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