SQL Server

Download HTML from URL and Convert to XML

See more HTML-to-XML/Text Examples

Downloads an HTML page from a URL and converts it to XML. SQL Server

-- Important: See this note about string length limitations for strings returned by sp_OAMethod calls.
--
CREATE PROCEDURE ChilkatSample
AS
BEGIN
    DECLARE @hr int
    DECLARE @iTmp0 int
    -- Important: Do not use nvarchar(max).  See the warning about using nvarchar(max).
    DECLARE @sTmp0 nvarchar(4000)
    DECLARE @success int
    SELECT @success = 0

    -- Note: This example requires the Chilkat Bundle license.

    -- Any string argument automatically begins the 30-day trial.
    DECLARE @glob int
    EXEC @hr = sp_OACreate 'Chilkat.Global', @glob OUT
    IF @hr <> 0
    BEGIN
        PRINT 'Failed to create ActiveX component'
        RETURN
    END

    EXEC sp_OAMethod @glob, 'UnlockBundle', @success OUT, '30-day trial'
    IF @success <> 1
      BEGIN
        EXEC sp_OAGetProperty @glob, 'LastErrorText', @sTmp0 OUT
        PRINT @sTmp0
        EXEC @hr = sp_OADestroy @glob
        RETURN
      END

    DECLARE @http int
    EXEC @hr = sp_OACreate 'Chilkat.Http', @http OUT

    DECLARE @html nvarchar(4000)

    EXEC sp_OAMethod @http, 'QuickGetStr', @html OUT, 'http://www.intel.com/'
    EXEC sp_OAGetProperty @http, 'LastMethodSuccess', @iTmp0 OUT
    IF @iTmp0 <> 1
      BEGIN
        EXEC sp_OAGetProperty @http, 'LastErrorText', @sTmp0 OUT
        PRINT @sTmp0
        EXEC @hr = sp_OADestroy @glob
        EXEC @hr = sp_OADestroy @http
        RETURN
      END

    DECLARE @htmlToXml int
    EXEC @hr = sp_OACreate 'Chilkat.HtmlToXml', @htmlToXml OUT

    -- Indicate the charset of the output XML we'll want.
    EXEC sp_OASetProperty @htmlToXml, 'XmlCharset', 'utf-8'

    -- Set the HTML:
    EXEC sp_OASetProperty @htmlToXml, 'Html', @html

    -- Convert to XML:
    DECLARE @xml nvarchar(4000)

    EXEC sp_OAMethod @htmlToXml, 'ToXml', @xml OUT

    -- Save the XML to a file.
    -- Make sure your charset here matches the charset
    -- used for the XmlCharset property.
    EXEC sp_OAMethod @htmlToXml, 'WriteStringToFile', @success OUT, @xml, 'qa_output/out.xml', 'utf-8'


    PRINT 'Finished.'

    EXEC @hr = sp_OADestroy @glob
    EXEC @hr = sp_OADestroy @http
    EXEC @hr = sp_OADestroy @htmlToXml


END
GO