| Submitter | Katsunori FUJIWARA |
|---|---|
| Date | Oct. 20, 2014, 1:27 p.m. |
| Message ID | <e5555e390871e982270f.1413811620@juju> |
| Download | mbox | patch |
| Permalink | /patch/6420/ |
| State | Changes Requested |
| Headers | show |
Comments
On 10/20/2014 03:27 PM, FUJIWARA Katsunori wrote: > # HG changeset patch > # User FUJIWARA Katsunori <foozy@lares.dti.ne.jp> > # Date 1413810488 -32400 > # Mon Oct 20 22:08:08 2014 +0900 > # Branch stable > # Node ID e5555e390871e982270f94cd62bfd771faa97f0b > # Parent da9791c9eb40fc5f2fdb89b97b487e80dcd1a04d > tests: use "sh -c 'script'" style to evaluate "$HG_ARGS" correctly anywhere > > Before this patch, a part of "test-alias.t" fails unexpectedly on > Windows environment, because "cmd.exe" can't evaluate "$HG_ARGS" > expression in shell alias. > > This patch uses "sh -c 'script'" style to evaluate "$HG_ARGS" > expression correctly anywhere. > > diff --git a/tests/test-alias.t b/tests/test-alias.t > --- a/tests/test-alias.t > +++ b/tests/test-alias.t > @@ -395,7 +395,7 @@ command provided extension, should be ab > > [extensions] > > hgext.rebase = > > [alias] > - > rebate = !echo this is \$HG_ARGS > + > rebate = !bash -c 'echo this is \$HG_ARGS' We do (unfortunately) not depend on bash. It would perhaps be more "correct" to test that native shell syntax works and use %HG_ARGS% when on windows. I guess this also means that the help for config.alias is incorrect. It is not ' "$HG" expands to ' but more like ' the HG environment variable is set to '. /Mads
At Mon, 20 Oct 2014 16:04:47 +0200, Mads Kiilerich wrote: > > On 10/20/2014 03:27 PM, FUJIWARA Katsunori wrote: > > # HG changeset patch > > # User FUJIWARA Katsunori <foozy@lares.dti.ne.jp> > > # Date 1413810488 -32400 > > # Mon Oct 20 22:08:08 2014 +0900 > > # Branch stable > > # Node ID e5555e390871e982270f94cd62bfd771faa97f0b > > # Parent da9791c9eb40fc5f2fdb89b97b487e80dcd1a04d > > tests: use "sh -c 'script'" style to evaluate "$HG_ARGS" correctly anywhere > > > > Before this patch, a part of "test-alias.t" fails unexpectedly on > > Windows environment, because "cmd.exe" can't evaluate "$HG_ARGS" > > expression in shell alias. > > > > This patch uses "sh -c 'script'" style to evaluate "$HG_ARGS" > > expression correctly anywhere. > > > > diff --git a/tests/test-alias.t b/tests/test-alias.t > > --- a/tests/test-alias.t > > +++ b/tests/test-alias.t > > @@ -395,7 +395,7 @@ command provided extension, should be ab > > > [extensions] > > > hgext.rebase = > > > [alias] > > - > rebate = !echo this is \$HG_ARGS > > + > rebate = !bash -c 'echo this is \$HG_ARGS' > > We do (unfortunately) not depend on bash. > > It would perhaps be more "correct" to test that native shell syntax > works and use %HG_ARGS% when on windows. I'll send revised one. > I guess this also means that the help for config.alias is incorrect. It > is not ' "$HG" expands to ' but more like ' the HG environment variable > is set to '. What about the notation "$FOO" instead of "environment variable FOO" in other help document ? Should we also replace them for understand-ability of users (especially on Windows) ? Of course, we should replace them after this code freeze period, because this may cause trivial mistakes in changes and/or overwork of translators in a short period, even if it is useful :-) > /Mads > > ---------------------------------------------------------------------- [FUJIWARA Katsunori] foozy@lares.dti.ne.jp
Patch
diff --git a/tests/test-alias.t b/tests/test-alias.t --- a/tests/test-alias.t +++ b/tests/test-alias.t @@ -395,7 +395,7 @@ command provided extension, should be ab > [extensions] > hgext.rebase = > [alias] - > rebate = !echo this is \$HG_ARGS + > rebate = !bash -c 'echo this is \$HG_ARGS' > EOF $ hg reba hg: command 'reba' is ambiguous: