Purebuilder Simply

Simplified website building system with Pandoc.

What is PureBuilder

PureBuilder is website building script.

What is ACCS

ACCS is site building script for serialized articles.

PB Simply ACCS make index.md with indexes build by PureBuilder.

Install

Install from rubygems.org

gem install pbsimply

Manually make Gem

• gem build pbsimply.gemspec
• gem install pbsimply-$version.gem

Manually from GitHub

• git clone https://github.com/reasonset/purebuilder-simply
• Copy under bin/ files on your PATH directory.
• Copy under lib/ files on your Ruby library directory.

Get start

• Make your document root
• Create and edit .pbsimply.yaml in your document root.
• Make directory and documents (Markdown or ReSTructured Text)
• Make template with pandoc -D html5 > template.html and edit it.

Dependency

• Ruby >= 3.0
• Pandoc >= 2.8

Usage

PureBuilder

Move your document root before running.

pbsimply directory

PureBuilder builds documents in the directory.

PureBuilder skip if filename start with draft- or ., or draft value in frontmatter is true.

Options

Option	Description
-f	Refresh all documents (force update mode.) This options useful when update template.
-I	Don't register to index database.
-A	Don't treat ACCS.
-o FILE	Output to FILE.
-m FILE	Additional meta (YAML) File.

Make ACCS index

pbsimply find ACCS documents directory automatically. You mark as "ACCS documents directory" with putting .accs.yaml file to the directory.

ACCS processor makes and puts index.html file.

Objects for eRuby in document

You can use these objects in eRuby on template or generated document.

@config

Loaded config YAML file (.pbsimply.yaml).

@indexes

Compiled index.

This object may be uncompleted when generating pbsimply.

@index

Document meta data that "default | indexed | frontmatter | current".

Index database

Documents metadata will be stored to index database on each document directory.

Ruby Marshal is used by default, the filename is .indexes.rbm.

If json or oj is set to dbstyle in configuration, JSON is used instead of Ruby Marshal and hhe filename is .indexes.json

Values in config.

Key	Type	Description
outdir	String	Base directory for output. Required
template	String	Path to Pandoc HTML template file. If not set, template is temaplte.html
css	String / Array	Path to CSS file(s)
toc	Boolean	Turn on TOC if true
pandoc_additional_options	Array	Extra pandoc options
post_eruby	Boolean	Process Pandoc output with eRuby if true
alt_frontmatter	Hash	Default frontmatter in ACCS index
default_meta	Hash	Default frontmatter
testserver_port	Fixnum	Port number of pbsimply-testserver (default 8000)
self_url_prefix	String	Absolute path prefix for URL of generated document itself instead of /.
self_url_external_prefix	String	Like self_url_prefix, but it only works on page_url_encoded_external.
dbstyle	String	If json given, Indexes database is used .indexes.json instead of .indexes.rbm. Also oj is same as it, but use Oj library instead of JSON standard library.
bless_style	String	If cmd given, blessing with command instead of Ruby procs.
bless_cmd	String / Array	Command for blessing.
bless_accscmd	String / Array	Command for ACCS blessing.
blessmethod_accs_rel	String	Automatic blessing method for find next/prev article.
accs_order	String	If desc is set, ACCS article list is sorted by descending order (on default template.)
accs_across_category	Boolean	Don't separate ACCS article lists by category (on default template.)
accs_sort_by	String	Sorting method for ACCS article list. default (date, title, last update), title (title, date), name (filename, title, date) and serial (serial, date, filename) are avilable. It works on default template.
auto_delete	Boolean	Delete output file when source file is deleted or turned to draft.
detect_modification	String	Detecting modification method. changes looks change changes header. mtimesize looks mtime and file size. Otherwise, it looks mtime.
pandoc_command	String	Pandoc command. pandoc is default.

