Some articles are written in several parts, for example, I get articles from IBM Developer:
Data processing distributed with Hadoop, Part 1: Getting StartedDistributed Data Processing with Hydop, Part 2: Moving
Data Processing distributed with Hadoop, Part 3: Application Development
I will index these three articles separately and some keywords search some keywords, this Impossible that Part 3 is on the top of the hit Whales Part 1, the 32nd. For this, if I list the results page of the page, then Part 1 and Part 3 will appear on a separate page.
How can I ensure that the documents displayed together in the same series should be put together?
I guess in SQL, we can use "by group".
I believe what you are asking, which is currently a trunk feature in solar, and Will be included in the next solar version.
If you want to roll your own, there is a possible way to do this:
- Add "series id" field to each document This is a member of a series You have to make sure that it grows for every new series.
- Create an initial query for Lucene, and get the hit list.
- For each hit, check to see if it is a chain ID; If this happens, then create another query from the series ID to retrieve all the members of the series.
Alternatively, to store all the members' IDs in each series within each area, the member's document
Comments
Post a Comment