tmp/druid:ab123ab1234.xml
<?xml version="1.0" encoding="UTF-8" standalone="no"?>
<crawlObject>
<crawlId>test_crawl</crawlId>
<collectionId>test_collection</collectionId>
<files>
<file>
<name>WARC-Test.warc.gz</name>
<type>WARC</type>
<size>6608320</size>
<recordCount>4027</recordCount>
<mimeType>application/octet-stream</mimeType>
<checksumMD5>c7edbde066e4697b3f2d823ac42c3692</checksumMD5>
<checksumSHA1>3a9f2ffac1497c70291d93a8bc86c1469547d8f8</checksumSHA1>
<isPatchCrawl>false</isPatchCrawl>
<software> Heritrix/3.2.0-SNAPSHOT-20140108-2049 http://crawler.archive.org</software>
<organizationName>"Stanford University</organizationName>
<ip> 207.241.226.90</ip>
<accountType>SUBSCRIBER</accountType>
<creationDate>2014-01-19T22:37:40Z</creationDate>
<seedCount>68</seedCount>
<robotsPolicy> obey</robotsPolicy>
<recurrence>QUARTERLY</recurrence>
<isTestCrawl>false</isTestCrawl>
<accountId>159</accountId>
<hostname> wbgrp-crawl051.us.archive.org</hostname>
<datatype> WARC File Format 1.0</datatype>
<id>ARCHIVEIT-924-QUARTERLY-31501-20140119223740943-00015-wbgrp-crawl051.us.archive.org-6441.warc.gz</id>
<collectionId>924</collectionId>
<maxDuration>604800</maxDuration>
<user-agent> Mozilla/5.0 (compatible; archive.org_bot; Archive-It; +http://archive-it.org/files/site-owners.html)</user-agent>
</file>
</files>
</crawlObject>