Displaying template "file_copies" keyword without --copies

Patrick Mézard pmezard at gmail.com
Thu Dec 10 14:59:57 CST 2009


Le 04/12/09 20:12, Matt Mackall a écrit :
> On Wed, 2009-12-02 at 15:32 +0100, Patrick Mézard wrote:
>> As suggested in http://www.selenic.com/pipermail/mercurial-devel/2009-November/017148.html , I have a local patch which forces a template or style containing the {file_copies} keyword to always display it instead of doing so only if --copies is passed. The remaining failing test is:
>>
>> -------------------------
>> ERROR: test-command-template output changed
>> --- Expected output 
>> +++ Test output 
>> @@ -4,7 +4,29 @@
>>  # default style is like normal output
>>  #  normal
>>  #  verbose
>> +log.out style.out differ: char 141, line 6
>> +--- log.out	2009-12-02 14:17:26.000000000 +0000
>> ++++ style.out	2009-12-02 14:17:26.000000000 +0000
>> +@@ -3,6 +3,7 @@
>> + user:        test
>> + date:        Wed Jan 01 10:01:00 2020 +0000
>> + files:       fourth second third
>> ++copies:      fourth (second)
>> + description:
>> + third
>> + 
>>  #  debug
>> +log.out style.out differ: char 351, line 10
>> +--- log.out	2009-12-02 14:17:26.000000000 +0000
>> ++++ style.out	2009-12-02 14:17:26.000000000 +0000
>> +@@ -7,6 +7,7 @@
>> + date:        Wed Jan 01 10:01:00 2020 +0000
>> + files+:      fourth third
>> + files-:      second
>> ++copies:      fourth (second)
>> + extra:       branch=default
>> + description:
>> + third
>>  # revision with no copies (used to print a traceback)
>>  
>>  # compact style works
>> !
>> -------------------------
>>
>> This test compares the output of the default changelog and the styled one with a normal, --verbose and --debug output. With my patch, both --verbose and --debug differ because the related styles contain {file_copies} so the copies are displayed even if --copies is not set in the test. What should I do about it ?
>>
>> 1- Accept the styled changelog outputs are now equivalent only to the --copies non-styled ones?
>> 2- Remove the {file_copies} from the styles so they match their non --copies counterpart?
>> 3- Give up for this is not a tolerable compatibility breakage?
> 
> 4- Have two variables: one that matches the current behavior (and name)
> and one that always shows copies?

Not easy to come up with an elegant name. At least, it should draw the attention on {file_copies} special behaviour.

{file_copies_forced}
{file_copies_always}
{file_copyrecords}
?

--
Patrick Mézard



More information about the Mercurial-devel mailing list