hyde / hyde-old

Static website generator inspired by Jekyll

Version 0.8.9

Overview

Hyde starter kit by merlinrebrovic is a really nice way to get started with hyde.

Hyde layout for bootstrap by auzigog is also a good alternative if you like Twitter's bootstrap framework.

You can also take a look at Hyde Powered Websites for inspiration and reference.

Currently hyde is only supported on python 2.7.x. Python 3.x support is in progress, thanks to jonafato.

Active development of hyde had stalled for some time, but thanks to llonchj and jonafato there is now a slow but steady progress towards 1.0.

Mailing List: hyde-dev on google groups. IRC: #hyde on freenode

Installation

To get the latest released version:

pip install hyde

For the current trunk:

pip install -e git://github.com/hyde/hyde.git#egg=hyde

Creating a new hyde site

The following command:

hyde -s ~/test_site create

will create a new hyde site using the test layout.

Generating the hyde site

cd ~/test_site
hyde gen

Serving the website

cd ~/test_site
hyde serve
open http://localhost:8080

Publishing the website

cd ~/test_site
hyde publish -p github

Hyde supports extensible publishers.

Github

The hyde documentation is published to github pages using this command with the following configuration:

publisher:
    github:
        type: hyde.ext.publishers.dvcs.Git
        path: ../hyde.github.com
        url: [email protected]:hyde/hyde.github.com.git

Note

Currently, the initial path must have clone of the repository already in place for this command to work.

PyFS

Hyde also has a publisher that acts as a frontend to the awesome PyFS library (thanks to rfk). Here are a few configuration options for some PyFS backends:

publisher:
    zip:
        type: hyde.ext.publishers.pyfs.PyFS
        url: zip://~/deploy/hyde/docs.zip
    s3:
        type: hyde.ext.publishers.pyfs.PyFS
        url: s3://hyde/docs
    sftp:
        type: hyde.ext.publishers.pyfs.PyFS
        url: sftp:hydeuser:[email protected]

Note

PyFS is not installed with hyde. In order to use the PyFS publisher, you need to install pyfs separately.

Any PyFS dependencies (Example: boto for S3 publishing) need to be installed separately as well.

pip install fs
pip install boto

To get additional help on PyFS backends, you can run the following command once PyFS is installed:

fsls --listopeners

Examples

1. julien.danjou.info
2. luffy.cx
3. Cloudpanic
4. Hyde Documentation Source

A brief list of features

1. Evented Plugins: The Plugin hooks allow plugins to listen to events that occur during different times in the lifecycle and respond accordingly.
2. Metadata: Hyde now supports hierarchical metadata. You can specify and override variables at the site, node or the page level and access them in the templates.
3. Organization: The sorter, grouper and tagger plugins provide rich meta-data driven organizational capabilities to hyde sites.
4. Publishing: Hyde sites can be published to variety of targets including github pages, Amazon S3 & SFTP.

Links

1. Changelog
2. Authors

hyde-old's Issues

add cssmin to requirements.txt

With a clean environment, you need to install cssmin in order to generate the site. This can be done easily now that we telling users to install using pip.

Hyde does not work with markdown2

See http://github.com/lakshmivyas/hyde/issues/closed#issue/11/comment/127370 for background.

enable notification on linux with pynotify/libnotify

by loooking carefully at the default settings.py I noticed a setting to enable growl notification.

so i added a few lines here and there to enable notification via pynotify (which uses libnotify)

This is not really an important feature ;-)

one can use PYNOTIFY = True in settings.py and install pynotify
--- a/hydeengine/init.py
+++ b/hydeengine/init.py
@@ -25,6 +25,13 @@ from file_system import File, Folder
from path_util import PathUtil
from processor import Processor
from siteinfo import SiteInfo
+try:
+ import pynotify
+ if not pynotify.init('Hyde'):
+ PYNOTIFY = False
+ PYNOTIFY = True
+except ImportError:
+ PYNOTIFY = False

 class _HydeDefaults:

@@ -248,6 +255,8 @@ class Generator(object):
                 subprocess.call([settings.GROWL, "-n", "Hyde", "-t", title, "-m", message])        
             except:
                 pass    
+        if hasattr(settings, "PYNOTIFY") and settings.PYNOTIFY and PYNOTIFY:
+            pynotify.Notification(title, message, None).show()

     def pre_process(self, node):
         self.processor.pre_process(node)

None is for the icon as there is none for hyde ???

