[nutch] branch master updated: NUTCH-2817 Avoid check for equality of URL path and file part using ==/!= - replace check whether URL path and file are identical by check whether URL has a query - clean up code and improve log messages

[snagel]

This is an automated email from the ASF dual-hosted git repository.

snagel pushed a commit to branch master
in repository https://gitbox.apache.org/repos/asf/nutch.git


The following commit(s) were added to refs/heads/master by this push:
     new c68780d  NUTCH-2817 Avoid check for equality of URL path and file part 
using ==/!= - replace check whether URL path and file are identical   by check 
whether URL has a query - clean up code and improve log messages
     new 466cac5  Merge pull request #548 from 
sebastian-nagel/NUTCH-2817-spotbugs-object-equality
c68780d is described below

commit c68780d97f5410da5a46eb69005bba723f89ff55
Author: Sebastian Nagel <sna...@apache.org>
AuthorDate: Sat Aug 8 10:54:42 2020 +0200

    NUTCH-2817 Avoid check for equality of URL path and file part using ==/!=
    - replace check whether URL path and file are identical
      by check whether URL has a query
    - clean up code and improve log messages
---
 .../java/org/apache/nutch/protocol/file/FileResponse.java   | 13 +++++--------
 .../src/java/org/apache/nutch/protocol/ftp/FtpResponse.java |  9 ++++-----
 2 files changed, 9 insertions(+), 13 deletions(-)

diff --git 
a/src/plugin/protocol-file/src/java/org/apache/nutch/protocol/file/FileResponse.java
 
b/src/plugin/protocol-file/src/java/org/apache/nutch/protocol/file/FileResponse.java
index b2db228..0579d96 100644
--- 
a/src/plugin/protocol-file/src/java/org/apache/nutch/protocol/file/FileResponse.java
+++ 
b/src/plugin/protocol-file/src/java/org/apache/nutch/protocol/file/FileResponse.java
@@ -66,7 +66,6 @@ public class FileResponse {
   private final File file;
   private Configuration conf;
 
-  private MimeUtil MIME;
   private Tika tika;
 
   /** Returns the response code. */
@@ -106,23 +105,21 @@ public class FileResponse {
     this.file = file;
     this.conf = conf;
 
-    MIME = new MimeUtil(conf);
     tika = new Tika();
 
     if (!"file".equals(url.getProtocol()))
       throw new FileException("Not a file url:" + url);
 
     if (File.LOG.isTraceEnabled()) {
-      File.LOG.trace("fetching " + url);
+      File.LOG.trace("fetching {}", url);
     }
 
-    if (url.getPath() != url.getFile()) {
-      if (File.LOG.isWarnEnabled()) {
-        File.LOG.warn("url.getPath() != url.getFile(): " + url);
-      }
+    if (url.getQuery() != null) {
+      File.LOG.warn(
+          "file:// URL may not include a query (query part ignored): {}", url);
     }
 
-    String path = "".equals(url.getPath()) ? "/" : url.getPath();
+    String path = url.getPath().isEmpty() ? "/" : url.getPath();
 
     try {
       // specify the encoding via the config later?
diff --git 
a/src/plugin/protocol-ftp/src/java/org/apache/nutch/protocol/ftp/FtpResponse.java
 
b/src/plugin/protocol-ftp/src/java/org/apache/nutch/protocol/ftp/FtpResponse.java
index 0451201..aee44b5 100644
--- 
a/src/plugin/protocol-ftp/src/java/org/apache/nutch/protocol/ftp/FtpResponse.java
+++ 
b/src/plugin/protocol-ftp/src/java/org/apache/nutch/protocol/ftp/FtpResponse.java
@@ -85,13 +85,12 @@ public class FtpResponse {
     if (!"ftp".equals(url.getProtocol()))
       throw new FtpException("Not a ftp url:" + url);
 
-    if (url.getPath() != url.getFile()) {
-      if (Ftp.LOG.isWarnEnabled()) {
-        Ftp.LOG.warn("url.getPath() != url.getFile(): " + url);
-      }
+    if (url.getQuery() != null) {
+      Ftp.LOG.warn(
+          "ftp:// URL may not include a query (query part ignored): {}", url);
     }
 
-    String path = "".equals(url.getPath()) ? "/" : url.getPath();
+    String path = url.getPath().isEmpty() ? "/" : url.getPath();
 
     try {