13.1
Name:Word Segmentation Demo Script
Description: With the need of handling larger amount of data and effective text processing, we need to explore the domain of arrays (number and string) and more advanced features like regular expressions. These are not easy to work with in Praat because of its rather limited functionality as a scripting language and also because the lack of detailed documentation for many of these features. This script is a demo of using string arrays and regex for the task of segmenting a series of strings (phrases or sentences) into word units automatically. It also demonstrates the ways variable substitution works in Praat scripts, in conjunction with arrays (especially with string arrays, which involve some very tricky syntax to get the substitution right). It can be used as a procedure in a larger script with moderate amount of modification.For more information, consult the PSL manual and the Praat Scripting section of the Postings.

13.2
Description:this script differs from 13.1 in several aspects. (1)It demos the word segmentation using a procedure;(2)the arguments of the procedure include a total word count argument, which is necessary when processing multiple lines of segmentation; (3)the variable name 'div' is found to be reserved when being evaluated in a conditional expression, so that was changed;(4)the use of space character as the cue to word segmentation turned out to be not so robust due to the variation in spaces (there can be extra spaces). Therefore, while 13.1 is good for a regularly spaced English sentence, 13.2 takes the advantage of Chinese tone marks in this specific task. Each tone mark consistently and unambiguously marks the end of a word. Thus the detection of word boundary in 13.2 is done by checking tone marks and using regular expression as well (all tone marks are numbers).