Site is not generated under Windows

Steps to reproduce:

1. python hyde.py -i -s c:\mysite
2. python hyde.py -g s c:\mysite
   Infinite loop starts and process never ends

I fixed it changing hardcoded slashes to os.sep in url.py:
return (parent.rstrip("/") + os.sep + child.lstrip("/")).rstrip("/")
and
url = url.strip(os.sep)

Feature request: Faster site generation by compiling only modified source files

If the generator kept a list of compiled files, and only processed new and newer-modified files, the generation time after any small modification to the source files would be greatly reduced.

issue with hyde tags <!-- Hyde::...::Begin--> and restructured text (and possibly textile and asciidoc)

restructuredtext does not recognise html statement and treat them like text. you have to include them in a raw paragraph like that:

.. raw:: html
   

so that there are outputted like html.

so this cause problem when you use _restructuredtex_t and you have {% article %} or {%excerpt %} tag in the middle of a {% restructuredtext %} block because all the html comment HYDE tag are rendered in html.
For example:

{% block article %}
{% restructuredtext %}
{% article %}
{% excerpt %}
This is a test
{% endexcerpt %}
This should show how article and excerpt (django) tags normally rendered in html comment are shown in plain html while inside restructuredtext block
{% endarticle %}
{% endrestructuredtext %}
{% endblock article %}

is rendered in html as :

<p>&lt;!-- Hyde::Article::Begin&nbsp;--&gt;</p> 
<p>&lt;!-- Hyde::Excerpt::Begin&nbsp;--&gt;</p> 
<p>This is a test
&lt;!-- Hyde::Excerpt::End --&gt;
This should show how article and excerpt (django) tags normally rendered in html   comment are shown in plain html while inside restructuredtext block
&lt;!-- Hyde::Article::End&nbsp;--&gt;</p> 

one have to use {% restruturedtext %} block at lowest level possible and use article and excerpt blocks on the outer level of restructured text block.

i guess that the same problem could be find with textitle and asciidoc if they can't recognize html statement and do treat them like text.

Add support for injecting other material(text, Ad's etc) into generated RSS feeds.

hyde --init loses current directory or something similar

When running hyde --init, the files and directories are created in the current directory, but not visible before cd'ing out and into directory again. Presumably, the current directory is somehow lost or there is some other problem with file system object creation.

I can provide an strace if needed, but I can't seem to find a way to attach files here so I won't do so now.

excerpt property not working

Testing the presence of a post's excerpt is not working currently.

To reproduce the bug:

1. mkdir bugexcerpt && cd bugexcerpt
2. create a new hyde site hyde -i -t default
3. generate and webserve: hyde -g -s . && hyde -w -s .
4. browsing to http://localhost:8080/blog/blog.html won't show the excerpt although in the template there is a test on page.excerpt to display it when true.

The render_article tag fetch the entire page body from the start article block

I use the render_article on one template and unfortunately the entire content is fetch from to the end of the page and not

Why is pip needed?

pip seems new and fairly rare, compared to easy_install which I already have installed. I haven’t been able to figure out how to actually install pip.

I know I should just suck it up, and I’m going to plow through this, but I’m not sure it’s quite the friendliest beginning of my relationship with hyde.

using restructedtext failed with exception

hi.
i simply changed index.html to include some tests. i.e. using snippets on the main page of hyde http://github.com/lakshmivyas/hyde

so trying to use restructedtext i paste

 I love templates. 
{% restructuredtext %}
Render this **content all in reStructuredText**.

     Writing in reStructuredText is also quicker than
     writing in HTML.

#. Or at least that is my opinion.
#. What about you?

