path: root/bin/feed

#!/usr/bin/env python3
#
#   SPDX-License-Identifier: ISC
#
#   Copyright © 2019 Free Software Foundation of India.
#

import html
import datetime
import os
import os.path
import re
import sys

import mistune

URL = "http://fsf.org.in"
SECTIONS = ["news", "article"]

F_PH = {
    "name": "<!-- Feed Name -->",
    "link": "<!-- Feed Link -->",
    "updated": "<!-- Feed Updated -->",
    "entries": "<!-- Entries -->",
}

E_PH = {
    "id": "<!-- Entry Id -->",
    "title": "<!-- Entry Title -->",
    "link": "<!-- Entry Link -->",
    "updated": "<!-- Entry Updated -->",
    "content": "<!-- Entry Content -->",
}


def err(s):
    print("Error: {}".format(s))
    sys.exit(1)


def fok(f):
    p = os.path.basename(f.path)

    if re.search(r"(^[\.\#])|(~$)", p):
        return False

    return True


def files(sec):
    files = os.scandir("md" + "/" + sec)

    fs = []
    for f in files:
        if not fok(f):
            print("Ignoring {}".format(f.path))
        else:
            fs.append(f)

    return fs


def read(f):
    with open(f) as f:
        c = f.read()
    return c


def write(p, c):
    d = os.path.dirname(p)

    if not os.path.exists(d):
        os.makedirs(d)

    with open(p, "w") as f:
        f.write(c)


def slug(p):
    m = re.search(r"(([a-zA-Z\-]+)(\.([a-z]{2}))?)\.md", p)

    if not m:
        err("Unable to get slug")

    return m.group(1), m.group(2), m.group(4)


def template(type):
    return read("templates/atom/{}.atom".format(type))


def title(c):
    m = re.search(r"^\# (.+)$", c, re.M)

    if not m:
        err("Title not found")

    return m.group(1)


def elink(sec, s, l):
    if l is None:
        return "/".join([URL, sec, s])
    else:
        return "/".join([URL, sec, s, l])


def flink(sec):
    return "/".join([URL, sec, "feed.atom"])


def time(c):
    m = re.search(r"pubdate: ([0-9]{8})", c)

    if not m:
        err("Publication date not found")

    d = m.group(1)
    d = datetime.datetime.strptime(d, "%Y%m%d").strftime("%Y-%m-%d")

    return d + "T00:00:00Z"


def markdown(c):
    try:
        r = mistune.markdown(c, False, parse_block_html=True, parse_inline_html=True)

    except Exception as e:
        err("Markdown parsing failed for {}".format(e))

    return r


def massage(c):
    c = html.escape(c)
    c = c.replace("\n", "&#xA;")
    c = re.sub(r" +", " ", c)

    return c


def content(c):
    m = re.search(r"^\# (.+)$", c, re.M)

    if not m:
        err("Unable to slurp content")

    c = c[m.end() :]
    c = markdown(c)

    return massage(c)


def now():
    n = datetime.datetime.today()

    return n.strftime("%Y-%m-%dT%H:%M:%SZ")


def entry(sec, f):
    c = read(f.path)
    u, s, l = slug(f.path)

    t = time(c)
    id = t + ":" + u

    e = template("entry")
    e = e.replace(E_PH["id"], id, 1)
    e = e.replace(E_PH["title"], title(c), 1)
    e = e.replace(E_PH["link"], elink(sec, s, l), 1)
    e = e.replace(E_PH["updated"], t, 1)
    e = e.replace(E_PH["content"], content(c), 1)

    return id, e


def esort(esd):
    ids = sorted(esd.keys(), reverse=True)

    es = []
    for id in ids:
        es.append(esd[id])

    return es


def feed(sec, es):
    f = template("feed")
    f = f.replace(F_PH["name"], sec, 2)
    f = f.replace(F_PH["link"], flink(sec), 1)
    f = f.replace(F_PH["updated"], now(), 1)
    f = f.replace(F_PH["entries"], es, 1)

    return f


def process(sec):
    esd = {}
    for f in files(sec):
        id, e = entry(sec, f)
        esd[id] = e

    es = "".join(esort(esd))

    write("/".join(["_build", sec, "feed.atom"]), feed(sec, es))


def run():
    for sec in SECTIONS:
        process(sec)


if __name__ == "__main__":
    run()