This is an automated email from the ASF dual-hosted git repository.
snoopdave pushed a commit to branch roller-6.1.x
in repository https://gitbox.apache.org/repos/asf/roller.git
The following commit(s) were added to refs/heads/roller-6.1.x by this push:
new 931161efb Keep blogroll links to web and relative URLs (#221)
931161efb is described below
commit 931161efb103ba6345043b0ef4815b8e7a876bb4
Author: David M. Johnson <[email protected]>
AuthorDate: Thu Oct 29 22:40:08 2026 -0400
Keep blogroll links to web and relative URLs (#221)
* Keep blogroll links to web and relative URLs
* Log skipped OPML outlines and note the import change in CHANGES.md
* Note in CHANGES.md that OPML links must parse as URIs
* Read blogroll link schemes one way and escape bookmark names
---
CHANGES.md | 6 ++
.../business/jpa/JPABookmarkManagerImpl.java | 33 ++++++++++
.../pojos/wrapper/WeblogBookmarkWrapper.java | 27 +++++++-
.../apache/roller/weblogger/util/LinkScheme.java | 73 ++++++++++++++++++++++
app/src/main/webapp/WEB-INF/velocity/weblog.vm | 16 ++---
.../business/BookmarkImportParsingTest.java | 32 ++++++++++
.../pojos/wrapper/WeblogBookmarkWrapperTest.java | 56 +++++++++++++++++
.../roller/weblogger/util/LinkSchemeTest.java | 51 +++++++++++++++
8 files changed, 280 insertions(+), 14 deletions(-)
diff --git a/CHANGES.md b/CHANGES.md
index c0070846e..7dcfa9c18 100644
--- a/CHANGES.md
+++ b/CHANGES.md
@@ -64,6 +64,12 @@
a `<`, as in `a < b`, `x<y` or `C<C++ fans`, is kept, and no longer gets a
space added before the `<`. The Your Weblogs page shows a weblog's
description as text.
+- **OPML import keeps only web links.** An imported bookmark is kept only when
+ its link is an absolute `http` or `https` URL; others are skipped and named
+ in the log. A feed URL that is not one is dropped. Blogroll links shown on
+ weblog pages must be relative or use `http`, `https`, `ftp` or `mailto`;
+ other links, and links with `&` before the path, are shown without a
+ target. The blogroll escapes bookmark names and descriptions.
### Bug fixes
diff --git
a/app/src/main/java/org/apache/roller/weblogger/business/jpa/JPABookmarkManagerImpl.java
b/app/src/main/java/org/apache/roller/weblogger/business/jpa/JPABookmarkManagerImpl.java
index f1fa7478e..77c919fc1 100644
---
a/app/src/main/java/org/apache/roller/weblogger/business/jpa/JPABookmarkManagerImpl.java
+++
b/app/src/main/java/org/apache/roller/weblogger/business/jpa/JPABookmarkManagerImpl.java
@@ -32,6 +32,7 @@ import org.apache.roller.weblogger.business.Weblogger;
import org.apache.roller.weblogger.pojos.WeblogBookmark;
import org.apache.roller.weblogger.pojos.WeblogBookmarkFolder;
import org.apache.roller.weblogger.pojos.Weblog;
+import org.apache.roller.weblogger.util.LinkScheme;
import org.jdom2.Document;
import org.jdom2.Element;
import org.jdom2.input.JDOMParseException;
@@ -228,6 +229,19 @@ public class JPABookmarkManagerImpl implements
BookmarkManager {
desc = null!=desc ? desc : title;
xmlUrl = null!=xmlUrl ? xmlUrl : url;
url = null!=htmlUrl ? htmlUrl : url;
+
+ // keep only web links; an outline with anything else gets no url
+ String importedUrl = url;
+ String importedXmlUrl = xmlUrl;
+ url = webLinkOrNull(url);
+ xmlUrl = webLinkOrNull(xmlUrl);
+ if (importedUrl != null && url == null) {
+ log.warn("OPML import skipped outline '" + title
+ + "': its link is not an absolute http or https URL: " +
importedUrl);
+ } else if (importedXmlUrl != null && xmlUrl == null) {
+ log.warn("OPML import dropped the feed URL of outline '" + title
+ + "': it is not an absolute http or https URL: " +
importedXmlUrl);
+ }
// better to truncate imported OPML fields than to fail import or drop
whole bookmark
int maxLength = RollerConstants.TEXTWIDTH_255;
@@ -270,6 +284,25 @@ public class JPABookmarkManagerImpl implements
BookmarkManager {
}
}
+ /**
+ * Returns the value, trimmed, if it is an absolute http or https URL with
+ * a host, else null. The scheme is read as the blogroll reads it (see
+ * {@link LinkScheme#of(String)}), so links that browsers accept are kept.
