I didn't expected that test.py is run outside V8's source directory. In that case my tests will certainly not work. I will figure out how to handle this. Sorry for breaking the tests.
On Tue, Jul 7, 2009 at 17:01, Kasper Lund<[email protected]> wrote: > Hey Mikhail, > > The new tests are failing so I'm reverting them for now. I hope that's okay. > > Cheers, > Kasper > > On Tue, Jul 7, 2009 at 2:08 PM, <[email protected]> wrote: >> >> Thanks, Erik! >> >> >> http://codereview.chromium.org/149195/diff/1/7 >> File test/mjsunit/tools/tickprocessor.js (right): >> >> http://codereview.chromium.org/149195/diff/1/7#newcode29 >> Line 29: // Files: tools/splaytree.js tools/codemap.js >> tools/csvparser.js tools/consarray.js tools/profile.js >> tools/profile_view.js tools/logreader.js tools/tickprocessor.js >> On 2009/07/07 11:15:32, Erik Corry wrote: >>> Over-long line. >> >> This is an instruction to testcfg.py script to load these files. To fix >> this, I've modified the script to search for several lines of files. >> >> http://codereview.chromium.org/149195/diff/1/10 >> File tools/tickprocessor.js (right): >> >> http://codereview.chromium.org/149195/diff/1/10#newcode520 >> Line 520: 'Show only ticks from JS VM state'], >> On 2009/07/07 11:15:32, Erik Corry wrote: >>> It would be better if this was an object with keys instead of an >> array. Then >>> you could use dispatch.key, dispatch.default and dispatch.helpText >> instead of >>> dispatch[0], dispatch[1] etc. below. You can defer this change to >> another >>> changelist if you like. >> >> The downside of using a map here is that the definition will become too >> wordy: several lines containing repetitive 'key', 'default', etc. Array >> is just briefer. Since this is internal data, I would prefer to leave it >> as is. >> >> http://codereview.chromium.org/149195/diff/1/10#newcode596 >> Line 596: s = s + (new Array(len - s.length + 1).join(' ')); >> On 2009/07/07 11:15:32, Erik Corry wrote: >>> I hope this part isn't performance critical in any way! Couldn't you >> just keep >>> a long string of spaces around and use substring to chop off as many >> as you need >>> to pad. >> >> No, this is just for printing an usage message. >> >> But I also have a twin of this: padLeft which is used for printing >> statistics. I've changed it to use memoization, so previously generated >> paddings are reused. >> >> http://codereview.chromium.org/149195 >> >> >> >> > --~--~---------~--~----~------------~-------~--~----~ v8-dev mailing list [email protected] http://groups.google.com/group/v8-dev -~----------~----~----~----~------~----~------~--~---
