Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Issue #5303: Improve background task monitor #5304

Merged
merged 2 commits into from
Feb 20, 2025
Merged

Issue #5303: Improve background task monitor #5304

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
20c92d1
#5303 - Improve TaskMonitor API
reckart Feb 19, 2025
6a8fe34
#5303 - Improve TaskMonitor API
reckart Feb 20, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
134 changes: 62 additions & 72 deletions ...java/de/tudarmstadt/ukp/clarin/webanno/agreement/task/CalculatePairwiseAgreementTask.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -46,7 +46,6 @@
import de.tudarmstadt.ukp.clarin.webanno.api.casstorage.session.CasStorageSession;
import de.tudarmstadt.ukp.clarin.webanno.model.AnnotationDocument;
import de.tudarmstadt.ukp.clarin.webanno.model.AnnotationFeature;
import de.tudarmstadt.ukp.clarin.webanno.model.AnnotationLayer;
import de.tudarmstadt.ukp.clarin.webanno.model.SourceDocument;
import de.tudarmstadt.ukp.inception.documents.api.DocumentService;
import de.tudarmstadt.ukp.inception.scheduling.Task;
Expand Down Expand Up @@ -86,77 +85,76 @@ public void execute()
{
summary = new PairwiseAgreementResult(feature, traits);

var maxProgress = allAnnDocs.size();
var progress = 0;

var docs = allAnnDocs.keySet().stream() //
.sorted(comparing(SourceDocument::getName)) //
.toList();

for (var doc : docs) {
var monitor = getMonitor();
if (monitor.isCancelled()) {
break;
}

monitor.setProgressWithMessage(progress, maxProgress,
LogMessage.info(this, doc.getName()));

try (var session = CasStorageSession.openNested()) {
for (int m = 0; m < annotators.size(); m++) {
var annotator1 = annotators.get(m);
var maybeCas1 = LazyInitializer.<Optional<CAS>> builder()
.setInitializer(() -> loadCas(doc, annotator1, allAnnDocs)).get();

for (int n = 0; n < annotators.size(); n++) {
if (!(n < m)) {
// Triangle matrix mirrored
continue;
}

var annotator2 = annotators.get(n);

if ((CURATION_USER.equals(annotator1) || CURATION_USER.equals(annotator2))
&& !asList(CURATION_IN_PROGRESS, CURATION_FINISHED)
.contains(doc.getState())) {
LOG.trace("Skipping combination {}/{}@{}: {} not in a curation state",
annotator1, annotator2, doc, annotator1);
summary.mergeResult(annotator1, annotator2, AgreementSummary
.skipped(feature.getLayer().getName(), feature.getName()));
continue;
}

if (maybeCas1.get().isEmpty()) {
LOG.trace("Skipping combination {}/{}@{}: {} has no data", annotator1,
annotator2, doc, annotator1);
summary.mergeResult(annotator1, annotator2, AgreementSummary
.skipped(feature.getLayer().getName(), feature.getName()));
continue;
}

var maybeCas2 = LazyInitializer.<Optional<CAS>> builder()
.setInitializer(() -> loadCas(doc, annotator2, allAnnDocs)).get();
try (var progress = getMonitor().openScope("documents", allAnnDocs.size())) {
for (var doc : docs) {
if (getMonitor().isCancelled()) {
break;
}

if (maybeCas2.get().isEmpty()) {
LOG.trace("Skipping combination {}/{}@{}: {} has no data", annotator1,
annotator2, doc, annotator2);
summary.mergeResult(annotator1, annotator2, AgreementSummary
.skipped(feature.getLayer().getName(), feature.getName()));
continue;
progress.update(up -> up.increment() //
.addMessage(LogMessage.info(this, doc.getName())));

try (var session = CasStorageSession.openNested()) {
for (int m = 0; m < annotators.size(); m++) {
var annotator1 = annotators.get(m);
var maybeCas1 = LazyInitializer.<Optional<CAS>> builder()
.setInitializer(() -> loadCas(doc, annotator1, allAnnDocs)).get();

