Modify Charset Aliases : Stop treat ISO-8859-1 as Windows-1252's alias

apache · Dec 5, 2020 · 0e7b475 · 0e7b475
1 parent 813a6eb
commit 0e7b475
Show file tree

Hide file tree

Showing 2 changed files with 7 additions and 8 deletions.
diff --git a/...html-module/src/main/java/org/apache/tika/parser/html/charsetdetector/CharsetAliases.java b/...html-module/src/main/java/org/apache/tika/parser/html/charsetdetector/CharsetAliases.java
@@ -98,8 +98,9 @@ private static void addAll() {
         addCharset(charset("windows-1250"), "cp1250", "windows-1250", "x-cp1250");
         addCharset(charset("windows-1251"), "cp1251", "windows-1251", "x-cp1251");
         addCharset(charset("windows-1252"), "ansi_x3.4-1968", "ascii", "cp1252", "cp819", "csisolatin1",
-                "ibm819", "iso-8859-1", "iso-ir-100", "iso8859-1", "iso88591", "iso_8859-1", "iso_8859-1:1987",
-                "l1", "latin1", "us-ascii", "windows-1252", "x-cp1252");
+                "ibm819", "l1", "latin1", "us-ascii", "windows-1252", "x-cp1252");
+        addCharset(charset("ISO-8859-1"), "iso-8859-1", "iso-ir-100", "iso8859-1", "iso88591",
+                "iso_8859-1", "iso_8859-1:1987");
         addCharset(charset("windows-1253"), "cp1253", "windows-1253", "x-cp1253");
         addCharset(charset("windows-1254"), "cp1254", "csisolatin5", "iso-8859-9", "iso-ir-148",
                 "iso8859-9", "iso88599", "iso_8859-9", "iso_8859-9:1989", "l5", "latin5", "windows-1254", "x-cp1254");

diff --git a/...ml-module/src/test/java/org/apache/tika/parser/html/StandardHtmlEncodingDetectorTest.java b/...ml-module/src/test/java/org/apache/tika/parser/html/StandardHtmlEncodingDetectorTest.java
@@ -177,8 +177,7 @@ public void replacement() throws IOException {
 
     @Test
     public void iso88591() throws IOException {
-        // In the spec, iso-8859-1 is an alias for WINDOWS-1252
-        assertWindows1252("<meta charset='iso-8859-1'>");
+        assertCharset("<meta charset='iso-8859-1'>", StandardCharsets.ISO_8859_1);
     }
 
     @Test
@@ -294,10 +293,9 @@ public void withCompactComment() throws IOException {
     @Test
     public void withCharsetInContentType() throws IOException {
         metadata.set(Metadata.CONTENT_TYPE, "text/html; Charset=ISO-8859-1");
-        // ISO-8859-1 is an alias for WINDOWS-1252, even if it's set at the transport layer level
-        assertWindows1252("");
-        assertWindows1252("<meta charset='UTF-8'>");
-        assertWindows1252("<meta http-equiv='content-type' content='charset=utf-8'>");
+        assertCharset("", StandardCharsets.ISO_8859_1);
+        assertCharset("<meta charset='UTF-8'>", StandardCharsets.ISO_8859_1);
+        assertCharset("<meta http-equiv='content-type' content='charset=utf-8'>", StandardCharsets.ISO_8859_1);
         // if a BOM is present, it has precedence over transport layer information
         assertCharset("\ufeff<meta charset='WINDOWS-1252'>", StandardCharsets.UTF_8);
         assertCharset("\ufeff<meta charset='WINDOWS-1252'>", StandardCharsets.UTF_16LE);