+ */
+ static String webLinkOrNull(String value) {
+ if (value == null) {
+ return null;
+ }
+ String scheme = LinkScheme.of(value);
+ if ("http".equals(scheme) || "https".equals(scheme)) {
+ String rest = LinkScheme.compact(value).substring(scheme.length()
+ 1);
+ if (rest.startsWith("//") && rest.length() > 2 && rest.charAt(2)
!= '/') {
+ return value.trim();
+ }
+ }
+ return null;
+ }
+
/**
* @see org.apache.roller.weblogger.business.BookmarkManager#getBookmarks(
* org.apache.roller.weblogger.pojos.WeblogBookmarkFolder)
diff --git
a/app/src/main/java/org/apache/roller/weblogger/pojos/wrapper/WeblogBookmarkWrapper.java
b/app/src/main/java/org/apache/roller/weblogger/pojos/wrapper/WeblogBookmarkWrapper.java
index 66a6cccda..281e404ac 100644
---
a/app/src/main/java/org/apache/roller/weblogger/pojos/wrapper/WeblogBookmarkWrapper.java
+++
b/app/src/main/java/org/apache/roller/weblogger/pojos/wrapper/WeblogBookmarkWrapper.java
@@ -18,13 +18,18 @@
package org.apache.roller.weblogger.pojos.wrapper;
+import java.util.Set;
+
import org.apache.roller.weblogger.pojos.WeblogBookmark;
+import org.apache.roller.weblogger.util.LinkScheme;
/**
* Pojo safety wrapper for WeblogBookmark object.
*/
public final class WeblogBookmarkWrapper {
+
+ private static final Set<String> ALLOWED_SCHEMES = Set.of("http", "https",
"ftp", "mailto");
// keep a reference to the wrapped pojo
private final WeblogBookmark pojo;
@@ -59,7 +64,7 @@ public final class WeblogBookmarkWrapper {
public String getUrl() {
- return this.pojo.getUrl();
+ return linkOrEmpty(this.pojo.getUrl());
}
public Integer getPriority() {
@@ -67,12 +72,28 @@ public final class WeblogBookmarkWrapper {
}
public String getImage() {
- return this.pojo.getImage();
+ return linkOrEmpty(this.pojo.getImage());
}
public String getFeedUrl() {
- return this.pojo.getFeedUrl();
+ return linkOrEmpty(this.pojo.getFeedUrl());
+ }
+
+ /**
+ * Returns the link if it is relative or uses a scheme a blogroll link
+ * needs (http, https, ftp, mailto), else an empty string. See
+ * {@link LinkScheme#of(String)} for how the scheme is read.
+ */
+ static String linkOrEmpty(String link) {
+ if (link == null) {
+ return null;
+ }
+ String scheme = LinkScheme.of(link);
+ if (scheme == null || (!scheme.isEmpty() &&
!ALLOWED_SCHEMES.contains(scheme))) {
+ return "";
+ }
+ return link;
}
diff --git a/app/src/main/java/org/apache/roller/weblogger/util/LinkScheme.java
b/app/src/main/java/org/apache/roller/weblogger/util/LinkScheme.java
new file mode 100644
index 000000000..761fd2074
--- /dev/null
+++ b/app/src/main/java/org/apache/roller/weblogger/util/LinkScheme.java
@@ -0,0 +1,73 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) under one or more
+ * contributor license agreements. See the NOTICE file distributed with
+ * this work for additional information regarding copyright ownership.
