We have a project with approximately 20,000 documents which require searching with hit highlighting on the content. The content is of variable size. My question is which option to take to support hit highlighting:
1. To store the content as a field in the lucene document and to highlight hits from there. 2. To store the path to the document in the lucene document, open up the document and then run hit highlighting. Is there a "best-practice" for this? Has anybody tried either of these approaches? Are there any advantages or drawbacks of either approach? Many thanks, Hugh