Issue 101254 - Broadcaster areas insertion performance bottleneck
Summary: Broadcaster areas insertion performance bottleneck
Status: CLOSED FIXED
Alias: None
Product: Calc
Classification: Application
Component: open-import (show other issues)
Version: DEV300m42
Hardware: All All
: P3 Trivial (vote)
Target Milestone: ---
Assignee: oc
QA Contact: issues@sc
URL:
Keywords: performance
Depends on:
Blocks:
 
Reported: 2009-04-21 20:09 UTC by ooo
Modified: 2017-05-20 11:42 UTC (History)
2 users (show)

See Also:
Issue Type: DEFECT
Latest Confirmation in: ---
Developer Difficulty: ---

Attachments
Test case with one sheet and 64k ranges to listen to, all different but starting at A1. (218.86 KB, application/vnd.oasis.opendocument.spreadsheet)
2009-04-23 14:43 UTC, ooo 		no flags Details
Test case with two sheets and 128k ranges to listen to (64k each sheet), all different but starting at A1 of the containing sheet. (400.79 KB, application/vnd.oasis.opendocument.spreadsheet)
2009-04-23 14:47 UTC, ooo 		no flags Details
Add an attachment (proposed patch, testcase, etc.)

Note You need to log in before you can comment on or make changes to this issue.
Description ooo 2009-04-21 20:09:41 UTC 
Setting up broadcaster areas for formula cells listening to cell ranges has a
performance bottleneck due to the underlying unique sorted associative container
::std:set. Details will be lined out in
http://wiki.services.openoffice.org/wiki/Calc/Performance/Refactoring_Area_Broadcasters
Comment 1 ooo 2009-04-21 20:11:01 UTC 
Working on it.
Comment 2 ooo 2009-04-22 12:54:16 UTC 
In cws calcperf04:

revision 271104
sc/inc/address.hxx
sc/source/core/data/bcaslot.cxx
sc/source/core/data/document.cxx
sc/source/core/inc/bcaslot.hxx
Comment 3 ooo 2009-04-23 14:43:48 UTC 
Created attachment 61758 [details]
Test case with one sheet and 64k ranges to listen to, all different but starting at A1.
Comment 4 ooo 2009-04-23 14:47:45 UTC 
Created attachment 61759 [details]
Test case with two sheets and 128k ranges to listen to (64k each sheet), all different but starting at A1 of the containing sheet.
Comment 5 ooo 2009-04-24 16:48:25 UTC 
Reassigning to QA for verification.
Details and measurements at the wiki page mentioned above.
Comment 6 joerg.skottke 2009-05-19 12:50:16 UTC 
OOo 3.1 (single): 29:6 seconds
CWS: 16,0 seconds
Gain: 46,3 %

OOo 3.1 (double): 59:9 seconds
CWS: 28:6 seconds
Gain: 52,3

Regression testing pending
Comment 7 oc 2009-05-26 14:35:53 UTC 
verified in internal build cws_calcperf04