#!/usr/bin/env python

import argparse
import datetime
import logging
import os
import re
import subprocess
import sys
import tempfile
import time
from collections import defaultdict
from html import unescape
from urllib.error import URLError
from urllib.parse import quote, urlparse, urlsplit, urlunsplit
from urllib.request import urlretrieve

import dateutil.parser

# because logging.setLoggerClass has to be called before logging.getLogger
from pelican.log import init
from pelican.settings import DEFAULT_CONFIG
from pelican.utils import SafeDatetime, slugify

logger = logging.getLogger(__name__)


def decode_wp_content(content, br=True):
    pre_tags = {}
    if content.strip() == "":
        return ""

    content += "\n"
    if "<pre" in content:
        pre_parts = content.split("</pre>")
        last_pre = pre_parts.pop()
        content = ""
        pre_index = 0

        for pre_part in pre_parts:
            start = pre_part.find("<pre")
            if start == -1:
                content = content + pre_part
                continue
            name = f"<pre wp-pre-tag-{pre_index}></pre>"
            pre_tags[name] = pre_part[start:] + "</pre>"
            content = content + pre_part[0:start] + name
            pre_index += 1
        content = content + last_pre

    content = re.sub(r"<br />\s*<br />", "\n\n", content)
    allblocks = (
        "(?:table|thead|tfoot|caption|col|colgroup|tbody|tr|"
        "td|th|div|dl|dd|dt|ul|ol|li|pre|select|option|form|"
        "map|area|blockquote|address|math|style|p|h[1-6]|hr|"
        "fieldset|noscript|samp|legend|section|article|aside|"
        "hgroup|header|footer|nav|figure|figcaption|details|"
        "menu|summary)"
    )
    content = re.sub(r"(<" + allblocks + r"[^>]*>)", "\n\\1", content)
    content = re.sub(r"(</" + allblocks + r">)", "\\1\n\n", content)
    #    content = content.replace("\r\n", "\n")
    if "<object" in content:
        # no <p> inside object/embed
        content = re.sub(r"\s*<param([^>]*)>\s*", "<param\\1>", content)
        content = re.sub(r"\s*</embed>\s*", "</embed>", content)
        #    content = re.sub(r'/\n\n+/', '\n\n', content)
    pgraphs = filter(lambda s: s != "", re.split(r"\n\s*\n", content))
    content = ""
    for p in pgraphs:
        content = content + "<p>" + p.strip() + "</p>\n"
    # under certain strange conditions it could create
    # a P of entirely whitespace
    content = re.sub(r"<p>\s*</p>", "", content)
    content = re.sub(r"<p>([^<]+)</(div|address|form)>", "<p>\\1</p></\\2>", content)
    # don't wrap tags
    content = re.sub(r"<p>\s*(</?" + allblocks + r"[^>]*>)\s*</p>", "\\1", content)
    # problem with nested lists
    content = re.sub(r"<p>(<li.*)</p>", "\\1", content)
    content = re.sub(r"<p><blockquote([^>]*)>", "<blockquote\\1><p>", content)
    content = content.replace("</blockquote></p>", "</p></blockquote>")
    content = re.sub(r"<p>\s*(</?" + allblocks + "[^>]*>)", "\\1", content)
    content = re.sub(r"(</?" + allblocks + r"[^>]*>)\s*</p>", "\\1", content)
    if br:

        def _preserve_newline(match):
            return match.group(0).replace("\n", "<WPPreserveNewline />")

        content = re.sub(r"/<(script|style).*?<\/\\1>/s", _preserve_newline, content)
        # optionally make line breaks
        content = re.sub(r"(?<!<br />)\s*\n", "<br />\n", content)
        content = content.replace("<WPPreserveNewline />", "\n")
    content = re.sub(r"(</?" + allblocks + r"[^>]*>)\s*<br />", "\\1", content)
    content = re.sub(
        r"<br />(\s*</?(?:p|li|div|dl|dd|dt|th|pre|td|ul|ol)[^>]*>)", "\\1", content
    )
    content = re.sub(r"\n</p>", "</p>", content)

    if pre_tags:

        def _multi_replace(dic, string):
            pattern = r"|".join(map(re.escape, dic.keys()))
            return re.sub(pattern, lambda m: dic[m.group()], string)

        content = _multi_replace(pre_tags, content)

