https://bz.apache.org/bugzilla/show_bug.cgi?id=58896
--- Comment #3 from Javen O'Neal <[email protected]> --- To test out the times, I wrote a unit test (r1730997) that creates a file with the following contents: 160 rows, 139 columns, with the content of each cell as "Cell[r=159, c=138]" using the workbook default style. The time required to calculate the best fit width is a function of the number of characters and number of styles/fonts. Likely due to the JVM's JIT compiler, the time spent calculating the best fit width for each column decreased over time. With the simplicity of this test, this is obviously a best case scenario for auto-sizing (hence the reason I didn't get 30 seconds). I have personally seen the 30 second turtle pace described by Gili with more complicated spreadsheets (numbers, number formatting, styles) [1] [2] [3] HSSF XSSF SXSSF Populate sheet: 21ms 344ms 9ms Autosize first column: 36ms 4ms 10ms Autosize first ten columns: 217ms 39ms 84ms Autosize all columns: 776ms 447ms 674ms Total (populate+autosize): 797ms 791ms 683ms There isn't a significant difference in the amount of time required to auto-size columns for each spreadsheet implementation. [1] http://pasted.co/af5f61a5 [2] http://pasted.co/5e0cd6700 [3] http://pasted.co/41a51c4e -- 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]