Special values in @index

Key	Set by	Used by	Description
title	frontmatter	Pandoc/System	Document title. required.
author	frontmatter	Default template	Author.
date	frontmatter/system	System	Date of written
lang	frontmatter	Pandoc template	lang/xml:lang
keywords	frontmatter	Pandoc template	An array, used as keywords in meta tag.
description	frontmatter	Sample template	Used as description in meta tag.
draft	frontmatter	System	Draft status. Skip process document if true.
_last_proced	system	system	Integer. DateTime of last processed by PureBuilder. 0 if this document is processed first.
last_updated	system		String. DateTime of last processed by Pandoc.
_size	system		File size (byte)
_mtime	system		Integer. mtime of this file.
_filename	system		File name
_docformat	system		Document Format. Markdown or ReST.
categories	frontmatter	ACCS	Document category. Sort documents by this value.
pagetype	frontmatter/config	ACCS	Document type of this page. accsindex is set if processed by ACCS, set post by default.
source_directory	system		Source directory string.
source_file	system		Source Filename.
source_path	system		Source path string.
dest_path	system	system	Output file path.
normalized_docdir	system		Normalized source document directory path, begin with /.
normalized_docpath	system		Normalized source document path, begin with /.
page_url	system		This (generated) page's URL.
page_url_encoded	system		This (generated) page's URI encoded URL.
page_url_encoded_external	system		This (generated) page's URI encoded URL with self_url_external_prefix.
title_encoded	system		URI encoded document title.
timestamp	frontmatter	system	The date and time of the document which is more detailed than date.
timestamp_xmlschema	system		XML Schema formatted Timestamp. Use date instead of timestamp if timestamp isn't defined.
timestamp_jplocal	system		Japanese local formatted Timestamp. Use date instead of timestamp if timestamp isn't defined.
timestamp_rubytimestr	system		Ruby's Time#to_s like formatted Timestamp. Use date instead of timestamp if timestamp isn't defined.
timestamp_str	system		%Y-%m-%d[ %H:%M:%S %Z]. Use date instead of timestamp if timestamp isn't defined.

Environment variables

Environment variables that able to use in Pre Plugins, Post plugins, Hooks or Blessing command.

|Name|Pre|Process|Delete|Post|Bless|Description| |---------|---|---|---|--------------------| |pbsimply_outdir|Yes|Yes|Yes|Yes|Yes|Path for output directory root.| |pbsimply_subdir|Yes|Yes|Yes|Yes|Yes|Path for document directory from document root.| |pbsimply_indexes|Yes|Yes|Yes|Yes|Yes|Path for index database.| |pbsimply_frontmatter|Yes|Yes|Yes|Yes|Yes|Path for current document's frontmatter (JSON).| |pbsimply_working_dir|Yes|Yes|Yes|Yes|Yes|Temporary directory path for putting processing data.| |pbsimply_currentdoc|Yes|Yes|No|No|No|Temporary filepath for processing document.| |pbsimply_filename|Yes|Yes|No|No|No|Original source filename.|

Testing

CSS, image or link locations should not be local place, path to web URL, so you cannot test generated documents normally case of suppose to put to WWW.

You can test in case like it with pbsimply-testserver.

How to use is very simple.

1. Move your document root
2. run it.
3. access http://localhost:port

You can config port with testserver_port in config file. default is 8000.

If you think to put subdirectory like http://example.com/site/index.html, I recommend that you put document in site subdirectory, and sync under there.

Blessing

With Ruby

You can modify Frontmatter with Ruby script with .pbsimply-bless.rb file.

If you want to use it, you should PureBuilder::BLESS Proc object in the file. It will be called with PureBuilder::BLESS.call(frontmatter, self).

This proc will be called after system paramaters were set.

It don't need to return something. You can modify frontmatter Hash object directly.

When processing directory is an ACCS document directory, PureBuilder::ACCS::BLESS is also called after PureBuilder::BLESS if defined.

You can add keys and Proc values to PureBuilder::ACCS::DEFINTIONS Hash. They are used for setting special value.

Key	Function
:next	Set returned value to frontmatter["next_article"]
:prev	Set returned value to frontmatter["prev_article"]

For example, this is Chienomi's blessing script.

#!/usr/bin/ruby

load "./.lib/categories.rb"

TOPICPATH = {
  "" => ["TOP", "/"],
  "/articles" => ["Articles", "/#Category"],
  "/override" => ["Override", "/"],
  "/archives" => ["Old Archives", "/articlelist-wp.html"]
}

ARTICLE_CATS.each do |k,v|
  TOPICPATH[["/articles", k].join("/")] = [v, ["", "articles", k, ""].join("/")]
end

PureBuilder::BLESS = ->(frontmatter, pb) {
  content = nil
  filetype = nil
  content = File.read(frontmatter["source_path"])
  filetype = File.extname(frontmatter["_filename"])

  url = frontmatter["page_url"].sub(/^\.?\/?/, "/")
  frontmatter["topicpath"] = []
  url = url.split("/")
  (1 .. url.length).each do |i|
    path = url[0, i].join("/")
    if v = TOPICPATH[path]
      frontmatter["topicpath"].push({"title" => v[0], "url" => v[1]})
    else
      frontmatter["topicpath"].push({"title" => frontmatter["title"]})
      break
    end
  end