    # convert [caption] tags into <figure>
    content = re.sub(
        r"\[caption(?:.*?)(?:caption=\"(.*?)\")?\]"
        r"((?:\<a(?:.*?)\>)?(?:\<img.*?\>)(?:\<\/a\>)?)\s?(.*?)\[\/caption\]",
        r"<figure>\n\2\n<figcaption>\1\3</figcaption>\n</figure>",
        content,
    )

    return content


def _import_bs4():
    """Import and return bs4, otherwise sys.exit."""
    try:
        import bs4
    except ImportError:
        error = (
            'Missing dependency "BeautifulSoup4" and "lxml" required to '
            "import XML files."
        )
        sys.exit(error)
    return bs4


def file_to_soup(xml, features="xml"):
    """Reads a file, returns soup."""
    bs4 = _import_bs4()
    with open(xml, encoding="utf-8") as infile:
        xmlfile = infile.read()
    soup = bs4.BeautifulSoup(xmlfile, features)
    return soup


def get_filename(post_name, post_id):
    if post_name is None or post_name.isspace():
        return post_id
    else:
        return post_name


def wp2fields(xml, wp_custpost=False):
    """Opens a wordpress XML file, and yield Pelican fields"""

    soup = file_to_soup(xml)
    items = soup.rss.channel.findAll("item")
    for item in items:
        if item.find("status").string in ["publish", "draft"]:
            try:
                # Use HTMLParser due to issues with BeautifulSoup 3
                title = unescape(item.title.contents[0])
            except IndexError:
                title = "No title [%s]" % item.find("post_name").string
                logger.warning('Post "%s" is lacking a proper title', title)

            post_name = item.find("post_name").string
            post_id = item.find("post_id").string
            filename = get_filename(post_name, post_id)

            content = item.find("encoded").string
            raw_date = item.find("post_date").string
            if raw_date == "0000-00-00 00:00:00":
                date = None
            else:
                date_object = SafeDatetime.strptime(raw_date, "%Y-%m-%d %H:%M:%S")
                date = date_object.strftime("%Y-%m-%d %H:%M")
            author = item.find("creator").string

            categories = [
                cat.string for cat in item.findAll("category", {"domain": "category"})
            ]

            tags = [
                tag.string for tag in item.findAll("category", {"domain": "post_tag"})
            ]
            # To publish a post the status should be 'published'
            status = (
                "published"
                if item.find("status").string == "publish"
                else item.find("status").string
            )

            kind = "article"
            post_type = item.find("post_type").string
            if post_type == "page":
                kind = "page"
            elif wp_custpost:
                if post_type == "post":
                    pass
                # Old behaviour was to name everything not a page as an
                # article.Theoretically all attachments have status == inherit
                # so no attachments should be here. But this statement is to
                # maintain existing behaviour in case that doesn't hold true.
                elif post_type == "attachment":
                    pass
                else:
                    kind = post_type
            yield (
                title,
                content,
                filename,
                date,
                author,
                categories,
                tags,
                status,
                kind,
                "wp-html",
            )


def blogger2fields(xml):
    """Opens a blogger XML file, and yield Pelican fields"""

    soup = file_to_soup(xml)
    entries = soup.feed.findAll("entry")
    for entry in entries:
        raw_kind = entry.find(
            "category", {"scheme": "http://schemas.google.com/g/2005#kind"}
        ).get("term")
        if raw_kind == "http://schemas.google.com/blogger/2008/kind#post":
            kind = "article"
        elif raw_kind == "http://schemas.google.com/blogger/2008/kind#comment":
            kind = "comment"
        elif raw_kind == "http://schemas.google.com/blogger/2008/kind#page":
            kind = "page"
        else:
            continue

        try:
            assert kind != "comment"
            filename = entry.find("link", {"rel": "alternate"})["href"]
            filename = os.path.splitext(os.path.basename(filename))[0]
        except (AssertionError, TypeError, KeyError):
            filename = entry.find("id").string.split(".")[-1]

        title = entry.find("title").string or ""

        content = entry.find("content").string
        raw_date = entry.find("published").string
        if hasattr(SafeDatetime, "fromisoformat"):
            date_object = SafeDatetime.fromisoformat(raw_date)
        else:
            date_object = SafeDatetime.strptime(raw_date[:23], "%Y-%m-%dT%H:%M:%S.%f")
        date = date_object.strftime("%Y-%m-%d %H:%M")
        author = entry.find("author").find("name").string

        # blogger posts only have tags, no category
        tags = [
            tag.get("term")
            for tag in entry.findAll(
                "category", {"scheme": "http://www.blogger.com/atom/ns#"}
            )
        ]

