I've changed string operation codes to use makeString().
Created attachment 96918 [details] Patch
LGTM.
Informal r+ from my side.
Comment on attachment 96918 [details] Patch View in context: https://bugs.webkit.org/attachment.cgi?id=96918&action=review > Source/WebKit2/ChangeLog:6 > + [EFL][WK2] Change string operation to makeString() > + https://bugs.webkit.org/show_bug.cgi?id=62527 Please add one line explaining why, like why that makes sense, why that is better etc? like 'use makeStrings() instead of creating individual strings and concatenating them'
Created attachment 97109 [details] Patch
(In reply to comment #4) > (From update of attachment 96918 [details]) > View in context: https://bugs.webkit.org/attachment.cgi?id=96918&action=review > > > Source/WebKit2/ChangeLog:6 > > + [EFL][WK2] Change string operation to makeString() > > + https://bugs.webkit.org/show_bug.cgi?id=62527 > > Please add one line explaining why, like why that makes sense, why that is better etc? > > like 'use makeStrings() instead of creating individual strings and concatenating them' Done. I've added one line for explaining.
Comment on attachment 97109 [details] Patch Patch looks OK; changelog could use more work.
Comment on attachment 97109 [details] Patch Looks OK to me.
Comment on attachment 97109 [details] Patch Informal r+
Created attachment 97207 [details] Patch
(In reply to comment #10) > Created an attachment (id=97207) [details] > Patch modify "use" to "Use"
Comment on attachment 97207 [details] Patch Internal r+
Comment on attachment 97207 [details] Patch Clearing flags on attachment: 97207 Committed r88907: <http://trac.webkit.org/changeset/88907>
All reviewed patches have been landed. Closing bug.
This patch is wrong. makeString is deprecated, using String operator+ gives exactly the same performance. This has been changed recently. use "osVersion = name.sysname + " " + name.machine;" You are not required to cast to Strings first. The StringAppend template solution will reduce this String concatting operation to just a single malloc of the final size, no appending is done at all.
(In reply to comment #15) > This patch is wrong. makeString is deprecated, using String operator+ gives > exactly the same performance. This has been changed recently. > > use "osVersion = name.sysname + " " + name.machine;" > You are not required to cast to Strings first. > > The StringAppend template solution will reduce this String concatting operation > to just a single malloc of the final size, no appending is done at all. You're right. I've rolled this out as r88944. Changing this bug status to 'invalid'.