inspirehep / inspire-utils Goto Github PK

View Code? Open in Web Editor NEW

2.0 5.0 19.0 179 KB

INSPIRE-specific utils.

License: GNU General Public License v3.0

Python 99.30% Shell 0.70%

inspirehep python

inspire-utils's Introduction

INSPIRE-Utils

About

INSPIRE-specific utils.

inspire-utils's People

Contributors

Stargazers

Watchers

Forkers

michamos chris-asl david-caro szymonlopaciuk vbalbp iulianav drjova harunurhan pazembrz monaawi ammirate miguelgrc mjedr hanadikaleel oguzdemirbasci nooraangelva yigitsasmazs jiangfy-ihep

inspire-utils's Issues

name: capitalization, should we use the ``force`` (parameter)?

HumanName supports capitalization of names (related docs).

Currently, we have capitalization enabled, which does a best effort to capitalize names.
Although, without the force=True it doesn't do anything on mixed case names.

In addition to the example from their docs, the same happens for names like: Ellis, john, the result will be the same, and not Ellis, John.
For the aforementioned case it seems logical to use it, but for the case of their example Shirley Maclaine (converting it to Shirley MacLaine) I don't know.

@michamos @jacquerie
It's not something that I need for authors search, but something to keep in mind and decide upon.

autor normalization: concatenated initials

From @jacquerie on August 26, 2017 23:9

From @annetteholtkamp on August 18, 2017 15:13

Expected Behavior

A name with concatenated capital initials like "Fitzakerley, DW" should be converted to
"Fitzakerley, D.W."
ex: https://inspirehep.net/record/1488707

This conversion should only happen, if the family name is not all caps since we may encounter Chinese names like "SHI, YU"

Context

Currently, cataloguers have to change this manually which is painful for long author lists and may easily be overlooked. In the current system I believe this may generate superfluous new author profiles.

Copied from original issue: inspirehep/inspire-next#2662

Copied from original issue: inspirehep/inspire-schemas#221

name: bad parses found in the tests of inspire-dojson

>>> from inspire_utils import normalize_name
>>> normalize_name('W, Schoutens.')
'W, Schoutens'  # should be 'Schoutens, W.'
>>> normalize_name(u'Erdélyi,A.')
'A.'            # should be u''Erdélyi, A.'

get_value needs to have a dict as first argument

it would be nice to be able to call it on a list of dicts also, e.g. https://github.com/michamos/inspire-next/blob/6481afdb3bf52e25c6b99b52ecf534d3cca7cdd3/tests/unit/editor/test_editor_views.py#L54 instead of using the workaround of wrapping it in a dict.

Line 196 in a3d3022

def normalize_date(date, **kwargs):