[
https://issues.apache.org/jira/browse/NUTCH-2445?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16214819#comment-16214819
]
Sebastian Nagel commented on NUTCH-2445:
----------------------------------------
+1
Two trivial points:
- there is a compiler warning because Set isn't used as Set<Integer>
- good idea to use a set of integers to reduce the memory footprint, but could
be worth a comment. In the rare case of hash collisions, some URLs may be not
immediately followed.
> Fetcher following outlinks to keep track of already fetched items
> -----------------------------------------------------------------
>
> Key: NUTCH-2445
> URL: https://issues.apache.org/jira/browse/NUTCH-2445
> Project: Nutch
> Issue Type: Improvement
> Affects Versions: 1.13
> Reporter: Markus Jelsma
> Assignee: Markus Jelsma
> Fix For: 1.14
>
> Attachments: NUTCH-2445.patch
>
>
> When fetcher.follow.outlinks.depth is non-zero, fetcher follows outlinks.
> This patch keeps track of already fetched URL's and thus avoid fetching the
> same URL twice.
> A Set is used to keep track of them, hashcodes to reduce memory usage. This
> is not used if fetcher doesn't follow outlinks.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
