:
+ stmt = stmt.where(_needs_summary_condition())
+
+ if needs_speakers:
+ stmt = stmt.where(_needs_speakers_condition())
+
# Apply search filters if provided
if search_query:
# Extract date filters
diff --git a/static/js/app.modular.js b/static/js/app.modular.js
index 88236b92..da6a5686 100644
--- a/static/js/app.modular.js
+++ b/static/js/app.modular.js
@@ -277,6 +277,9 @@ document.addEventListener('DOMContentLoaded', async () => {
const filterTextQuery = ref('');
const filterStarred = ref(false);
const filterInbox = ref(false);
+ const filterNeedsTranscription = ref(false);
+ const filterNeedsSummary = ref(false);
+ const filterNeedsSpeakers = ref(false);
const showArchivedRecordings = ref(false);
const showSharedWithMe = ref(false);
@@ -1575,6 +1578,7 @@ document.addEventListener('DOMContentLoaded', async () => {
// Filters
showAdvancedFilters, filterTags, filterSpeakers, filterTagSearch, filterSpeakerSearch,
filterDateRange, filterDatePreset, filterTextQuery, filterStarred, filterInbox,
+ filterNeedsTranscription, filterNeedsSummary, filterNeedsSpeakers,
showArchivedRecordings, showSharedWithMe, sortBy, selectedTagFilter,
// Pagination
@@ -3277,6 +3281,18 @@ document.addEventListener('DOMContentLoaded', async () => {
recordingsComposable.loadRecordings(1, false, searchQuery.value);
});
+ watch(filterNeedsTranscription, () => {
+ recordingsComposable.loadRecordings(1, false, searchQuery.value);
+ });
+
+ watch(filterNeedsSummary, () => {
+ recordingsComposable.loadRecordings(1, false, searchQuery.value);
+ });
+
+ watch(filterNeedsSpeakers, () => {
+ recordingsComposable.loadRecordings(1, false, searchQuery.value);
+ });
+
watch(filterFolder, (newValue) => {
// Persist folder selection to localStorage
if (newValue) {
diff --git a/static/js/modules/composables/recordings.js b/static/js/modules/composables/recordings.js
index 57658b3f..15c549f3 100644
--- a/static/js/modules/composables/recordings.js
+++ b/static/js/modules/composables/recordings.js
@@ -11,7 +11,8 @@ export function useRecordings(state, utils, reprocessComposable) {
currentPage, perPage, totalRecordings, totalPages, hasNextPage, hasPrevPage,
showSharedWithMe, showArchivedRecordings, searchQuery, searchDebounceTimer,
filterTags, filterSpeakers, filterDatePreset, filterDateRange, filterTextQuery,
- filterStarred, filterInbox, filterFolder, sortBy,
+ filterStarred, filterInbox, filterNeedsTranscription, filterNeedsSummary, filterNeedsSpeakers,
+ filterFolder, sortBy,
availableTags, availableSpeakers, availableFolders, selectedTagIds, uploadLanguage, uploadMinSpeakers, uploadMaxSpeakers, uploadHotwords, uploadInitialPrompt,
useAsrEndpoint, connectorSupportsDiarization, globalError, uploadQueue, isProcessingActive, currentView, showUploadModal, uploadDeepLinkPending,
isMobileScreen, isSidebarCollapsed, isRecording, audioBlobURL,
@@ -61,6 +62,15 @@ export function useRecordings(state, utils, reprocessComposable) {
if (filterInbox.value) {
params.set('inbox', 'true');
}
+ if (filterNeedsTranscription.value) {
+ params.set('needs_transcription', 'true');
+ }
+ if (filterNeedsSummary.value) {
+ params.set('needs_summary', 'true');
+ }
+ if (filterNeedsSpeakers.value) {
+ params.set('needs_speakers', 'true');
+ }
// Add folder filter
if (filterFolder && filterFolder.value) {
@@ -439,6 +449,9 @@ export function useRecordings(state, utils, reprocessComposable) {
filterTextQuery.value = '';
filterStarred.value = false;
filterInbox.value = false;
+ filterNeedsTranscription.value = false;
+ filterNeedsSummary.value = false;
+ filterNeedsSpeakers.value = false;
// Note: filterFolder is NOT cleared here - it's a navigation element, not a filter
searchQuery.value = '';
};
