Carolina Rau asked 2 weeks ago
Hi, I am new to the platform and I am trying to parse a PDF corpus. I have uploaded a zip file with my pdfs in it and followed the parsing instructions. I have received a error message, which tells me to provide you with this message
Data Parsing did not finish successfully. See the job log for additional information. To obtain help with this issue, ask us at the forum and include the following text:
job id: 423028
Lionel Staff answered 2 weeks ago
Dear Carolina,
Yes, thanks and sorry for that. Apparently it is just a matter of accents in the two pdf file names.
Just replace
- SISTEMATIZACIÓN-2022.pdf -> Sistematizacion-2022.pdf
- Sistematización-2018.pdf -> Sistematizacion-2018.pdf
And it will help!
L
Nathanael Jeune
replied 1 week ago Hi, I have also tried parsing a PDF corpus without any accent, all file names are like “Hemsley-Brown_Sharp_2003.pdf”
I get the following error when I look into the debug : “A filename in your dataset seems to include invalid characters, please remove any accent in the names of the files you are uploading, etc.”
Could you please add in the documentation what are the requirements in terms of file names (or even better, make the parsing more permisive in terms of names)? Thank you very much!
job id: 424616
Lionel Staff answered 1 week ago
Dear Carolina,
Yes, we should definitively improve the parsing of PDF files and/or update the documentation.
In your new dataset, just replace “-” by “_” in your pdf filenames (for Grima-Farrell_et_al_2011.pdf | Hemsley-Brown_Sharp_2003.pdf | Rycroft‐Smith_2022.pdf …)
I hope it helps!
L