+ * The ASF licenses this file to You under the Apache License, Version 2.0
+ * (the "License"); you may not use this file except in compliance with
+ * the License. You may obtain a copy of the License at
+ *
+ * http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+package org.apache.roller.weblogger.util;
+
+import java.util.Locale;
+import java.util.regex.Pattern;
+
+/**
+ * Reads the scheme of a link the way a browser would.
+ */
+public final class LinkScheme {
+
+ private static final Pattern SCHEME =
Pattern.compile("[a-zA-Z][a-zA-Z0-9+.-]*");
+
+ private LinkScheme() {
+ }
+
+ /**
+ * Removes what browsers ignore in a link: tabs and line breaks anywhere,
+ * and spaces and control characters at either end.
+ */
+ public static String compact(String link) {
+ return link.replaceAll("[\\t\\n\\r]", "").trim();
+ }
+
+ /**
+ * Returns the lower-case scheme of the link, an empty string if the link
+ * is relative, or null if the part before the path cannot be read. That
+ * part is everything before the first '/', '?' or '#'; it is refused if
+ * it holds a character reference ('&') or a scheme that is not valid.
+ */
+ public static String of(String link) {
+ if (link == null) {
+ return null;
+ }
+ String compact = compact(link);
+ int end = compact.length();
+ for (char c : new char[] {'/', '?', '#'}) {
+ int i = compact.indexOf(c);
+ if (i >= 0 && i < end) {
+ end = i;
+ }
+ }
+ String head = compact.substring(0, end);
+ if (head.indexOf('&') >= 0) {
+ return null;
+ }
+ int colon = head.indexOf(':');
+ if (colon < 0) {
+ return "";
+ }
+ String scheme = head.substring(0, colon);
+ if (!SCHEME.matcher(scheme).matches()) {
+ return null;
+ }
+ return scheme.toLowerCase(Locale.ROOT);
+ }
+}
diff --git a/app/src/main/webapp/WEB-INF/velocity/weblog.vm
b/app/src/main/webapp/WEB-INF/velocity/weblog.vm
index b80804dac..2d9886c2e 100644
--- a/app/src/main/webapp/WEB-INF/velocity/weblog.vm
+++ b/app/src/main/webapp/WEB-INF/velocity/weblog.vm
@@ -478,22 +478,16 @@ Show bookmarks as a HTML ul list.
#foreach($bookmark in $bookmarks)
<li class="rFolderItem">
#if($utils.isNotEmpty($bookmark.Image))
- #if( $bookmark.Image.startsWith("http://") )
- <a href="$bookmark.url">
- <img alt="$bookmark.description" border="0"
- src="$bookmark.image" class="rBookmark" /></a>
- #else
- <a href="$bookmark.url">
- <img alt="$bookmark.description" border="0"
- src="$bookmark.image" class="rBookmark" /></a>
- #end
+ <a href="$utils.escapeHTML($bookmark.url)">
+ <img alt="$utils.escapeHTML($bookmark.description)" border="0"
+ src="$utils.escapeHTML($bookmark.image)" class="rBookmark"
/></a>
#else
#if($utils.isNotEmpty( $bookmark.FeedUrl))
- <a href="$bookmark.feedUrl"><img class="smrssbadge"
+ <a href="$utils.escapeHTML($bookmark.feedUrl)"><img
class="smrssbadge"
src="$url.absoluteSite/images/smrssbadge.gif"
alt="$text.get("macro.bookmark.urlFeed")"/></a>
#end
- <a href="$bookmark.url"
title="$bookmark.description">$bookmark.name</a>
+ <a href="$utils.escapeHTML($bookmark.url)"
title="$utils.escapeHTML($bookmark.description)">$utils.escapeHTML($bookmark.name)</a>
#end
</li>
#end
diff --git
a/app/src/test/java/org/apache/roller/weblogger/business/BookmarkImportParsingTest.java
b/app/src/test/java/org/apache/roller/weblogger/business/BookmarkImportParsingTest.java
index ce8e8e29a..1d2530d5b 100644
---
a/app/src/test/java/org/apache/roller/weblogger/business/BookmarkImportParsingTest.java
+++
b/app/src/test/java/org/apache/roller/weblogger/business/BookmarkImportParsingTest.java
@@ -32,12 +32,15 @@ import org.junit.jupiter.api.AfterEach;
import org.junit.jupiter.api.BeforeEach;
import org.junit.jupiter.api.Test;
+import static org.junit.jupiter.api.Assertions.assertEquals;
import static org.junit.jupiter.api.Assertions.assertFalse;
+import static org.junit.jupiter.api.Assertions.assertNull;
import static org.junit.jupiter.api.Assertions.assertTrue;
/**
* Covers the OPML bookmark import's handling of document type declarations:
* documents that carry one are refused, while ordinary OPML still imports.
