marcel/familienarchiv
1
0
Fork 0
Code Issues 114 Pull Requests 3 Actions 4 Packages Projects Releases Wiki Activity

feat(search): enrich searchDocuments with per-document match data

Browse Source 
DocumentService.searchDocuments now returns DocumentSearchResult with matchData
populated from findEnrichmentData. Title highlights are parsed from chr(1)/chr(2)
delimiters into MatchOffset lists; transcription snippet and sender/receiver/tag
match flags are extracted from the same native SQL row.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
This commit is contained in:
Marcel
2026-04-15 17:53:57 +02:00
committed by marcel
parent c235151075
commit f723a83011
5 changed files with 157 additions and 27 deletions
Download Patch File Download Diff File Expand all files Collapse all files

3
backend/src/main/java/org/raddatz/familienarchiv/controller/DocumentController.java
View File

@@ -208,8 +208,7 @@ public class DocumentController {
if (!"ASC".equalsIgnoreCase(dir) && !"DESC".equalsIgnoreCase(dir)) {
throw new ResponseStatusException(HttpStatus.BAD_REQUEST, "dir must be ASC or DESC");
}
List<Document> results = documentService.searchDocuments(q, from, to, senderId, receiverId, tags, tagQ, status, sort, dir);
return ResponseEntity.ok(DocumentSearchResult.of(results));
return ResponseEntity.ok(documentService.searchDocuments(q, from, to, senderId, receiverId, tags, tagQ, status, sort, dir));
}
// --- TRAINING LABELS ---

87
backend/src/main/java/org/raddatz/familienarchiv/service/DocumentService.java
View File

