So, I just sat here looking at the latest release notes, and once again marvel at how the team at Logos/Faithlife continues to expand the metadata tagging and associated search/query facilities that are in Logos. My ability to search my library in complex ways just grows and grows. Thank you.
That said, I still find times where it doesn't seem like Logos will allow me to do what I want wrt searching. Of course, I freely acknowledge that in spite of having used Logos since the 2.x days, I don't know it all. So, I may be ignorant of something that is already present.
Most of what I find I can't do seems to center around building searches that are dependent on previous search results or those that could easily be defined with boolean expressions that are not expressable with the Logos search syntax. Let me provide a prototypical example:
- I perform a search "X greek lemma WITHIN 0 WORDS Y english word" within my tagged KJV. This works fine.
- Now, I want to redo my search to include all of those results that would be included with a search on only the first (greek lemma) part of my search above, but NOT those in the second (english word) part.
- It seems like what I want from a boolean expression standpoint is term 1 INTERSECTION NOT term 2.. Basically, I want the results that do not meet the 2nd term's criteria (including the WITHIN 0 WORDS constraint), but do meet the first term's.
- I can get close by doing first term ANDNOT second term...but this doesn't enforce the "WITHIN 0 WORDS" constraint on the atomic search terms.
Similar to this is just a general capability to build a new search off of an already executed one. For example, in my steps above, I could perform this query by first doing a search on "X greek lemma" and then executing a filtering search that refines those results to not include those in the search "term 2" (but still matching the term1 criteria.). Basically, instead of running my search against a resource or range within a resource, I run it against previous search results.
Again, I may be missing something, or thinking about the problem incorrectly, but it doesn't seem like I can do such things with the current feature set. Am I wrong?
Dave
