#!/usr/bin/env python3
import fnmatch
import locale
import os
from dataclasses import dataclass
from datetime import datetime, timedelta
from html import escape
from pathlib import Path
from textwrap import dedent
from time import perf_counter
import mistune
from jinja2 import Environment as Env
from jinja2 import FileSystemLoader
from minicli import cli, run, wrap
from mistune.directives import DirectiveInclude
from PIL import Image
from slugify import slugify
# Useful for dates rendering within Jinja2.
locale.setlocale(locale.LC_ALL, "fr_FR.UTF-8")
HERE = Path(".")
DAVID = HERE / "david"
STATIC = HERE / ".." / "larlet-fr-static"
DOMAIN = "https://larlet.fr"
LOCAL_DOMAIN = "http://larlet.test:3579"
# Hardcoding publication at 12 in Paris timezone.
NORMALIZED_STRFTIME = "%Y-%m-%dT12:00:00+01:00"
TODAY = datetime.today() + timedelta(hours=6)
class MarkParser(mistune.InlineParser):
"""Parses `==foo==` as `foo`."""
MARK = (
r"(\={2})(?=[^\s*])("
r"(?:\\[\\*]|[^*])*"
r"(?:" + mistune.InlineParser.ESCAPE + r"|[^\s*]))\1"
)
RULE_NAMES = mistune.InlineParser.RULE_NAMES + ("mark",)
def parse_mark(self, m, state):
marker = m.group(1)
text = m.group(2)
return "mark", self.render(text, state)
class MarkRenderer(mistune.HTMLRenderer):
"""To use in conjunction with `MarkParser`."""
def mark(self, text):
return "" + text + ""
class BlockquoteLanguageRenderer(MarkRenderer):
"""Sets the English language attribute for blockquotes with `[en]` prefix."""
def _get_language(self, text):
if text.startswith("
[en] "):
return "en", text.replace("
[en] ", "
")
else:
return None, text
def block_quote(self, text):
language, text = self._get_language(text)
if language:
return f'\n
\n{text}
\n'
else:
return f"\n
\n{text}
\n"
class ImgsWithSizesRenderer(BlockquoteLanguageRenderer):
"""Renders images as s and add sizes (useful for lazy loading)."""
def paragraph(self, text):
# In case of a figure, we do not want the (non-standard) paragraph.
if text.strip().startswith(""):
return text
return f"
{text}
\n"
def image(self, src, alt="", title=None):
full_path = STATIC / Path(src[1:])
image = Image.open(full_path)
width, height = image.size
return dedent(
f"""\
{title}
"""
)
class H2AnchorsRenderer(ImgsWithSizesRenderer):
"""Custom renderer for H2 titles with anchors."""
def heading(self, text, level):
if level == 2:
slug = slugify(text)
return (
f'
"
)
else:
return super().heading(text, level)
# We want a custom renderer to create a hash/link for each H2 headings.
markdown_with_h2_anchors = mistune.Markdown(
renderer=H2AnchorsRenderer(escape=False),
inline=MarkParser(H2AnchorsRenderer(escape=False)),
plugins=[DirectiveInclude()],
)
# The second markdown is pertinent to generate articles for the feed,
# we do not need anchors in that case.
markdown_with_img_sizes = mistune.Markdown(
renderer=ImgsWithSizesRenderer(escape=False),
inline=MarkParser(ImgsWithSizesRenderer(escape=False)),
plugins=[DirectiveInclude()],
)
# This is the jinja2 configuration to locate templates.
environment = Env(loader=FileSystemLoader(str(DAVID / "templates")))
def neighborhood(iterable, first=None, last=None):
"""
Yield the (previous, current, next) items given an iterable.
You can specify a `first` and/or `last` item for bounds.
"""
iterator = iter(iterable)
previous = first
current = next(iterator) # Throws StopIteration if empty.
for next_ in iterator:
yield (previous, current, next_)
previous = current
current = next_
yield (previous, current, last)
def each_markdown_from(source_dir, file_name="*.md"):
"""Walk across the `source_dir` and return the md file paths."""
for filename in fnmatch.filter(os.listdir(source_dir), file_name):
yield os.path.join(source_dir, filename)
@dataclass
class Page:
title: str
content: str
file_path: str
lang: str = "fr"
def __post_init__(self):
suffix = len(".md")
prefix = len("YYYY/MM-DD") + suffix
date_str = self.file_path[-prefix:-suffix].replace("-", "/")
self.url = f"/david/{date_str}/"
self.date = datetime.strptime(date_str, "%Y/%m/%d").date()
self.full_url = f"{DOMAIN}{self.url}"
self.normalized_date = self.date.strftime(NORMALIZED_STRFTIME)
self.escaped_title = escape(self.title)
self.escaped_content = escape(
self.content.replace('href="/', f'href="{DOMAIN}/')
.replace('src="/', f'src="{DOMAIN}/')
.replace('href="#', f'href="{self.full_url}#')
+ '