【error】There is already a relation named obl:agent!

[unikcc]

First I start a local server via StanfordNLP. Then I send a post resquest by python packages named requests to parse sentence. It throwed an error when I sent a sentence. 【sentence】 Technical problems will include user manipulation of agents through and agent editor , activation of agents in a shared context , presentation of agents , and creation of shared context through specification , construction , task representations , and interaction histories . 【code】

import requests as rq
url = 'http://172.16.133.173:9000/?properties={"annotators":"tokenize, ssplit, pos, lemma, ner,     depparse","outputFormat":"json"}'
res = rq.post(url, data=sentence).text

【error】 java.util.concurrent.ExecutionException: java.lang.IllegalArgumentException: There is already a relation named obl:agent!

[J38]

What version of CoreNLP are you using? I was able to run on that sentence and not get an error.

[AngledLuffa]

Fascinating. First it should be pointed out that StanfordNLP is deprecated. Please use stanza instead:

github.com/stanfordnlp/stanza

The next thing I would ask is to explain what you are doing to get this error. As @j38 says, we are not able to reproduce this error.

[unikcc]

@J38 @AngledLuffa Thanks for your kind reply~ I just want to get the dependency parsing tree of each sentences. And I'm used to do this via StanfordCoreNLP. It's easy to use and works very well. Here are some details:

【platform】: Ubuntu 18.04 LTS 
【Java vesion】: 1.8.0_192
【StanfordCoreNLP version】: 4.0.0
【Start server code】: nohup java -mx4g -cp "*" edu.stanford.nlp.pipeline.StanfordCoreNLPServer -port 9000 -timeout 15000 > 1.log 2>&1 &

So I'll also go to learn how to use stanza~

[AngledLuffa]

What I meant is, how did you parse the text to trigger this error?

[unikcc]

What I meant is, how did you parse the text to trigger this error?

All of the contexts are here:

import requests as rq
url = 'http://172.16.133.173:9000/?properties={"annotators":"tokenize, ssplit, pos, lemma, ner,     depparse","outputFormat":"json"}'
sentence = "Technical problems will include user manipulation of agents through and agent editor , activation of agents in a shared context , presentation of agents , and creation of shared context through specification , construction , task representations , and interaction histories ."
res = rq.post(url, data=sentence).text

Then it throws an error.

[AngledLuffa]

Sorry, but that still doesn't trigger it (when sent to my own local server).

What if you use the new 4.1.0 release? Just in case something was fixed without us realizing it. http://nlp.stanford.edu/software/corenlp.shtml

[unikcc]

Sorry, but that still doesn't trigger it (when sent to my own local server). What if you use the new 4.1.0 release? Just in case something was fixed without us realizing it. http://nlp.stanford.edu/software/corenlp.shtml

Magically. The new 4.1.0 release works perfectly. I'll use the lastest version🚀🚀🚀. Thank you!