mirror of
https://github.com/ilri/csv-metadata-quality.git
synced 2024-11-26 07:38:18 +01:00
Alan Orth
196bb434fa
I'm only concerned with validating issue dates here. In DSpace they are generally always YYYY, YYY-MM, or YYYY-MM-DD (though in theory they could be any valid ISO8601 format). This also checks for cases where the date is missing and where the metadata has specified multiple dates like "1990||1991", as this is valid, but there is no practical value for it in our system.
33 lines
1.1 KiB
Python
33 lines
1.1 KiB
Python
import csv_metadata_quality.check as check
|
|
import csv_metadata_quality.fix as fix
|
|
import pandas as pd
|
|
import re
|
|
|
|
def main():
|
|
# Read all fields as strings so dates don't get converted from 1998 to 1998.0
|
|
#df = pd.read_csv('/home/aorth/Downloads/2019-07-26-Bioversity-Migration.csv', dtype=str)
|
|
#df = pd.read_csv('/tmp/quality.csv', dtype=str)
|
|
df = pd.read_csv('data/test.csv', dtype=str)
|
|
|
|
# Fix whitespace in all columns
|
|
for column in df.columns.values.tolist():
|
|
# Run whitespace fix on all columns
|
|
df[column] = df[column].apply(fix.whitespace)
|
|
|
|
# Run invalid multi-value separator check on all columns
|
|
df[column] = df[column].apply(check.separators)
|
|
|
|
if column == 'dc.identifier.issn':
|
|
df[column] = df[column].apply(check.issn)
|
|
|
|
if column == 'dc.identifier.isbn':
|
|
df[column] = df[column].apply(check.isbn)
|
|
|
|
# check if column is a date column like dc.date.issued
|
|
match = re.match(r'^.*?date.*$', column)
|
|
if match is not None:
|
|
df[column] = df[column].apply(check.date)
|
|
|
|
# Write
|
|
df.to_csv('/tmp/test.fixed.csv', index=False)
|