DIP/src/main/java/at/procon/dip/search/service/DocumentLexicalIndexService.java

package at.procon.dip.search.service;

import java.util.List;
import java.util.UUID;
import lombok.RequiredArgsConstructor;
import lombok.extern.slf4j.Slf4j;
import org.springframework.jdbc.core.JdbcTemplate;
import org.springframework.jdbc.core.namedparam.MapSqlParameterSource;
import org.springframework.jdbc.core.namedparam.NamedParameterJdbcTemplate;
import org.springframework.stereotype.Service;
import org.springframework.transaction.annotation.Transactional;

@Service
@RequiredArgsConstructor
@Slf4j
public class DocumentLexicalIndexService {

    private final NamedParameterJdbcTemplate namedParameterJdbcTemplate;
    private final JdbcTemplate jdbcTemplate;

    /**
     * New Slice 2 name kept for current code.
     */
    @Transactional
    public void indexRepresentation(UUID representationId) {
        refreshRepresentationLexicalIndex(representationId);
    }

    /**
     * Backward-compatible Slice 1 method name.
     */
    @Transactional
    public void refreshRepresentationLexicalIndex(UUID representationId) {
        if (!isLexicalSearchSchemaAvailable()) {
            log.debug("Skipping lexical indexing for representation {} because search_vector columns are not present yet", representationId);
            return;
        }

        MapSqlParameterSource params = new MapSqlParameterSource();
        params.addValue("representationId", representationId);
        namedParameterJdbcTemplate.update("""
                UPDATE doc.doc_text_representation
                SET search_config = CASE
                        WHEN lower(coalesce(language_code, '')) = 'de' THEN 'german'
                        WHEN lower(coalesce(language_code, '')) = 'en' THEN 'english'
                        ELSE 'simple'
                    END,
                    search_vector = to_tsvector(
                        CASE
                            WHEN lower(coalesce(language_code, '')) = 'de' THEN 'german'::regconfig
                            WHEN lower(coalesce(language_code, '')) = 'en' THEN 'english'::regconfig
                            ELSE 'simple'::regconfig
                        END,
                        coalesce(text_body, '')
                    )
                WHERE id = :representationId
                """, params);
    }

    /**
     * New Slice 2 method kept for current startup runner.
     */
    @Transactional
    public int backfillMissingVectors(int limit) {
        if (!isLexicalSearchSchemaAvailable()) {
            return 0;
        }
        List<UUID> ids = jdbcTemplate.query("""
                SELECT id
                FROM doc.doc_text_representation
                WHERE search_vector IS NULL
                ORDER BY created_at ASC
                LIMIT ?
                """, (rs, rowNum) -> rs.getObject(1, UUID.class), limit);
        ids.forEach(this::refreshRepresentationLexicalIndex);
        return ids.size();
    }

    /**
     * Backward-compatible Slice 1 method name.
     */
    @Transactional
    public void refreshAllMissingLexicalIndexes() {
        if (!isLexicalSearchSchemaAvailable()) {
            log.info("Lexical search columns are not available yet. Skipping startup backfill for DOC lexical indexes.");
            return;
        }
        jdbcTemplate.update("""
                UPDATE doc.doc_text_representation
                SET search_config = CASE
                        WHEN lower(coalesce(language_code, '')) = 'de' THEN 'german'
                        WHEN lower(coalesce(language_code, '')) = 'en' THEN 'english'
                        ELSE 'simple'
                    END,
                    search_vector = to_tsvector(
                        CASE
                            WHEN lower(coalesce(language_code, '')) = 'de' THEN 'german'::regconfig
                            WHEN lower(coalesce(language_code, '')) = 'en' THEN 'english'::regconfig
                            ELSE 'simple'::regconfig
                        END,
                        coalesce(text_body, '')
                    )
                WHERE search_vector IS NULL
                """);
    }

    /**
     * New Slice 2 name kept for current code.
     */
    public boolean searchVectorColumnsPresent() {
        return isLexicalSearchSchemaAvailable();
    }

    /**
     * Backward-compatible Slice 1 method name.
     */
    public boolean isLexicalSearchSchemaAvailable() {
        Integer count = jdbcTemplate.queryForObject("""
                SELECT COUNT(*)
                FROM information_schema.columns
                WHERE table_schema = 'doc'
                  AND table_name = 'doc_text_representation'
                  AND column_name IN ('search_vector', 'search_config')
                """, Integer.class);
        return count != null && count >= 2;
    }
}