        # Drafts have <app:control><app:draft>yes</app:draft></app:control>
        status = "published"
        try:
            if entry.find("control").find("draft").string == "yes":
                status = "draft"
        except AttributeError:
            pass

        yield (title, content, filename, date, author, None, tags, status, kind, "html")


def dc2fields(file):
    """Opens a Dotclear export file, and yield pelican fields"""
    try:
        from bs4 import BeautifulSoup
    except ImportError:
        error = (
            "Missing dependency "
            '"BeautifulSoup4" and "lxml" required '
            "to import Dotclear files."
        )
        sys.exit(error)

    in_cat = False
    in_post = False
    category_list = {}
    posts = []

    with open(file, encoding="utf-8") as f:
        for line in f:
            # remove final \n
            line = line[:-1]

            if line.startswith("[category"):
                in_cat = True
            elif line.startswith("[post"):
                in_post = True
            elif in_cat:
                fields = line.split('","')
                if not line:
                    in_cat = False
                else:
                    # remove 1st and last ""
                    fields[0] = fields[0][1:]
                    # fields[-1] = fields[-1][:-1]
                    category_list[fields[0]] = fields[2]
            elif in_post:
                if not line:
                    in_post = False
                    break
                else:
                    posts.append(line)

    print("%i posts read." % len(posts))

    subs = DEFAULT_CONFIG["SLUG_REGEX_SUBSTITUTIONS"]
    for post in posts:
        fields = post.split('","')

        # post_id = fields[0][1:]
        # blog_id = fields[1]
        # user_id = fields[2]
        cat_id = fields[3]
        # post_dt = fields[4]
        # post_tz = fields[5]
        post_creadt = fields[6]
        # post_upddt = fields[7]
        # post_password = fields[8]
        # post_type = fields[9]
        post_format = fields[10]
        # post_url = fields[11]
        # post_lang = fields[12]
        post_title = fields[13]
        post_excerpt = fields[14]
        post_excerpt_xhtml = fields[15]
        post_content = fields[16]
        post_content_xhtml = fields[17]
        # post_notes = fields[18]
        # post_words = fields[19]
        # post_status = fields[20]
        # post_selected = fields[21]
        # post_position = fields[22]
        # post_open_comment = fields[23]
        # post_open_tb = fields[24]
        # nb_comment = fields[25]
        # nb_trackback = fields[26]
        post_meta = fields[27]
        # redirect_url = fields[28][:-1]

        # remove seconds
        post_creadt = ":".join(post_creadt.split(":")[0:2])

        author = ""
        categories = []
        tags = []

        if cat_id:
            categories = [category_list[id].strip() for id in cat_id.split(",")]

        # Get tags related to a post
        tag = (
            post_meta.replace("{", "")
            .replace("}", "")
            .replace('a:1:s:3:\\"tag\\";a:', "")
            .replace("a:0:", "")
        )
        if len(tag) > 1:
            if int(len(tag[:1])) == 1:
                newtag = tag.split('"')[1]
                tags.append(
                    BeautifulSoup(newtag, "xml")
                    # bs4 always outputs UTF-8
                    .decode("utf-8")
                )
            else:
                i = 1
                j = 1
                while i <= int(tag[:1]):
                    newtag = tag.split('"')[j].replace("\\", "")
                    tags.append(
                        BeautifulSoup(newtag, "xml")
                        # bs4 always outputs UTF-8
                        .decode("utf-8")
                    )
                    i = i + 1
                    if j < int(tag[:1]) * 2:
                        j = j + 2

        """
        dotclear2 does not use markdown by default unless
        you use the markdown plugin
        Ref: http://plugins.dotaddict.org/dc2/details/formatting-markdown
        """
        if post_format == "markdown":
            content = post_excerpt + post_content
        else:
            content = post_excerpt_xhtml + post_content_xhtml
            content = content.replace("\\n", "")
            post_format = "html"

        kind = "article"  # TODO: Recognise pages
        status = "published"  # TODO: Find a way for draft posts

        yield (
            post_title,
            content,
            slugify(post_title, regex_subs=subs),
            post_creadt,
            author,
            categories,
            tags,
            status,
            kind,
            post_format,
        )


def _get_tumblr_posts(api_key, blogname, offset=0):
    import json
    import urllib.request as urllib_request

    url = (
        "https://api.tumblr.com/v2/blog/%s.tumblr.com/"
        "posts?api_key=%s&offset=%d&filter=raw"
    ) % (blogname, api_key, offset)
    request = urllib_request.Request(url)
    handle = urllib_request.urlopen(request)
    posts = json.loads(handle.read().decode("utf-8"))
    return posts.get("response").get("posts")


