| Submitter | Pierre-Yves David |
|---|---|
| Date | June 21, 2015, 3:38 a.m. |
| Message ID | <da62bc192d6302180309.1434857923@marginatus.alto.octopoid.net> |
| Download | mbox | patch |
| Permalink | /patch/9738/ |
| State | Accepted |
| Commit | ae22d3048075d0a4e4daece4bdb231a1d939f97e |
| Delegated to: | Augie Fackler |
| Headers | show |
Comments
On Sat, Jun 20, 2015 at 08:38:43PM -0700, Pierre-Yves David wrote: > # HG changeset patch > # User Pierre-Yves David <pierre-yves.david@fb.com> > # Date 1434842530 25200 > # Sat Jun 20 16:22:10 2015 -0700 > # Node ID da62bc192d63021803097a267d48e66f9f64e9fe > # Parent 75305531b0c92e31b35a3357774597ac78b1a1fd > revsetbenchmark: do not abort on failure to run a revset > > Instead of aborting the whole process, we just skip entry for revset that > failed to run. I'm not super-thrilled with this. Could we be capturing the error output somehow? > > diff --git a/contrib/revsetbenchmarks.py b/contrib/revsetbenchmarks.py > --- a/contrib/revsetbenchmarks.py > +++ b/contrib/revsetbenchmarks.py > @@ -60,11 +60,11 @@ def perf(revset, target=None): > print >> sys.stderr, 'abort: cannot run revset benchmark: %s' % exc.cmd > if exc.output is None: > print >> sys.stderr, '(no ouput)' > else: > print >> sys.stderr, exc.output > - sys.exit(exc.returncode) > + return None > > outputre = re.compile(r'! wall (\d+.\d+) comb (\d+.\d+) user (\d+.\d+) ' > 'sys (\d+.\d+) \(best of (\d+)\)') > > def parseoutput(output): > @@ -158,12 +158,17 @@ def formattiming(value): > > _marker = object() > def printresult(variants, idx, data, maxidx, verbose=False, reference=_marker): > """print a line of result to stdout""" > mask = '%%0%ii) %%s' % idxwidth(maxidx) > + > out = [] > for var in variants: > + if data[var] is None: > + out.append('error ') > + out.append(' ' * 4) > + continue > out.append(formattiming(data[var]['wall'])) > if reference is not _marker: > factor = None > if reference is not None: > factor = getfactor(reference[var], data[var], 'wall') > _______________________________________________ > Mercurial-devel mailing list > Mercurial-devel@selenic.com > https://selenic.com/mailman/listinfo/mercurial-devel
On 06/22/2015 07:23 AM, Augie Fackler wrote: > On Sat, Jun 20, 2015 at 08:38:43PM -0700, Pierre-Yves David wrote: >> # HG changeset patch >> # User Pierre-Yves David <pierre-yves.david@fb.com> >> # Date 1434842530 25200 >> # Sat Jun 20 16:22:10 2015 -0700 >> # Node ID da62bc192d63021803097a267d48e66f9f64e9fe >> # Parent 75305531b0c92e31b35a3357774597ac78b1a1fd >> revsetbenchmark: do not abort on failure to run a revset >> >> Instead of aborting the whole process, we just skip entry for revset that >> failed to run. > > I'm not super-thrilled with this. Could we be capturing the error > output somehow? Meh, the error output is alreay forwarded to the stderr when we try to run the revsets, and this is "good enough" for now. We will not reused the error output in the result summary because it is will never fit in any column. As we currently have no user to capture it, we don't. It is simple enough to capture it when we need it later.
On Mon, Jun 22, 2015 at 10:12:46AM -0700, Pierre-Yves David wrote: > > > On 06/22/2015 07:23 AM, Augie Fackler wrote: > >On Sat, Jun 20, 2015 at 08:38:43PM -0700, Pierre-Yves David wrote: > >># HG changeset patch > >># User Pierre-Yves David <pierre-yves.david@fb.com> > >># Date 1434842530 25200 > >># Sat Jun 20 16:22:10 2015 -0700 > >># Node ID da62bc192d63021803097a267d48e66f9f64e9fe > >># Parent 75305531b0c92e31b35a3357774597ac78b1a1fd > >>revsetbenchmark: do not abort on failure to run a revset > >> > >>Instead of aborting the whole process, we just skip entry for revset that > >>failed to run. > > > >I'm not super-thrilled with this. Could we be capturing the error > >output somehow? > > Meh, the error output is alreay forwarded to the stderr when we try to run > the revsets, and this is "good enough" for now. We will not reused the error > output in the result summary because it is will never fit in any column. Ah, okay, I didn't have that context. Queued with delight. > As we currently have no user to capture it, we don't. It is simple enough to > capture it when we need it later. > > -- > Pierre-Yves David
Patch
diff --git a/contrib/revsetbenchmarks.py b/contrib/revsetbenchmarks.py --- a/contrib/revsetbenchmarks.py +++ b/contrib/revsetbenchmarks.py @@ -60,11 +60,11 @@ def perf(revset, target=None): print >> sys.stderr, 'abort: cannot run revset benchmark: %s' % exc.cmd if exc.output is None: print >> sys.stderr, '(no ouput)' else: print >> sys.stderr, exc.output - sys.exit(exc.returncode) + return None outputre = re.compile(r'! wall (\d+.\d+) comb (\d+.\d+) user (\d+.\d+) ' 'sys (\d+.\d+) \(best of (\d+)\)') def parseoutput(output): @@ -158,12 +158,17 @@ def formattiming(value): _marker = object() def printresult(variants, idx, data, maxidx, verbose=False, reference=_marker): """print a line of result to stdout""" mask = '%%0%ii) %%s' % idxwidth(maxidx) + out = [] for var in variants: + if data[var] is None: + out.append('error ') + out.append(' ' * 4) + continue out.append(formattiming(data[var]['wall'])) if reference is not _marker: factor = None if reference is not None: factor = getfactor(reference[var], data[var], 'wall')