You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Hi, it is clear that BoboBrowse.Net works only with strings, but how can you deal with this limitation during a search of numbers and dates?
I try to make a clear example.
Suppose to index a field such as
Now, you would like to retrieve the indexed value (19). You can do that only writing a (lucene) query like fileld='number' text='000019', which assume that the user knows in advance the correct format for the field number. In fact, if you write fileld='number' text='19' (which is much more intuitive) it returns no results. Thus, the only option you have is to format the user query as done for the indexed value. Is that right?
I ask you that because I am used to build Lucene queries exploiting MultiFieldQueryParser that cannot be used if you format the input number or dates before indexing.
The text was updated successfully, but these errors were encountered:
The simplest solution is to store 2 fields with the same data in 2 different formats. An index is not a database, there is no rule that says that duplicating the data is bad. You just have to keep in mind that BoboBrowse.Net has these requirements and design the index to work with these formatted string data types as well as any other types you require. If you need to make them native data types for some other reason, you just need to add another field to the document.
In the above case, you would use the "numberBobo" field with one (or more than one) facet handlers. But the index is still tracking the numeric field "number" in case you need to utilize it in another way, such as for search and/or feeding it back to the application after a hit is made.
Hi, it is clear that BoboBrowse.Net works only with strings, but how can you deal with this limitation during a search of numbers and dates?
I try to make a clear example.
Suppose to index a field such as
using a standard analyzer (which is the best choice to analyze a number as text) into the index writer
Now, you would like to retrieve the indexed value (19). You can do that only writing a (lucene) query like
fileld='number' text='000019'
, which assume that the user knows in advance the correct format for the field number. In fact, if you writefileld='number' text='19'
(which is much more intuitive) it returns no results. Thus, the only option you have is to format the user query as done for the indexed value. Is that right?I ask you that because I am used to build Lucene queries exploiting
MultiFieldQueryParser
that cannot be used if you format the input number or dates before indexing.The text was updated successfully, but these errors were encountered: