BLAG

NAME

blag - blag 1.4.1

blag is a blog-aware, static site generator, written in Python. An example "deployment" can be found here.

blag is named after the blag of the webcomic xkcd.

FEATURES

	•		Write content in Markdown
	•		Theming support using Jinja2 templates
	•		Generation of Atom feeds for blog content
	•		Fenced code blocks and syntax highlighting using Pygments
	•		Integrated devserver
	•		Available on PyPI

blag runs on Linux, Mac and Windows and requires Python >= 3.8

MANUAL

Quickstart

Install blag from PyPI

$ pip install blag

Run blag's quickstart command to create the configuration needed

$ blag quickstart

Create some content

$ mkdir content
$ edit content/hello-world.md

Generate the website

$ blag build

By default, blag will search for content in content and the output will be generated in build. All markdown files in content will be converted to html, all other files (i.e. static files) will be copied over).

If you want more separation between the static files and the markdown content, you can put all static files into the static directory. Blag will copy them over to the build directory.

If you want to customize the looks of the generated site, create a template directory and put your jinja2 templates here.

Those directories can be changed via command line arguments. See

$ blag --help

Manual

Pages and Articles

Internally, blag differentiates between pages and articles. Intuitively, pages are simple pages and articles are blog posts. The decision whether a document is a page or an article is made depending on the presence of the date metadata element: Any document that contains the date metadata element is an article, everything else a page.

This differentiation has consequences:

	•		blag uses different templates: page.html and article.html
	•		only articles are collected in the Atom feed
	•		only articles are aggregated in the tag pages

blag does not enforce a certain directory structure for pages and articles. You can mix and match them freely or structure them in different directories. blag will mirror the structure found in the content directory

content/
article1.md
article2.md
page1.md

results in:

build/
article1.html
article2.html
page1.html

Arbitrary complex structures are possible too:

content/
posts/
2020/
2020-01-01-foo.md
2020-02-01-foo.md
pages/
foo.md
bar.md

results in:

build/
posts/
2020/
2020-01-01-foo.html
2020-02-01-foo.html
pages/
foo.html
bar.html

Static Files

Static files can be put into the content directory and will be copied over to the build directory as well. If you want better separation between content and static files, you can create a static directory and put the files there. All files and directories found in the static directory will be copied over to build.

content/
foo.md
bar.md
kitty.jpg

results in:

build/
foo.html
bar.html
kitty.jpg

Alternatively:

content/
foo.md
bar.md
static/
kitty.jpg

results in:

build/
foo.html
bar.html
kitty.jpg

Internal Links

In contrast to most other static blog generators, blag will automatically convert relative markdown links. That means you can link you content using relative markdown links and blag will convert them to html automatically. The advantage is that your content tree in markdown is consistent and self-contained even if you don't generate html from it.

[...]
this is a [link](foo.md) to an internal page foo.

becomes

<p>this is a <a href="foo.html">link</a> to an internal page foo.</p>

Templating

Custom templates are optional and stored by default in the templates directory. blag will search the templates directory first, and fall back to blag's default built-in templates.

If you make use of Jinja2's template inheritance, you can of course have more template files in the templates directory.

site

This dictionary contains the site configuration, namely: base_url, title, description and author. Don't confuse the site-title and -description with the title and description of individual pages or articles.

content

HTML, converted from markdown.

meta

meta stands for all metadata elements available in the article or page. Please be aware that those are not wrapped in a dictionary, but directly available as variables.

archive

A list of [destination path, context] tuples, where the context are the respective variables that would be provided to the individual page or article.

	tags		List of tags.
	tag		A tag.

Metadata

blag supports metadata elements in the markdown files. They must come before the content and should be separated from the content with a blank line:

title: foo
date: 2020-02-02
tags: this, is, a, test
description: some subtitle

this is my content.
[...]

blag supports arbitrary metadata in your documents, and you can use them freely in you templates. However, some metadata elements are treated special:

	date		If a document contains the date element, it is treated as an article, otherwise as a page. Additionally, date elements are expected to be in ISO format (e.g. 1980-05-05 21:58). They are automatically converted into datetime objects with the local timezone attached.
	tags		Tags are interpreted as a comma separated list. All elements are stripped and converted to lower-case: tags: foo, Foo Bar, BAZ becomes: [foo, foo bar, baz].

Tags in articles are also used to generate the tag-pages, that aggregate all articles per tag.

title and description

The title and description are used in the html header and in the atom feed.

Devserver

blag provides a devserver which you can use for local web-development. The devserver provides a simple web server, serving your site in http://localhost:8000 and will automatically rebuild the project when it detects modifications in one of the content, static and templates directories.

$ blag serve

API

blag.__init__

blag.__init__ = <method-wrapper '__init__' of module object>

Initialize self. See help(type(self)) for accurate signature.

blag.version

blag.blag

blag's core methods.

Functions

blag.blag.build(args: argparse.Namespace) -> None

Build the site.

This is blag's main method that builds the site, generates the feed etc.
Parameters

args --

blag.blag.environment_factory(template_dir: Optional[str] = None, globals_:
Optional[dict[str, object]] = None) -> jinja2.environment.Environment

Environment factory.

Creates a Jinja2 Environment with the default templates and additional templates from template_dir loaded. If globals are provided, they are attached to the environment and thus available to all contexts.
Parameters

	•		template_dir -- directory containing the templates
	•		globals --

Return type

jinja2.Environment

blag.blag.generate_archive(articles: list[tuple[str, dict[str, Any]]],
template: jinja2.environment.Template, output_dir: str) -> None

Generate the archive page.
Parameters

	•		articles -- List of articles. Each article has the destination path and a dictionary with the content.
	•		template --
	•		output_dir --

blag.blag.generate_feed(articles: list[tuple[str, dict[str, Any]]],
output_dir: str, base_url: str, blog_title: str, blog_description: str,
blog_author: str) -> None

Generate Atom feed.
Parameters

	•		articles -- list of relative output path and article dictionary
	•		output_dir -- where the feed is stored
	•		base_url -- base url
	•		blog_title -- blog title
	•		blog_description -- blog description
	•		blog_author -- blog author

blag.blag.generate_tags(articles: list[tuple[str, dict[str, Any]]],
tags_template: jinja2.environment.Template, tag_template:
jinja2.environment.Template, output_dir: str) -> None

Generate the tags page.
Parameters

	•		articles -- List of articles. Each article has the destination path and a dictionary with the content.
	•		tags_template --
	•		tag_template --
	•		output_dir --

blag.blag.get_config(configfile: str) -> configparser.SectionProxy

Load site configuration from configfile.
Parameters

configfile -- path to configuration file

Return type

configparser.SectionProxy

blag.blag.main(arguments: Optional[list[str]] = None) -> None

Main entrypoint for the CLI.

This method parses the CLI arguments and executes the respective commands.
Parameters

arguments -- optional parameters, used for testing

blag.blag.parse_args(args: Optional[list[str]] = None) ->
argparse.Namespace

Parse command line arguments.
Parameters

args -- optional parameters, used for testing

Return type

arparse.Namespace

blag.blag.process_markdown(convertibles: list[tuple[str, str]], input_dir:
str, output_dir: str, page_template: jinja2.environment.Template,
article_template: jinja2.environment.Template) -> tuple[list[tuple[str,
dict[str, Any]]], list[tuple[str, dict[str, Any]]]]

Process markdown files.

This method processes the convertibles, converts them to html and saves them to the respective destination paths.

If a markdown file has a date metadata field it will be recognized as article otherwise as page.
Parameters

	•		convertibles -- relative paths to markdown- (src) html- (dest) files
	•		input_dir --
	•		output_dir --
	•		page_template -- templats for pages and articles
	•		archive_template -- templats for pages and articles

Returns

articles and pages

Return type

list[tuple[str, dict[str, Any]]], list[tuple[str, dict[str, Any]]]

blag.markdown

Markdown Processing.

This module contains the methods responsible for blag's markdown processing.

Functions

Classes

class blag.markdown.MarkdownLinkExtension(**kwargs)

markdown.extension that converts relative .md- to .html-links.
__module__ = 'blag.markdown'
extendMarkdown(md: markdown.core.Markdown) -> None

Add the various processors and patterns to the Markdown Instance.

This method must be overridden by every extension.

Keyword arguments:

•

md: The Markdown instance.

class blag.markdown.MarkdownLinkTreeprocessor(md=None)

Converts relative links to .md files to .html
__module__ = 'blag.markdown'
convert(url: str) -> str
run(root: xml.etree.ElementTree.Element) ->
xml.etree.ElementTree.Element

Subclasses of Treeprocessor should implement a run method, which takes a root ElementTree. This method can return another ElementTree object, and the existing root ElementTree will be replaced, or it can modify the current tree and return None.

blag.markdown.convert_markdown(md: markdown.core.Markdown, markdown: str)
-> tuple[str, dict[str, str]]

Convert markdown into html and extract meta data.
Some meta data is treated special:

	•		date is converted into datetime with local timezone
	•		tags is interpreted as a comma-separeted list of strings. All strings are stripped and converted to lower case.

Parameters

	•		md -- the Markdown instance
	•		markdown -- the markdown text that should be converted

Returns

html and metadata

Return type

str, dict[str, str]

blag.markdown.markdown_factory() -> markdown.core.Markdown

Create a Markdown instance.

This method exists only to ensure we use the same Markdown instance for tests as for the actual thing.
Return type

markdown.Markdown

blag.devserver

Development Server.

This module provides functionality for blag's development server. It automatically detects changes in certain directories and rebuilds the site if necessary.

Functions

blag.devserver.autoreload(args: argparse.Namespace) -> NoReturn

Start the autoreloader.

This method monitors the given directories for changes (i.e. the last modified time). If the last modified time has changed, a rebuild is triggered.

A rebuild is also performed immediately when this method is called to avoid serving stale contents.
Parameters

args -- contains the input-, template- and static dir

blag.devserver.get_last_modified(dirs: list[str]) -> float

Get the last modified time.

This method recursively goes through dirs and returns the most recent modification time time found.
Parameters

dirs -- list of directories to search

Returns

most recent modification time found in dirs

Return type

float

blag.devserver.serve(args: argparse.Namespace) -> None

Start the webserver and the autoreloader.
Parameters

args -- contains the input-, template- and static dir

blag.quickstart

Helper methods for blag's quickstart command.

Functions

blag.quickstart.get_input(question: str, default: str) -> str

Prompt for user input.

This is a wrapper around the input-builtin. It will show the default answer in the prompt and -- if no answer was given -- use the default.
Parameters

	•		question -- the question the user is presented
	•		default -- the default value that will be used if no answer was given

Returns

the answer

Return type

str

blag.quickstart.quickstart(args: argparse.Namespace | None) -> None

Quickstart.

This method asks the user some questions and generates a configuration file that is needed in order to run blag.
Parameters

args -- not used

	•		Index
	•		Module Index
	•		Search Page

AUTHOR

Bastian Venthur

COPYRIGHT

2022, Bastian Venthur