  if frontmatter["category"] && url.include?("articles")
    frontmatter["category_spec"] = [ARTICLE_CATS[url[-2]], frontmatter["category"]].join("::")
  end

  if content
    if((filetype == ".md" && content =~ %r:\!\[.*\]\(/img/thumb/:) || (filetype == ".rst" || filetype == ".rest") && content =~ %r!\.\. image:: .*?/img/thumb!)
      frontmatter["lightbox"] = true
    end
  end
}

article_order = nil
rev_article_order_index = {}

PureBuilder::ACCS::BLESS = -> (frontmatter, pb) {
  frontmatter["ACCS"] = true
  unless article_order
    article_order = pb.indexes.to_a.sort_by {|i| i[1]["date"]}
    article_order.each_with_index {|x,i| rev_article_order_index[x[0]] = i }
  end
}

PureBuilder::ACCS::DEFINITIONS[:next] = ->(frontmatter, pb) {
  index = rev_article_order_index[frontmatter["_filename"]] or next nil
  if article_order[index + 1]
    {"url" => article_order[index + 1][1]["page_url"],
     "title" => article_order[index + 1][1]["title"]}
  end
}

PureBuilder::ACCS::DEFINITIONS[:prev] = ->(frontmatter, pb) {
  index = rev_article_order_index[frontmatter["_filename"]] or next nil
  if index > 0
    {"url" => article_order[index - 1][1]["page_url"],
     "title" => article_order[index - 1][1]["title"]}
  end
}

With other language or programs

If cmd is set to bless_style in configuration, use external command instead of Ruby procs.

bless_cmd is command for blessing. bless_accscmd is command for ACCS blessing.

You can read document metadata from pbsimply-frontmatter.json file on directory $pbsimply_working_dir environment variable, and you can apply changes with write to the file.

Automatic blessing

Some configuration applies prepared blessing method.

ACCS Relation

blessmethod_accs_rel makes next_article and prev_article property. They are an associative array that have keys url and title.

numbering (heading number of filename), lexical (filename), date and timestamp are avilable.

Files

In this repository

Filename	Description
pbsimply-testserver	Satrt web server for testing.
accsindex.erb	Sample for .accsindex.erb. Normaly you don't need to edit template.
docroot-sample	For (part) copy to your document root.
template.html	Sample for PureBuilder template.
postgenerate	Sample files for post generate script.

You put or generated

Filename	Place	Description
.pbsimply.yaml	root	Configuration file. Put in document root.
.indexes.rbm	each	metadata Ruby marshal file generated by PureBuilder Simply.
.indexes.json	each	metadata JSON file generated by PureBuilder Simply.
.index.md	each	Index source generated by ACCS.
.accsindex.erb	root or each ACCS	Markdown eRuby template for ACCS index.
.accs.yaml	each	@index for the ACCS index.
.pbsimply-bless.rb	root	Ruby script for blessing.

Document Sample

Template

This template is associeted to basic blog theme.

I customized:

• Two Headers. #TopHeader for banner (default display: none), #TitleHeader for title.
• Main section is wrapped by #MainContainer section.
• Article body is wrapped by #MainArticle article.
• Side bar is available as #SideBar section. You can write content with include before function.
• Author section is deleted.
• Embedded CSS is deleted.
• Syntax highlightning theme is deleted. If you need it, get themes from pandoc-goodies or other.

CSS

Basic sample CSS.

file	description
layout.css	Layouting.
base.css	Basic design.
skylightning.css	Pandoc default Source code CSS
lightbox.css	Lightbox theme for lightbox plugin.

Post

ACCS template directory.

Post Generate

Sample Post Plugins.

Configuration file

.accsindex.rb and .pbsimply.yaml are avialable.

JavaScript

JavaScript Plugins.

See README on each file.

Themes

You can use sample Themes with merge content to copied sample document directory.

Base

Warm

Practical

Bloggy

Use a processor instead of Pandoc

Basics

PureBuilder Simply is a very powerful tool when used with Pandoc, but if you do not prefer Pandoc, you can use other document processors.

However, it is limited in its functionality.

The processor to use is specified with the value of pbsimply_processor in .pbsimply.yaml.

Processor	pbsimply_processor
RDoc	rdoc
RDoc/Markdown	rdoc_markdown
Kramdown	kramdown
Redcarpet	redcarpet
CommonMarker (cmark-gfm)	cmark

Also, for those templates that evaluate as eRuby templates, you can use the following values on the template (in most cases, use frontmatter and article_body.)

Variable name	Description
dir	Relative directory path from document root.
filename	Source file name.
frontmatter	Blessed metadata.
orig_filepath	Original source filepath.
procdoc	Actual (treated) source filepath.
article_body	Generated document body.