+ * Also checks that imported links are kept only when they are web links.
*/
public class BookmarkImportParsingTest {
@@ -133,4 +136,33 @@ public class BookmarkImportParsingTest {
assertTrue(importedBookmarks().isEmpty(),
"a document carrying a DOCTYPE was still imported");
}
+
+ /** Only absolute http and https links are imported. */
+ @Test
+ public void importKeepsOnlyWebLinks() throws Exception {
+ String opml = "<?xml version=\"1.0\"?><opml
version=\"1.1\"><head><title>t</title></head><body>"
+ + "<outline text=\"web\" htmlUrl=\"https://example.com/a\"
xmlUrl=\"http://example.com/feed\"/>"
+ + "<outline text=\"ftp\" htmlUrl=\"ftp://example.com/a\"/>"
+ + "<outline text=\"mail\"
htmlUrl=\"mailto:[email protected]\"/>"
+ + "<outline text=\"relative\" htmlUrl=\"/a/b\"/>"
+ + "<outline text=\"spaces\" htmlUrl=\"http://example.com/a
b\"/>"
+ + "<outline text=\"underscore\" htmlUrl=\"
https://my_blog.example.com/ \"/>"
+ + "<outline text=\"nohost\" htmlUrl=\"http:///a\"/>"
+ + "<outline text=\"feed\" htmlUrl=\"https://example.com/b\"
xmlUrl=\"file:///etc/feed\"/>"
+ + "</body></opml>";
+
bookmarkManager().importBookmarks(TestUtils.getManagedWebsite(testWeblog),
folderName, opml);
+ TestUtils.endSession(true);
+
+ java.util.Map<String, WeblogBookmark> byName = new
java.util.HashMap<>();
+ for (WeblogBookmark b : importedBookmarks()) {
+ byName.put(b.getName(), b);
+ }
+ // links that browsers accept are kept even when java.net.URI refuses
them
+ assertEquals(java.util.Set.of("web", "feed", "spaces", "underscore"),
byName.keySet());
+ assertEquals("https://example.com/a", byName.get("web").getUrl());
+ assertEquals("http://example.com/feed",
byName.get("web").getFeedUrl());
+ assertEquals("https://example.com/b", byName.get("feed").getUrl());
+ assertNull(byName.get("feed").getFeedUrl());
+ assertEquals("https://my_blog.example.com/",
byName.get("underscore").getUrl());
+ }
}
diff --git
a/app/src/test/java/org/apache/roller/weblogger/pojos/wrapper/WeblogBookmarkWrapperTest.java
b/app/src/test/java/org/apache/roller/weblogger/pojos/wrapper/WeblogBookmarkWrapperTest.java
new file mode 100644
index 000000000..a1b41206b
--- /dev/null
+++
b/app/src/test/java/org/apache/roller/weblogger/pojos/wrapper/WeblogBookmarkWrapperTest.java
@@ -0,0 +1,56 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) under one or more
+ * contributor license agreements. The ASF licenses this file to You
+ * under the Apache License, Version 2.0 (the "License"); you may not
+ * use this file except in compliance with the License.
+ * You may obtain a copy of the License at
+ *
+ * http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License. For additional information regarding
+ * copyright in this work, please see the NOTICE file in the top level
+ * directory of this distribution.
+ */
+package org.apache.roller.weblogger.pojos.wrapper;
+
+import org.junit.jupiter.api.Test;
+
+import static org.junit.jupiter.api.Assertions.assertEquals;
+import static org.junit.jupiter.api.Assertions.assertNull;
+
+/**
+ * Blogroll links reach templates only when they are relative or use a scheme
+ * a blogroll link needs.
