From 9468cb0c1fcd9415b1a86ec5a558efb70c53199a Mon Sep 17 00:00:00 2001 From: nlevitt Date: Wed, 9 Feb 2011 01:17:26 +0000 Subject: [PATCH] Report total bytes, as well as breakdown, in xml version of job page, to match html. * CrawlJob.java sizeTotalsReportData() - add "total" (bytes) and "total-count" (urls) to map * CrawledBytesHistotable.java new method getTotalUrls() to complement getTotalBytes() --- .../main/java/org/archive/crawler/framework/CrawlJob.java | 8 +++++++- .../org/archive/crawler/util/CrawledBytesHistotable.java | 4 ++++ 2 files changed, 11 insertions(+), 1 deletion(-) diff --git a/engine/src/main/java/org/archive/crawler/framework/CrawlJob.java b/engine/src/main/java/org/archive/crawler/framework/CrawlJob.java index 1a19f3f2..cb63ee68 100644 --- a/engine/src/main/java/org/archive/crawler/framework/CrawlJob.java +++ b/engine/src/main/java/org/archive/crawler/framework/CrawlJob.java @@ -31,6 +31,7 @@ import java.util.LinkedHashMap; import java.util.LinkedList; import java.util.List; import java.util.Map; +import java.util.TreeMap; import java.util.logging.FileHandler; import java.util.logging.Formatter; import java.util.logging.Handler; @@ -797,7 +798,12 @@ public class CrawlJob implements Comparable, ApplicationListener { if(stats==null) { return null; } - return stats.getCrawledBytes(); + + // stats.crawledBytesSummary() also includes totals, so add those in here + TreeMap map = new TreeMap(stats.getCrawledBytes()); + map.put("total", stats.getCrawledBytes().getTotalBytes()); + map.put("total-count", stats.getCrawledBytes().getTotalUrls()); + return map; } public String sizeTotalsReport() { diff --git a/engine/src/main/java/org/archive/crawler/util/CrawledBytesHistotable.java b/engine/src/main/java/org/archive/crawler/util/CrawledBytesHistotable.java index 0bb0fddb..6797bd4b 100644 --- a/engine/src/main/java/org/archive/crawler/util/CrawledBytesHistotable.java +++ b/engine/src/main/java/org/archive/crawler/util/CrawledBytesHistotable.java @@ -79,4 +79,8 @@ implements CoreAttributeConstants { public long getTotalBytes() { return get(NOVEL) + get(DUPLICATE) + get(NOTMODIFIED); } + + public long getTotalUrls() { + return get(NOVELCOUNT) + get(DUPLICATECOUNT) + get(NOTMODIFIEDCOUNT); + } }