diff --git a/static/locales/en.json b/static/locales/en.json
index e7d13eae..96d57e4a 100644
--- a/static/locales/en.json
+++ b/static/locales/en.json
@@ -1409,6 +1409,9 @@
"loadMore": "Load More",
"markAsRead": "Mark as read",
"moveToInbox": "Move to Inbox",
+ "needsSpeakers": "Needs Speakers",
+ "needsSummary": "Needs Summary",
+ "needsTranscription": "Needs Transcription",
"noRecordings": "No recordings found",
"older": "Older",
"removeFromHighlighted": "Remove from starred",
diff --git a/templates/components/sidebar.html b/templates/components/sidebar.html
index d5b30246..cc793151 100644
--- a/templates/components/sidebar.html
+++ b/templates/components/sidebar.html
@@ -62,16 +62,16 @@
class="flex-1 h-7 flex items-center justify-between pl-2 pr-2 bg-[var(--bg-input)] border border-[var(--border-secondary)] rounded text-xs hover:bg-[var(--bg-tertiary)] transition-colors">
<span class="flex items-center">
<i class="fas fa-filter mr-1.5 text-[var(--text-muted)] text-[10px]"></i>
- <span v-if="!searchQuery && filterTags.length === 0 && filterSpeakers.length === 0 && !filterDatePreset && !filterDateRange.start && !filterDateRange.end && !filterTextQuery && !filterStarred && !filterInbox"
+ <span v-if="!searchQuery && filterTags.length === 0 && filterSpeakers.length === 0 && !filterDatePreset && !filterDateRange.start && !filterDateRange.end && !filterTextQuery && !filterStarred && !filterInbox && !filterNeedsTranscription && !filterNeedsSummary && !filterNeedsSpeakers"
v-text="t('sidebar.searchRecordings')">
</span>
<span v-else class="text-[var(--text-accent)]">
- Active filters (${ (filterTags.length > 0 ? 1 : 0) + (filterSpeakers.length > 0 ? 1 : 0) + (filterDatePreset || filterDateRange.start || filterDateRange.end ? 1 : 0) + (filterTextQuery ? 1 : 0) + (filterStarred ? 1 : 0) + (filterInbox ? 1 : 0) })
+ Active filters (${ (filterTags.length > 0 ? 1 : 0) + (filterSpeakers.length > 0 ? 1 : 0) + (filterDatePreset || filterDateRange.start || filterDateRange.end ? 1 : 0) + (filterTextQuery ? 1 : 0) + (filterStarred ? 1 : 0) + (filterInbox ? 1 : 0) + (filterNeedsTranscription ? 1 : 0) + (filterNeedsSummary ? 1 : 0) + (filterNeedsSpeakers ? 1 : 0) })
</span>
</span>
<i :class="['fas fa-chevron-down transition-transform text-[var(--text-muted)] text-[10px]', showAdvancedFilters ? 'rotate-180' : '']"></i>
</button>
- <button v-if="searchQuery || filterTags.length > 0 || filterSpeakers.length > 0 || filterDatePreset || filterDateRange.start || filterDateRange.end || filterTextQuery || filterStarred || filterInbox"
+ <button v-if="searchQuery || filterTags.length > 0 || filterSpeakers.length > 0 || filterDatePreset || filterDateRange.start || filterDateRange.end || filterTextQuery || filterStarred || filterInbox || filterNeedsTranscription || filterNeedsSummary || filterNeedsSpeakers"
@click="clearAllFilters"
class="w-7 h-7 flex items-center justify-center bg-[var(--bg-input)] border border-[var(--border-secondary)] rounded text-xs hover:bg-[var(--bg-tertiary)] hover:text-[var(--text-primary)] transition-colors"
:title="t('buttons.clearAllFilters')">
@@ -123,6 +123,46 @@
</button>
</div>
+ <!-- Needs Filters -->
+ <div class="grid grid-cols-3 gap-2">
+ <button @click="filterNeedsTranscription = !filterNeedsTranscription"
+ :aria-pressed="filterNeedsTranscription"
+ :aria-label="t('sidebar.needsTranscription')"
+ :class="[
+ 'min-h-[3.75rem] min-w-0 flex flex-col items-center justify-center gap-1 px-1 py-2 rounded-md text-[11px] font-medium leading-[1.15] text-center transition-all border overflow-hidden',
+ filterNeedsTranscription
+ ? 'bg-red-500/20 border-red-500/50 text-red-400'
+ : 'bg-[var(--bg-input)] border-[var(--border-secondary)] text-[var(--text-muted)] hover:border-red-500/30 hover:text-red-400'
