RFR: CODETOOLS-7901819 hg tstatus does not work well with pager

[Erik Joelsson]

I think the solution suggested by John Coomes looks good and would like 
to see it pushed.

Patch repeated here for reference:

diff --git a/trees.py b/trees.py
--- a/trees.py
+++ b/trees.py
@@ -241,6 +241,7 @@
          if o[1] in cmdopts:
              del cmdopts[o[1]]
      trc = cmd(ui, repo, *args, **cmdopts)
+    ui.flush()
      rc = trc != None and trc or 0
      for subtree in _subtreelist(ui, repo, opts):
          ui.status('\n')
@@ -256,6 +257,7 @@

      ui.status('[%s]:\n' % repo.root)
      trc = cmd(ui, repo, remote, **opts)
+    ui.flush()
      rc = trc != None and trc or 0
      for subtree in _subtreelist(ui, repo, opts):
          ui.status('\n')

/Erik


On 2017-03-07 14:47, Magnus Ihse Bursie wrote:
>
>
> On 2017-01-16 14:24, Magnus Ihse Bursie wrote:
>>
>> On 2016-12-07 13:53, Magnus Ihse Bursie wrote:
>>> On 2016-11-23 19:51, John Coomes wrote:
>>>>> On November 23, 2016 at 1:16 AM Magnus Ihse Bursie 
>>>>> <magnus.ihse.bursie at oracle.com> wrote:
>>>>>
>>>>> Any reviewers on this?
>>>> Hi Magnus,
>>>>
>>>> I think it's better to call ui.flush() after running a command to 
>>>> get the right ordering.  The  patch below does that in both 
>>>> _docmd1() and _docmd2().  Feel free to push w/me as reviewer.
>>>
>>> I tried that but I do still not have write access to the hg repo. :-(
>
> Can anyone with reviewer status please push this fix?
>
> Also, can anyone with the correct status try to make me a reviewer so 
> simple patches just don't get stuck for months upon months? The 
> criteria for getting the proper authorization in these simple tools 
> project should be much lower than the jdk projects. If there's a 
> formal problem, can someone point it out to me so we can work on 
> getting that fixed?
>
> /Magnus