asimokby / cv-parser-huggingface

A CV parser built with Hugging Face.

cv-parser-huggingface 🤗

A CV parser built with Hugging Face. The tool parses the following sections: Job History, Education History, Skills, Name, Email, Phone Numbers, and address. Here is a (NOT up to date as this repo) demo 🚀 on HuggingFace Spaces.

Installation

1. Clone the Project

git clone https://github.com/asimokby/cv-parser-huggingface.git

2. Create the environment

• You may use environment.yml or requirements.txt to setup the environment. For environment.yml, run the following commands. Replace <env_name> with the name you choose.

  conda env create --name <env_name> --file=environment.yml 

3. Activate the environment

  conda activate <env_name>

Usage

You can find the following use case in this example

from parcv import parcv

parser = parcv.Parser(pickle=True, load_pickled=True)
json_output = parser.parse('your_cv.pdf')
print(json_output)

To save the output in a json file

file_name = "output.json"
parser.save_as_json(file_name)

You can get a list of the lines in the CV:

lines = parser.get_resume_lines()
print(lines)

Or the segments/sections of the CV:

segments = parser.get_resume_segments()
print(segments)