def tumblr2fields(api_key, blogname):
    """Imports Tumblr posts (API v2)"""
    offset = 0
    posts = _get_tumblr_posts(api_key, blogname, offset)
    subs = DEFAULT_CONFIG["SLUG_REGEX_SUBSTITUTIONS"]
    while len(posts) > 0:
        for post in posts:
            title = (
                post.get("title")
                or post.get("source_title")
                or post.get("type").capitalize()
            )
            slug = post.get("slug") or slugify(title, regex_subs=subs)
            tags = post.get("tags")
            timestamp = post.get("timestamp")
            date = SafeDatetime.fromtimestamp(
                int(timestamp), tz=datetime.timezone.utc
            ).strftime("%Y-%m-%d %H:%M:%S%z")
            slug = (
                SafeDatetime.fromtimestamp(
                    int(timestamp), tz=datetime.timezone.utc
                ).strftime("%Y-%m-%d-")
                + slug
            )
            format = post.get("format")
            content = post.get("body")
            type = post.get("type")
            if type == "photo":
                if format == "markdown":
                    fmtstr = "![%s](%s)"
                else:
                    fmtstr = '<img alt="%s" src="%s" />'
                content = "\n".join(
                    fmtstr
                    % (photo.get("caption"), photo.get("original_size").get("url"))
                    for photo in post.get("photos")
                )
            elif type == "quote":
                if format == "markdown":
                    fmtstr = "\n\n&mdash; %s"
                else:
                    fmtstr = "<p>&mdash; %s</p>"
                content = post.get("text") + fmtstr % post.get("source")
            elif type == "link":
                if format == "markdown":
                    fmtstr = "[via](%s)\n\n"
                else:
                    fmtstr = '<p><a href="%s">via</a></p>\n'
                content = fmtstr % post.get("url") + post.get("description")
            elif type == "audio":
                if format == "markdown":
                    fmtstr = "[via](%s)\n\n"
                else:
                    fmtstr = '<p><a href="%s">via</a></p>\n'
                content = (
                    fmtstr % post.get("source_url")
                    + post.get("caption")
                    + post.get("player")
                )
            elif type == "video":
                if format == "markdown":
                    fmtstr = "[via](%s)\n\n"
                else:
                    fmtstr = '<p><a href="%s">via</a></p>\n'
                source = fmtstr % post.get("source_url")
                caption = post.get("caption")
                players = [
                    # If embed_code is False, couldn't get the video
                    player.get("embed_code") or None
                    for player in post.get("player")
                ]
                # If there are no embeddable players, say so, once
                if len(players) > 0 and all(player is None for player in players):
                    players = "<p>(This video isn't available anymore.)</p>\n"
                else:
                    players = "\n".join(players)
                content = source + caption + players
            elif type == "answer":
                title = post.get("question")
                content = (
                    "<p>"
                    '<a href="%s" rel="external nofollow">%s</a>'
                    ": %s"
                    "</p>\n"
                    " %s"
                    % (
                        post.get("asking_name"),
                        post.get("asking_url"),
                        post.get("question"),
                        post.get("answer"),
                    )
                )

            content = content.rstrip() + "\n"
            kind = "article"
            status = "published"  # TODO: Find a way for draft posts

            yield (
                title,
                content,
                slug,
                date,
                post.get("blog_name"),
                [type],
                tags,
                status,
                kind,
                format,
            )

        offset += len(posts)
        posts = _get_tumblr_posts(api_key, blogname, offset)


def strip_medium_post_content(soup) -> str:
    """Strip some tags and attributes from medium post content.

    For example, the 'section' and 'div' tags cause trouble while rendering.

    The problem with these tags is you can get a section divider (--------------)
    that is not between two pieces of content.  For example:

      Some text.

      .. container:: section-divider

         --------------

      .. container:: section-content

      More content.

    In this case, pandoc complains: "Unexpected section title or transition."

    Also, the "id" and "name" attributes in tags cause similar problems.  They show
    up in .rst as extra junk that separates transitions.
    """
    # Remove tags
    # section and div cause problems
    # footer also can cause problems, and has nothing we want to keep
    # See https://stackoverflow.com/a/8439761
    invalid_tags = ["section", "div", "footer"]
    for tag in invalid_tags:
        for match in soup.findAll(tag):
            match.replaceWithChildren()

    # Remove attributes
    # See https://stackoverflow.com/a/9045719
    invalid_attributes = ["name", "id", "class"]
    bs4 = _import_bs4()
    for tag in soup.descendants:
        if isinstance(tag, bs4.element.Tag):
            tag.attrs = {
                key: value
                for key, value in tag.attrs.items()
                if key not in invalid_attributes
            }

    # Get the string of all content, keeping other tags
    all_content = "".join(str(element) for element in soup.contents)
    return all_content


def mediumpost2fields(filepath: str) -> tuple:
    """Take an HTML post from a medium export, return Pelican fields."""

    soup = file_to_soup(filepath, "html.parser")
    if not soup:
        raise ValueError(f"{filepath} could not be parsed by beautifulsoup")
    kind = "article"

    content = soup.find("section", class_="e-content")
    if not content:
        raise ValueError(f"{filepath}: Post has no content")

    title = soup.find("title").string or ""

    raw_date = soup.find("time", class_="dt-published")
    date = None
    if raw_date:
        # This datetime can include timezone, e.g., "2017-04-21T17:11:55.799Z"
        # python before 3.11 can't parse the timezone using datetime.fromisoformat
        # See also https://docs.python.org/3.10/library/datetime.html#datetime.datetime.fromisoformat
        # "This does not support parsing arbitrary ISO 8601 strings"
        # So, we use dateutil.parser, which can handle it.
        date_object = dateutil.parser.parse(raw_date.attrs["datetime"])
        date = date_object.strftime("%Y-%m-%d %H:%M")
        status = "published"
    else:
        status = "draft"
    author = soup.find("a", class_="p-author h-card")
    if author:
        author = author.string

