#!/usr/bin/env python3
#
# RSS integration for Zulip
#

import argparse
import calendar
import errno
import hashlib
import logging
import os
import re
import sys
import time
import urllib.parse
from html.parser import HTMLParser
from typing import Any, Dict, List, Tuple

import feedparser

import zulip

VERSION = "0.9"  # type: str
RSS_DATA_DIR = os.path.expanduser(os.path.join("~", ".cache", "zulip-rss"))  # type: str
OLDNESS_THRESHOLD = 30  # type: int

usage = """Usage: Send summaries of RSS entries for your favorite feeds to Zulip.

This bot requires the feedparser module.

To use this script:

1. Create an RSS feed file containing 1 feed URL per line (default feed
   file location: ~/.cache/zulip-rss/rss-feeds)
2. Subscribe to the stream that will receive RSS updates (default stream: rss)
3. create a ~/.zuliprc as described on https://zulip.com/api/configuring-python-bindings
4. Test the script by running it manually, like this:

/usr/local/share/zulip/integrations/rss/rss-bot

You can customize the location on the feed file and recipient stream, e.g.:

/usr/local/share/zulip/integrations/rss/rss-bot --feed-file=/path/to/my-feeds --stream=my-rss-stream

4. Configure a crontab entry for this script. A sample crontab entry for
processing feeds stored in the default location and sending to the default
stream every 5 minutes is:

*/5 * * * * /usr/local/share/zulip/integrations/rss/rss-bot"""

parser = zulip.add_default_arguments(
    argparse.ArgumentParser(usage)
)  # type: argparse.ArgumentParser
parser.add_argument(
    "--stream",
    dest="stream",
    help="The stream to which to send RSS messages.",
    default="rss",
    action="store",
)
parser.add_argument(
    "--data-dir",
    dest="data_dir",
    help="The directory where feed metadata is stored",
    default=os.path.join(RSS_DATA_DIR),
    action="store",
)
parser.add_argument(
    "--feed-file",
    dest="feed_file",
    help="The file containing a list of RSS feed URLs to follow, one URL per line",
    default=os.path.join(RSS_DATA_DIR, "rss-feeds"),
    action="store",
)
parser.add_argument(
    "--unwrap",
    dest="unwrap",
    action="store_true",
    help="Convert word-wrapped paragraphs into single lines",
    default=False,
)
parser.add_argument(
    "--math",
    dest="math",
    action="store_true",
    help="Convert $ to $$ (for KaTeX processing)",
    default=False,
)

opts = parser.parse_args()  # type: Any


def mkdir_p(path: str) -> None:
    # Python doesn't have an analog to `mkdir -p` < Python 3.2.
    try:
        os.makedirs(path)
    except OSError as e:
        if e.errno == errno.EEXIST and os.path.isdir(path):
            pass
        else:
            raise


try:
    mkdir_p(opts.data_dir)
except OSError:
    # We can't write to the logfile, so just print and give up.
    print("Unable to store RSS data at %s." % (opts.data_dir,), file=sys.stderr)
    exit(1)

log_file = os.path.join(opts.data_dir, "rss-bot.log")  # type: str
log_format = "%(asctime)s: %(message)s"  # type: str
logging.basicConfig(format=log_format)

formatter = logging.Formatter(log_format)  # type: logging.Formatter
file_handler = logging.FileHandler(log_file)  # type: logging.FileHandler
file_handler.setFormatter(formatter)

logger = logging.getLogger(__name__)  # type: logging.Logger
logger.setLevel(logging.DEBUG)
logger.addHandler(file_handler)


def log_error_and_exit(error: str) -> None:
    logger.error(error)
    logger.error(usage)
    exit(1)


class MLStripper(HTMLParser):
    def __init__(self) -> None:
        super().__init__()
        self.reset()
        self.fed = []  # type: List[str]

    def handle_data(self, data: str) -> None:
        self.fed.append(data)

    def get_data(self) -> str:
        return "".join(self.fed)


def strip_tags(html: str) -> str:
    stripper = MLStripper()
    stripper.feed(html)
    return stripper.get_data()


