Empty _site generated when using utf-8 characters

Question

Empty _site generated when using utf-8 characters

mmozuras opened this issue 13 years ago · 13 comments

Steps to reproduce:

1 Step one:

mkdir jekyll
cd jekyll
mkdir _layouts

2 Create index.html with this content:


---
layout: default
title: Test

---
<div></div>

3 Create _config.yml with this content:

markdown: rdiscount
auto: true
permalink: pretty

4 Create default.html in _layouts with this content:

<!doctype html>
<!--[if lt IE 7 ]> <html lang="en" class="no-js ie6"> <![endif]-->
<!--[if IE 7 ]>    <html lang="en" class="no-js ie7"> <![endif]-->
<!--[if IE 8 ]>    <html lang="en" class="no-js ie8"> <![endif]-->
<!--[if IE 9 ]>    <html lang="en" class="no-js ie9"> <![endif]-->
<!--[if (gt IE 9)|!(IE)]><!--> <html lang="en" class="no-js"> <!--<![endif]-->
<head>
  <meta charset="utf-8">
  <meta http-equiv="Content-Type" content="text/html;charset=utf-8" />
  <title>{{ page.title }}</title>
</head>
    <body>
        <h1>Mindaugas Mozūras</h1>
    </body>
</html>

5 Run:

jekyll --server

6 Observe that _site is completely empty.
7 Delete 'Mozūras' from default layout. Repeat step 5 and find that this time _site is not empty.

I'm using:

MacVim/vim. Made sure it doesn't add any BOM characters.
Ruby 1.9.2p290
Jekyll 0.11.2
RDiscount 1.6.8

I also added this to my ~/.profile:

LANG=en_US.UTF-8
LC_ALL=en_US.UTF-8

GitHub Pages also failed to generate new version of my site when I added an utf-8 character.

What am I missing?
Thanks in advance for your help.

Answer 1 · 2012-02-05T20:59:21.000Z

That is really weird. Are you sure that _site does not contain index.html and that this isn't a problem with your browser?

P.S. I noticed that you declared the character set twice. You only need to declare it once (<meta charset="utf-8" />).

Answer 2 · 2012-02-05T21:55:41.000Z

Yes, I'm sure. I encountered this issue on a slightly bigger site, this is just a repro example. At first I was jus surprised that my site is not updating with new content (jekyll doesn't delete old _site by default). Then I looked at generated content and was surprised to find old content. I was even more surprised after deleting _site and finding absolutely no content in _site. I needed some time to find what was causing it, googling and (unsuccessfully) trying a couple of things. After that I came here and posted this issue.

P.S. Forgot to delete it from example (I didn't expect that to do anything, but I tried it anyway).

Answer 3 · 2012-02-06T00:41:02.000Z

If you delete the content in _site of course it wouldn't appear in there. The only thing I can think of at this point is to reinstall Jekyll.

Answer 4 · 2012-02-06T06:25:58.000Z

What I meant - I delete _site, run jekyll --server, empty _site gets created. I remove all utf-8 characters, delete _site, run jekyll --server, non-empty _site gets created.
Also, I have the same result with GitHub Pages.

Answer 5 · 2012-02-06T08:42:09.000Z

run with --no-auto option you will see error message.

If message not clear try following: edit Gemfile

gem "jekyll", :git => "git://github.com/stereobooster/jekyll.git"

run

bundle install
bundle exec jekyll

Also check: there is no BOM

Answer 6 · 2012-02-06T18:23:10.000Z

Here's the error message:

/Users/mindaugasmozuras/.rbenv/versions/1.9.2-p290/lib/ruby/gems/1.9.1/gems/jekyll-0.11.2/lib/jekyll/convertible.rb:29:in `read_yaml': invalid byte sequence in US-ASCII (ArgumentError)
    from /Users/mindaugasmozuras/.rbenv/versions/1.9.2-p290/lib/ruby/gems/1.9.1/gems/jekyll-0.11.2/lib/jekyll/layout.rb:31:in `initialize'
    from /Users/mindaugasmozuras/.rbenv/versions/1.9.2-p290/lib/ruby/gems/1.9.1/gems/jekyll-0.11.2/lib/jekyll/site.rb:113:in `new'
    from /Users/mindaugasmozuras/.rbenv/versions/1.9.2-p290/lib/ruby/gems/1.9.1/gems/jekyll-0.11.2/lib/jekyll/site.rb:113:in `block in read_layouts'
    from /Users/mindaugasmozuras/.rbenv/versions/1.9.2-p290/lib/ruby/gems/1.9.1/gems/jekyll-0.11.2/lib/jekyll/site.rb:111:in `each'
    from /Users/mindaugasmozuras/.rbenv/versions/1.9.2-p290/lib/ruby/gems/1.9.1/gems/jekyll-0.11.2/lib/jekyll/site.rb:111:in `read_layouts'
    from /Users/mindaugasmozuras/.rbenv/versions/1.9.2-p290/lib/ruby/gems/1.9.1/gems/jekyll-0.11.2/lib/jekyll/site.rb:97:in `read'
    from /Users/mindaugasmozuras/.rbenv/versions/1.9.2-p290/lib/ruby/gems/1.9.1/gems/jekyll-0.11.2/lib/jekyll/site.rb:38:in `process'
    from /Users/mindaugasmozuras/.rbenv/versions/1.9.2-p290/lib/ruby/gems/1.9.1/gems/jekyll-0.11.2/bin/jekyll:250:in `<top (required)>'
    from /Users/mindaugasmozuras/.rbenv/versions/1.9.2-p290/bin/jekyll:19:in `load'
    from /Users/mindaugasmozuras/.rbenv/versions/1.9.2-p290/bin/jekyll:19:in `<main>'

29th line in convertible.rb:

      if self.content =~ /^(---\s*\n.*?\n?)^(---\s*$\n?)/m

Answer 7 · 2012-02-06T19:53:24.000Z

Again: check there is no BOM. Post your file.

Answer 8 · 2012-02-06T20:15:15.000Z

No BOM. I understand why you're suggesting it - I tried setting BOM (:set bomb with Vim) and deleting utf-8 character ('ū'), I get exactly the same error.

Answer 9 · 2012-02-06T21:37:25.000Z

One more idea. You set

LANG=en_US.UTF-8
LC_ALL=en_US.UTF-8

Try locale which corresponds to this letter 'ū'

Answer 10 · 2012-02-06T22:48:41.000Z

That was not it, but thanks to you I found what I missed. I'm using zsh, which doesn't load .profile. I was not aware of that and until now I stupidly never bothered to run locale to check if I actually change LANG/LC_ALL.

Thanks and sorry for wasting everyone's valuable time.

Answer 11 · 2012-02-06T23:53:41.000Z

It's okay. I still love you. :D

Answer 12 · 2012-03-09T01:50:17.000Z

Just want to say that I had the exact same problem, and this thread was a big help. Thanks! :)

Answer 13 · 2012-08-03T23:50:50.000Z

Thanks for this issue I could make jekyll work in my zsh!