Added a dictionary to keep a count of all periodical titles found in reference lines during an extraction job. (This dictionary of totals could be used to produce some kind of statistic about most commonly found titles, for example.)