Skipping text

We skip over text using a skip element. We use them in sequenced extractors only. When used there, they skip over up to a certain number of words, sentences or paragraphs. In that sense, they share the grammatical sequence type of the parent sequenced extractor. Skips can either be greedy or non-greedy – it is typically the non-greedy one you want to work with. The non-greedy skip will match the smallest valid sequence it can find, whereas the non-greedy will do the opposite, and match the largest valid sequence it can find. Usually, it is the smallest, tightest, sequence you wish to match with. But the requirements ultimately dictate which you use.

You can never start an extractor’s sequence with a skip or end an extractor’s sequence with a skip (as a skip always requires an anchor or a fact to skip from or up to).

You can have two or more skips in sequence – they will just add up their counts. This might be useful in certain situations where you have certain textual blocks in your content that you would like to treat separately for whatever reasons (governance) – as it technically will not affect performance.

See Skips for information regarding Skip Properties.