miniflux_v2/internal/reader/atom/atom_10_adapter.go

// SPDX-FileCopyrightText: Copyright The Miniflux Authors. All rights reserved.
// SPDX-License-Identifier: Apache-2.0

package atom // import "miniflux.app/v2/internal/reader/atom"

import (
	"log/slog"
	"slices"
	"sort"
	"strconv"
	"strings"
	"time"

	"miniflux.app/v2/internal/crypto"
	"miniflux.app/v2/internal/model"
	"miniflux.app/v2/internal/reader/date"
	"miniflux.app/v2/internal/reader/sanitizer"
	"miniflux.app/v2/internal/urllib"
)

type atom10Adapter struct {
	atomFeed *atom10Feed
}

func NewAtom10Adapter(atomFeed *atom10Feed) *atom10Adapter {
	return &atom10Adapter{atomFeed}
}

func (a *atom10Adapter) BuildFeed(baseURL string) *model.Feed {
	feed := &model.Feed{
		FeedURL: baseURL,
		SiteURL: baseURL,
	}

	// Populate the feed URL.
	feedURL := a.atomFeed.Links.firstLinkWithRelation("self")
	if feedURL != "" {
		if absoluteFeedURL, err := urllib.ResolveToAbsoluteURL(baseURL, feedURL); err == nil {
			feed.FeedURL = absoluteFeedURL
		}
	}

	// Populate the site URL.
	siteURL := a.atomFeed.Links.originalLink()
	if siteURL != "" {
		if absoluteSiteURL, err := urllib.ResolveToAbsoluteURL(baseURL, siteURL); err == nil {
			feed.SiteURL = absoluteSiteURL
		}
	}

	// Populate the feed title.
	feed.Title = a.atomFeed.Title.body()
	if feed.Title == "" {
		feed.Title = feed.SiteURL
	}

	// Populate the feed description.
	feed.Description = a.atomFeed.Subtitle.body()

	// Populate the feed icon.
	for _, value := range []string{a.atomFeed.Icon, a.atomFeed.Logo} {
		if value = strings.TrimSpace(value); value == "" {
			continue
		}

		if iconURL, err := urllib.ResolveToAbsoluteURL(feed.SiteURL, value); err == nil {
			feed.IconURL = iconURL
			break
		}
	}

	feed.Entries = a.populateEntries(feed.SiteURL)
	return feed
}

func (a *atom10Adapter) populateEntries(siteURL string) model.Entries {
	entries := make(model.Entries, 0, len(a.atomFeed.Entries))

	for _, atomEntry := range a.atomFeed.Entries {
		entry := model.NewEntry()

		// Populate the entry URL.
		entry.URL = atomEntry.Links.originalLink()
		if entry.URL != "" {
			if absoluteEntryURL, err := urllib.ResolveToAbsoluteURL(siteURL, entry.URL); err == nil {
				entry.URL = absoluteEntryURL
			}
		}

		// Populate the entry content.
		entry.Content = atomEntry.Content.body()
		if entry.Content == "" {
			entry.Content = atomEntry.Summary.body()
			if entry.Content == "" {
				entry.Content = atomEntry.FirstMediaDescription()
			}
		}

		// Populate the entry title.
		entry.Title = atomEntry.Title.title()
		if entry.Title == "" {
			entry.Title = sanitizer.TruncateHTML(entry.Content, 100)
			if entry.Title == "" {
				entry.Title = entry.URL
			}
		}

		// Populate the entry author.
		authors := atomEntry.Authors.personNames()
		if len(authors) == 0 {
			authors = a.atomFeed.Authors.personNames()
		}
		sort.Strings(authors)
		authors = slices.Compact(authors)
		entry.Author = strings.Join(authors, ", ")

		// Populate the entry date.
		for _, value := range []string{atomEntry.Published, atomEntry.Updated} {
			if value = strings.TrimSpace(value); value == "" {
				continue
			}

			parsedDate, err := date.Parse(value)
			if err != nil {
				slog.Debug("Unable to parse date from Atom 1.0 feed",
					slog.String("date", value),
					slog.String("url", entry.URL),
					slog.Any("error", err),
				)
				continue
			}

			entry.Date = parsedDate
			break
		}

		if entry.Date.IsZero() {
			entry.Date = time.Now()
		}

		// Populate categories.
		categories := atomEntry.Categories.CategoryNames()
		if len(categories) == 0 {
			categories = a.atomFeed.Categories.CategoryNames()
		}

		// Sort and deduplicate categories.
		sort.Strings(categories)
		entry.Tags = slices.Compact(categories)