+ ]">
+ <i class="fas fa-keyboard text-[11px] leading-none"></i>
+ <span class="block max-w-full text-[10px] leading-[1.15] whitespace-normal break-normal" v-text="t('sidebar.needsTranscription')"></span>
+ </button>
+ <button @click="filterNeedsSummary = !filterNeedsSummary"
+ :aria-pressed="filterNeedsSummary"
+ :aria-label="t('sidebar.needsSummary')"
+ :class="[
+ 'min-h-[3.75rem] min-w-0 flex flex-col items-center justify-center gap-1 px-1 py-2 rounded-md text-[11px] font-medium leading-[1.15] text-center transition-all border overflow-hidden',
+ filterNeedsSummary
+ ? 'bg-emerald-500/20 border-emerald-500/50 text-emerald-400'
+ : 'bg-[var(--bg-input)] border-[var(--border-secondary)] text-[var(--text-muted)] hover:border-emerald-500/30 hover:text-emerald-400'
+ ]">
+ <i class="fas fa-list-alt text-[11px] leading-none"></i>
+ <span class="block max-w-full whitespace-normal break-normal" v-text="t('sidebar.needsSummary')"></span>
+ </button>
+ <button @click="filterNeedsSpeakers = !filterNeedsSpeakers"
+ :aria-pressed="filterNeedsSpeakers"
+ :aria-label="t('sidebar.needsSpeakers')"
+ :class="[
+ 'min-h-[3.75rem] min-w-0 flex flex-col items-center justify-center gap-1 px-1 py-2 rounded-md text-[11px] font-medium leading-[1.15] text-center transition-all border overflow-hidden',
+ filterNeedsSpeakers
+ ? 'bg-violet-500/20 border-violet-500/50 text-violet-400'
+ : 'bg-[var(--bg-input)] border-[var(--border-secondary)] text-[var(--text-muted)] hover:border-violet-500/30 hover:text-violet-400'
+ ]">
+ <i class="fas fa-user-edit text-[11px] leading-none"></i>
+ <span class="block max-w-full whitespace-normal break-normal" v-text="t('sidebar.needsSpeakers')"></span>
+ </button>
+ </div>
+
<!-- Tag Filter -->
<div class="bg-[var(--bg-secondary)] rounded-lg p-2.5 border border-[var(--border-primary)]">
<div class="flex items-center justify-between gap-2 mb-2">
diff --git a/tests/test_cov_recordings_read.py b/tests/test_cov_recordings_read.py
index e7e13082..2dc1aa82 100644
--- a/tests/test_cov_recordings_read.py
+++ b/tests/test_cov_recordings_read.py
@@ -156,7 +156,7 @@ def make_user():
def make_recording(user, *, status="COMPLETED", transcription="hello world this is a transcript",
audio_path="local://covrec/audio.mp3", title="rec",
summary="a summary", original_filename="audio.mp3",
- mime_type="audio/mpeg", folder_id=None):
+ mime_type="audio/mpeg", folder_id=None, participants=None):
rec = Recording(
audio_path=audio_path,
original_filename=original_filename,
@@ -164,6 +164,7 @@ def make_recording(user, *, status="COMPLETED", transcription="hello world this
status=status,
transcription=transcription,
summary=summary,
+ participants=participants,
user_id=user.id,
mime_type=mime_type,
folder_id=folder_id,
@@ -385,6 +386,166 @@ def test_paginated_speaker_filter(owner):
assert nomatch_id not in ids
+def test_paginated_needs_transcription_filter(owner):
+ with _db():
+ u = db.session.get(User, owner)
+ missing = make_recording(u, title="needs-transcription", transcription=None).id
+ error = make_recording(
+ u,
+ title="failed-transcription",
+ status="FAILED",
+ transcription="Transcription failed: provider unavailable",
+ ).id
+ ready = make_recording(u, title="already-transcribed", transcription="valid transcript").id
+ active = make_recording(
+ u,
+ title="active-transcription",
+ status="QUEUED",
+ transcription=None,
+ ).id
+
+ c = new_client()
+ login(c, owner)
+ resp = c.get("/api/recordings?needs_transcription=true&per_page=100")
+ assert resp.status_code == 200
+ ids = {r["id"] for r in resp.get_json()["recordings"]}
+ assert missing in ids
+ assert error in ids
+ assert ready not in ids