for (int n = 0; n < annotators.size(); n++) {
if (!(n < m)) {
// Triangle matrix mirrored
continue;
}

var annotator2 = annotators.get(n);

if ((CURATION_USER.equals(annotator1)
|| CURATION_USER.equals(annotator2))
&& !asList(CURATION_IN_PROGRESS, CURATION_FINISHED)
.contains(doc.getState())) {
LOG.trace(
"Skipping combination {}/{}@{}: {} not in a curation state",
annotator1, annotator2, doc, annotator1);
summary.mergeResult(annotator1, annotator2, AgreementSummary
.skipped(feature.getLayer().getName(), feature.getName()));
continue;
}

if (maybeCas1.get().isEmpty()) {
LOG.trace("Skipping combination {}/{}@{}: {} has no data",
annotator1, annotator2, doc, annotator1);
summary.mergeResult(annotator1, annotator2, AgreementSummary
.skipped(feature.getLayer().getName(), feature.getName()));
continue;
}

var maybeCas2 = LazyInitializer.<Optional<CAS>> builder()
.setInitializer(() -> loadCas(doc, annotator2, allAnnDocs))
.get();

if (maybeCas2.get().isEmpty()) {
LOG.trace("Skipping combination {}/{}@{}: {} has no data",
annotator1, annotator2, doc, annotator2);
summary.mergeResult(annotator1, annotator2, AgreementSummary
.skipped(feature.getLayer().getName(), feature.getName()));
continue;
}

var casMap = new LinkedHashMap<String, CAS>();
casMap.put(annotator1, maybeCas1.get().get());
casMap.put(annotator2, maybeCas2.get().get());
var res = AgreementSummary.of(measure.getAgreement(casMap));
summary.mergeResult(annotator1, annotator2, res);
}

var casMap = new LinkedHashMap<String, CAS>();
casMap.put(annotator1, maybeCas1.get().get());
casMap.put(annotator2, maybeCas2.get().get());
var res = AgreementSummary.of(measure.getAgreement(casMap));
summary.mergeResult(annotator1, annotator2, res);
}
}

progress++;
}
catch (Exception e) {
LOG.error("Unable to load data", e);
catch (Exception e) {
LOG.error("Unable to load data", e);
}
}
}
}
Expand Down Expand Up @@ -229,7 +227,6 @@ public static class Builder<T extends Builder<?>>
{
private List<String> annotators;
private DefaultAgreementTraits traits;
private AnnotationLayer layer;
private AnnotationFeature feature;
private AgreementMeasure<?> measure;
private Map<SourceDocument, List<AnnotationDocument>> allAnnDocs;
Expand All @@ -253,13 +250,6 @@ public T withTraits(DefaultAgreementTraits aTraits)
return (T) this;
}

@SuppressWarnings("unchecked")
public T withLayer(AnnotationLayer aLayer)
{
layer = aLayer;
return (T) this;
}

@SuppressWarnings("unchecked")
public T withFeature(AnnotationFeature aFeature)
{
Expand Down
64 changes: 26 additions & 38 deletions ...a/de/tudarmstadt/ukp/clarin/webanno/agreement/task/CalculatePerDocumentAgreementTask.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -48,7 +48,6 @@
import de.tudarmstadt.ukp.clarin.webanno.model.AnnotationDocument;
import de.tudarmstadt.ukp.clarin.webanno.model.AnnotationDocumentState;
import de.tudarmstadt.ukp.clarin.webanno.model.AnnotationFeature;
import de.tudarmstadt.ukp.clarin.webanno.model.AnnotationLayer;
import de.tudarmstadt.ukp.clarin.webanno.model.SourceDocument;
import de.tudarmstadt.ukp.inception.documents.api.DocumentService;
import de.tudarmstadt.ukp.inception.scheduling.Task;
Expand Down Expand Up @@ -88,45 +87,42 @@ public void execute()
{
summary = new PerDocumentAgreementResult(feature, traits);

var maxProgress = allAnnDocs.size();
var progress = 0;

