package main

import (
	"bytes"
	"context"
	"errors"
	"fmt"
	"io"
	"log"
	"net/http"
	"sort"
	"strings"
	"sync"
	"time"

	"github.com/PuerkitoBio/goquery"
	"golang.org/x/net/html"
	"golang.org/x/net/html/atom"
)

const maxResponseSize = 1e6 // 1MB

// https://docs.joinmastodon.org/spec/microformats/

// Root elements (h-*)
//
// h-feed
//     Represents a stream of entries. Attached to a profile's toots. Also
//     attached to the parent thread within detailed status views.
//
// h-entry
//     Represents episodic or date stamped online content. Attached to a status.
//
// URL properties (u-*)
//
// u-photo
//     Within h-card, represents the profile picture. Attached to the avatar image.
//
// u-uid
//     Within h-entry or h-cite, represents a universally unique identifier.
//     Attached to timestamp link.
//
// u-url
//     Within h-entry or h-cite, represents the status permalink. Attached to
//     timestamp link. Within h-card, represents the profile permalink.
//     Attached to display name link.
//
//
//
// Datetime properties (dt-*)
//
// dt-published
//     Within h-entry or h-cite, represents the date and time at which the
//     status was published. Attached to data element with value attribute.
//
// Element tree (e-*)
//
// e-content
//     Within h-entry or h-cite, represents the content of the status. Attached to status content.
//

type MastoSource struct {
	Items     []*MastoItem
	Title     string
	URL       string
	LastFetch time.Time
	Error     error

	LastStatusCode int
	LastStatus     string

	mu sync.Mutex
}

type MastoFeed struct {
	Title string
	Items []*MastoItem
}

type MastoItem struct {
	Title           string `json:"title,omitempty"`
	Content         string `json:"content,omitempty"`
	Link            string `json:"link,omitempty"`
	PublishedString string `json:"published,omitempty"`
	Author          string `json:"author,omitempty"`
}

func NewMastoSource(url string) *MastoSource {
	return &MastoSource{
		URL: url,
	}
}

func (src *MastoSource) update(ctx context.Context) {
	src.mu.Lock()
	defer src.mu.Unlock()

	req, err := http.NewRequestWithContext(ctx, "GET", src.URL, nil)
	if err != nil {
		src.Error = fmt.Errorf("error fetching %q: %w", src.URL, err)
		log.Println(src.Error)
		return // return err?
	}
	req.Header.Set("User-Agent", UserAgent)
	// TODO: If-Modified-Since, Etag

	resp, err := http.DefaultClient.Do(req)
	if err != nil {
		err := fmt.Errorf("error fetching %q: %w", src.URL, err)
		log.Println(err)
		src.Error = err
		return // return err?
	}

	if resp != nil && resp.Body != nil {
		defer func() {
			err := resp.Body.Close()
			if err != nil {
				log.Printf("error closing response body for %q: %v", src.URL, err)
			}
		}()
	}

	src.LastStatusCode = resp.StatusCode
	src.LastStatus = resp.Status
	if resp.StatusCode != 200 {
		src.Error = fmt.Errorf("error fetching %q: status %s", src.URL, resp.Status)
		log.Println(src.Error)
		return
	}

	// TODO: cache body
	body := MaxBytesReader(resp.Body, maxResponseSize)

	feed, err := parseMicroformats(body)
	if err != nil {
		err := fmt.Errorf("error parsing %q: %w", src.URL, err)
		log.Println(err)
		src.Error = err
		return // return err?
	}

	items := feed.Items
	sort.Slice(items, func(i, j int) bool {
		return items[i].PublishedString >= items[j].PublishedString
	})
	src.Title = feed.Title
	src.Items = items
	src.LastFetch = time.Now()
	src.Error = nil
}

func parseMicroformats(r io.Reader) (*MastoFeed, error) {
	doc, err := goquery.NewDocumentFromReader(r)
	if err != nil {
		return nil, err
	}
	feed := new(MastoFeed)
	if doc.Find(".h-feed").Length() == 0 {
		return nil, fmt.Errorf("no feed content found")
	}
	feed.Title = doc.Find(".h-feed > .p-name").First().AttrOr("value", "")
	doc.Find(".h-feed .h-entry").Each(func(i int, elem *goquery.Selection) {
		cw := strings.TrimSpace(text(elem.Find(".p-summary")))
		if cw != "" {
			cw = "[" + cw + "] "
		}
		feed.Items = append(feed.Items, &MastoItem{
			Title:           "",
			Content:         cw + text(elem.Find(".e-content")),
			Link:            elem.Find("a.u-url.u-uid").AttrOr("href", ""),
			Author:          text(elem.Find(".p-author .p-name")),
			PublishedString: elem.Find("data.dt-published").AttrOr("value", ""),
		})
	})
	return feed, nil
}

// Text gets the combined text contents of each element in the set of matched
// elements, including their descendants.
func text(s *goquery.Selection) string {
	var buf bytes.Buffer

	// Slightly optimized vs calling Each: no single selection object created
	var f func(*html.Node)
	f = func(n *html.Node) {
		if n.Type == html.TextNode {
			// Keep newlines and spaces, like jQuery
			buf.WriteString(n.Data)
		} else if n.Type == html.ElementNode && n.DataAtom == atom.Br {
			//buf.WriteString("\n")
			buf.WriteString(" ")
		} else if n.Type == html.ElementNode && n.DataAtom == atom.P && n.PrevSibling != nil {
			//buf.WriteString("\n\n")
			buf.WriteString(" ")
		}
		if n.FirstChild != nil {
			for c := n.FirstChild; c != nil; c = c.NextSibling {
				f(c)
			}
		}
	}
	for _, n := range s.Nodes {
		f(n)
	}

	return buf.String()
}

// MaxBytesReader is similar to io.LimitReader but is intended for
// limiting the size of incoming request bodies. In contrast to
// io.LimitReader, MaxBytesReader's result is a ReadCloser, returns a
// non-EOF error for a Read beyond the limit, and closes the
// underlying reader when its Close method is called.
//
// MaxBytesReader prevents clients from accidentally or maliciously
// sending a large request and wasting server resources.
//
// Based on http.MaxBytesReader
func MaxBytesReader(r io.ReadCloser, n int64) io.ReadCloser {
	if n < 0 { // Treat negative limits as equivalent to 0.
		n = 0
	}
	return &maxBytesReader{r: r, n: n}
}

type maxBytesReader struct {
	r   io.ReadCloser // underlying reader
	n   int64         // max bytes remaining
	err error         // sticky error
}

func (l *maxBytesReader) Read(p []byte) (n int, err error) {
	if l.err != nil {
		return 0, l.err
	}
	if len(p) == 0 {
		return 0, nil
	}
	// If they asked for a 32KB byte read but only 5 bytes are
	// remaining, no need to read 32KB. 6 bytes will answer the
	// question of the whether we hit the limit or go past it.
	if int64(len(p)) > l.n+1 {
		p = p[:l.n+1]
	}
	n, err = l.r.Read(p)

	if int64(n) <= l.n {
		l.n -= int64(n)
		l.err = err
		return n, err
	}

	n = int(l.n)
	l.n = 0

	l.err = errors.New("http: response body too large")
	return n, l.err
}

func (l *maxBytesReader) Close() error {
	return l.r.Close()
}