RDoc

Summery

It uses RDoc, the standard documentation system of Ruby. Source files are treated as RDoc, and the target is limited to *.rdoc files.

Templates are handled as eRuby templates.

Dependency

• rdoc library

Disabled configurations

• css
• toc
• pandoc_additional_options
• post_eruby

RDoc/Markdown

Summery

Use the Markdown processor of RDoc, Ruby's standard document system. Treat source files as if they were Markdown, limited to *.md files.

Templates are handled as eRuby templates.

Dependency

• rdoc library

Disabled configurations

• css
• toc
• pandoc_additional_options
• post_eruby

Kramdown

Summery

It is generated using Kramdown, Ruby's Markdown library. Treat source files as if they were Markdown, limited to *.md files.

Templates are handled as eRuby templates.

Dependency

• kramdown library

Disabled configurations

• css
• toc
• pandoc_additional_options
• post_eruby

Additional configurations

Key	Type	Description
kramdown_features	Hash	An associative array passed as the second argument to Kramdown::Document.new. See the [API documentation for details]. (https://kramdown.gettalong.org/rdoc/Kramdown/Options.html)

Redcarpet

Summery

It is generated using Redcarpet, Ruby's Markdown library. Treat source files as if they were Markdown, limited to *.md files.

Templates are handled as eRuby templates.

Dependency

• redcarpet library

Disabled configurations

• css
• toc
• pandoc_additional_options
• post_eruby

Additional configurations

Key	Type	Description
redcarpet_extensions	Hash	An associative array showing the extensions to Redcarpet. See the Redcarpet page for details.

Redcarpet

Summery

Generated using CommonMarker, a Ruby wrapper for libcmark-gfm. Source files are treated as Markdown, and the target is limited to *.md files.

The table and strikethrough extensions are enabled.

Templates are handled as eRuby templates.

Dependency

• libcmark-gfm
• commonmarker library

Disabled configurations

• css
• toc
• pandoc_additional_options
• post_eruby

Docutils

説明

Generated using Docutils ReSTructured Text processor written by Python. Source files are treated as ReSTructured Text, and the target is limited to *.rst files.

Dependency

• Docutils (rst2html5)

Disabled configurations

• toc
• pandoc_additional_options

Additional configurations

Key	Type	Description
docutils_options	Array	Command-line option arguments for rst2html5 command.

Hooks

Overview

You can tweak PureBuilder Simply behavior with hooks feature.

write .pbsimply-hooks.rb on document root to use Hooks. It should define PBSimply::Hooks.load_hooks.

It called with PBSimply::Hooks object. PBSimply::Hooks object has "timing" methods. You can add Proc to timing object with << method.

#!/bin/ruby

def (PBSimply::Hooks).load_hooks h
  h.process << ->(v) {
    db[v["normalized_docpath"]] = v
  }

  h.post << ->(v) {
    db.delete_if do |dbk, dbv|
      not File.exist? dbv["dest_path"]
    end
  }
end

Hook called with one argument. Ordinary it is a Hash, but keys and values are different between timing object.

Timing methods

Methods of argument of PBSimply::Hooks#load_hooks.

#add {|arg| ... }

Add block to timing object.

#<< proc

Add proc to timing object.

#cmd(*cmdarg)

Call command with system(*cmdarg).

You can modify content of $pbsimply_currentdoc on pre.

#filter(*cmdarg)

Call command with IO.popen(cmdarg, "w+").

Command is given document content from STDIN, and overwrite document content with command's output.

you can use this method only on pre.

Timing object

pre

PBSimply::Hooks#pre is called before processing document.

Argument are frontmatter and procdoc.

pre is called after BLESSing.

procdoc is processing document's temporary path. Its content same as source document but without frontmatter.

process

PBSimply::Hooks#process is called just after generating.

Arguments are frontmatter, procdoc, outpath.

frontmatter and procdoc are same as #pre, but after other process. outpath is output file path.

delete

PBSimply::Hooks#delete is called when document is lost include it turn to draft.

Arguments are target_file_path and source_file_path.

target_file_path is output file path (existing or non-existing.)

source_file_path is source file path (existing or non-existing.)

post

PBsimply::Hooks#post is called after all document processed.

Argument is this_time_processed.

this_time_processed is an Array of Hash actual processed documents in this time.

It has source (original source file path,) dest (output file path,) frontmatter.

accs

PBSimply::Hooks#accs is called before generating ACCS index.

Arguments are index and indexes. They are same as @index and @indexes in .accsindex.erb.