+ */
+class WeblogBookmarkWrapperTest {
+
+ @Test
+ void keepsWebAndRelativeLinks() {
+ assertEquals("https://example.com/a",
WeblogBookmarkWrapper.linkOrEmpty("https://example.com/a"));
+ assertEquals("HTTP://example.com/",
WeblogBookmarkWrapper.linkOrEmpty("HTTP://example.com/"));
+ assertEquals("ftp://example.com/f",
WeblogBookmarkWrapper.linkOrEmpty("ftp://example.com/f"));
+ assertEquals("mailto:[email protected]",
WeblogBookmarkWrapper.linkOrEmpty("mailto:[email protected]"));
+ assertEquals("/blog/page",
WeblogBookmarkWrapper.linkOrEmpty("/blog/page"));
+ assertEquals("page?a=b:c",
WeblogBookmarkWrapper.linkOrEmpty("page?a=b:c"));
+ assertNull(WeblogBookmarkWrapper.linkOrEmpty(null));
+ }
+
+ @Test
+ void dropsOtherSchemes() {
+ assertEquals("",
WeblogBookmarkWrapper.linkOrEmpty("data:text/plain,x"));
+ assertEquals("", WeblogBookmarkWrapper.linkOrEmpty("FILE:///etc/x"));
+ assertEquals("", WeblogBookmarkWrapper.linkOrEmpty("
fi\tle:///etc/x"));
+ }
+
+ @Test
+ void dropsCharacterReferencesBeforeThePath() {
+ assertEquals("", WeblogBookmarkWrapper.linkOrEmpty("x:y"));
+ assertEquals("", WeblogBookmarkWrapper.linkOrEmpty("x:y"));
+ assertEquals("", WeblogBookmarkWrapper.linkOrEmpty("x:y"));
+ assertEquals("/a?b=1&c=2",
WeblogBookmarkWrapper.linkOrEmpty("/a?b=1&c=2"));
+ }
+}
diff --git
a/app/src/test/java/org/apache/roller/weblogger/util/LinkSchemeTest.java
b/app/src/test/java/org/apache/roller/weblogger/util/LinkSchemeTest.java
new file mode 100644
index 000000000..71631d378
--- /dev/null
+++ b/app/src/test/java/org/apache/roller/weblogger/util/LinkSchemeTest.java
@@ -0,0 +1,51 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) under one or more
+ * contributor license agreements. See the NOTICE file distributed with
+ * this work for additional information regarding copyright ownership.
+ * The ASF licenses this file to You under the Apache License, Version 2.0
+ * (the "License"); you may not use this file except in compliance with
+ * the License. You may obtain a copy of the License at
+ *
+ * http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+package org.apache.roller.weblogger.util;
+
+import org.junit.jupiter.api.Test;
+
+import static org.junit.jupiter.api.Assertions.assertEquals;
+import static org.junit.jupiter.api.Assertions.assertNull;
+
+class LinkSchemeTest {
+
+ @Test
+ void readsSchemes() {
+ assertEquals("https", LinkScheme.of("HTTPS://example.com/"));
+ assertEquals("mailto", LinkScheme.of("mailto:[email protected]"));
+ assertEquals("file", LinkScheme.of(" fi\tle:///etc/x"));
+ }
+
+ @Test
+ void relativeLinksHaveNoScheme() {
+ assertEquals("", LinkScheme.of("/a/b"));
+ assertEquals("", LinkScheme.of("page?a=b:c"));
+ assertEquals("", LinkScheme.of("#top"));
+ assertEquals("", LinkScheme.of("/a?b=1&c=2"));
+ }
+
+ @Test
+ void refusesWhatCannotBeRead() {
+ assertNull(LinkScheme.of(null));
+ assertNull(LinkScheme.of("x:y"));
+ assertNull(LinkScheme.of("x:y"));
+ assertNull(LinkScheme.of("x:y"));
+ assertNull(LinkScheme.of(":y"));
+ assertNull(LinkScheme.of("1x:y"));
+ }
+}