Metrics: Overview

Here you can see a simplified overview of the main stages building the metrical analysis process, from the text to the observed data with their rich query and reporting features.

Text extraction and parsing

Reading this schema from top left to bottom right, first of all we need texts to analyze. This system components can get text from several sources, mainly Packard Humanities cd-roms (by automatically converting their Beta code into Unicode: for a demo cf. Ibis, anyway I have created a specialized application which browses, extracts and converts texts from such cd-roms); of course, other digital formats are allowed (whatever their encoding is) or even from direct typing (cf. the Word addin Theuth).

Once we have a text, a smart parsing removes all the irrelevant noise characters from it: some of these characters (like line numbers) are used to collect metadata, others (like non-textual characters as diplai, parentheses, etc.) are just discarded (but they will be restored at their place when generating an output). Finally, the text filtered in this way is normalized so that the phonemic analysis can get an input where all the confusing or irrelevant variants (uppercase, lowercase, final or lunate sigma, etc.) have been removed.