https://bz.apache.org/bugzilla/show_bug.cgi?id=57840

Javen O'Neal <[email protected]> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
         Resolution|---                         |FIXED
             Status|NEW                         |RESOLVED

--- Comment #39 from Javen O'Neal <[email protected]> ---
(In reply to Greg Woolsey from comment #38)
> Perhaps we still need some comments as to why the explicit boxing is 
> happening,
> so they don't get undone?
Applied in r1748479.

I kept this bug open in case you had anything else you discovered from your
testing. I'll close it now.
These changes will be available in POI 3.15 beta 2, to be released in the next
couple months.

I'm assuming that general formula evaluation is faster as a whole with
r1747840, not just structured references.

In future bugs, we should look at:
 * comment 5: get Vaadin Spreadsheet working with POI versions higher than
3.12.
 * comment 12: When evaluating multiple formulas in a sheet, cache the formula
result of cells with the same expression
 * comment 21: add to POI documentation that "-XX:AutoBoxCacheMax=1048704" may
improve performance of working with large worksheets (perhaps this is no longer
true now that we're using explicit boxing).
 * comment 23: use HashMap instead of TreeMap to hold XSSFRows and XSSFCells in
XSSFSheet and XSSFRows, respectively. Obviously this would make moving rows or
getting all rows in a range extremely expensive.
 * comment 34: trigger a cache reset, either with an explicit public method or
implicitly with an internal method (this could make regular workbook operations
slower since any cell that is part of a table or any cell in a table that
references a value that changes in the workbook). Several methods in XSSFTable
still use caches that may contain stale data.
 * comment 38: continue to improve formula evaluation performance

-- 
You are receiving this mail because:
You are the assignee for the bug.

---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to