SQL Server Requires Chilkat v11.7.0+

Expand Ligatures and Normalize Unicode in Extracted PDF Text

See more PDF Text Extraction Examples

Demonstrates the "expandLigatures" and "normalize" clean-up options. Typeset PDFs use single glyphs for "fi" and "fl"; by default they are expanded to their letters. Unicode normalization is NFC by default; NFKC also folds compatibility characters such as circled digits, the trademark sign and vulgar fractions into plain characters, and "none" leaves the text as the PDF has it. SQL Server

--
CREATE PROCEDURE ChilkatSample
AS
BEGIN
    DECLARE @hr int
    DECLARE @sTmp0 nvarchar(4000)
    DECLARE @success int
    SELECT @success = 0

    --  This example requires the Chilkat API to have been previously unlocked.
    --  See Global Unlock Sample for sample code.

    --  The PDF used in this example is available at ligatures.pdf

    DECLARE @pdf int
    EXEC @hr = sp_OACreate 'Chilkat.Pdf', @pdf OUT
    IF @hr <> 0
    BEGIN
        PRINT 'Failed to create ActiveX component'
        RETURN
    END

    EXEC sp_OAMethod @pdf, 'LoadFile', @success OUT, 'qa_data/pdf/ligatures.pdf'
    IF @success = 0
      BEGIN
        EXEC sp_OAGetProperty @pdf, 'LastErrorText', @sTmp0 OUT
        PRINT @sTmp0
        EXEC @hr = sp_OADestroy @pdf
        RETURN
      END

    DECLARE @json int
    EXEC @hr = sp_OACreate 'Chilkat.JsonObject', @json OUT

    DECLARE @jsonOut int
    EXEC @hr = sp_OACreate 'Chilkat.JsonObject', @jsonOut OUT

    --  The first line is set with the fi and fl ligature glyphs; the second has circled digits,
    --  a trademark sign and a vulgar fraction.
    EXEC sp_OAMethod @pdf, 'ExtractText', @success OUT, 'text', @json, @jsonOut
    IF @success = 0
      BEGIN
        EXEC sp_OAGetProperty @pdf, 'LastErrorText', @sTmp0 OUT
        PRINT @sTmp0
        EXEC @hr = sp_OADestroy @pdf
        EXEC @hr = sp_OADestroy @json
        EXEC @hr = sp_OADestroy @jsonOut
        RETURN
      END

    PRINT '---- default (expandLigatures: true, normalize: NFC) ----'
    EXEC sp_OAMethod @jsonOut, 'StringOf', @sTmp0 OUT, 'text'
    PRINT @sTmp0

    --  Keep the ligature characters as they are in the PDF.
    EXEC sp_OAMethod @json, 'UpdateBool', @success OUT, 'expandLigatures', 0
    EXEC sp_OAMethod @json, 'UpdateString', @success OUT, 'normalize', 'none'
    EXEC sp_OAMethod @pdf, 'ExtractText', @success OUT, 'text', @json, @jsonOut
    IF @success = 0
      BEGIN
        EXEC sp_OAGetProperty @pdf, 'LastErrorText', @sTmp0 OUT
        PRINT @sTmp0
        EXEC @hr = sp_OADestroy @pdf
        EXEC @hr = sp_OADestroy @json
        EXEC @hr = sp_OADestroy @jsonOut
        RETURN
      END

    PRINT '---- expandLigatures: false, normalize: none ----'
    EXEC sp_OAMethod @jsonOut, 'StringOf', @sTmp0 OUT, 'text'
    PRINT @sTmp0

    --  NFKC folds the compatibility characters too (and expands ligatures on its own).
    EXEC sp_OAMethod @json, 'UpdateString', @success OUT, 'normalize', 'NFKC'
    EXEC sp_OAMethod @pdf, 'ExtractText', @success OUT, 'text', @json, @jsonOut
    IF @success = 0
      BEGIN
        EXEC sp_OAGetProperty @pdf, 'LastErrorText', @sTmp0 OUT
        PRINT @sTmp0
        EXEC @hr = sp_OADestroy @pdf
        EXEC @hr = sp_OADestroy @json
        EXEC @hr = sp_OADestroy @jsonOut
        RETURN
      END

    PRINT '---- normalize: NFKC ----'
    EXEC sp_OAMethod @jsonOut, 'StringOf', @sTmp0 OUT, 'text'
    PRINT @sTmp0

    --  Sample output (shown here in ASCII: (fi) and (fl) stand for the single ligature characters, (1) and (2) for the circled digits,
    --  (TM) for the trademark sign and 1/2 for the vulgar fraction; the NFKC output has plain "1", "2", "TM" and "1", fraction slash, "2"):

    --  ---- default (expandLigatures: true, normalize: NFC) ----
    --  Typeset Text
    --  
    --  The final flight of the season was officially confirmed on the fifth. Rooms (1) and (2) are reserved; the Harbor Cafe(TM) serves 1/2 portions.
    --  ---- expandLigatures: false, normalize: none ----
    --  Typeset Text
    --  
    --  The (fi)nal (fl)ight of the season was of(fi)cially con(fi)rmed on the (fi)fth. Rooms (1) and (2) are reserved; the Harbor Cafe(TM) serves 1/2 portions.
    --  ---- normalize: NFKC ----
    --  Typeset Text
    --  
    --  The final flight of the season was officially confirmed on the fifth. Rooms 1 and 2 are reserved; the Harbor CafeTM serves 1/2 portions.

    EXEC @hr = sp_OADestroy @pdf
    EXEC @hr = sp_OADestroy @json
    EXEC @hr = sp_OADestroy @jsonOut


END
GO