Closed jakelandis closed 7 years ago
@jordansissel - Could you please review ? I am unsure if forcing UTF-8 is appropriate here, or if we should just better understand the environment in which this error was seen.
Also, this change is setting a global value from the merge command, which is kinda odd. Is there better place for this ?
It's probably best to put the Encoding globals to somewhere more ... global (I don't have words for this).. Maybe in the lita config?
Either way, I think this change is fine, and I am not worried about where the Encoding global setting lives.
Hmm. Jarvis can't merge this:
{:timestamp=>"2017-07-18T04:51:09.049197+0000", :message=>"error: patch failed: docs/index.asciidoc:35", :level=>:error, :file=>"cabin/mixins/pipe.rb", :line=>"47", :method=>"block in pipe", :context=>"logstash-filter-tld#6", :branch=>"master"}
{:timestamp=>"2017-07-18T04:51:09.049333+0000", :message=>"error: docs/index.asciidoc: patch does not apply", :level=>:error, :file=>"cabin/mixins/pipe.rb", :line=>"47", :method=>"block in pipe", :context=>"logstash-filter-tld#6", :branch=>"master"}
{:timestamp=>"2017-07-18T04:51:09.049612+0000", :message=>"Patch failed at 0001 asciidocs update", :level=>:info, :file=>"cabin/mixins/pipe.rb", :line=>"47", :method=>"block in pipe", :context=>"logstash-filter-tld#6", :branch=>"master"}
I'll merge manually.
Fixes #81
I had a hard time testing the full workflow, so am only 90% sure this fixes the underlying issue. I was able to reproduce the error on an Ubuntu box by breaking the local setting with
export LC_ALL=en_US
(which is invalid), and then watch this change resolve that theinvalid byte sequence in US-ASCII
error.However, I could not get a full workflow tested due to
Also, the comment to the gemspec was necessary to get bundler to run, and missing require showed up in my testing the single merge command.
FWIW, I believe that this issue can also be fixed locally by ensuring the locale is setup for UTF-8. This change should allow the environment to be what ever, and will always enforce UTF-8. Note - this change may actually beak UTF-16 users.