{% endrestructuredtext %}

and when i run

$ ./hyde.py -g -s /tmp/test/

INFO:Pre processing /tmp/test[2010-01-28 22:03:56,732]
INFO:Processing /index.html[2010-01-28 22:03:56,733]
***********************
Error while rendering page /index.html
***********************
Generation Failed
(, AttributeError("class _HydeDefaults has no attribute 'RST_SETTINGS_OVERRIDES'",), <traceback object at 0x8e68b6c>)

moreover some things have been quickly written.
because line 73 of ./hydeengine/templatetags/aym.py must be
print u"Requires docutils library to use reStructuredText tag."

i am using docutils-0.6

if i comment line 76-77 of ./hydeengine/templatetags/aym.py, it seems to work.

issue with render_excerpt

hi.
i just used the default template, and from the code of content/blog/blog.html and _listing.html and _innerlisting.html, i expect to see some excerpts in deploy/blog/blog.html once i run ./hyde.py -g
but there is none.

in fact, if i look at _layout/skeleton/innerlisting.html, this seems wrong

{% if page.excerpt %}
    {% render_excerpt list_page %}
{%endif%}

shouldn't it be {% if list_page.excerpt %} ? but this does not work either. in fact list_page does not seems to have excerpt_title or excerpt_url either. niether page.

the only way to make it work is to simple remove the if test keeping only

{% render_excerpt list_page %}

and this, at last, works as expected.

However, by looking at hydeengine/templatetags/hydetags.py, i have easily spotted an error:
line 170:

@register.tag(name="render_article")    
def render_excerpt(parser, token):

should be

@register.tag(name="render_article")    
def render_article(parser, token):

but this does not seems to cause problem at all...

HTML tags encoded in page.variable string

I abstracted some data points to page variables, so I can easily change the presentation of a large table: change the text in these vars, style that text, substitute an image for the text, etc.

When a page.var contains only a simple string, all is good, but any HTML I include in the string is entity-encoded before being inserted into the output. Not what I expected. I've tried escaping the affected chars, but no luck.

How can I get Hyde to leave the contents of these string variables alone? Thanks.

LQ

Markdown Extension Documentation

Markdown extensions add a fair bit of complexity and must be documented in the wiki.

Archives generator should not modify the source folder

Category archives should be generated in the deploy_tmp folder and later moved to the deploy folder.

Update version on ringce website.

http://ringce.com/hyde

Still lists hyde as version 0.4

directory constants should be independent of settings.py location

When you specify LAYOUT_DIR, CONTENT_DIR, and MEDIA_DIR values that are not in the same directory with settings.py, you get errors during generate. For instance, I want the root directory of my project to contain two directories: src and deploy, deploy being a git ignored throwaway directory, src containing everything I need to generate deploy. To accomplish this, I first tried to put a settings.py at the root of my project with directory settings like this:

ROOT_PATH = os.path.abspath(os.path.dirname(file))
SRC_DIR = os.path.join(ROOT_PATH,'src')
LAYOUT_DIR = os.path.join(SRC_DIR, 'layout')
CONTENT_DIR = os.path.join(SRC_DIR, 'content')
MEDIA_DIR = os.path.join(SRC_DIR, 'media')
DEPLOY_DIR = os.path.join(ROOT_PATH, 'deploy')
TMP_DIR = os.path.join(ROOT_PATH, 'deploy_tmp')

This yielded an error on line 58 of processor.py, in extract_processors:
AttributeError: 'SiteNode' object has no attribute 'fragment'

Next, I put settings.py at the root of my src directory, and was able to generate deploy as a sibling of src like this:

ROOT_PATH = os.path.abspath(os.path.dirname(file))
PROJ_PATH, whatevs = os.path.split(ROOT_PATH)
LAYOUT_DIR = os.path.join(ROOT_PATH, 'layout')
CONTENT_DIR = os.path.join(ROOT_PATH, 'content')
MEDIA_DIR = os.path.join(ROOT_PATH, 'media')
DEPLOY_DIR = os.path.join(PROJ_PATH, 'deploy')
TMP_DIR = os.path.join(PROJ_PATH, 'deploy_tmp')

but I wasn't super happy with settings.py being there, so i tried creating a src/build directory and putting settings.py in there, using these settings:

ROOT_PATH, foo = os.path.split(os.path.abspath(os.path.dirname(file)))
PROJ_PATH, bar = os.path.split(ROOT_PATH)
LAYOUT_DIR = os.path.join(ROOT_PATH, 'layout')
CONTENT_DIR = os.path.join(ROOT_PATH, 'content')
MEDIA_DIR = os.path.join(ROOT_PATH, 'media')
DEPLOY_DIR = os.path.join(PROJ_PATH, 'deploy')
TMP_DIR = os.path.join(PROJ_PATH, 'deploy_tmp')

This yields a recursion overflow during some kind of directory traversal in siteinfo.py, line 521.