    # Now that we're done with classes, we can strip the content
    content = strip_medium_post_content(content)

    # medium HTML export doesn't have tag or category
    # RSS feed has tags, but it doesn't have all the posts.
    tags = ()

    slug = medium_slug(filepath)

    # TODO: make the fields a python dataclass
    return (
        title,
        content,
        slug,
        date,
        author,
        None,
        tags,
        status,
        kind,
        "html",
    )


def medium_slug(filepath: str) -> str:
    """Make the filepath of a medium exported file into a slug."""
    # slug: filename without extension
    slug = os.path.basename(filepath)
    slug = os.path.splitext(slug)[0]
    # A medium export filename looks like date_-title-...html
    # But, RST doesn't like "_-" (see https://github.com/sphinx-doc/sphinx/issues/4350)
    # so get rid of it
    slug = slug.replace("_-", "-")
    # drop the hex string medium puts on the end of the filename, why keep it.
    # e.g., "-a8a8a8a8" or "---a9a9a9a9"
    # also: drafts don't need "--DRAFT"
    slug = re.sub(r"((-)+([0-9a-f]+|DRAFT))+$", "", slug)
    return slug


def mediumposts2fields(medium_export_dir: str):
    """Take HTML posts in a medium export directory, and yield Pelican fields."""
    for file in os.listdir(medium_export_dir):
        filename = os.fsdecode(file)
        yield mediumpost2fields(os.path.join(medium_export_dir, filename))


def feed2fields(file):
    """Read a feed and yield pelican fields"""
    import feedparser

    d = feedparser.parse(file)
    subs = DEFAULT_CONFIG["SLUG_REGEX_SUBSTITUTIONS"]
    for entry in d.entries:
        date = (
            time.strftime("%Y-%m-%d %H:%M", entry.updated_parsed)
            if hasattr(entry, "updated_parsed")
            else None
        )
        author = entry.author if hasattr(entry, "author") else None
        tags = [e["term"] for e in entry.tags] if hasattr(entry, "tags") else None

        slug = slugify(entry.title, regex_subs=subs)
        kind = "article"
        yield (
            entry.title,
            entry.description,
            slug,
            date,
            author,
            [],
            tags,
            None,
            kind,
            "html",
        )


def build_header(
    title, date, author, categories, tags, slug, status=None, attachments=None
):
    """Build a header from a list of fields"""

    from docutils.utils import column_width

    header = "{}\n{}\n".format(title, "#" * column_width(title))
    if date:
        header += ":date: %s\n" % date
    if author:
        header += ":author: %s\n" % author
    if categories:
        header += ":category: %s\n" % ", ".join(categories)
    if tags:
        header += ":tags: %s\n" % ", ".join(tags)
    if slug:
        header += ":slug: %s\n" % slug
    if status:
        header += ":status: %s\n" % status
    if attachments:
        header += ":attachments: %s\n" % ", ".join(attachments)
    header += "\n"
    return header


def build_asciidoc_header(
    title, date, author, categories, tags, slug, status=None, attachments=None
):
    """Build a header from a list of fields"""

    header = "= %s\n" % title
    if author:
        header += "%s\n" % author
        if date:
            header += "%s\n" % date
    if categories:
        header += ":category: %s\n" % ", ".join(categories)
    if tags:
        header += ":tags: %s\n" % ", ".join(tags)
    if slug:
        header += ":slug: %s\n" % slug
    if status:
        header += ":status: %s\n" % status
    if attachments:
        header += ":attachments: %s\n" % ", ".join(attachments)
    header += "\n"
    return header


def build_markdown_header(
    title, date, author, categories, tags, slug, status=None, attachments=None
):
    """Build a header from a list of fields"""
    header = "Title: %s\n" % title
    if date:
        header += "Date: %s\n" % date
    if author:
        header += "Author: %s\n" % author
    if categories:
        header += "Category: %s\n" % ", ".join(categories)
    if tags:
        header += "Tags: %s\n" % ", ".join(tags)
    if slug:
        header += "Slug: %s\n" % slug
    if status:
        header += "Status: %s\n" % status
    if attachments:
        header += "Attachments: %s\n" % ", ".join(attachments)
    header += "\n"
    return header


def get_ext(out_markup, in_markup="html"):
    if out_markup == "asciidoc":
        ext = ".adoc"
    elif in_markup == "markdown" or out_markup == "markdown":
        ext = ".md"
    else:
        ext = ".rst"
    return ext


def get_out_filename(
    output_path,
    filename,
    ext,
    kind,
    dirpage,
    dircat,
    categories,
    wp_custpost,
    slug_subs,
):
    filename = os.path.basename(filename)