+ assert active not in ids
+
+
+def test_paginated_needs_summary_filter(owner):
+ with _db():
+ u = db.session.get(User, owner)
+ missing = make_recording(
+ u,
+ title="needs-summary",
+ transcription="valid transcript with enough words",
+ summary=None,
+ ).id
+ failed = make_recording(
+ u,
+ title="failed-summary",
+ status="FAILED",
+ transcription="valid transcript with enough words",
+ summary="[Summary generation failed: quota exceeded]",
+ ).id
+ no_transcript = make_recording(
+ u,
+ title="no-transcript-summary",
+ transcription=None,
+ summary=None,
+ ).id
+ active = make_recording(
+ u,
+ title="active-summary",
+ status="SUMMARIZING",
+ transcription="valid transcript with enough words",
+ summary=None,
+ ).id
+ ready = make_recording(
+ u,
+ title="summary-ready",
+ transcription="valid transcript with enough words",
+ summary="finished summary",
+ ).id
+
+ c = new_client()
+ login(c, owner)
+ resp = c.get("/api/recordings?needs_summary=true&per_page=100")
+ assert resp.status_code == 200
+ ids = {r["id"] for r in resp.get_json()["recordings"]}
+ assert missing in ids
+ assert failed in ids
+ assert no_transcript not in ids
+ assert active not in ids
+ assert ready not in ids
+
+
+def test_paginated_needs_speakers_filter(owner):
+ with _db():
+ u = db.session.get(User, owner)
+ unresolved = make_recording(
+ u,
+ title="needs-speakers",
+ transcription=json.dumps([
+ {"speaker": "SPEAKER_00", "sentence": "Hello."},
+ {"speaker": "Alice", "sentence": "Hi."},
+ ]),
+ ).id
+ assigned = make_recording(
+ u,
+ title="speakers-assigned",
+ transcription=json.dumps([
+ {"speaker": "Alice", "sentence": "Hello."},
+ {"speaker": "Bob", "sentence": "Hi."},
+ ]),
+ participants="Alice, Bob",
+ ).id
+ active = make_recording(
+ u,
+ title="active-speakers",
+ status="PROCESSING",
+ transcription=json.dumps([
+ {"speaker": "SPEAKER_00", "sentence": "Still running."},
+ ]),
+ ).id
+ no_transcript = make_recording(u, title="no-speaker-transcript", transcription=None).id
+
+ c = new_client()
+ login(c, owner)
+ resp = c.get("/api/recordings?needs_speakers=true&per_page=100")
+ assert resp.status_code == 200
+ ids = {r["id"] for r in resp.get_json()["recordings"]}
+ assert unresolved in ids
+ assert assigned not in ids
+ assert active not in ids
+ assert no_transcript not in ids
+
+
+def test_paginated_needs_filters_and_with_existing_filters(owner):
+ tagname = f"needstag{uuid.uuid4().hex[:6]}"
+ with _db():
+ u = db.session.get(User, owner)
+ tag = Tag(name=tagname, user_id=u.id)
+ db.session.add(tag)
+ db.session.commit()
+ tagged_need = make_recording(
+ u,
+ title="tagged-needs-summary",
+ transcription="valid transcript with enough words",
+ summary=None,
+ )
+ db.session.add(RecordingTag(recording_id=tagged_need.id, tag_id=tag.id, order=1))
+ tagged_ready = make_recording(
+ u,
+ title="tagged-summary-ready",
+ transcription="valid transcript with enough words",
+ summary="finished summary",
+ )
+ db.session.add(RecordingTag(recording_id=tagged_ready.id, tag_id=tag.id, order=1))
+ untagged_need = make_recording(
+ u,
+ title="untagged-needs-summary",
+ transcription="valid transcript with enough words",
+ summary=None,
+ )
+ db.session.commit()
+ tagged_need_id = tagged_need.id
+ tagged_ready_id = tagged_ready.id
+ untagged_need_id = untagged_need.id
+
+ c = new_client()
+ login(c, owner)
+ resp = c.get(f"/api/recordings?needs_summary=true&q=tag:{tagname}&per_page=100")
+ assert resp.status_code == 200
+ ids = {r["id"] for r in resp.get_json()["recordings"]}
+ assert tagged_need_id in ids
+ assert tagged_ready_id not in ids
+ assert untagged_need_id not in ids
+
+
def test_paginated_requires_login():
c = new_client()
resp = c.get("/api/recordings")