I admit I could just be doing it wrong - but my experiences here seem to indicate that there is some dependency on python.py being in the same directory with the layout, media, and deploy directories, which I wouldn't think there should be. It prevents, for instance, projects that use a standard Maven directory structure.

One extra bit of information, in case it's useful: in each of the failures mentioned above, the deploy and deploy_temp directories got created successfully in the desired place before the script crashed.

Thanks so much for such an awesome project! I'm happy to have found it.

AttributeError with specific _atom.xml

I'm using the following code for my _atom.xml (https://gist.github.com/755575), however I get the following error when generating my site.

***********************
Error while rendering page /feed/
***********************
Generation Failed
(<type 'exceptions.AttributeError'>, AttributeError("'str' object has no attribute 'walk_pages'",), <traceback object at 0x0217E990>)

Paths with spaces silently fail when using YUI Compressor (Mac OS X)

If the path to the file to be compressed contains a space on Mac OS X, the YUI compressor step silently fails because the shell can't accept spaces in paths.

Replacing line 61 in media_processors.py with the code below fixes this, but there's surely a more graceful way to do this (additionally, the problem will affect any other command-line arguments used by Hyde, like those for HSS, and this step might not be necessary for all OSes or shells; I don't know much beyond bash, myself).

u"java -jar %s %s > %s" % (compress.replace(' ', '\ '), 
    resource.source_file.path.replace(' ', '\ '),
    tmp_file.path.replace(' ', '\ ')
))

At the very minimum, Hyde should print an error to the command line when generating a site that has spaces in its path to let the user know that YUI has failed to compress the file and why.

Templates reference the wrong yuicompressor version

The templates point to version 2.4.1 of the YUI-Compressor, but the included version is 2.4.2.

The affected files are:
templates/test/settings.py:YUI_COMPRESSOR = "./lib/yuicompressor-2.4.1.jar"
templates/simple/settings.py:YUI_COMPRESSOR = "./lib/yuicompressor-2.4.1.jar"
templates/default/settings.py:YUI_COMPRESSOR = "./lib/yuicompressor-2.4.1.jar"

Files in the content dir that start with and '_' should not be processed at all

As far as I understood files in the content dir that start with an underscore are supposed to be complelety ignored. In fact they won't be copied to the deploy folder. However, when adding hyde attributes to an underscore-prefixed file they are processed wich is not intuitive.

Here's an example:

_template.html

{% extends "_post.html" %}
{%hyde
    title: Title
    created: 2010-01-01
    updated: 2010-01-01
    categories: [dummy1, dummy2]
    tags: []
%}
{% block excerpt %}

Put here the excerpt / introduction. This is going to be included in the real
post, too.

{% endblock %}
{% block article %}

Put here the main content

{% endblock %}

When using the category manager it lists the categories dummy1 and dummy2. Although you can easily circumvent the issue by omitting any categories in _template.html it is probably conceptionally cleaner if underscore-prefixed files would be totally ignored.

Syntax in combination with textile

Hi,

I can't seem to get syntax highlighting working with textile.

If I use the syntax tag inside a textile tag, the syntax highlighted html gets escaped. How should I use the syntax tag so it works?

Thanks.

What is the correct way to create contextual navigation for within a module rather than just a node?

When creating contextual navigation for a page, the wiki says to use page.next and page.prev to get the next and previous pages within the node that the page belongs to. But I have a blog layout like this:

blog
    - 2010
        - 9
            - 6
                - article1.html
        - 10
            - 9
                - article2.html

And I need the contextual navigation for article1 to point its next article to article2. page.next will give me null because there are no other articles in the blog/2010/9/6 node.

How should I go about doing this? Thanks!

On Mac Os X 10.4, ImportError: No module named markdown

I had to install markdown with easy_install. This should be precised in the doc, I guess !

