Only skip tags with case considered

pombredanne · pombredanne · commit c3408122500e · 2024-09-11T23:20:37.000+02:00
Do not lower tags first

Signed-off-by: Philippe Ombredanne &lt;pombredanne@nexb.com&gt;
diff --git a/src/textcode/markup.py b/src/textcode/markup.py
@@ -1931,7 +1931,9 @@ def strip_known_markup_from_text(text):
 
 SKIP_ATTRIBUTES = (
     "href=",
+    "HREF=",
     "class=",
+    "CLASS=",
     "width=",
     "@end",
     "@group",
@@ -1940,7 +1942,8 @@ def strip_known_markup_from_text(text):
     "xmlns:",
     "xml:",
     "lang=",
-    "<windows"
+    "<Windows",
+    "<windows",
 )
 
 
@@ -1960,7 +1963,7 @@ def keep_tag(token, skips_tags=ALL_TAGS, skip_attributes=SKIP_ATTRIBUTES, kept_t
     if any(k in tlow for k in kept_tags):
         return True
 
-    if tlow.startswith(skip_attributes):
+    if token.startswith(skip_attributes):
         return False
 
     if tlow in skips_tags or tlow == ">":
diff --git a/tests/formattedcode/data/common/manifests-expected.yaml b/tests/formattedcode/data/common/manifests-expected.yaml
@@ -29,10 +29,10 @@ headers:
       system_environment:
         operating_system: linux
         cpu_architecture: 64
-        platform: Linux-5.15.0-116-generic-x86_64-with-glibc2.35
-        platform_version: '#126-Ubuntu SMP Mon Jul 1 10:14:24 UTC 2024'
-        python_version: 3.10.12 (main, Mar 22 2024, 16:50:05) [GCC 11.4.0]
-      spdx_license_list_version: '3.24'
+        platform: Linux-6.5.0-1025-oem-x86_64-with-glibc2.35
+        platform_version: '#26-Ubuntu SMP PREEMPT_DYNAMIC Tue Jun 18 12:35:22 UTC 2024'
+        python_version: 3.11.8 (main, Feb 12 2024, 14:25:06) [GCC 11.4.0]
+      spdx_license_list_version: '3.25'
       files_count: 4
 summary:
   declared_license_expression: apache-2.0 AND cddl-1.0 AND lgpl-3.0 AND mit
@@ -1399,7 +1399,7 @@ license_references:
     minimum_coverage: '0'
     standard_notice:
     ignorable_copyrights:
-      - Copyright (c) 2007 Free Software Foundation, Inc. <https://fsf.org/>
+      - Copyright (c) 2007 Free Software Foundation, Inc. https://fsf.org
     ignorable_holders:
       - Free Software Foundation, Inc.
     ignorable_authors: []
diff --git a/tests/textcode/data/markup_expected/PDL.html.stripmarkup.json b/tests/textcode/data/markup_expected/PDL.html.stripmarkup.json
@@ -126,7 +126,7 @@
   "})();",
   "",
   "",
-  "<body lang=\"en-US\"",
+  "<body",
   "id=\"bannera\"",
   "id=\"languagesdiv\"",
   "English:",
diff --git a/tests/textcode/data/markup_expected/java.jsp.stripmarkup.json b/tests/textcode/data/markup_expected/java.jsp.stripmarkup.json
@@ -15,7 +15,7 @@
   "<%--!DOCTYPE html PUBLIC \"-//W3C//DTD XHTML 1.1//EN\"",
   "\"http://www.w3.org/TR/xhtml11/DTD/xhtml11.dtd\"--%",
   "",
-  "xml:lang=\"en\"",
+  "",
   "<% IServerInfo serverInfos = (IServerInfo) request.getAttribute(\"serverInfos\");",
   "String webFilesRoot = (String) request.getAttribute(\"WebFilesRoot\"); %",
   "<%--c:url value=\"${pageContext.request.servletPath}\" var=\"submitUrl\" scope=\"page\"/--% <%-- can use value=\"${pageContext.request.servletPath}\" because this JSP is include()'ed --%",
diff --git a/tests/textcode/data/markup_expected/services.htm.stripmarkup.json b/tests/textcode/data/markup_expected/services.htm.stripmarkup.json
@@ -1,6 +1,6 @@
 [
   "<!DOCTYPE html",
-  "lang=\"en-US\"",
+  "",
   "",
   "charset=\"utf-8\"",
   "",

Original file line number	Diff line number	Diff line change
`@@ -1,6 +1,6 @@`
`1`	`1`	`[`
`2`	`2`	`"<!DOCTYPE html",`
`3`		`- "lang=\"en-US\"",`
	`3`	`+ "",`
`4`	`4`	`"",`
`5`	`5`	`"charset=\"utf-8\"",`
`6`	`6`	`"",`