    # Enforce filename restrictions for various filesystems at once; see
    # https://en.wikipedia.org/wiki/Filename#Reserved_characters_and_words
    # we do not need to filter words because an extension will be appended
    filename = re.sub(r'[<>:"/\\|?*^% ]', "-", filename)  # invalid chars
    filename = filename.lstrip(".")  # should not start with a dot
    if not filename:
        filename = "_"
    filename = filename[:249]  # allow for 5 extra characters

    out_filename = os.path.join(output_path, filename + ext)
    # option to put page posts in pages/ subdirectory
    if dirpage and kind == "page":
        pages_dir = os.path.join(output_path, "pages")
        if not os.path.isdir(pages_dir):
            os.mkdir(pages_dir)
        out_filename = os.path.join(pages_dir, filename + ext)
    elif not dirpage and kind == "page":
        pass
    # option to put wp custom post types in directories with post type
    # names. Custom post types can also have categories so option to
    # create subdirectories with category names
    elif kind != "article":
        if wp_custpost:
            typename = slugify(kind, regex_subs=slug_subs)
        else:
            typename = ""
            kind = "article"
        if dircat and (len(categories) > 0):
            catname = slugify(categories[0], regex_subs=slug_subs, preserve_case=True)
        else:
            catname = ""
        out_filename = os.path.join(output_path, typename, catname, filename + ext)
        if not os.path.isdir(os.path.join(output_path, typename, catname)):
            os.makedirs(os.path.join(output_path, typename, catname))
    # option to put files in directories with categories names
    elif dircat and (len(categories) > 0):
        catname = slugify(categories[0], regex_subs=slug_subs, preserve_case=True)
        out_filename = os.path.join(output_path, catname, filename + ext)
        if not os.path.isdir(os.path.join(output_path, catname)):
            os.mkdir(os.path.join(output_path, catname))

    return out_filename


def get_attachments(xml):
    """returns a dictionary of posts that have attachments with a list
    of the attachment_urls
    """
    soup = file_to_soup(xml)
    items = soup.rss.channel.findAll("item")
    names = {}
    attachments = []

    for item in items:
        kind = item.find("post_type").string
        post_name = item.find("post_name").string
        post_id = item.find("post_id").string

        if kind == "attachment":
            attachments.append(
                (item.find("post_parent").string, item.find("attachment_url").string)
            )
        else:
            filename = get_filename(post_name, post_id)
            names[post_id] = filename
    attachedposts = defaultdict(set)
    for parent, url in attachments:
        try:
            parent_name = names[parent]
        except KeyError:
            # attachment's parent is not a valid post
            parent_name = None

        attachedposts[parent_name].add(url)
    return attachedposts


def download_attachments(output_path, urls):
    """Downloads WordPress attachments and returns a list of paths to
    attachments that can be associated with a post (relative path to output
    directory). Files that fail to download, will not be added to posts"""
    locations = {}
    for url in urls:
        path = urlparse(url).path
        # teardown path and rebuild to negate any errors with
        # os.path.join and leading /'s
        path = path.split("/")
        filename = path.pop(-1)
        localpath = ""
        for item in path:
            if sys.platform != "win32" or ":" not in item:
                localpath = os.path.join(localpath, item)
        full_path = os.path.join(output_path, localpath)

        # Generate percent-encoded URL
        scheme, netloc, path, query, fragment = urlsplit(url)
        if scheme != "file":
            path = quote(path)
            url = urlunsplit((scheme, netloc, path, query, fragment))

        if not os.path.exists(full_path):
            os.makedirs(full_path)
        print(f"downloading {filename}")
        try:
            urlretrieve(url, os.path.join(full_path, filename))
            locations[url] = os.path.join(localpath, filename)
        except (URLError, OSError) as e:
            # Python 2.7 throws an IOError rather Than URLError
            logger.warning("No file could be downloaded from %s\n%s", url, e)
    return locations


def is_pandoc_needed(in_markup):
    return in_markup in ("html", "wp-html")


def get_pandoc_version():
    cmd = ["pandoc", "--version"]
    try:
        output = subprocess.check_output(cmd, text=True)
    except (subprocess.CalledProcessError, OSError) as e:
        logger.warning("Pandoc version unknown: %s", e)
        return ()

    return tuple(int(i) for i in output.split()[1].split("."))