Thanks for this very useful software,

Charles

Feature Request: Git-style CLI + plugin interface

I'd like Hyde to take Git/webgen style commands as well, e.g. hyde.py init and hyde.py serve.

This could also be a beginning of a plugin system - hyde.py foo could try to run a script called foo.py in a plugins directory inside the main Hyde directory.

Add support for related items

Essentially this would mean adding support for Categories/Tags. Can be integrated as a part of hyde by directly supporting Categories property on Page and a method for retrieving related pages in SiteNode or as a content processor.

Context variables not inherited?

Context variables defined using the {% hyde %} tag are not being inherited by an extending template. Is that by design, or is it a bug, or am I doing something wrong?

What is the hyde 0.8 equivalent of generate_clean_urls?

The 0.4/0.5-era directions for GENERATE_CLEAN_URLS don’t seem to apply to v0.8. What’s to be done?

Non-root site not browseable under hyde.py webserver

I have a site that sits at my home directory, not the server root. This works fine after deployment with the settings:

SITE_ROOT = "/~sdavis/"
SITE_WWW_URL = "http://watson.nci.nih.gov/"

However, when I use the webserver to preview the site, I end up with the content sitting at the server root but the URLs being generated with the SITE_ROOT included. It seems to me that the cherrypy webserver should be serving on the complete URL and not at the server root when SITE_ROOT is not "/" but is something else. Perhaps I have missed something, though.

I am using hyde version 0.4.

few question about hyde

first of all , I'm very newbie for python and hyde.
I can running the hyde on mac osx with out big problem, just few small questions :
1 . how I can render {% markdown %} as plant text, for situation I wanna explain how to use hyde
2. for {% syntax %} , seems doesn't work with php or I missing some thing ? I check the pagments have php lexer .

Document the proper way to exclude an .htaccess file

I have a .htaccess file at the root of my content directory to rewrite some old URLs.

I also have a feed.html file to serve http://stevelosh.com/feed/ as an atom feed. In feed.xml I use {% for node_page in page.node.walk_pages %} to walk over all the posts to find the content to serve in the feed.

The problem is that the feed then includes the .htaccess file.

I tried adding a {% hyde %} tag in the .htaccess file in sjl/stevelosh@ef485be but it did not work. It was just passed through and rendered in the deploy/.htaccess file, and Apache does not like that at all.

I ended up just checking the filename in the atom template in sjl/stevelosh@5c09be3 for now, but that seems like a hacky solution. Why doesn't the hyde tag in the .htaccess file get processed (and removed)?

Site renders on Windows 7 when I remove the png file form prerendered. Some issue with utf8 ?

INFO:Processing /prerendered/hyde-icon.png[2010-12-26 17:04:00,822]

---

Error while rendering page /prerendered/hyde-icon.png

---

Generation Failed
(<type 'exceptions.UnicodeDecodeError'>, UnicodeDecodeError('utf8', '\x89PNG\n',
0, 1, 'invalid start byte'), <traceback object at 0x02103CB0>)

Site Pre/Post Processors on Windows

The pre and post site processors look like they're not being used on Windows, even when using them in the settings.py

Compass support?

I see Hyde works with sass, but does it work with compass? I can’t see it being very hard, given how dead simple it is to use Compass on the command line.

YAML tags fail when their text contains a colon

If I have this at the top of one of my pages:

{%hyde
     title: Python: An Interactive Tutorial
%}

It will fail because of the colon after "Python". How am I supposed to generate that particular title?

Execution of TemplateProcessor as ordinary CONTENT_PROCESSOR

If I understood the code of processor.py correctly, TemplateProcessor.process(resource) is executed explicitly for all documents in the content/-tree. I think it is more logical not to do so, but to add TemplateProcessor to the default CONTENT_PROCESSOR list.

This would e.g. enable you to limit TemplateProcessing to specific file extensions or subfolders of the content/-directory.

What do you think?

Point me in the right direction

This isn't really an issue.

I'm looking to be pointed in the right direction regarding two customizations I'd like to make.

1. How to change the order of navigation items.
2. How to have the index display the content of a static page, like about.html instead of the latest blog entry.

Regards,
-SL

Support for RSS feeds per category.

