From ff0fb55d60f4c29ffb6b8a69c2be3469f794990a Mon Sep 17 00:00:00 2001
From: Shulhan <ms@kilabit.info>
Date: Sun, 1 Jun 2025 16:02:46 +0700
Subject: all: brokenlinks should scan only URL on given path

Previously, if we pass the URL with path to brokenlinks, for example
"web.tld/path" it will scan all of the pages in the website "web.tld".

Now, it only scan the "/path" and its sub paths.
---
 README                | 35 ++++++++++++++++++++++++++++++++++-
 brokenlinks_test.go   | 25 ++-----------------------
 brokenlinks_worker.go |  2 +-
 3 files changed, 37 insertions(+), 25 deletions(-)

diff --git a/README b/README
index a26722f..2d8b273 100644
--- a/README
+++ b/README
@@ -20,8 +20,41 @@ Available commands,
         Links will be scanned on anchor href attribute ("<a href=...>") or
         on the image src attribute ("<img src=...").
 
+        The URL can be start from base or from sub path.
+        Scanning from path only report brokenlinks on that path and their
+        sub paths.
+        For example, given a website that have the following pages,
+
+                - web.tld (base)
+                - web.tld/page1
+                - web.tld/page1/sub1
+                - web.tld/page2
+                - web.tld/page2/sub1
+
+        Invoking brokenlinks with
+
+                $ jarink brokenlinks https://web.tld
+
+        will scan all of the pages, but invoking brokenlinks on path
+        "/page2"
+
+                $ jarink brokenlinks https://web.tld/page2
+
+        only scan "/page2" and "/page2/sub1".
+
         Once finished it will print the page and list of broken links in
-        JSON format to standard output.
+        JSON format to standard output,
+
+                {
+                        "$PAGE": [{
+                                "Link": <string>,
+                                "Error": <string>,
+                                "Code": <integer>
+                        },
+                        ...
+                        ],
+                        ...
+                }
 
         This command accept the following options,
 
diff --git a/brokenlinks_test.go b/brokenlinks_test.go
index c1a607f..1c43937 100644
--- a/brokenlinks_test.go
+++ b/brokenlinks_test.go
@@ -126,31 +126,10 @@ func TestBrokenlinks(t *testing.T) {
 			},
 		},
 	}, {
+		// Scanning on "/path" should not scan the the "/" or other
+		// pages other than below of "/path" itself.
 		scanUrl: testUrl + `/page2`,
 		exp: map[string][]jarink.Broken{
-			testUrl: []jarink.Broken{
-				{
-					Link: testUrl + `/broken.png`,
-					Code: http.StatusNotFound,
-				}, {
-					Link: testUrl + `/brokenPage`,
-					Code: http.StatusNotFound,
-				}, {
-					Link:  `http://127.0.0.1:abc`,
-					Error: `parse "http://127.0.0.1:abc": invalid port ":abc" after host`,
-					Code:  jarink.StatusBadLink,
-				}, {
-					Link:  `http:/127.0.0.1:11836`,
-					Error: `Head "http:/127.0.0.1:11836": http: no Host in request URL`,
-					Code:  jarink.StatusBadLink,
-				},
-			},
-			testUrl + `/broken.html`: []jarink.Broken{
-				{
-					Link: testUrl + `/brokenPage`,
-					Code: http.StatusNotFound,
-				},
-			},
 			testUrl + `/page2`: []jarink.Broken{
 				{
 					Link: testUrl + `/broken.png`,
diff --git a/brokenlinks_worker.go b/brokenlinks_worker.go
index a196d2f..5cc8c25 100644
--- a/brokenlinks_worker.go
+++ b/brokenlinks_worker.go
@@ -303,7 +303,7 @@ func (wrk *brokenlinksWorker) scan(linkq linkQueue) {
 		}
 		_, seen := resultq[nodeLink.url]
 		if !seen {
-			if !strings.HasPrefix(nodeLink.url, wrk.baseUrl.String()) {
+			if !strings.HasPrefix(nodeLink.url, wrk.scanUrl.String()) {
 				nodeLink.isExternal = true
 			}
 			resultq[nodeLink.url] = *nodeLink
-- 
cgit v1.3