Latest scan data.

ukwa · Dec 2, 2015 · 68a6a11 · 68a6a11
1 parent 1d0154d
commit 68a6a11
Show file tree

Hide file tree

Showing 13 changed files with 57,981 additions and 6,001 deletions.
diff --git a/archive-sample/ldwa.2015.12-total-urls-per-year.json b/archive-sample/ldwa.2015.12-total-urls-per-year.json
@@ -0,0 +1,77 @@
+{
+  "responseHeader":{
+    "status":0,
+    "QTime":86,
+    "params":{
+      "facet.date.start":"2000-01-01T00:00:00Z",
+      "facet":"true",
+      "facet.mincount":"1",
+      "indent":"true",
+      "q":"*:*",
+      "facet.date":"crawl_date",
+      "facet.date.gap":"+1YEAR",
+      "wt":"json",
+      "facet.date.end":"2050-01-01T00:00:00Z",
+      "rows":"0"}},
+  "response":{"numFound":3058205547,"start":0,"maxScore":1.0,"docs":[]
+  },
+  "facet_counts":{
+    "facet_queries":{},
+    "facet_fields":{},
+    "facet_dates":{
+      "crawl_date":{
+        "2000-01-01T00:00:00Z":0,
+        "2001-01-01T00:00:00Z":0,
+        "2002-01-01T00:00:00Z":0,
+        "2003-01-01T00:00:00Z":0,
+        "2004-01-01T00:00:00Z":0,
+        "2005-01-01T00:00:00Z":0,
+        "2006-01-01T00:00:00Z":0,
+        "2007-01-01T00:00:00Z":0,
+        "2008-01-01T00:00:00Z":0,
+        "2009-01-01T00:00:00Z":0,
+        "2010-01-01T00:00:00Z":0,
+        "2011-01-01T00:00:00Z":0,
+        "2012-01-01T00:00:00Z":0,
+        "2013-01-01T00:00:00Z":1354258438,
+        "2014-01-01T00:00:00Z":1538695978,
+        "2015-01-01T00:00:00Z":164171513,
+        "2016-01-01T00:00:00Z":0,
+        "2017-01-01T00:00:00Z":0,
+        "2018-01-01T00:00:00Z":0,
+        "2019-01-01T00:00:00Z":0,
+        "2020-01-01T00:00:00Z":0,
+        "2021-01-01T00:00:00Z":0,
+        "2022-01-01T00:00:00Z":0,
+        "2023-01-01T00:00:00Z":0,
+        "2024-01-01T00:00:00Z":0,
+        "2025-01-01T00:00:00Z":0,
+        "2026-01-01T00:00:00Z":0,
+        "2027-01-01T00:00:00Z":0,
+        "2028-01-01T00:00:00Z":0,
+        "2029-01-01T00:00:00Z":0,
+        "2030-01-01T00:00:00Z":0,
+        "2031-01-01T00:00:00Z":0,
+        "2032-01-01T00:00:00Z":0,
+        "2033-01-01T00:00:00Z":0,
+        "2034-01-01T00:00:00Z":0,
+        "2035-01-01T00:00:00Z":0,
+        "2036-01-01T00:00:00Z":0,
+        "2037-01-01T00:00:00Z":0,
+        "2038-01-01T00:00:00Z":0,
+        "2039-01-01T00:00:00Z":0,
+        "2040-01-01T00:00:00Z":0,
+        "2041-01-01T00:00:00Z":0,
+        "2042-01-01T00:00:00Z":0,
+        "2043-01-01T00:00:00Z":0,
+        "2044-01-01T00:00:00Z":0,
+        "2045-01-01T00:00:00Z":0,
+        "2046-01-01T00:00:00Z":0,
+        "2047-01-01T00:00:00Z":0,
+        "2048-01-01T00:00:00Z":0,
+        "2049-01-01T00:00:00Z":0,
+        "gap":"+1YEAR",
+        "start":"2000-01-01T00:00:00Z",
+        "end":"2050-01-01T00:00:00Z"}},
+    "facet_ranges":{},
+    "facet_intervals":{}}}
diff --git a/archive-sample/sample-of-2000/ldwa.2015.09-sample-for-2013.csv b/archive-sample/sample-of-2000/ldwa.2015.09-sample-for-2013.csv
diff --git a/archive-sample/sample-of-2000/ldwa.2015.09-sample-for-2014.csv b/archive-sample/sample-of-2000/ldwa.2015.09-sample-for-2014.csv
diff --git a/archive-sample/sample-of-2000/ldwa.2015.09-sample-for-2015.csv b/archive-sample/sample-of-2000/ldwa.2015.09-sample-for-2015.csv
diff --git a/archive-sample/sample-of-2000/ldwa.2015.12-sample-for-2013.csv b/archive-sample/sample-of-2000/ldwa.2015.12-sample-for-2013.csv
diff --git a/archive-sample/sample-of-2000/ldwa.2015.12-sample-for-2013.json b/archive-sample/sample-of-2000/ldwa.2015.12-sample-for-2013.json
diff --git a/archive-sample/sample-of-2000/ldwa.2015.12-sample-for-2014.csv b/archive-sample/sample-of-2000/ldwa.2015.12-sample-for-2014.csv
diff --git a/archive-sample/sample-of-2000/ldwa.2015.12-sample-for-2014.json b/archive-sample/sample-of-2000/ldwa.2015.12-sample-for-2014.json
diff --git a/archive-sample/sample-of-2000/ldwa.2015.12-sample-for-2015.csv b/archive-sample/sample-of-2000/ldwa.2015.12-sample-for-2015.csv
diff --git a/archive-sample/sample-of-2000/ldwa.2015.12-sample-for-2015.json b/archive-sample/sample-of-2000/ldwa.2015.12-sample-for-2015.json
diff --git a/sample-scan-results/2015-12-Explorer/ldwa-sample-of-2000-scan-data.csv b/sample-scan-results/2015-12-Explorer/ldwa-sample-of-2000-scan-data.csv
diff --git a/sample-scan-results/2015-12-Explorer/ldwa-sample-of-2000-scan-results.csv b/sample-scan-results/2015-12-Explorer/ldwa-sample-of-2000-scan-results.csv
diff --git a/tools/halflife/yearwise_sampler.py b/tools/halflife/yearwise_sampler.py
@@ -20,7 +20,8 @@
 #prefix="open"
 
 #solr_endpoint = "http://192.168.45.17:8983/solr/ldwa/select"
-solr_endpoint = "http://192.168.1.65:8983/solr/ldwa/select"
+#solr_endpoint = "http://192.168.1.65:8983/solr/ldwa/select"
+solr_endpoint = "http://192.168.1.54:8983/solr/ldukwadev/select"
 date_field = "crawl_date"
 url_field = "url"
 wayback_date_field = "wayback_date"