Fix handling of multiple attributes with the same name

Closes #8.
jsdom · Feb 23, 2020 · cb04b18 · cb04b18
1 parent 9fd7d44
commit cb04b18
Show file tree

Hide file tree

Showing 3 changed files with 38 additions and 1 deletion.
diff --git a/lib/html-encoding-sniffer.js b/lib/html-encoding-sniffer.js
@@ -51,14 +51,16 @@ function prescanMetaCharset(buffer) {
          (isSpaceCharacter(c5) || c5 === 0x2F)) {
         // "meta" + space or /
         i += 6;
+        const attributeList = new Set();
         let gotPragma = false;
         let needPragma = null;
         let charset = null;
 
         let attrRes;
         do {
           attrRes = getAttribute(buffer, i, l);
-          if (attrRes.attr) {
+          if (attrRes.attr && !attributeList.has(attrRes.attr.name)) {
+            attributeList.add(attrRes.attr.name);
             if (attrRes.attr.name === "http-equiv") {
               gotPragma = attrRes.attr.value === "content-type";
             } else if (attrRes.attr.name === "content" && !charset) {

diff --git a/test/no-bom-charset-http-equiv-refresh.html b/test/no-bom-charset-http-equiv-refresh.html
@@ -0,0 +1,7 @@
+<!DOCTYPE html>
+<html>
+    <head>
+        <meta http-equiv=refresh http-equiv=Content-Type content="text/html;charset=iso-8859-2">
+    </head>
+    <body></body>
+</html>
diff --git a/test/tests.js b/test/tests.js
@@ -289,6 +289,34 @@ describe("A file with no BOM and a <meta http-equiv> with 'charsetcharset'", ()
   });
 });
 
+describe("A file with no BOM and a <meta http-equiv=refresh> with another http-equiv", () => {
+  const buffer = read("no-bom-charset-http-equiv-refresh.html");
+
+  it("should sniff as windows-1252, given no options", () => {
+    const sniffedEncoding = htmlEncodingSniffer(buffer);
+
+    assert.strictEqual(sniffedEncoding, "windows-1252");
+  });
+
+  it("should sniff as the transport layer encoding, given that", () => {
+    const sniffedEncoding = htmlEncodingSniffer(buffer, {
+      transportLayerEncodingLabel: "windows-1251",
+      defaultEncoding: "ISO-8859-16"
+    });
+
+    assert.strictEqual(sniffedEncoding, "windows-1251");
+  });
+
+
+  it("should sniff as the default encoding, given that", () => {
+    const sniffedEncoding = htmlEncodingSniffer(buffer, {
+      defaultEncoding: "ISO-8859-16"
+    });
+
+    assert.strictEqual(sniffedEncoding, "ISO-8859-16");
+  });
+});
+
 for (const utf16Encoding of ["utf-16be", "utf-16", "utf-16le"]) {
   describe(`A file with a BOM and a <meta charset> of ${utf16Encoding}`, () => {
     const buffer = read(`no-bom-charset-${utf16Encoding}.html`);