From cc0ded7abdcd1006ca0327229806c3789b55df9e Mon Sep 17 00:00:00 2001 From: Michael Peter Christen Date: Mon, 4 Aug 2014 14:44:20 +0200 Subject: [PATCH] set process type of web graph according to fields as defined in the schema --- source/net/yacy/search/schema/WebgraphConfiguration.java | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) diff --git a/source/net/yacy/search/schema/WebgraphConfiguration.java b/source/net/yacy/search/schema/WebgraphConfiguration.java index dc01e6093..e30e4535e 100644 --- a/source/net/yacy/search/schema/WebgraphConfiguration.java +++ b/source/net/yacy/search/schema/WebgraphConfiguration.java @@ -51,7 +51,6 @@ import net.yacy.cora.protocol.ResponseHeader; import net.yacy.cora.util.CommonPattern; import net.yacy.cora.util.ConcurrentLog; import net.yacy.crawler.HostBalancer; -import net.yacy.document.parser.html.ImageEntry; import net.yacy.search.schema.CollectionConfiguration.Subgraph; public class WebgraphConfiguration extends SchemaConfiguration implements Serializable { @@ -271,9 +270,11 @@ public class WebgraphConfiguration extends SchemaConfiguration implements Serial } } - if ((allAttr || contains(WebgraphSchema.process_sxt)) && processTypes.size() > 0) { + if ((allAttr || contains(WebgraphSchema.process_sxt)) && + ((this.contains(WebgraphSchema.source_id_s) && this.contains(WebgraphSchema.source_cr_host_norm_i)) || + (this.contains(WebgraphSchema.target_id_s) && this.contains(WebgraphSchema.target_cr_host_norm_i))) && processTypes.contains(ProcessType.CITATION)) { List pr = new ArrayList(); - for (ProcessType t: processTypes) pr.add(t.name()); + pr.add(ProcessType.CITATION.name()); add(edge, WebgraphSchema.process_sxt, pr); if (allAttr || contains(CollectionSchema.harvestkey_s)) { add(edge, CollectionSchema.harvestkey_s, sourceName);