def update_links_to_attached_files(content, attachments):
    for old_url, new_path in attachments.items():
        # url may occur both with http:// and https://
        http_url = old_url.replace("https://", "http://")
        https_url = old_url.replace("http://", "https://")
        for url in [http_url, https_url]:
            content = content.replace(url, "{static}" + new_path)
    return content


def fields2pelican(
    fields,
    out_markup,
    output_path,
    dircat=False,
    strip_raw=False,
    disable_slugs=False,
    dirpage=False,
    filename_template=None,
    filter_author=None,
    wp_custpost=False,
    wp_attach=False,
    attachments=None,
):
    pandoc_version = get_pandoc_version()
    posts_require_pandoc = []

    slug_subs = DEFAULT_CONFIG["SLUG_REGEX_SUBSTITUTIONS"]

    for (
        title,
        content,
        filename,
        date,
        author,
        categories,
        tags,
        status,
        kind,
        in_markup,
    ) in fields:
        if filter_author and filter_author != author:
            continue
        if is_pandoc_needed(in_markup) and not pandoc_version:
            posts_require_pandoc.append(filename)

        slug = not disable_slugs and filename or None
        assert slug is None or filename == os.path.basename(
            filename
        ), f"filename is not a basename: {filename}"

        if wp_attach and attachments:
            try:
                urls = attachments[filename]
                links = download_attachments(output_path, urls)
            except KeyError:
                links = None
        else:
            links = None

        ext = get_ext(out_markup, in_markup)
        if ext == ".adoc":
            header = build_asciidoc_header(
                title, date, author, categories, tags, slug, status, attachments
            )
        elif ext == ".md":
            header = build_markdown_header(
                title,
                date,
                author,
                categories,
                tags,
                slug,
                status,
                links.values() if links else None,
            )
        else:
            out_markup = "rst"
            header = build_header(
                title,
                date,
                author,
                categories,
                tags,
                slug,
                status,
                links.values() if links else None,
            )

        out_filename = get_out_filename(
            output_path,
            filename,
            ext,
            kind,
            dirpage,
            dircat,
            categories,
            wp_custpost,
            slug_subs,
        )
        print(out_filename)

        if in_markup in ("html", "wp-html"):
            with tempfile.TemporaryDirectory() as tmpdir:
                html_filename = os.path.join(tmpdir, "pandoc-input.html")
                # Replace newlines with paragraphs wrapped with <p> so
                # HTML is valid before conversion
                if in_markup == "wp-html":
                    new_content = decode_wp_content(content)
                else:
                    paragraphs = content.splitlines()
                    paragraphs = [f"<p>{p}</p>" for p in paragraphs]
                    new_content = "".join(paragraphs)
                with open(html_filename, "w", encoding="utf-8") as fp:
                    fp.write(new_content)

                if pandoc_version < (2,):
                    parse_raw = "--parse-raw" if not strip_raw else ""
                    wrap_none = (
                        "--wrap=none" if pandoc_version >= (1, 16) else "--no-wrap"
                    )
                    cmd = (
                        "pandoc --normalize {0} --from=html"
                        ' --to={1} {2} -o "{3}" "{4}"'
                    )
                    cmd = cmd.format(
                        parse_raw,
                        out_markup if out_markup != "markdown" else "gfm",
                        wrap_none,
                        out_filename,
                        html_filename,
                    )
                else:
                    from_arg = "-f html+raw_html" if not strip_raw else "-f html"
                    cmd = 'pandoc {0} --to={1}-smart --wrap=none -o "{2}" "{3}"'
                    cmd = cmd.format(
                        from_arg,
                        out_markup if out_markup != "markdown" else "gfm",
                        out_filename,
                        html_filename,
                    )

                try:
                    rc = subprocess.call(cmd, shell=True)
                    if rc < 0:
                        error = "Child was terminated by signal %d" % -rc
                        sys.exit(error)

                    elif rc > 0:
                        error = "Please, check your Pandoc installation."
                        sys.exit(error)
                except OSError as e:
                    error = "Pandoc execution failed: %s" % e
                    sys.exit(error)

            with open(out_filename, encoding="utf-8") as fs:
                content = fs.read()
                if out_markup == "markdown":
                    # In markdown, to insert a <br />, end a line with two
                    # or more spaces & then a end-of-line
                    content = content.replace("\\\n ", "  \n")
                    content = content.replace("\\\n", "  \n")

            if wp_attach and links:
                content = update_links_to_attached_files(content, links)

        with open(out_filename, "w", encoding="utf-8") as fs:
            fs.write(header + content)

    if posts_require_pandoc:
        logger.error(
            "Pandoc must be installed to import the following posts:\n  {}".format(
                "\n  ".join(posts_require_pandoc)
            )
        )

    if wp_attach and attachments and None in attachments:
        print("downloading attachments that don't have a parent post")
        urls = attachments[None]
        download_attachments(output_path, urls)


def main():
    parser = argparse.ArgumentParser(
        description="Transform feed, Blogger, Dotclear, Tumblr, or "
        "WordPress files into reST (rst) or Markdown (md) files. "
        "Be sure to have pandoc installed.",
        formatter_class=argparse.ArgumentDefaultsHelpFormatter,
    )

