From e7b2fad441fd6092354f47764d340e1223da290e Mon Sep 17 00:00:00 2001
From: alex <alex@cloudware.io>
Date: Sat, 10 Dec 2022 17:58:47 +0100
Subject: [PATCH] verifyEventLink: check cache first to allow more valid URLs

previously, a request to an open graph image reported from say
https://github.com/example/some.repo would result in 404 not found, a
false negative because "is link in event?" verification would fail:
"some.repo" looks like a file with .repo extension, not an html page.

in this commit, if a client already requested example/some.repo
before, and it is cached, the response to the OGP image is successful.
---
 noxy.go | 52 ++++++++++++++++++++++++++++++++++++++--------------
 1 file changed, 38 insertions(+), 14 deletions(-)

diff --git a/noxy.go b/noxy.go
index f2d321e..bbc5269 100644
--- a/noxy.go
+++ b/noxy.go
@@ -283,25 +283,28 @@ func (x *Noxer) verifyEventLink(ctx context.Context, eventID, relayURL, link str
 
 	// link not found in the event text/json.
 	// check URLs in OGP metadata for each suitable link found in the event.
-	for _, urlStr := range eventURLs {
-		u, err := url.Parse(urlStr)
-		if err != nil {
-			continue // invalid url
-		}
-		if ext := path.Ext(u.Path); ext != "" {
-			if !strings.HasSuffix(ext, "html") && !strings.HasSuffix(ext, "htm") {
-				continue // assume not an html page
+	for _, urlInEvent := range eventURLs {
+		// try only cache first. a client may have already requested /meta
+		// with this URL. if so, need no further parsing and network roundtrips.
+		var cachedMeta LinkMeta
+		if x.Cache.GetJSON(ctx, MakeCacheKey(urlInEvent, CacheKeyURLPreview), &cachedMeta) == nil {
+			if nonSortedSliceContains(cachedMeta.ImageURLs, link) {
+				return nil // ok; found
 			}
+			continue // move on to the next url in the event
+		}
+
+		// cached failed or miss; possibly fetch from remote and parse.
+		if !looksLikeHTMLPage(urlInEvent) {
+			continue
 		}
-		meta, err := x.slurpLinkMeta(ctx, urlStr)
+		meta, err := x.slurpLinkMeta(ctx, urlInEvent)
 		if err != nil {
-			log.Printf("verifyEventLink slurpLinkMeta(%s): %v", u, err)
+			log.Printf("verifyEventLink slurpLinkMeta(%s): %v", urlInEvent, err)
 			continue
 		}
-		for _, imgURL := range meta.ImageURLs {
-			if imgURL == link {
-				return nil
-			}
+		if nonSortedSliceContains(meta.ImageURLs, link) {
+			return nil // ok; found
 		}
 	}
 	return ErrNotFound
@@ -591,3 +594,24 @@ func validURL(urlStr string) bool {
 	}
 	return u.Scheme == "" || u.Scheme == "http" || u.Scheme == "https"
 }
+
+// O(N) lookup of elem in a
+func nonSortedSliceContains(a []string, elem string) bool {
+	for _, v := range a {
+		if v == elem {
+			return true
+		}
+	}
+	return false
+}
+
+// reports whether urlStr looks like a URL of an html webpage.
+func looksLikeHTMLPage(urlStr string) bool {
+	u, err := url.Parse(urlStr)
+	if err != nil {
+		return false
+	}
+	ext := path.Ext(u.Path)
+	// any .xxxhtml is ok
+	return ext == "" || strings.HasSuffix(ext, "html") || strings.HasSuffix(ext, "htm")
+}