Блог

У меня следующий вопрос.

 _SEARCH_TSQUERY AS (
        SELECT NULLIF(string_agg(n || ':*', ' | '), ':*')
        FROM (SELECT TRIM(TRIM(n), '''') n
              FROM unnest(regexp_split_to_array
              (plainto_tsquery('simple' :: REGCONFIG, 'washing hair' :: TEXT) :: TEXT, 'amp;')) AS n
                 WHERE 'washing hair' :: TEXT IS NOT NULL
             ) t
        WHERE length(n) > 2
    ), _PRE_SEARCH AS ( SELECT N."@Nomenclature",
                    N."Title",
                   N."Folder",
                   N."Folder@",
                   True                            is_found,
                   (row_number() OVER ())   200 AS rank
            FROM "Nomenclature" N,
                 to_tsquery('simple' :: REGCONFIG, (TABLE _SEARCH_TSQUERY)) as Query
            WHERE (TABLE _SEARCH_TSQUERY) :: TEXT IS NOT NULL
              AND to_tsvector('simple' :: REGCONFIG, "Title") @@ Query
              AND "Type" IN (TABLE _TYPE_IDS)
    
              AND CASE
                      WHEN '{}' :: INT[] IS NOT NULL
                          THEN
                          NOT "@Nomenclature" = ANY ('{}' :: INT[])
                      ELSE
                          TRUE
                END
              AND CASE
                      WHEN true :: BOOLEAN IS TRUE THEN
                          "Folder@" IS NOT TRUE
                      ELSE
                          TRUE
                END
              AND CASE
                      WHEN false :: BOOLEAN IS TRUE THEN
                          "Folder@" IS TRUE
                      ELSE
                          TRUE
                END
              AND "Folder" = ANY ((SELECT array_agg("@Nomenclature") FROM _F_TREE) ::int[])
              AND "Type" = ANY ('{10,11}' :: INT[])  LIMIT 100)) TABLE _PRE_SEARCH;
 

В результате у меня много нерелевантных результатов, как организовать ORDER BY сортировку результатов по первому слову в строке поиска, в данном примере первое слово washing ?