Closed J4bbi closed 1 year ago
This happens for a valid reasons.
Whereas before, there would be a pre-populated free text field with names, synonyms, roles now, each field is independently searched.
In the example mentioned here, searching on authors starts with new results in a SQL query like so:
SELECT "cofk_union_work"."work_id", "cofk_union_work"."description", "cofk_union_work"."date_of_work_as_marked", "cofk_union_work"."original_calendar", "cofk_union_work"."date_of_work_std", "cofk_union_work"."date_of_work_std_gregorian", "cofk_union_work"."date_of_work_std_year", "cofk_union_work"."date_of_work_std_month", "cofk_union_work"."date_of_work_std_day", "cofk_union_work"."date_of_work2_std_year", "cofk_union_work"."date_of_work2_std_month", "cofk_union_work"."date_of_work2_std_day", "cofk_union_work"."date_of_work_std_is_range", "cofk_union_work"."date_of_work_inferred", "cofk_union_work"."date_of_work_uncertain", "cofk_union_work"."date_of_work_approx", "cofk_union_work"."authors_as_marked", "cofk_union_work"."addressees_as_marked", "cofk_union_work"."authors_inferred", "cofk_union_work"."authors_uncertain", "cofk_union_work"."addressees_inferred", "cofk_union_work"."addressees_uncertain", "cofk_union_work"."destination_as_marked", "cofk_union_work"."origin_as_marked", "cofk_union_work"."destination_inferred", "cofk_union_work"."destination_uncertain", "cofk_union_work"."origin_inferred", "cofk_union_work"."origin_uncertain", "cofk_union_work"."abstract", "cofk_union_work"."keywords", "cofk_union_work"."work_is_translation", "cofk_union_work"."incipit", "cofk_union_work"."explicit", "cofk_union_work"."ps", "cofk_union_work"."original_catalogue", "cofk_union_work"."accession_code", "cofk_union_work"."work_to_be_deleted", "cofk_union_work"."iwork_id", "cofk_union_work"."editors_notes", "cofk_union_work"."edit_status", "cofk_union_work"."relevant_to_cofk", "cofk_union_work"."creation_timestamp", "cofk_union_work"."creation_user", "cofk_union_work"."change_timestamp", "cofk_union_work"."change_user", "cofk_union_work"."uuid" FROM "cofk_union_work" WHERE EXISTS(SELECT (1) AS "a" FROM "cofk_union_work" U0 INNER JOIN "cofk_work_person_map" U1 ON (U0."work_id" = U1."work_id") LEFT OUTER JOIN "cofk_union_person" U2 ON (U1."person_id" = U2."person_id") WHERE (U1."relationship_type" IN (created) AND (UPPER(U2."date_of_birth_year"::text) LIKE UPPER(new%) OR UPPER(U2."date_of_death_year"::text) LIKE UPPER(new%) OR UPPER(U2."date_of_death_is_range"::text) LIKE UPPER(new%) OR UPPER(U2."date_of_birth_is_range"::text) LIKE UPPER(new%) OR UPPER(U2."foaf_name"::text) LIKE UPPER(new%) OR UPPER(U2."skos_altlabel"::text) LIKE UPPER(new%) OR UPPER(U2."person_aliases"::text) LIKE UPPER(new%)) AND U0."work_id" = ("cofk_union_work"."work_id")) LIMIT 1) ORDER BY "cofk_union_work"."date_of_work_std" ASC
Miranda is used to starts with searching on the block of text that would contain the sum of all these fields, invariably with the person's name first.
In the search works, some of the searches are calling up strange results, e.g. Author/sender starts with "New" brings up, as the 4th letter result, a letter with the author ‘Locke, Thomas’ (which is obviously not a name that begis with ‘New’):
Or, if you do an ‘Addressee starts with Lewis’ search you get Thiery, Louis. This means it’s not just searching on the primary name, but rather on the synonyms, so all sorts of strange results are thrown up. e.g.
Rather than in EMLO-Edit-Old:
Incidentally, this isn’t the case with ‘Author starts with …’ searches, so it’s drawing on the primary name field in that search.