var docs = allAnnDocs.keySet().stream() //
.sorted(comparing(SourceDocument::getName)) //
.toList();

for (var doc : docs) {
var monitor = getMonitor();
if (monitor.isCancelled()) {
break;
}
try (var progress = getMonitor().openScope("documents", allAnnDocs.size())) {
for (var doc : docs) {
if (getMonitor().isCancelled()) {
break;
}

progress.update(up -> up.increment() //
.addMessage(LogMessage.info(this, doc.getName())));

monitor.setProgressWithMessage(progress, maxProgress,
LogMessage.info(this, doc.getName()));
try (var session = CasStorageSession.openNested()) {
var casMap = new LinkedHashMap<String, CAS>();
for (var annDoc : allAnnDocs.get(doc)) {
var dataOwner = annDoc.getUser();
if (!annotators.contains(dataOwner)) {
continue;
}

try (var session = CasStorageSession.openNested()) {
var casMap = new LinkedHashMap<String, CAS>();
for (var annDoc : allAnnDocs.get(doc)) {
var dataOwner = annDoc.getUser();
if (!annotators.contains(dataOwner)) {
continue;
casMap.put(dataOwner, loadCas(annDoc.getDocument(), dataOwner));
}

casMap.put(dataOwner, loadCas(annDoc.getDocument(), dataOwner));
}
if (annotators.contains(CURATION_USER)) {
casMap.put(CURATION_USER, loadCas(doc, CURATION_USER));
}

if (annotators.contains(CURATION_USER)) {
casMap.put(CURATION_USER, loadCas(doc, CURATION_USER));
LOG.trace("Calculating agreement on {} for [{}] annotators", doc,
casMap.size());
var agreementResult = AgreementSummary.of(measure.getAgreement(casMap));
summary.mergeResult(doc, agreementResult);
}
catch (Exception e) {
LOG.error("Unable to load data", e);
}

LOG.trace("Calculating agreement on {} for [{}] annotators", doc, casMap.size());
var agreementResult = AgreementSummary.of(measure.getAgreement(casMap));
summary.mergeResult(doc, agreementResult);

progress++;
}
catch (Exception e) {
LOG.error("Unable to load data", e);
}
}
}
Expand Down Expand Up @@ -184,7 +180,6 @@ public static class Builder<T extends Builder<?>>
{
private List<String> annotators;
private DefaultAgreementTraits traits;
private AnnotationLayer layer;
private AnnotationFeature feature;
private AgreementMeasure<?> measure;
private Map<SourceDocument, List<AnnotationDocument>> allAnnDocs;
Expand All @@ -209,13 +204,6 @@ public T withTraits(DefaultAgreementTraits aTraits)
return (T) this;
}

@SuppressWarnings("unchecked")
public T withLayer(AnnotationLayer aLayer)
{
layer = aLayer;
return (T) this;
}

@SuppressWarnings("unchecked")
public T withFeature(AnnotationFeature aFeature)
{
Expand Down
91 changes: 41 additions & 50 deletions ...c/main/java/de/tudarmstadt/ukp/inception/assistant/documents/UpdateDocumentIndexTask.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -27,9 +27,9 @@
import static de.tudarmstadt.ukp.inception.assistant.documents.DocumentQueryService.FIELD_SOURCE_DOC_COMPLETE;
import static de.tudarmstadt.ukp.inception.assistant.documents.DocumentQueryService.FIELD_SOURCE_DOC_ID;
import static de.tudarmstadt.ukp.inception.assistant.documents.DocumentQueryService.FIELD_TEXT;
import static de.tudarmstadt.ukp.inception.scheduling.ProgressScope.SCOPE_DOCUMENTS;
import static de.tudarmstadt.ukp.inception.scheduling.ProgressScope.SCOPE_UNITS;
import static de.tudarmstadt.ukp.inception.scheduling.TaskState.CANCELLED;
import static de.tudarmstadt.ukp.inception.scheduling.TaskState.COMPLETED;
import static de.tudarmstadt.ukp.inception.scheduling.TaskState.RUNNING;
import static java.lang.Math.floorDiv;
import static java.time.Duration.ofSeconds;
import static org.apache.commons.collections4.ListUtils.partition;
Expand Down Expand Up @@ -138,41 +138,38 @@ public void execute() throws Exception
}
}