Strange issue with markdown and BracketNode's usage of HTML comments

I'm having a strange problem with recent versions of Markdown (2.0.3 or git-head) giving up on processing posts/articles when they contain HTML comments produced by BracketNode ( <!-- Hyde::Excerpt::Begin --> for example). What happens is that some or all of the markdown syntax to be processed isn't processed. This happens even when markdown is run externally from Hyde, with the command line markdown_py (sometimes just markdown) command. Maybe Hyde could consider some other way of bracketing content for future versions, as it would appear that this HTML comment approach also has issues in other text processors (see other issues).

My super-nasty hackish patch to work around this is here:
https://github.com/hugovincent/hyde/commit/bb9638312712352f9bd47e3603e2cf186658264e

It would be great if a better solution could be arrived at.

issue with yuicompressor and clevercss

hi.
i was trying to use a clevercss template in media/css
and after ./hyde.py -g, i noticed only a .ccss and a .ccss.z-tmp in my deploy/media/css

it seems the command running the yuicompressor failed because the file deploy/media/css/base.ccss.z-tmp contains:

Usage: java -jar yuicompressor-x.y.z.jar [options] [input file]

Global Options
  -h, --help                Displays this information
  --type            Specifies the type of the input file
  --charset        Read the input file using 
  --line-break      Insert a line break after the specified column number
  -v, --verbose             Display informational messages and warnings
  -o                  Place the output into . Defaults to stdout.

JavaScript Options
  --nomunge                 Minify only, do not obfuscate
  --preserve-semi           Preserve all semicolons
  --disable-optimizations   Disable all micro optimizations

If no input file is specified, it defaults to stdin. In this case, the 'type'
option is required. Otherwise, the 'type' option is required only if the input
file extension is neither 'js' nor 'css'.

it seems to fail because the input file has .ccss extension
one need to use --type css if the extension is not .css

there seems to be an issue with the pipe processing of media processor.

base.ccss is converted to base.ccss by clevercss (wouldn't it be better to be base.css), then yuicompressor use base.ccss but fails because the extension is not .js or .css.

also, if i remove the yuicompressor from the pipe for ccss in my settings.py, i end up with a base.ccss in my deploy/media/css, that file containing css and not clevercss, and then should have been called base.css.

render_excerpt fails (inc patch)

The following issue:

INFO:Processing /blog/[2010-02-21 22:09:24,468]
***********************
Error while rendering page /blog/
***********************
Generation Failed
(<type 'exceptions.AttributeError'>, AttributeError("'SafeUnicode' object has no attribute 'render'",), <traceback object at 0x1013c9f38>)

Was fixed with the following patch:
diff --git a/hydeengine/templatetags/hydetags.py b/hydeengine/templatetags/hydetags.py
index aec60ed..88a33e4 100644
--- a/hydeengine/templatetags/hydetags.py
+++ b/hydeengine/templatetags/hydetags.py
@@ -39,7 +39,7 @@ def excerpt(parser, token):
return BracketNode("Excerpt", nodelist)

 @register.tag(name="article")
-def excerpt(parser, token):
+def article(parser, token):
     nodelist = parser.parse(('endarticle',))
     parser.delete_first_token()
     return BracketNode("Article", nodelist)
@@ -63,8 +63,6 @@ class LatestExcerptNode(template.Node):

     def render(self, context):
         sitemap_node = None
-        if not self.words == 50:
-            self.words = self.words.render(context)
         self.path = self.path.render(context).strip('"')
         sitemap_node = context["site"].find_node(Folder(self.path))
         if not sitemap_node:
@@ -154,7 +152,7 @@ def latest_excerpt(parser, token):
     if len(tokens) > 1:
         path = Template(tokens[1])
     if len(tokens) > 2:
-        words = Template(tokens[2])
+        words = int(tokens[2])
     return LatestExcerptNode(path, words)

 @register.tag(name="render_excerpt")    
@@ -165,7 +163,7 @@ def render_excerpt(parser, token):
     if len(tokens) > 1:
         path = parser.compile_filter(tokens[1])
     if len(tokens) > 2:
-        words = Template(tokens[2])
+        words = int(tokens[2])
     return RenderExcerptNode(path, words)

