DOCS/tools/kanban_extract.py

"""Extraction unique : kanban.html -> DOCS/kanban/ + kanban-template.html.

Une carte = un fichier, un « fait recemment » = un fichier. Ensuite kanban.html
est regenere par build_kanban.py et ne s'edite plus a la main.
"""
import io
import json
import os
import re
import unicodedata

DOCS = os.path.dirname(os.path.dirname(os.path.abspath(__file__)))
SRC = os.path.join(DOCS, "kanban.html")
KANBAN = os.path.join(DOCS, "kanban")
CARDS_DIR = os.path.join(KANBAN, "cards")
DONE_DIR = os.path.join(KANBAN, "done")
BOARD_OUT = os.path.join(KANBAN, "board.json")
TPL_OUT = os.path.join(DOCS, "kanban-template.html")


def write(path, text):
    io.open(path, "w", encoding="utf-8", newline="\n").write(text)


def slug(text, maxlen=45):
    text = re.sub(r"<[^>]+>", "", text)
    text = unicodedata.normalize("NFKD", text).encode("ascii", "ignore").decode()
    text = re.sub(r"[^a-zA-Z0-9]+", "-", text).strip("-").lower()
    return text[:maxlen].rstrip("-") or "sans-titre"


html = io.open(SRC, encoding="utf-8", newline="").read()

masthead = re.search(r'<header class="masthead">(.*?)</header>', html, re.S).group(1)
intro = [p.strip() for p in re.findall(r"<p>(.*?)</p>", masthead, re.S)]
stamp = re.search(r'<div class="stamp">(.*?)</div>', masthead, re.S).group(1).strip()

summary_block = re.search(r'<section class="summary".*?</section>', html, re.S).group(0)
stats = re.findall(r'<span class="n([^"]*)">(\d+)</span><span class="k">([^<]*)</span>', summary_block)
stat_class = {k: cls.strip() for cls, n, k in stats}

done_start = html.index('<section class="done">')
board_start = html.index('<div class="board">')
board_end = html.rindex("</section>", 0, done_start)
board_end = html.index("</div>", board_end) + len("</div>")
board = html[board_start:board_end]

columns = []
n_cards = 0
for pos, col in enumerate(
    re.finditer(
        r'<!--\s*([^>]*?)\s*-->\s*<section class="col" style="--stripe: var\(--([\w-]+)\)">\s*'
        r'<div class="col-head"><h2>([^<]*)</h2><span class="count">\d+</span></div>\s*'
        r'<div class="stack">(.*?)</div>\s*</section>',
        board,
        re.S,
    ),
    start=1,
):
    comment, stripe, title, stack = col.groups()
    dirname = "%d-%s" % (pos, slug(title))
    os.makedirs(os.path.join(CARDS_DIR, dirname), exist_ok=True)

    for rank, card in enumerate(re.finditer(r'<article class="card"([^>]*)>(.*?)</article>', stack, re.S), start=1):
        attrs, inner = card.groups()
        horizon = re.search(r'data-horizon="([^"]*)"', attrs)
        meta = re.search(r'<div class="card-meta">(.*?)</div>', inner, re.S).group(1)
        flag = re.search(r'<span class="flag f-(\w+)">(.*?)</span>', meta, re.S)
        card_title = re.search(r"<h3>(.*?)</h3>", inner, re.S).group(1).strip()

        front = ["title: " + card_title, "area: " + re.search(r'data-area="([^"]*)"', attrs).group(1)]
        if horizon:
            front.append("horizon: " + horizon.group(1))
        front.append("tags: " + ", ".join(re.findall(r'<span class="tag">(.*?)</span>', meta, re.S)))
        if flag:
            front.append("flag: %s | %s" % (flag.group(1), flag.group(2).strip()))
        front.append("src: " + re.search(r'<span class="src">(.*?)</span>', inner, re.S).group(1).strip())

        body = "\n".join("<p>%s</p>" % p.strip() for p in re.findall(r"<p>(.*?)</p>", inner, re.S))
        write(
            os.path.join(CARDS_DIR, dirname, "%03d-%s.md" % (rank * 10, slug(card_title))),
            "---\n" + "\n".join(front) + "\n---\n" + body + "\n",
        )
        n_cards += 1

    columns.append(
        {"dir": dirname, "title": title, "stripe": stripe, "statClass": stat_class.get(title, ""), "comment": comment}
    )

done_section = html[done_start : html.index("</section>", html.index('<div class="done-grid">', done_start))]
os.makedirs(DONE_DIR, exist_ok=True)
n_done = 0
for rank, item in enumerate(re.finditer(r'<div class="done-item">\s*(.*?)\s*</div>', done_section, re.S), start=1):
    inner = item.group(1)
    title = re.search(r"<strong>(.*?)</strong>", inner, re.S).group(1).strip()
    body = inner[inner.index("</strong>") + len("</strong>") :].strip()
    write(
        os.path.join(DONE_DIR, "%03d-%s.md" % (rank * 10, slug(title))),
        "---\ntitle: " + title + "\n---\n" + body + "\n",
    )
    n_done += 1

board_meta = {
    "updated": re.search(r"Mise à jour · ([\d-]+)", stamp).group(1),
    "stampNote": stamp.split("<br>")[-1].strip(),
    "intro": intro,
    "columns": columns,
    "done": {
        "heading": re.search(r"<h2>(.*?)</h2>", done_section, re.S).group(1).strip(),
        "statLabel": next((k for k in stat_class if "Fait" in k), "Fait récemment"),
        "statClass": next((v for k, v in stat_class.items() if "Fait" in k), "n-good"),
        "intro": re.search(r"<p>(.*?)</p>", done_section, re.S).group(1).strip(),
    },
}
write(BOARD_OUT, json.dumps(board_meta, ensure_ascii=False, indent=2) + "\n")

template = (
    html[: html.index('<header class="masthead">')]
    + '<header class="masthead">\n    <div>\n      '
    + re.search(r"<h1>.*?</h1>", masthead, re.S).group(0)
    + "\n{{INTRO}}\n"
    + '    <div class="stamp">\n      {{STAMP}}\n    </div>\n  </header>\n\n'
    + "{{SUMMARY}}\n\n"
    + html[html.index('<div class="filters"') : board_start]
    + "{{BOARD}}\n\n  {{DONE}}\n"
    + html[html.index("</section>", html.index('<div class="done-grid">')) + len("</section>") :]
)
write(TPL_OUT, template)

print("cartes : %d   fait-recemment : %d" % (n_cards, n_done))
for c in columns:
    print("  %-14s %2d" % (c["title"], len(os.listdir(os.path.join(CARDS_DIR, c["dir"])))))
print("board.json %d o   template %d o" % (os.path.getsize(BOARD_OUT), os.path.getsize(TPL_OUT)))