segment_notebooks.py

import json
import itertools
from datasets import load_dataset


def segment_cells(content):
    
    # segment notebooks into lists of individual cells
    cells = []
    cell_types = []
    for cell in content['cells']:
        # select only non-empty cells
        if len(cell['source']) != 0:
            cells.append(' '.join(cell['source']))
            cell_types.append(cell['cell_type'])
            
    return cells, cell_types


def parse_notebook(batch):
    try:
        cells, types = segment_cells(json.loads(batch['content']))
        
        # get cell types and group them into lists
        cell_type_groups = [list(g) for k,g in itertools.groupby(types)]
        cell_types = [k for k,g in itertools.groupby(types)]
        cell_groups = []
        
        group_start = 0
        for g in cell_type_groups:
            cell_groups.append(cells[group_start:group_start+len(g)])
            group_start += len(g)
        
        batch['cells'] = cell_groups
        batch['cell_types'] = cell_types
        batch['cell_type_groups'] = cell_type_groups
        
    except:
        # if json.loads() returns error, skip and add a placeholder
        batch['cells'] = [['empty']]
        batch['cell_types'] = ['empty']
        batch['cell_type_groups'] = [['empty']]
        
    del batch['content']
    return batch


if __name__ == "__main__":
    
    # load dataset
    dataset = load_dataset("bigcode/the-stack",data_dir="data/jupyter-notebook", split="train",use_auth_token=True)
    # segment notebooks
    dataset = dataset.map(segment) 
    # filter out erronous cells via placeholders
    dataset = dataset.filter(lambda entry: entry['cell_types']!=['empty'])
    # push to hub
    dataset.push_to_hub("bigcode/jupyter-parsed")