WIP Convert Spotlight PA embeds to native

internal/iterx/iterx.go

@@ -3,7 +3,11 @@
 // Package iterx has iteration utilities.
 package iterx
 
-import "iter"
+import (
+	"cmp"
+	"iter"
+	"slices"
+)
 
 // Filter returns a sequence of matching items.
 func Filter[T any](seq iter.Seq[T], match func(T) bool) iter.Seq[T] {
@@ -32,3 +36,19 @@ func First[T any](seq iter.Seq[T]) (v T) {
 	}
 	return
 }
+
+func Keys[M ~map[K]V, K comparable, V any](m M) iter.Seq[K] {
+	return func(yield func(K) bool) {
+		for k := range m {
+			if !yield(k) {
+				return
+			}
+		}
+	}
+}
+
+func Sorted[T cmp.Ordered](seq iter.Seq[T]) []T {
+	s := Collect(seq)
+	slices.Sort(s)
+	return s
+}

pkg/almanack/service-gdocs.go

@@ -10,6 +10,7 @@ import (
 	"errors"
 	"fmt"
 	"net/http"
+	"net/url"
 	"slices"
 	"strconv"
 	"strings"
@@ -21,6 +22,7 @@ import (
 	"github.com/spotlightpa/almanack/internal/blocko"
 	"github.com/spotlightpa/almanack/internal/db"
 	"github.com/spotlightpa/almanack/internal/gdocs"
+	"github.com/spotlightpa/almanack/internal/iterx"
 	"github.com/spotlightpa/almanack/internal/must"
 	"github.com/spotlightpa/almanack/internal/stringx"
 	"github.com/spotlightpa/almanack/internal/xhtml"
@@ -234,7 +236,7 @@ func (svc Services) ProcessGDocsDoc(ctx context.Context, dbDoc db.GDocsDoc) (err
 	blocko.RemoveMarks(docHTML)
 
 	// Warn about fake headings
-	for n := range xhtml.All(docHTML) {
+	for n := range xhtml.Children(docHTML) {
 		// <p> with only b/i/strong/em for a child
 		if n.DataAtom != atom.P {
 			continue
@@ -306,6 +308,46 @@ func removeTail(n *html.Node) {
 	}
 }
 
+func replaceSpotlightEmbeds(s string) string {
+	n, err := html.Parse(strings.NewReader(s))
+	if err != nil {
+		// l := almlog.FromContext(ctx)
+		// l.WarnContext(ctx, "invalid HTML in embed", "html", s)
+		return s
+	}
+	// data-spl-embed-version="1"
+	div := xhtml.Select(n, func(n *html.Node) bool {
+		return n.DataAtom == atom.Div && xhtml.Attr(n, "data-spl-embed-version") == "1"
+	})
+	if div == nil {
+		return s
+	}
+	netloc := xhtml.Attr(div, "data-spl-src")
+	u, err := url.Parse(netloc)
+	if err != nil {
+		// l := almlog.FromContext(ctx)
+		// l.WarnContext(ctx, "invalid URL in embed", "html", s, "url", netloc)
+		return s
+	}
+	tag := strings.Trim(u.Path, "/")
+	q := u.Query()
+	var buf strings.Builder
+	buf.WriteString("{{<")
+	buf.WriteString(tag)
+	for _, k := range iterx.Sorted(iterx.Keys(q)) {
+		vv := q[k]
+		for _, v := range vv {
+			buf.WriteString(" ")
+			buf.WriteString(k)
+			buf.WriteString("=\"")
+			buf.WriteString(html.EscapeString(v))
+			buf.WriteString("\"")
+		}
+	}
+	buf.WriteString(">}}\n")
+	return buf.String()
+}
+
 func (svc Services) replaceImageEmbed(
 	ctx context.Context,
 	tbl *html.Node,
@@ -592,9 +634,10 @@ func fixMarkdownPlaceholders(rawHTML *html.Node) {
 		case db.PartnerRawEmbedTag, db.PartnerTextTag:
 			dataEl.Parent.RemoveChild(dataEl)
 		case db.RawEmbedTag, db.SpotlightRawEmbedOrTextTag:
+			data := replaceSpotlightEmbeds(embed.Value.(string))
 			xhtml.ReplaceWith(dataEl, &html.Node{
 				Type: html.RawNode,
-				Data: embed.Value.(string),
+				Data: data,
 			})
 		case db.ToCEmbedTag:
 			container := xhtml.New("div")

pkg/almanack/service-gdocs_test.go

@@ -1,9 +1,11 @@
 package almanack
 
 import (
+	"context"
 	"testing"
 
 	"github.com/carlmjohnson/be"
+	"github.com/spotlightpa/almanack/pkg/almlog"
 )
 
 func TestImageCAS(t *testing.T) {
@@ -15,7 +17,6 @@ func TestImageCAS(t *testing.T) {
 		{"Hello, World!", "image/jpeg", "cas/cpme-4zc8-f4m3-gcdp.jpeg"},
 	}
 	for _, tc := range cases {
-		tc := tc
 		t.Run("", func(t *testing.T) {
 			got := makeCASaddress([]byte(tc.body), tc.ct)
 			be.Equal(t, tc.want, got)
@@ -30,3 +31,29 @@ func TestImageCAS(t *testing.T) {
 		}
 	}
 }
+
+func TestReplaceSpotlightEmbeds(t *testing.T) {
+	cases := []struct {
+		in, want string
+	}{
+		{"", ""},
+		{"Hello, World!", "Hello, World!"},
+		{
+			`<script src="https://www.spotlightpa.org/embed.js" async></script><div data-spl-embed-version="1" data-spl-src="https://www.spotlightpa.org/embeds/cta/"></div>`,
+			"",
+		},
+		{`
+		<script src="https://www.spotlightpa.org/embed.js" async></script>
+
+		<div data-spl-embed-version="1" data-spl-src="https://www.spotlightpa.org/embeds/cta/"></div>
+		`,
+			`
+		`},
+	}
+	ctx := context.Background()
+	almlog.UseDevLogger()
+	for _, tc := range cases {
+		t := be.Relaxed(t)
+		be.Equal(t, tc.want, replaceSpotlightEmbeds(ctx, tc.in))
+	}
+}