Merge pull request #175 from chalin/chalin-ignoreurl-regex-for-intern…

…al-too-2021-09-17
wjdp · Mar 28, 2022 · 6aa3355 · 6aa3355
2 parents 54089cb + 72bd8a3
commit 6aa3355
Show file tree

Hide file tree

Showing 4 changed files with 60 additions and 10 deletions.
diff --git a/README.md b/README.md
@@ -158,7 +158,7 @@ htmltest uses a YAML configuration file. Put `.htmltest.yml` in the same directo
 | `EnforceHTML5` | Fails when the doctype isn't `<!DOCTYPE html>`. | `false` |
 | `EnforceHTTPS` | Fails when encountering an `http://` link. Useful to prevent mixed content errors when serving over HTTPS. | `false` |
 | `IgnoreURLs` | Array of regexs of URLs to ignore. | empty |
-| `IgnoreInternalURLs` | Array of strings of internal URLs to ignore. | empty |
+| `IgnoreInternalURLs` | Array of strings of internal URLs to ignore. Exact matches only. ⚠ Likely to be deprecated, use `IgnoreURLs` instead. | empty |
 | `IgnoreDirs` | Array of regexs of directories to ignore when scanning for HTML files. | empty |
 | `IgnoreInternalEmptyHash` | When true prevents raising an error for links with `href="#"`. | `false` |
 | `IgnoreEmptyHref` | When true prevents raising an error for links with `href=""`. | `false` |
@@ -189,8 +189,7 @@ DirectoryPath: "_site"
 EnforceHTTPS: true
 IgnoreURLs:
 - "example.com"
-IgnoreInternalURLs:
-- "/misc/js/script.js"
+- "^/misc/js/script.js$"
 IgnoreDirs:
 - "lib"
 CacheExpires: "6h"

diff --git a/htmltest/check-link.go b/htmltest/check-link.go
@@ -277,11 +277,11 @@ func (hT *HTMLTest) checkInternal(ref *htmldoc.Reference) {
 		return
 	}
 
-	// Solve #168
 	urlStr := ref.URLString()
 
-	// Does this internal url match an internal url ignore rule?
-	if hT.opts.isInternalURLIgnored(urlStr) {
+	// Does this internal url match either a standard URL ignore rule or internal
+	// url ignore rule?
+	if hT.opts.isInternalURLIgnored(urlStr) || hT.opts.isURLIgnored(urlStr) {
 		return
 	}
 

diff --git a/htmltest/check-link_test.go b/htmltest/check-link_test.go
@@ -324,6 +324,24 @@ func TestAnchorInternalBrokenIgnore(t *testing.T) {
 	tExpectIssueCount(t, hT, 0)
 }
 
+func TestAnchorInternalIgnoreUrl(t *testing.T) {
+	// ignores internal links in IgnoreURLs
+	hT := tTestFileOpts("fixtures/links/brokenLinkInternal.html",
+		map[string]interface{}{
+			"IgnoreURLs": []interface{}{"no\\w+.html"},
+		})
+	tExpectIssueCount(t, hT, 0)
+}
+
+func TestAnchorFileIgnoreUrl(t *testing.T) {
+	// ignores file links in IgnoreURLs
+	hT := tTestFileOpts("fixtures/links/brokenLinkFile.html",
+		map[string]interface{}{
+			"IgnoreURLs": []interface{}{"no\\w+.html"},
+		})
+	tExpectIssueCount(t, hT, 0)
+}
+
 func TestAnchorInternalRelativeLinksBase(t *testing.T) {
 	// passes for relative links with a base
 	hT := tTestFile("fixtures/links/relativeLinksWithBase.html")
@@ -498,20 +516,43 @@ func TestAnchorInternalHashWeird(t *testing.T) {
 	tExpectIssueCount(t, hT, 0)
 }
 
-func TestAnchorInternalUrl(t *testing.T) {
-	// fails for internal linking writen not in IgnoreInternalURLs (#168)
+func TestAnchorInternalUrlDoesNotExist(t *testing.T) {
+	// fails for internal link not in IgnoreInternalURLs (#168)
 	hT := tTestFile("fixtures/links/link_directory_internal_invalid.html")
 	tExpectIssueCount(t, hT, 1)
 	tExpectIssue(t, hT, "target does not exist", 1)
 }
 
-func TestAnchorInternalUrlOption(t *testing.T) {
-	// passes for internal linking writen in IgnoreInternalURLs option (#168)
+func TestAnchorInternalUrlIgnoreUsingIgnoreInternalURLs(t *testing.T) {
+	// passes for internal link in IgnoreInternalURLs option (#168)
 	hT := tTestFileOpts("fixtures/links/link_directory_internal_valid.html",
 		map[string]interface{}{"IgnoreInternalURLs": []interface{}{"/misc/js/script.js"}})
 	tExpectIssueCount(t, hT, 0)
 }
 
+func TestAnchorInternalUrlIgnoreInternalURLsIsStrict(t *testing.T) {
+	// fails as IgnoreInternalURLs requires a string match
+	hT := tTestFileOpts("fixtures/links/link_directory_internal_valid.html",
+		map[string]interface{}{"IgnoreInternalURLs": []interface{}{"misc/js/script.js"}})
+	tExpectIssueCount(t, hT, 1)
+	tExpectIssue(t, hT, "target does not exist", 1)
+}
+
+func TestAnchorInternalUrlIgnoreUsingIgnoreURLs(t *testing.T) {
+	// passes for internal link in IgnoreURLs option using regex
+	hT := tTestFileOpts("fixtures/links/link_directory_internal_valid.html",
+		map[string]interface{}{"IgnoreURLs": []interface{}{"^/misc/js/script.js$"}})
+	tExpectIssueCount(t, hT, 0)
+}
+
+func TestAnchorInternalUrlNotIgnoreUsingIgnoreURLs(t *testing.T) {
+	// fails for internal link in when doesn't match strict regex in IgnoreURLs
+	hT := tTestFileOpts("fixtures/links/link_directory_internal_invalid.html",
+		map[string]interface{}{"IgnoreURLs": []interface{}{"^/misc/js/script.js$"}})
+	tExpectIssueCount(t, hT, 1)
+	tExpectIssue(t, hT, "target does not exist", 1)
+}
+
 func TestAnchorMultipleProblems(t *testing.T) {
 	// finds a mix of broken and unbroken links
 	t.Skip("Only single problem, and an hash which is not yet supported.")

diff --git a/htmltest/fixtures/links/brokenLinkFile.html b/htmltest/fixtures/links/brokenLinkFile.html
@@ -0,0 +1,10 @@
+<html>
+
+<body>
+
+	<p>Blah blah blah. <a href="file://a/b/c/notreal.html">Not a real link!</a></p>
+<p id="safeHash">Blah blah blah. <a href="./missingLinkHref.html">A real link!</a></p>
+
+</body>
+
+</html>