 @register.tag(name="render_article")    
@@ -182,8 +180,6 @@ class RenderExcerptNode(template.Node):
         self.words = words

     def render(self, context):
-        if not self.words == 50:
-            self.words = self.words.render(context)
         page = self.page.resolve(context)
         context["excerpt_url"] = page.url
         context["excerpt_title"] = page.title

"date" filter doesn't support 'B' - locale's full month name (e.g., January)

Tag: {{ post.created|date:"B" }}
fails with:
Generation Failed
(<type 'exceptions.NotImplementedError'>, NotImplementedError(), <traceback object at 0x1c30b90>)

Support styles(CSS) in RSS/Atom Feeds

0.5.3 is broken

Downloads ❯ hyde.py -i -s test
Traceback (most recent call last):
File "/usr/local/bin/hyde.py", line 5, in <module>
    pkg_resources.run_script('hyde==0.5.3', 'hyde.py')
File "/System/Library/Frameworks/Python.framework/Versions/2.6/Extras/lib/python/pkg_resources.py", line 442, in run_script
    self.require(requires)[0].run_script(script_name, ns)
File "/System/Library/Frameworks/Python.framework/Versions/2.6/Extras/lib/python/pkg_resources.py", line 1160, in run_script
    execfile(script_filename, namespace, namespace)
File "/Library/Python/2.6/site-packages/hyde-0.5.3-py2.6.egg/EGG-INFO/scripts/hyde.py", line 110, in <module>
    main()
File "/Library/Python/2.6/site-packages/hyde-0.5.3-py2.6.egg/EGG-INFO/scripts/hyde.py", line 71, in main
    options.template, options.force_init)
File "/Library/Python/2.6/site-packages/hyde-0.5.3-py2.6.egg/hydeengine/__init__.py", line 579, in initialize
    "Cannot find the specified template[%s]." % template_dir)
ValueError: Cannot find the specified template[/Library/Python/2.6/site-packages/hyde-0.5.3-py2.6.egg/EGG-INFO/scripts/templates/default].

ImportError: No module named cherrypy when launching web server (add easy_install cherrypy to doc)

You may probably add "sudo easy_install cherrypy" to the Basic Installation section

Add 'cssmin' to requirements.txt

$ python hyde.py  -g -s ~/blog -d ~/blog-deploy 
INFO:Pre processing /Users/user/blog[2011-02-05 16:51:40,922]
INFO:Processing /index.html[2011-02-05 16:51:40,923]
INFO:Processing /about/about.html[2011-02-05 16:51:41,385]
INFO:Processing /archives/firstpost.html[2011-02-05 16:51:41,413]
INFO:Processing /archives/archives.html[2011-02-05 16:51:41,426]
INFO:Processing /media/css/base.css[2011-02-05 16:51:41,438]
Generation Failed
Traceback (most recent call last):
  File "/Users/sridharr/code/hyde/hydeengine/__init__.py", line 417, in process_all
    self.process(resource)
  File "/Users/sridharr/code/hyde/hydeengine/__init__.py", line 377, in process
    return self.processor.process(item)
  File "/Users/sridharr/code/hyde/hydeengine/processor.py", line 111, in process
    processor.process(resource)
  File "/Users/sridharr/code/hyde/hydeengine/media_processors.py", line 102, in process
    import cssmin
ImportError: No module named cssmin

Categories Manager will overwrite context['categories'] with the last processed node.

The line context['categories'] = categories assumes that the pre-processor will run for only one node. It should check if there are categories defined in the context already and merge if needed.

IncludeTextNode assumes markdown

The IncludeTextNode template tag assumes that markdown library as opposed to markdown2 library is used. It also does not factor in the Markdown extension settings.

Bug in new Categories Manager

There is a bug in the new categories manager. When clicking on a category on an archives page the category urls do not match the category names anymore.

For example, have a look at this website: http://www.poxd.org/blog/index.html

When hovering over the category links everything seems fine and the names match the links. However, when clicking on 'open source' and inspecting the categories on that page (http://www.poxd.org/archives/open+source.html) almost all categories link to http://www.poxd.org/archives/open+source.html although their correct name is shown.