Skip to content

Latest commit

 

History

History
105 lines (72 loc) · 2.9 KB

readme.md

File metadata and controls

105 lines (72 loc) · 2.9 KB
Raw

pandas-marc

pandas-marc is a lightweight Python library for working with MARC 21 bibliographic metadata using pandas dataframes. It uses pymarc for serializing to and deserializing from MARC.

Quickstart

Let's say we have a CSV file with some tabular bibliographic metadata:

MARC 007 title
ta Woman in the nineteenth century
ta The fire next time

Using pandas, we can load in the CSV and prepare it for processing into MARC:

import pandas as pd

dataframe = pd.read_csv('marc_data.csv')
dataframe.columns = ['m007', 'm245']
dataframe['m245'] = dataframe['m245'].map(lambda title: f'$a{title}.' )
dataframe['m245_indicators'] = '10'

Our dataframe now looks like this:

m007 m245 m245_indicators
ta $aWoman in the nineteenth century. 10
ta $aThe fire next time. 10

To convert the dataframe into a series of MARC records, we load it into an instance of pandas-marc's MARCDataFrame class. This allows us to generate pymarc Record objects from the dataframe rows:

from pandas_marc import MARCDataFrame

mdf = MARCDataFrame(dataframe)

for record in mdf.records:
    print(record.title())

Output:

Woman in the nineteenth century.
The fire next time.

Record objects may be readily serialized using pymarc's MARCWriter class:

from pymarc import MARCWriter

with open('marc_data.mrc', 'wb') as marc_file:
    writer = MARCWriter(marc_file)
    for record in mdf.records:
        writer.write(record)

Delimiters

pandas-marc uses the following delimiters by default:

  • Subfields are delimited with a dollar sign, $. For example, here is a MARC 100 personal name field comprising an author name and relator term under subfields a and e: $aZitkála-Šá, $eauthor.
  • Multiple field occurrences may be delimited with a backslash, \. For example, here is a series of delimited MARC 650 subject fields: $aRhetoric.\$aSpeeches.

Alternate delimiters can be specified using arguments to MARCDataFrame:

mdf = MARCDataFrame(
    dataframe=dataframe,
    occurrence_delimiter='|',
    subfield_delimiter='‡'
)

Installation

Download the latest stable release from the master branch and install with pip:

pip install pandas-marc.tar.gz

Tests

Run pandas-marc's test suite using pytest:

pytest tests/test_pandas_marc.py

# Or just...
pytest