    parser.add_argument(dest="input", help="The input file to read")
    parser.add_argument(
        "--blogger", action="store_true", dest="blogger", help="Blogger XML export"
    )
    parser.add_argument(
        "--dotclear", action="store_true", dest="dotclear", help="Dotclear export"
    )
    parser.add_argument(
        "--medium", action="store_true", dest="medium", help="Medium export"
    )
    parser.add_argument(
        "--tumblr", action="store_true", dest="tumblr", help="Tumblr export"
    )
    parser.add_argument(
        "--wpfile", action="store_true", dest="wpfile", help="Wordpress XML export"
    )
    parser.add_argument(
        "--feed", action="store_true", dest="feed", help="Feed to parse"
    )
    parser.add_argument(
        "-o", "--output", dest="output", default="content", help="Output path"
    )
    parser.add_argument(
        "-m",
        "--markup",
        dest="markup",
        default="rst",
        help="Output markup format (supports rst & markdown)",
    )
    parser.add_argument(
        "--dir-cat",
        action="store_true",
        dest="dircat",
        help="Put files in directories with categories name",
    )
    parser.add_argument(
        "--dir-page",
        action="store_true",
        dest="dirpage",
        help=(
            'Put files recognised as pages in "pages/" sub-directory'
            " (blogger and wordpress import only)"
        ),
    )
    parser.add_argument(
        "--filter-author",
        dest="author",
        help="Import only post from the specified author",
    )
    parser.add_argument(
        "--strip-raw",
        action="store_true",
        dest="strip_raw",
        help="Strip raw HTML code that can't be converted to "
        "markup such as flash embeds or iframes (wordpress import only)",
    )
    parser.add_argument(
        "--wp-custpost",
        action="store_true",
        dest="wp_custpost",
        help="Put wordpress custom post types in directories. If used with "
        "--dir-cat option directories will be created as "
        "/post_type/category/ (wordpress import only)",
    )
    parser.add_argument(
        "--wp-attach",
        action="store_true",
        dest="wp_attach",
        help="(wordpress import only) Download files uploaded to wordpress as "
        "attachments. Files will be added to posts as a list in the post "
        "header. All files will be downloaded, even if "
        "they aren't associated with a post. Files will be downloaded "
        "with their original path inside the output directory. "
        "e.g. output/wp-uploads/date/postname/file.jpg "
        "-- Requires an internet connection --",
    )
    parser.add_argument(
        "--disable-slugs",
        action="store_true",
        dest="disable_slugs",
        help="Disable storing slugs from imported posts within output. "
        "With this disabled, your Pelican URLs may not be consistent "
        "with your original posts.",
    )
    parser.add_argument(
        "-b", "--blogname", dest="blogname", help="Blog name (Tumblr import only)"
    )

    args = parser.parse_args()

    input_type = None
    if args.blogger:
        input_type = "blogger"
    elif args.dotclear:
        input_type = "dotclear"
    elif args.medium:
        input_type = "medium"
    elif args.tumblr:
        input_type = "tumblr"
    elif args.wpfile:
        input_type = "wordpress"
    elif args.feed:
        input_type = "feed"
    else:
        error = (
            "You must provide one of --blogger, --dotclear, "
            "--medium, --tumblr, --wpfile or --feed options"
        )
        sys.exit(error)

    if not os.path.exists(args.output):
        try:
            os.mkdir(args.output)
        except OSError:
            error = "Unable to create the output folder: " + args.output
            sys.exit(error)

    if args.wp_attach and input_type != "wordpress":
        error = "You must be importing a wordpress xml to use the --wp-attach option"
        sys.exit(error)

    if input_type == "blogger":
        fields = blogger2fields(args.input)
    elif input_type == "dotclear":
        fields = dc2fields(args.input)
    elif input_type == "medium":
        fields = mediumposts2fields(args.input)
    elif input_type == "tumblr":
        fields = tumblr2fields(args.input, args.blogname)
    elif input_type == "wordpress":
        fields = wp2fields(args.input, args.wp_custpost or False)
    elif input_type == "feed":
        fields = feed2fields(args.input)
    else:
        raise ValueError(f"Unhandled input_type {input_type}")

    if args.wp_attach:
        attachments = get_attachments(args.input)
    else:
        attachments = None

    # init logging
    init()
    fields2pelican(
        fields,
        args.markup,
        args.output,
        dircat=args.dircat or False,
        dirpage=args.dirpage or False,
        strip_raw=args.strip_raw or False,
        disable_slugs=args.disable_slugs or False,
        filter_author=args.author,
        wp_custpost=args.wp_custpost or False,
        wp_attach=args.wp_attach or False,
        attachments=attachments or None,
    )