--datastet

[squigzzz]

Even with this option for some reason the client still wants to send a request to the annotatesoftwareTEI. I just want to use datastet on my xml documents though. Here is the error I am getting.

INFO:root:blacklist size: 533
DEBUG:urllib3.connectionpool:Starting new HTTP connection (1): localhost:8060
DEBUG:urllib3.connectionpool:http://localhost:8060 "GET /service/isalive HTTP/1.1" 200 4
INFO:root:dataset mention extraction is up and running
DEBUG:urllib3.connectionpool:Starting new HTTP connection (1): localhost:8060
DEBUG:urllib3.connectionpool:http://localhost:8060 "POST /service/annotateSoftwareTEI HTTP/1.1" 404 259
ERROR:root:[404] URL not found: [http://localhost:8060/service/annotateSoftwareTEI]

Is it a bug or am I doing something wrong ? Here is my module call

python3 -m software_mentions_client.client --datastet --file-in input_files/33_4_samples.grobid.tei.xml --file-out results.json

[kermitt2]

Hello @squigzzz

Datastet takes as input only PDF files, contrary to Softcite Software Mention Recognizer, which supports PDF and TEI XML.

So if we input a .tei.xml file, the client thinks that the service to use is the Softcite Software Mention Recognizer - the only possible in this case. If datastet is selected, it should better output an error message... this is something to be done indeed !

When a PDF is selected, everything works fine:

python3 -m software_mentions_client.client --datastet --file-in 
~/Downloads/PEER_stage2_10.1177_1465116507085955.pdf --file-out toto.json

INFO:root:blacklist size: 533
DEBUG:urllib3.connectionpool:Starting new HTTP connection (1): 
localhost:8060
DEBUG:urllib3.connectionpool:http://localhost:8060/ "GET /service/isalive 
HTTP/1.1" 200 4
INFO:root:dataset mention extraction is up and running
DEBUG:urllib3.connectionpool:Starting new HTTP connection (1): 
localhost:8060
DEBUG:urllib3.connectionpool:http://localhost:8060/ "POST 
/service/annotateDatasetPDF HTTP/1.1" 200 None

[squigzzz]

Got it, thank you. Is there a way to pass a grobid config file to control which deep learning models are being called or is that config file locked ?.

[kermitt2]

Hi @squigzzz
See the Grobid documentation for this: https://grobid.readthedocs.io/en/latest/Grobid-service/#configure-the-server
https://grobid.readthedocs.io/en/latest/Grobid-docker/#configure-using-the-yaml-config-file