Skip to main content
ExLibris
  • Subscribe by RSS
  • Ex Libris Knowledge Center

    Check for Duplicate Record in Cataloging

    • Article Type: General
    • Product: Aleph
    • Product Version: 20

    Description:
    When only Title information is entered, system is not able to point us to the correct duplicated record when there is a duplicate.
    Currently, the system is able to check and point us to the correct record when:
    (a) Only ISBN is present
    (b) Both ISBN and Title is present.

    We need the system to be able to do so in all 3 scenarios:
    (a) Only ISBN is present
    (b) Both ISBN and Title is present.
    (c) Only Title is present.

    Can this be achieved?

    Resolution:
    tab00.eng:
    H WTI W-002 00 02 W-titles
    H ISBN IND 02 00 00 ISBN

    tab_locate:
    NPL01 020## a isbn= locate_str_3
    NPL01 022## a issn= locate_str_3
    NPL01 245## a wti= locate_str_1

    locate_str_1 – Runs "build_filing_key" on subfield and takes the 2 longest words. A word must be at least 2 characters in order to be considered as a "word". If the subfield has only one word, the one word will be taken.
    locate_str_3 – Works similarly to locate_str_1, but takes the number of longest words specified in the Column 6 parameters (for example, the three longest words).

    from tab_locate header:
    1) The query includes all lines for which there is data in the source document, with Boolean "and" between them.
    2) Word breaking procedure "97" is used when creating the 'FIND' query, and therefore must be defined in the tab_word_breaking configuration table.
    3) COL4: Find command (WRD code) that is used in the target base for searching similar records; [that is, can only use "Find/Word/tab11_word index, not Browse/ACC/tab11_acc or tab11_ind index]

    check_doc:
    CATALOG-INSERT check_doc_locate

    check_doc_locate
    This program checks if there are records in the database that are similar to the record currently being updated. The mechanism used by this program is determined by the definitions in the tab_locate table of the library's tab directory.

    CONCLUSION: The setup above cannot work because the ISBN and ISSN are IND indexes.

    Open duplicate record and click Save on server button or Check Record (CTRL-U).

    Suggested Scenario: Use Match Instead of Locate
    Match is more flexible and can use the IND indexes, as requested.

    Setup

    tab11_ind
    020 STIDN az
    022# STIDN almyz

    check_doc
    !CATALOG-INSERT check_doc_locate
    CATALOG-INSERT check_doc_match

    tab_match: Match specifications for the check_doc_match checking routine (and others)
    CAT match_doc_uid I- STIDN
    CAT match_doc_acc tab_match_acc

    - match_doc_uid: Match is based on a direct index (Z11). The parameters column (column 3) should contain either the index name (column 5 in tab11_ind) or the tag code (column 1 in tab11_ind). For example, if tab11_ind is defined as follows for the ISBN direct index:
    1 2 3 4 5 6 7 8
    !!!!!-!!!!!-!-!!!!!!!!!!-!!!!!-!!!!!!!!!!!!!!!!!!!!-!-!
    020 ISBN az
    the parameters for a match based on the ISBN can be defined as follows:
    XXX match_doc_uid I-ISBN
    or
    XXX match_doc_uid T-020
    Use I-<index code> or T-<tag code>. When using T-<tag code> there must be an exact match. If tab11_ind col.1 uses 020##, this table must have T-020## as well.
    from SKB 8192-8439: "the match_doc_uid routine reads only the first line of an index's definition from tab11_ind"
    - match_doc_acc: Match is based on a headings (ACC) index. The argument defined in column 3 is a table name. This table lists the tags in the record that should be checked against the headings index.

    tab_match_acc
    245##

    Test 1: Only ISBN is a match. Title is not.
    unsaved = 0898677882, Nonsense
    000077571 = 0898677882, Concrete pressure pipe.

    Result: Document is duplicate in the database (Matched against System No. 000077571 by MATCH_DOC command).

    Test 2: Only Title (Concrete pressure pipe) is a match. ISBN is not.
    unsaved = 1583215484, Concrete pressure pipe
    000077571 = 0898677882, Concrete pressure pipe.

    Result: Document is duplicate in the database (Matched against System No. 000077571 by MATCH_DOC command).

    Test 3: Both Title and ISBN match
    unsaved = 0898677882, Concrete pressure pipe
    000077571 = 0898677882, Concrete pressure pipe.

    Result: Document is duplicate in the database (Matched against System No. 000077571 by MATCH_DOC command).

    Test 4: Neither Title nor ISBN match.
    unsaved = 1583215484, Nonsense
    000077571 = 0898677882, Concrete pressure pipe.

    Result: [no duplicate message]


    • Article last edited: 10/8/2013