feat: support IPFS/IPNS paths, Gateways, etc (#4)

* feat: support IPFS/IPNS paths, Gateways, etc Adds support for fetching resources in the form: - IPFS Path, e.g. `/ipfs/<CID>` - IPNS Path, e.g. `/ipns/<PeerId>` - IPFS Gateway Path, e.g. `http://example.com/ipfs/<CID>` - IPNS Gateway Path, e.g. `http://example.com/ipns/<PeerId>` - IPFS Subdomain Gateway Path, e.g. `http://<CID>.ipfs.example.com` - IPNS Subdomain Gateway Path, e.g. `http://<PeerId>.ipns.example.com` * chore: simplify regex and test https urls too * chore: apply suggestions from code review Co-authored-by: Daniel Norman <1992255+2color@users.noreply.github.com> * chore: fix tests * chore: more tests * chore: remove $ * chore: https? test group name --------- Co-authored-by: Daniel Norman <1992255+2color@users.noreply.github.com> Co-authored-by: Russell Dempsey <1173416+SgtPooki@users.noreply.github.com>
ipfs · Mar 3, 2024 · e7f1816 · e7f1816
1 parent 76485a4
commit e7f1816
Show file tree

Hide file tree

Showing 2 changed files with 625 additions and 147 deletions.
diff --git a/packages/verified-fetch/src/utils/parse-url-string.ts b/packages/verified-fetch/src/utils/parse-url-string.ts
@@ -30,7 +30,22 @@ export interface ParsedUrlStringResults {
   query: ParsedUrlQuery
 }
 
-const URL_REGEX = /^(?<protocol>ip[fn]s):\/\/(?<cidOrPeerIdOrDnsLink>[^/$?]+)\/?(?<path>[^$?]*)\??(?<queryString>.*)$/
+const URL_REGEX = /^(?<protocol>ip[fn]s):\/\/(?<cidOrPeerIdOrDnsLink>[^/?]+)\/?(?<path>[^?]*)\??(?<queryString>.*)$/
+const PATH_REGEX = /^\/(?<protocol>ip[fn]s)\/(?<cidOrPeerIdOrDnsLink>[^/?]+)\/?(?<path>[^?]*)\??(?<queryString>.*)$/
+const PATH_GATEWAY_REGEX = /^https?:\/\/(.*[^/])\/(?<protocol>ip[fn]s)\/(?<cidOrPeerIdOrDnsLink>[^/?]+)\/?(?<path>[^?]*)\??(?<queryString>.*)$/
+const SUBDOMAIN_GATEWAY_REGEX = /^https?:\/\/(?<cidOrPeerIdOrDnsLink>[^/.?]+)\.(?<protocol>ip[fn]s)\.([^/?]+)\/?(?<path>[^?]*)\??(?<queryString>.*)$/
+
+function matchURLString (urlString: string): Record<string, string> {
+  for (const pattern of [URL_REGEX, PATH_REGEX, PATH_GATEWAY_REGEX, SUBDOMAIN_GATEWAY_REGEX]) {
+    const match = urlString.match(pattern)
+
+    if (match?.groups != null) {
+      return match.groups
+    }
+  }
+
+  throw new TypeError(`Invalid URL: ${urlString}, please use ipfs://, ipns://, or gateway URLs only`)
+}
 
 /**
  * A function that parses ipfs:// and ipns:// URLs, returning an object with easily recognizable properties.
@@ -41,13 +56,7 @@ const URL_REGEX = /^(?<protocol>ip[fn]s):\/\/(?<cidOrPeerIdOrDnsLink>[^/$?]+)\/?
  */
 export async function parseUrlString ({ urlString, ipns, logger }: ParseUrlStringInput, options?: ParseUrlStringOptions): Promise<ParsedUrlStringResults> {
   const log = logger.forComponent('helia:verified-fetch:parse-url-string')
-  const match = urlString.match(URL_REGEX)
-
-  if (match == null || match.groups == null) {
-    throw new TypeError(`Invalid URL: ${urlString}, please use ipfs:// or ipns:// URLs only.`)
-  }
-
-  const { protocol, cidOrPeerIdOrDnsLink, path: urlPath, queryString } = match.groups
+  const { protocol, cidOrPeerIdOrDnsLink, path: urlPath, queryString } = matchURLString(urlString)
 
   let cid: CID | undefined
   let resolvedPath: string | undefined