		// Populate the commentsURL if defined.
		// See https://tools.ietf.org/html/rfc4685#section-4
		// If the type attribute of the atom:link is omitted, its value is assumed to be "application/atom+xml".
		// We accept only HTML or XHTML documents for now since the intention is to have the same behavior as RSS.
		commentsURL := atomEntry.Links.firstLinkWithRelationAndType("replies", "text/html", "application/xhtml+xml")
		if urllib.IsAbsoluteURL(commentsURL) {
			entry.CommentsURL = commentsURL
		}

		// Generate the entry hash.
		for _, value := range []string{atomEntry.ID, atomEntry.Links.originalLink()} {
			if value != "" {
				entry.Hash = crypto.SHA256(value)
				break
			}
		}

		// Populate the entry enclosures.
		uniqueEnclosuresMap := make(map[string]bool)

		for _, mediaThumbnail := range atomEntry.AllMediaThumbnails() {
			mediaURL := strings.TrimSpace(mediaThumbnail.URL)
			if mediaURL == "" {
				continue
			}

			if _, found := uniqueEnclosuresMap[mediaURL]; found {
				continue
			}

			mediaAbsoluteURL, err := urllib.ResolveToAbsoluteURL(siteURL, mediaURL)
			if err != nil {
				slog.Debug("Unable to build absolute URL for media thumbnail",
					slog.String("url", mediaThumbnail.URL),
					slog.String("site_url", siteURL),
					slog.Any("error", err),
				)
				continue
			}

			uniqueEnclosuresMap[mediaAbsoluteURL] = true

			entry.Enclosures = append(entry.Enclosures, &model.Enclosure{
				URL:      mediaAbsoluteURL,
				MimeType: mediaThumbnail.MimeType(),
				Size:     mediaThumbnail.Size(),
			})
		}

		for _, link := range atomEntry.Links.findAllLinksWithRelation("enclosure") {
			absoluteEnclosureURL, err := urllib.ResolveToAbsoluteURL(siteURL, link.Href)
			if err != nil {
				slog.Debug("Unable to resolve absolute URL for enclosure",
					slog.String("enclosure_url", link.Href),
					slog.String("entry_url", entry.URL),
					slog.Any("error", err),
				)
				continue
			}

			if _, found := uniqueEnclosuresMap[absoluteEnclosureURL]; found {
				continue
			}

			uniqueEnclosuresMap[absoluteEnclosureURL] = true

			length, _ := strconv.ParseInt(link.Length, 10, 0)
			entry.Enclosures = append(entry.Enclosures, &model.Enclosure{
				URL:      absoluteEnclosureURL,
				MimeType: link.Type,
				Size:     length,
			})
		}

		for _, mediaContent := range atomEntry.AllMediaContents() {
			mediaURL := strings.TrimSpace(mediaContent.URL)
			if mediaURL == "" {
				continue
			}

			mediaAbsoluteURL, err := urllib.ResolveToAbsoluteURL(siteURL, mediaURL)
			if err != nil {
				slog.Debug("Unable to build absolute URL for media content",
					slog.String("url", mediaContent.URL),
					slog.String("site_url", siteURL),
					slog.Any("error", err),
				)
				continue
			}

			if _, found := uniqueEnclosuresMap[mediaAbsoluteURL]; found {
				continue
			}

			uniqueEnclosuresMap[mediaAbsoluteURL] = true

			entry.Enclosures = append(entry.Enclosures, &model.Enclosure{
				URL:      mediaAbsoluteURL,
				MimeType: mediaContent.MimeType(),
				Size:     mediaContent.Size(),
			})
		}

		for _, mediaPeerLink := range atomEntry.AllMediaPeerLinks() {
			mediaURL := strings.TrimSpace(mediaPeerLink.URL)
			if mediaURL == "" {
				continue
			}

			mediaAbsoluteURL, err := urllib.ResolveToAbsoluteURL(siteURL, mediaURL)
			if err != nil {
				slog.Debug("Unable to build absolute URL for media peer link",
					slog.String("url", mediaPeerLink.URL),
					slog.String("site_url", siteURL),
					slog.Any("error", err),
				)
				continue
			}

			if _, found := uniqueEnclosuresMap[mediaAbsoluteURL]; found {
				continue
			}

			uniqueEnclosuresMap[mediaAbsoluteURL] = true

			entry.Enclosures = append(entry.Enclosures, &model.Enclosure{
				URL:      mediaAbsoluteURL,
				MimeType: mediaPeerLink.MimeType(),
				Size:     mediaPeerLink.Size(),
			})
		}

		entries = append(entries, entry)
	}

	return entries
}