webweav.ing/feed/feed.go

package feed

import (
	"fmt"
	"io"
	"net/http"
	"strings"
	"time"

	"golang.org/x/net/html"
)

const feedfmtopen = `<?xml version="1.0" encoding="utf-8"?>
<rss version="2.0">
<channel>
<title>%s</title>
<link>%s</link>
<description>%s</description>`

const feedfmtclose = `</channel>
</rss>`

const itemfmt = `<item>
<title>%s</title>
<link>%s</link>
<guid>%s</guid>
<pubDate>%s</pubDate>
<description>
%s
</description>
</item>`

type FeedInfo struct {
    SiteName    string   
    SiteUrl     string
    SiteDesc    string
    PageUrls    []string
    Items       []*FeedItem
    errors      map[string]error
} 

type FeedItem struct {
    Url         string
    Title       string
    Author      string
    PubTime     time.Time
    RawText     string
}

func fetchPage(url string) (string, error) {
    resp, err := http.Get(url)
    if err != nil {
        return "", fmt.Errorf("Error sending Get request: %w", err)
    }
    defer resp.Body.Close()
    body, err := io.ReadAll(resp.Body)

    return string(body), nil
}

func parseTime(timestr string) (time.Time, error) {
    var formats = []string {
        time.ANSIC,
        time.UnixDate,
        time.RubyDate,
        time.RFC822,
        time.RFC822Z,
        time.RFC850,
        time.RFC1123,
        time.RFC1123Z,
        "2006-01-02 15:04:05 -0700 -0700",
        time.RFC3339,
        time.RFC3339Nano,
        time.DateTime,
        time.DateOnly,
    }
    var pagetime time.Time
    var err error
    for _, f := range formats {
        pagetime, err = time.Parse(f, timestr)
        if err == nil {
            return pagetime, err
        }
    }
    return pagetime, fmt.Errorf("Error parsing time: invalid format")
}

func getHtmlElement(doc *html.Node, tag string) (*html.Node, error) {
    var f func(*html.Node, string)
    var element *html.Node
    f = func(n *html.Node, s string) {
        if n.Type == html.ElementNode && n.Data == s{
            element = n
            return
        }
        for c := n.FirstChild; c != nil; c = c.NextSibling {
            f(c, tag)
        }
    }
    f(doc, tag)
    if element == nil {
        return nil, fmt.Errorf("no <%s> element found", tag)
    }
    return element, nil
}

func (f *FeedItem) ParseContent(content string) error {
    doc, err := html.Parse(strings.NewReader(content))
    if err != nil {
        return fmt.Errorf("Error parsing HTML: %w", err)
    }
    earticle, err := getHtmlElement(doc, "article")
    if err != nil {
        return err
    }
    etitle, err := getHtmlElement(doc, "title")
    if err != nil {
        return err
    }
    f.Title = etitle.FirstChild.Data

    var articleBuilder strings.Builder
    html.Render(&articleBuilder, earticle)
    f.RawText = articleBuilder.String()

    etime, err := getHtmlElement(earticle, "time")
    if err != nil {
        return err
    }
    var pubTime time.Time
    for _, d := range etime.Attr {
        if d.Key == "datetime" {
            pubTime, err = parseTime(d.Val)
        }
        if err != nil {
            return  fmt.Errorf("Error parsing time: %w", err)
        }
        f.PubTime = pubTime
    }
    return nil 
}

func NewFeedItem(url string) (*FeedItem, error) {
    rawhtml, err := fetchPage(url)
    if err != nil {
        return nil, fmt.Errorf("Could not fetch page '%s': %w", url, err)
    }
    item := FeedItem{
        Url: url,
    }
    item.ParseContent(rawhtml);
    if err != nil {
        return nil, fmt.Errorf("Could not parse feed item: %w", err)
    }
    return &item, nil
}

func NewFeedInfo(name, base_url, desc, author string, page_urls...string) (*FeedInfo, error) {
    info := FeedInfo{
        SiteName: name,
        SiteUrl: base_url,
        SiteDesc: desc,
        PageUrls: page_urls,
    }
    for _,url := range info.PageUrls {
        item, err := NewFeedItem(url)
        if err != nil {
            info.errors[url] = err
        } else {
            info.Items = append(info.Items, item)
        }
    }
    return &info, nil
}

func (info *FeedInfo) format(raw string) string {
    var formatBuilder strings.Builder
    depth := 0
    oldDepth := 0
    for _,line := range strings.Split(raw, "\n") {
        tmp := strings.TrimSpace(line)
        if tmp == "" {
            continue
        }
        oldDepth = depth
        for i,s := range line {
            if i < len(line) - 1 {
                t := line[i + 1]
                if s == '<' && t != '?' && t != '/' {
                    depth += 1
                }
                if s == '<' && t == '/' {
                    depth -= 1
                }
                if s == '/' && t == '>' {
                    depth -= 1
                }
            }
        }
        for i := 0; i < depth; i++ {
            if (i == depth - 1 && oldDepth < depth) {
                continue
            }
            formatBuilder.WriteString("    ")
        }
        formatBuilder.WriteString(html.EscapeString(tmp))
        formatBuilder.WriteString("\n")
    }
    return formatBuilder.String()
}

func (info *FeedInfo) GenerateRSS() string {
    var outputBuilder strings.Builder
    outputBuilder.WriteString(fmt.Sprintf(feedfmtopen, info.SiteName, info.SiteUrl, info.SiteDesc))
    outputBuilder.WriteString("\n")
    for _, item := range info.Items {
        outputBuilder.WriteString(fmt.Sprintf(
            itemfmt,
            item.Title,
            item.Url,
            item.Url,
            item.PubTime.Format("Mon, 2 Jan 2006 15:04:05 MST"), 
            item.RawText,
            ))
        outputBuilder.WriteString("\n")
    }
    outputBuilder.WriteString(feedfmtclose)
    return info.format(outputBuilder.String())
}