Solr 突出显示返回空结果

Question

我有一个搜索系统，它在 Solr 4.9 上运行良好，具有开箱即用的配置和模式。

在 Solr 6.2 中，结果中返回了高亮条目，但它们仅包含文档 ID - 没有高亮文本。

起初我以为是因为“内容”不在默认(托管)架构中，但添加它并没有任何区别。无论如何，在默认模式(作者、主题、标题等)中还有其他字段可以返回突出显示的文本，但我也没有得到任何关于这些的信息。

在默认配置中似乎启用了突出显示，而且我的查询似乎没问题(从管理界面运行时它也不会返回任何 HL 文本，因此排除了我的代码):

q=mike&rows=10&start=0&sort=score desc, last_modified desc&wt=json&fl=content_type,fsb_doctype,author,id,last_modified,subject,title,score,url&hl=true&hl.fragsize=250&hl.fl=content,author,subject ,title&hl.simple.pre=%3Cb%3E&hl.simple.post=%3C%2Fb%3E&facet=true&facet.field=fsb_doctype&facet.field=fsb_origin&facet.field=fsb_mission

我所做的唯一架构更改是添加了 3 个用于分面的自定义字段。

这是 Solr 4.9 系统返回的 json 示例(只是突出显示的部分):

"highlighting":{"/event_20060718.cfm":{"content":[" \n \n \n \n \n \n \n Code 582 - Events \n \n July 18, 2006 - Flight Software Branch <b>CMMI</b> <b>Appraisal</b> \n\n \nThe Flight Software Branch was successful in its latest <b>CMMI</b> <b>appraisal</b> - a SCAMPI Class A <b>Appraisal</b> of the Supplier Agreement Management (SAM"]},"/SDODocs/":{"content":[" \n \n \n \n \n \n \n SDO Flight Software \n \n \t \t SDO Flight Software Baselined Documents \n(Restricted access - contains only assets used in the <b>CMMI</b> <b>appraisal</b>. For access to up-to-date SDO FSW documents, see M W.) \n \t \t SDO Flight"]},"/LRO/":{"content":[" \n \n \n \n \n \n \n LRO Flight Software \n \n \t \t LRO Flight Software Baselined Documents \n(Restricted access - contains only assets used in the <b>CMMI</b> <b>appraisal</b>. For access to up-to-date LRO FSW documents, see Mike B.) \n \t \t LRO Flight"]},"TDL_582 Web&id=501":{"content":[" \n \n \n \n \n \n \n Action item from 582 Web group - group id 501 - closed item \n \n Michael \n Mike Tilley \n Update Events - <b>CMMI</b> <b>appraisal</b> results are official today! \n 09/18/14 - updated/created & deployed:\n\n/default.cfm\n\n/events.cfm"]},"TDL_582 Web&id=203":{"content":[" \n \n \n \n \n \n \n Action item from 582 Web group - group id 203 - closed item \n \n Michael \n Mike Tilley \n <b>CMMI</b> <b>Appraisal</b> results are \"official\" today - post an event. \n 10/11/11 - updated/created & moved to production (& Linux)..\n\n/default.cfm"]},"TDL_BSR FPI&id=1":{"content":[" \n \n \n \n \n \n \n Action item from BSR FPI group - group id 1 - closed item \n \n Michael \n Mike Tilley \n Action: send a BSR template, customized for FPI, to Victor. \n 12/15/10 - done (distracted by <b>CMMI</b> <b>Appraisal</b>!) \n "]},"/event_20080516.cfm":{"content":[" \n \n \n \n \n \n \n Code 582 - Events \n \n May 16, 2008 - Center <b>CMMI</b> <b>Appraisal</b> \n\n\n A <b>CMMI</b> SCAMPI Class A <b>Appraisal</b> for GSFC was completed successfully on Friday, May 16th. GSFC is now compliant with Agency policy regarding <b>CMMI</b> Maturity Level 2 for"]},"/event_20140918.cfm":{"content":[" \n \n \n \n \n \n \n Code 582 - Events \n \n September 18, 2014 - Center <b>CMMI</b> <b>Appraisal</b> \n\n\n A <b>CMMI</b> SCAMPI Class A <b>Appraisal</b> for GSFC was completed successfully on Monday, September 15th. GSFC continues to be compliant with Agency policy regarding <b>CMMI</b>"]},"TDL_582 Web&id=353":{"content":[" & we made the mistake of trying to use the same repository for the team, and for <b>CMMI</b> <b>appraisal</b> evidence. As a result, the team never used it, and it contains sensitive <b>appraisal</b> data. Closing this AI. \n "]},"TDL_582 Web&id=67":{"content":[" \n \n \n \n \n \n \n Action item from 582 Web group - group id 67 - closed item \n \n Michael \n Mike Tilley \n Had to find the SEI certification page for the GSFC <b>CMMI</b> <b>appraisal</b> - it wasn't easy. Should probably post a link to this on the website"]}

这是 6.1 json 结果的示例(此服务器上的文档集略有不同，但两种情况下的第一个结果相同:

"highlighting":{"/event_20060718.cfm":{},"/event_20080516.cfm":{},"/event_20140918.cfm":{},"/event_20111011.cfm":{},"/LRO/":{},"/SDODocs/":{},"/event_20060201.cfm":{},"TDL_582 Web&id=353":{},"TDL_582 Web&id=67":{},"TDL_BSR JWST&id=94":{}

所以我得到的只是一个没有高亮文本的文档 ID 列表。

Answer 1

将您的代码更改为类似这样的内容应该可行:

select?q=I+WANT+PIZZA&wt=php&indent=true&fl=id,name&group=true&group.field=content&hl=on&hl.fl=*&hl.encoder=html&hl.fragmenter=regex&hl.regex.slop=100.0 &hl.fl=text_*&hl.bs.type=WHOLE&hl.defaultSummary=true&hl.offsetSource=POSTINGS