nutch-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Hasan Diwan" <hasan.di...@gmail.com>
Subject Re: [Nutch-dev] Checking if crawl dir exists ...
Date Sat, 26 Aug 2006 10:37:08 GMT
On 25/08/06, Michael Wechner <michael.wechner@wyona.com> wrote:

> >> Index: nutch-0.8/src/java/org/apache/nutch/searcher/NutchBean.java
> >> ===================================================================
> >> --- nutch-0.8/src/java/org/apache/nutch/searcher/NutchBean.java
> >> (Revision 436787)
> >> +++ nutch-0.8/src/java/org/apache/nutch/searcher/NutchBean.java
> >> (Arbeitskopie)
> >> @@ -95,6 +95,9 @@
> >>         if (dir == null) {
> >>             dir = new Path(this.conf.get("searcher.dir", "crawl"));
> >>         }
> >> +       if (!new java.io.File(dir.toString()).exists()) {
> >> +            LOG.warn("No such directory: " + new java.io.File
> >> (dir.toString()));
> >> +        }
> >>         Path servers = new Path(dir, "search-servers.txt");
> >>         if (fs.exists(servers)) {
> >>             if (LOG.isInfoEnabled())


Is this not what java.io.FileNotFoundException is for? In other words, in
addition to putting a log message, you should throw a FileNotFoundException
as well. You might also want to terminate the process altogether, as I see
no use in doing anything without a search directory. Another approach is to
throw an exception. The caller of the function then catches said exception
and generates the searcher.dir.
-- 
Cheers,
Hasan Diwan <hasan.diwan@gmail.com>

Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message