blevesearch / beer-search Goto Github PK

Steps to replicate:

1. Open http://localhost:8094/search/phrase/
2. Choose Category and add term

Error received:
error parsing query: json: cannot unmarshal object into Go struct field _mphraseQuery.terms of type []string {"query":{"terms":[{"term":"ipa","field":"description","boost":1}],"boost":1},"highlight":{},"explain":true,"size":10}

I have installed and run the example program.
The data set in /data is 2,5MB
After running the example the folder "beer-search.bleve" includes a generated index "store" of 211MB

I would like to use Bleve on 100 Gb of datas...this would mean a 10 TB index !!!
I am new to search engines, so excuse my question: isn t an index supposed to be much smaller than the data set ?

I have googled around and found this explanation from one year ago 👍 https://groups.google.com/forum/#!topic/bleve/xTxlOkTWrLQ

Are those statement still correct ? Is a search engine generating such big indexes is viable ?

Running main.go results in three errors:

./main.go:57:24: undefined: buildIndexMapping
./main.go:89:12: undefined: staticFileRouter
./main.go:99:29: undefined: docIDLookup

Is the problem because the functions are not exported by mapping.go and http_util.go?

$ beer-search
2014/10/01 10:04:21 GOMAXPROCS: 1
2014/10/01 10:04:21 Creating new index...
2014/10/01 10:04:21 no analyzer with name or type 'en' registered
$

using latest bleve and beer-search

Some of the alternate mapping examples use []string instead of []interface{} and therefor silently do not behave as expected.

We should remove these alternate mapping examples here as they do not get proper testing. Consider creating a new repo solely to host these examples, and include unit tests in these examples to ensure they don't break.

I've added n-gram token analyzer on "description" field. It outputs strange results. I mean, for the input "brew durin" how come it gives score of "brettanomyces. During" bigger than next "brewed during" ?

go build

# golang.org/x/sys/unix
../../go/pkg/mod/golang.org/x/[email protected]/unix/syscall_darwin.1_13.go:25:3: //go:linkname must refer to declared function or variable
../../go/pkg/mod/golang.org/x/[email protected]/unix/zsyscall_darwin_amd64.1_13.go:27:3: //go:linkname must refer to declared function or variable
../../go/pkg/mod/golang.org/x/[email protected]/unix/zsyscall_darwin_amd64.1_13.go:40:3: //go:linkname must refer to declared function or variable
../../go/pkg/mod/golang.org/x/[email protected]/unix/zsyscall_darwin_amd64.go:28:3: //go:linkname must refer to declared function or variable
../../go/pkg/mod/golang.org/x/[email protected]/unix/zsyscall_darwin_amd64.go:43:3: //go:linkname must refer to declared function or variable
../../go/pkg/mod/golang.org/x/[email protected]/unix/zsyscall_darwin_amd64.go:59:3: //go:linkname must refer to declared function or variable
../../go/pkg/mod/golang.org/x/[email protected]/unix/zsyscall_darwin_amd64.go:75:3: //go:linkname must refer to declared function or variable
../../go/pkg/mod/golang.org/x/[email protected]/unix/zsyscall_darwin_amd64.go:90:3: //go:linkname must refer to declared function or variable
../../go/pkg/mod/golang.org/x/[email protected]/unix/zsyscall_darwin_amd64.go:105:3: //go:linkname must refer to declared function or variable
../../go/pkg/mod/golang.org/x/[email protected]/unix/zsyscall_darwin_amd64.go:121:3: //go:linkname must refer to declared function or variable
../../go/pkg/mod/golang.org/x/[email protected]/unix/zsyscall_darwin_amd64.go:121:3: too many errors

I set the build tag to example2, which uses an edgengram tokenizer, but when I run and term search description on "eer" I get no hits. I was hoping to get a hit for beer as well.

I guess this would be an ngram, not an edgengram. But are there such examples?

Hello, I'd love to use Bleve, but I'm having a hard time figuring out how to use it. I would greatly appreciate if you could point me in the right direction. There are two things I'd like to solve:

1. Is there an easy way to ensure the data is automatically deleted after X time? For instance, with BadgerKV you can specify that the data should be deleted after 1 week.

2. Could you help give an example as to how this would be possible: let's say I want users to be able to filter on brewery_id and abv, while searching the description text.. for instance:

• brewery_id == rogue_ales OR route_66_brewery
• abv >= 2.5
• description contains text: imperial

Your help would be much appreciated :)