@@ -3,10 +3,13 @@ package org.raddatz.familienarchiv.service;
import lombok.RequiredArgsConstructor;
import lombok.extern.slf4j.Slf4j;
import org.raddatz.familienarchiv.dto.DocumentSearchResult;
import org.raddatz.familienarchiv.dto.DocumentSort;
import org.raddatz.familienarchiv.dto.DocumentUpdateDTO;
import org.raddatz.familienarchiv.dto.IncompleteDocumentDTO;
import org.raddatz.familienarchiv.dto.MatchOffset;
import org.raddatz.familienarchiv.dto.SearchMatchData;
import org.raddatz.familienarchiv.model.Document;
import org.raddatz.familienarchiv.dto.DocumentSort;
import org.raddatz.familienarchiv.model.DocumentStatus;
import org.raddatz.familienarchiv.model.ScriptType;
import org.raddatz.familienarchiv.model.TrainingLabel;
@@ -290,13 +293,13 @@ public class DocumentService {
}
// 1. Allgemeine Suche (für das Suchfeld im Frontend)
public List<Document> searchDocuments(String text, LocalDate from, LocalDate to, UUID sender, UUID receiver, List<String> tags, String tagQ, DocumentStatus status, DocumentSort sort, String dir) {
public DocumentSearchResult searchDocuments(String text, LocalDate from, LocalDate to, UUID sender, UUID receiver, List<String> tags, String tagQ, DocumentStatus status, DocumentSort sort, String dir) {
boolean hasText = StringUtils.hasText(text);
List<UUID> rankedIds = null;
if (hasText) {
rankedIds = documentRepository.findRankedIdsByFts(text);
if (rankedIds.isEmpty()) return List.of();
if (rankedIds.isEmpty()) return DocumentSearchResult.withMatchData(List.of(), Map.of());
}
Specification<Document> textSpec = hasText ? hasIds(rankedIds) : (root, query, cb) -> null;
@@ -312,11 +315,13 @@ public class DocumentService {
// generates an INNER JOIN that silently drops documents with null sender/receivers.
if (sort == DocumentSort.RECEIVER) {
List<Document> results = documentRepository.findAll(spec);
return sortByFirstReceiver(results, dir);
List<Document> sorted = sortByFirstReceiver(results, dir);
return DocumentSearchResult.withMatchData(sorted, enrichWithMatchData(sorted, text));
}
if (sort == DocumentSort.SENDER) {
List<Document> results = documentRepository.findAll(spec);
return sortBySender(results, dir);
List<Document> sorted = sortBySender(results, dir);
return DocumentSearchResult.withMatchData(sorted, enrichWithMatchData(sorted, text));
}
// RELEVANCE: default when text present and no explicit sort given
@@ -325,14 +330,16 @@ public class DocumentService {
List<Document> results = documentRepository.findAll(spec);
Map<UUID, Integer> rankMap = new HashMap<>();
for (int i = 0; i < rankedIds.size(); i++) rankMap.put(rankedIds.get(i), i);
return results.stream()
List<Document> sorted = results.stream()
.sorted(Comparator.comparingInt(
doc -> rankMap.getOrDefault(doc.getId(), Integer.MAX_VALUE)))
.toList();
return DocumentSearchResult.withMatchData(sorted, enrichWithMatchData(sorted, text));
}
Sort springSort = resolveSort(sort, dir);
return documentRepository.findAll(spec, springSort);
List<Document> results = documentRepository.findAll(spec, springSort);
return DocumentSearchResult.withMatchData(results, enrichWithMatchData(results, text));
}
private Sort resolveSort(DocumentSort sort, String dir) {
@@ -584,6 +591,72 @@ public class DocumentService {
return null;
}
/**
* Calls {@code findEnrichmentData} and converts the raw Object[] rows into a
* {@link SearchMatchData} per document. Short-circuits when the list is empty or
* the query is blank (no text search active).
*/
private Map<UUID, SearchMatchData> enrichWithMatchData(List<Document> docs, String query) {
if (docs.isEmpty() || !StringUtils.hasText(query)) return Map.of();
List<UUID> ids = docs.stream().map(Document::getId).toList();
Map<UUID, SearchMatchData> result = new HashMap<>();
for (Object[] row : documentRepository.findEnrichmentData(ids, query)) {
UUID docId = (UUID) row[0];
String titleHeadline = (String) row[1];
String transcriptionSnippet = (String) row[2];
Boolean senderMatched = (Boolean) row[3];
String receiverIdsStr = (String) row[4];
String tagIdsStr = (String) row[5];
result.put(docId, new SearchMatchData(
transcriptionSnippet,
parseTitleOffsets(titleHeadline),
senderMatched != null && senderMatched,
parseUUIDs(receiverIdsStr),
parseUUIDs(tagIdsStr)
));
}
return result;
}
/**
* Scans a {@code ts_headline} result that uses {@code chr(1)}/{@code chr(2)} as
* start/stop delimiters and converts each delimited span into a {@link MatchOffset}
* whose {@code start} and {@code length} are positions in the <em>clean</em> text
* (delimiters stripped). These values align with JavaScript {@code String} indexing.
*/
private static List<MatchOffset> parseTitleOffsets(String headline) {
if (headline == null) return List.of();
List<MatchOffset> offsets = new ArrayList<>();
int i = 0;
int pos = 0; // char position in the clean string (no delimiters)
while (i < headline.length()) {
char c = headline.charAt(i);
if (c == '\u0001') {
int start = pos;
i++;
while (i < headline.length() && headline.charAt(i) != '\u0002') {
i++;
pos++;
}
offsets.add(new MatchOffset(start, pos - start));
i++; // skip \u0002
} else {
i++;
pos++;
}
}
return offsets;
}
private static List<UUID> parseUUIDs(String csv) {
if (csv == null || csv.isBlank()) return List.of();
return Arrays.stream(csv.split(","))
.map(String::trim)
.filter(s -> !s.isEmpty())
.map(UUID::fromString)
.toList();
}
private static String sha256Hex(byte[] bytes) {
try {
MessageDigest digest = MessageDigest.getInstance("SHA-256");