Error about invalid byte sequence in read_yaml

[chartjes]

When running 'rake generate' I am getting the following error message:

/Users/chartjes/.rvm/gems/ruby-1.9.2-p290/gems/jekyll-0.11.0/lib/jekyll/convertible.rb:29:in `read_yaml': invalid byte sequence in US-ASCII (ArgumentError)

How can I go about fixing this problem? I have 427 posts in markdown and having gone through all of them I cannot find the offending character.

[imathis]

Could this be a utf8 issue?

http://stackoverflow.com/questions/3140111/jekyll-does-not-parse-utf-8

[chartjes]

It could very well be. I'm using vim on OS-X and I'd bet it's set to encode everything as UTF8. So how would I go about re-incoding all those files?

Chris Hartjes
http://www.littlehart.net/atthekeyboard
@chartjes

On Saturday, 10 September, 2011 at 5:35 PM, Brandon Mathis wrote:

Could this be a utf8 issue?

http://stackoverflow.com/questions/3140111/jekyll-does-not-parse-utf-8

Reply to this email directly or view it on GitHub:
#144 (comment)

[imathis]

I use MacVim on OS X too, that's not the issue. Didn't you convert from Wordpress? Do find in page for 'ASCII' on http://zanshin.net/2011/08/11/switching-to-octopress/ that might help you

[chartjes]

Yes, I was converting from WordPress. That post doesn't offer any insight into how to actually re-encode the files so that Jekyll / Octopress will be happy

Chris Hartjes
http://www.littlehart.net/atthekeyboard
@chartjes

On Saturday, 10 September, 2011 at 5:39 PM, Brandon Mathis wrote:

I use MacVim on OS X too, that's not the issue. Didn't you convert from Wordpress? Do find in page for 'ASCII' on http://zanshin.net/2011/08/11/switching-to-octopress/ that might help you

Reply to this email directly or view it on GitHub:
#144 (comment)

[imathis]

:( I was hoping it'd set you on a course. I'll look further.

[imathis]

Here's the script he used: https://gist.github.com/1133266 note the 'encode' bit on line 44. Hope this helps.

[chartjes]

That would be cool if my stuff was exported as XML…I had to use the Jekyll migration script that pulled stuff in from the database.

Chris Hartjes
http://www.littlehart.net/atthekeyboard
@chartjes

On Saturday, 10 September, 2011 at 5:46 PM, Brandon Mathis wrote:

Here's the script he used: https://gist.github.com/1133266 note the 'encode' bit on line 44. Hope this helps.

Reply to this email directly or view it on GitHub:
#144 (comment)

[chartjes]

Using the file command that is suggested in that blog post tells me I have "HTML document text" and "ASCII English Text" documents in my .markdown files

Chris Hartjes
http://www.littlehart.net/atthekeyboard
@chartjes

On Saturday, 10 September, 2011 at 5:44 PM, Brandon Mathis wrote:

:( I was hoping it'd set you on a course. I'll look further.

Reply to this email directly or view it on GitHub:
#144 (comment)

[imathis]

Your avatar is intimidating me into helping you figure this out, but I don't know what to suggest :/

[zanshin]

Why not just create a small ruby script to read each .markdown file that was output from your Jekyll conversion, do the encode bit from my script to clean things up, and re-save the file?

[imathis]

@chartjes I think you solved this right? Closing, but you're welcome to reopen.

[chartjes]

Yeah, problem was installed by reinstalling Octopress and following the directions about RVM

[imathis]

Great, thanks.

[pfleidi]

Had the same problem. For me, exporting the right locale setting in my shell environment fixed that:

 export LC_CTYPE=en_US.UTF-8
 export LANG=en_US.UTF-8

You can validate your settings by executing locale in your current shell session.

[aetherwu]

@pfleidi is right. thank you!

[lscott3]

@pfleidi Thanks! This fixed my issue as well!

[andrewmichaelsmith]

@pfleidi Thanks, that did the job

[PHironaka]

Should I write blog posts in .textile to avid UTF-8 issues?

[dearprakash]

@PHironaka not necessary. If you encounter this issue, just use the tip from @pfleidi

I use Coda2 and i had issues even when i set encoding in the editor. But the shell is the right place to correct.

[imathis]

@dearprakash I've seen times where Coda2 has issues with UTF8. Recently I helped someone over Twitter who resolved that his encoding problems were because Coda2 wasn't respecting the setting and was outputting mixed content.

@PHironaka if I were you, I'd use Markdown. My experience with Textile hasn't been too good. It's really hard to extend compared to Markdown and features like backtick code blocks have been a real pain to get working in Textile. Also there's not a lot of community support for Textile as Markdown seems to be what everyone is developing around, building Markdown support in lots of writing applications (and GitHub's comment system). I think Textile is on the way out.

[PHironaka]

Thank you @imathis and @dearprakash for the help! I'm guessing it's just an error in my _config.yml file. Is there some sort of validation tool to check for errors in YAML?

[imathis]

@PHironaka a tip: check make sure your strings are quoted. Some unescaped characters freak out the parser.

[itsPG]

Thanks, this helped me a lot

[jeremyckahn]

If Ubuntu users are also having this issue, this thread helped me figure it out.

I'm having another issue that is preventing me from generating my site, but I believe that issue is unrelated.

[zlx]

@pfleidi Perfect, can not thanks any more

[peterwillcn]

just add export RUBYOPT="-KU -E utf-8:utf-8" into ~/.bash_profile

[jiangxiaopeng]

@pfleidi thanks very much!

[zosiu]

@pfleidi thanks!