var toProcess = documentsToUnindex.size() + documentsToIndex.size() * 100;
var processed = 0;
monitor.setStateAndProgress(RUNNING, processed * 100, toProcess);

for (var sourceDocumentId : documentsToUnindex) {
if (monitor.isCancelled()) {
monitor.setState(CANCELLED);
break;
}

monitor.setProgressWithMessage(processed * 100, toProcess,
LogMessage.info(this, "Unindexing..."));
unindexDocument(index, sourceDocumentId);
processed++;
}

try (var session = CasStorageSession.openNested()) {
for (var sourceDocument : documentsToIndex.values()) {
try (var progress = getMonitor().openScope(SCOPE_DOCUMENTS,
documentsToUnindex.size() + documentsToIndex.size())) {
progress.update(up -> up.addMessage(LogMessage.info(this, "Unindexing...")));
for (var sourceDocumentId : documentsToUnindex) {
if (monitor.isCancelled()) {
monitor.setState(CANCELLED);
monitor.update(up -> up.setState(CANCELLED));
break;
}

monitor.setProgressWithMessage(processed * 100, toProcess,
LogMessage.info(this, "Indexing: %s", sourceDocument.getName()));
indexDocument(index, chunker, sourceDocument);
processed++;
progress.update(up -> up.increment());
unindexDocument(index, sourceDocumentId);
}

try (var session = CasStorageSession.openNested()) {
for (var sourceDocument : documentsToIndex.values()) {
if (monitor.isCancelled()) {
monitor.update(up -> up.setState(CANCELLED));
break;
}

progress.update(up -> up.increment() //
.addMessage(LogMessage.info(this, "Indexing: %s",
sourceDocument.getName())));
indexDocument(index, chunker, sourceDocument);
}
}
}

if (!monitor.isCancelled()) {
monitor.setStateAndProgress(COMPLETED, processed * 100, toProcess);
index.getIndexWriter().commit();
// We can probably live with a partial index, so we do not roll back if
// cancelled
if (!monitor.isCancelled()) {
index.getIndexWriter().commit();
// We can probably live with a partial index, so we do not roll back if
// cancelled
}
}
}
}
Expand Down Expand Up @@ -204,27 +201,21 @@ private void indexDocument(LuceneIndexPool.PooledIndex aIndex, Chunker<CAS> aChu
var chunks = aChunker.process(cas);
var batches = partition(chunks, properties.getEmbedding().getBatchSize());

var totalBatches = batches.size();
var processedBatches = 0;
var chunksSeen = 0;
var progressOffset = getMonitor().getProgress();
for (var batch : batches) {
try {
var docEmbeddings = embeddingService.embed(Chunk::text, batch);
for (var embedding : docEmbeddings) {
indexChunks(aIndex, aSourceDocument, embedding);
try (var progress = getMonitor().openScope(SCOPE_UNITS, chunks.size())) {
for (var batch : batches) {
progress.update(up -> up.increment(batch.size()));

try {
var docEmbeddings = embeddingService.embed(Chunk::text, batch);
for (var embedding : docEmbeddings) {
indexChunks(aIndex, aSourceDocument, embedding);
}
}
catch (IOException e) {
LOG.error("Error indexing document {} chunks {}-{} ", aSourceDocument,
progress.getProgress(), progress.getProgress() + batch.size(), e);
}
}
catch (IOException e) {
LOG.error("Error indexing document {} chunks {}-{} ", aSourceDocument,
chunksSeen, chunksSeen + batch.size(), e);
}

getMonitor().setStateAndProgress(RUNNING,
progressOffset + floorDiv(processedBatches * 100, totalBatches));

chunksSeen += batch.size();
processedBatches++;
}

markDocumentAsIndexed(aIndex, aSourceDocument);
Expand Down
Loading
Loading