Run & test web scrapers for assessment report directories

In [ ]:

    

import logging
import sys

import pandas as pd
from pandas import DataFrame, Series

sys.path.append('..')
from chemex.project import ChemexProject
from chemex import reports
from chemex.reports import dir_gsstore, gsstore_substances, dir_techstreet_gsas, hbn_format

logging.getLogger('chemex').handlers[0].setLevel('DEBUG')


    

In [ ]:

    

proj = ChemexProject('reports')


    

In [ ]:

    

gss = dir_gsstore(dedup=True)


    

In [ ]:

    

gss


    

In [ ]:

    

hbn_gss = hbn_format(gss, project=proj)


    

In [ ]:

    

hbn_gss['CASRN or Material ID'].isnull().value_counts()


    

In [ ]:

    

subst = gsstore_substances()


    

In [ ]:

    

tech = dir_techstreet_gsas()


    

In [ ]:

    

tech


    

In [ ]:

    

hbn_tech = hbn_format(tech, project=proj)


    

In [ ]:

    

hbn_tech


    

In [ ]:

    

hbn_tech['CASRN or Material ID'].isnull().value_counts()