def compute_entry_hash(entry: Dict[str, Any]) -> str:
    entry_time = entry.get("published", entry.get("updated"))
    entry_id = entry.get("id", entry.get("link"))
    return hashlib.md5((entry_id + str(entry_time)).encode()).hexdigest()


def unwrap_text(body: str) -> str:
    # Replace \n by space if it is preceded and followed by a non-\n.
    # Example: '\na\nb\nc\n\nd\n' -> '\na b c\n\nd\n'
    return re.sub("(?<=[^\n])\n(?=[^\n])", " ", body)


def elide_subject(subject: str) -> str:
    MAX_TOPIC_LENGTH = 60
    if len(subject) > MAX_TOPIC_LENGTH:
        subject = subject[: MAX_TOPIC_LENGTH - 3].rstrip() + "..."
    return subject


def send_zulip(entry: Any, feed_name: str) -> Dict[str, Any]:
    body = entry.summary  # type: str
    if opts.unwrap:
        body = unwrap_text(body)

    content = "**[%s](%s)**\n%s\n%s" % (
        entry.title,
        entry.link,
        strip_tags(body),
        entry.link,
    )  # type: str

    if opts.math:
        content = content.replace("$", "$$")

    message = {
        "type": "stream",
        "sender": opts.zulip_email,
        "to": opts.stream,
        "subject": elide_subject(feed_name),
        "content": content,
    }  # type: Dict[str, str]
    return client.send_message(message)


try:
    with open(opts.feed_file) as f:
        feed_urls = [feed.strip() for feed in f.readlines()]  # type: List[str]
except OSError:
    log_error_and_exit("Unable to read feed file at %s." % (opts.feed_file,))

client = zulip.Client(
    email=opts.zulip_email,
    api_key=opts.zulip_api_key,
    config_file=opts.zulip_config_file,
    site=opts.zulip_site,
    client="ZulipRSS/" + VERSION,
)  # type: zulip.Client

first_message = True  # type: bool

for feed_url in feed_urls:
    feed_file = os.path.join(opts.data_dir, urllib.parse.urlparse(feed_url).netloc)  # Type: str

    try:
        with open(feed_file) as f:
            old_feed_hashes = {
                line.strip(): True for line in f.readlines()
            }  # type: Dict[str, bool]
    except OSError:
        old_feed_hashes = {}

    new_hashes = []  # type: List[str]
    data = feedparser.parse(feed_url)  # type: feedparser.parse

    for entry in data.entries:
        entry_hash = compute_entry_hash(entry)  # type: str
        # An entry has either been published or updated.
        entry_time = entry.get(
            "published_parsed", entry.get("updated_parsed")
        )  # type: Tuple[int, int]
        if (
            entry_time is not None
            and (time.time() - calendar.timegm(entry_time)) > OLDNESS_THRESHOLD * 60 * 60 * 24
        ):
            # As a safeguard against misbehaving feeds, don't try to process
            # entries older than some threshold.
            continue
        if entry_hash in old_feed_hashes:
            # We've already seen this. No need to process any older entries.
            break
        if (not old_feed_hashes) and (len(new_hashes) >= 3):
            # On a first run, pick up the 3 most recent entries. An RSS feed has
            # entries in reverse chronological order.
            break

        feed_name = data.feed.title or feed_url  # type: str

        response = send_zulip(entry, feed_name)  # type: Dict[str, Any]
        if response["result"] != "success":
            logger.error("Error processing %s" % (feed_url,))
            logger.error(str(response))
            if first_message:
                # This is probably some fundamental problem like the stream not
                # existing or something being misconfigured, so bail instead of
                # getting the same error for every RSS entry.
                log_error_and_exit("Failed to process first message")
        # Go ahead and move on -- perhaps this entry is corrupt.
        new_hashes.append(entry_hash)
        first_message = False

    with open(feed_file, "a") as f:
        for hash in new_hashes:
            f.write(hash + "\n")

    logger.info("Sent zulips for %d %s entries" % (len(new_hashes), feed_url))