TIKA-4809: Fix welcome-page links, and two bad URLs from the README commit

Author: tballison

tika-server/README.md

@@ -33,8 +33,7 @@ $ java -jar tika-server/target/tika-server.jar --help
 ```
 
 Everything beyond host, port and id is configured in the tika-config JSON file
-passed with `-c`, not on the command line. See
-[Configuration](https://tika.apache.org/configuration/index.html).
+passed with `-c`, not on the command line.
 
 Running via Docker
 ------------------
@@ -50,8 +49,8 @@ Note the `127.0.0.1:` prefix. Unlike the jar, which binds `localhost` by default
 the Docker images start the server with `-h 0.0.0.0`, so publishing the port
 without an explicit interface exposes it on every interface of the host.
 tika-server performs no authentication and parses untrusted files; only expose it
-on a trusted, access-controlled network. See
-[Security](https://tika.apache.org/security.html).
+on a trusted, access-controlled network. See the
+[Tika Security Model](https://tika.apache.org/security-model.html).
 
 You may also be interested in the https://github.com/apache/tika-docker project
 which provides prebuilt Docker images.

tika-server/tika-server-core/src/main/java/org/apache/tika/server/core/resource/TikaWelcome.java

@@ -47,7 +47,12 @@
  */
 @Path("/")
 public class TikaWelcome {
-    private static final String DOCS_URL = "https://wiki.apache.org/tika/TikaJAXRS";
+    private static final String DOCS_URL =
+            "https://cwiki.apache.org/confluence/display/TIKA/TikaJAXRS";
+
+    /** Matches {@code {name : regex}} in a JAX-RS path, capturing the parameter name. */
+    private static final Pattern PATH_TEMPLATE_REGEX =
+            Pattern.compile("\\{\\s*(\\w+)\\s*:[^}]*}");
 
     private static final Map<Class<? extends Annotation>, String> HTTP_METHODS = new HashMap<>();
 
@@ -164,13 +169,21 @@ public String getWelcomeHTML() {
 
         h.append("<ul>\n");
         for (Endpoint e : identifyEndpoints()) {
+            String displayPath = simplifyPathTemplate(e.path);
             h.append("<li><b>");
             h.append(e.httpMethod);
-            h.append("</b> <i><a href=\"");
-            h.append(e.path);
-            h.append("\">");
-            h.append(e.path);
-            h.append("</a></i><br />");
+            h.append("</b> <i>");
+            // Only linkify concrete paths; one with a {param} is not fetchable as written.
+            if (displayPath.indexOf('{') < 0) {
+                h.append("<a href=\"");
+                h.append(displayPath);
+                h.append("\">");
+                h.append(displayPath);
+                h.append("</a>");
+            } else {
+                h.append(displayPath);
+            }
+            h.append("</i><br />");
             h.append("Class: ");
             h.append(e.className);
             h.append("<br />Method: ");
@@ -213,6 +226,17 @@ public String getWelcomePlain() {
         return text.toString();
     }
 
+    /**
+     * Strips the regex from a JAX-RS path template so {@code /rmeta/{handler : (\w+)?}}
+     * renders as {@code /rmeta/{handler}}.
+     */
+    static String simplifyPathTemplate(String path) {
+        if (path == null || path.indexOf('{') < 0) {
+            return path;
+        }
+        return PATH_TEMPLATE_REGEX.matcher(path).replaceAll("{$1}");
+    }
+
     protected static class Endpoint {
         public final String className;
         public final String methodName;