update workflows to have `is-update` extra_data flag by default to False

update-workflow.py

import time
from invenio_db import db
from invenio_workflows import workflow_object_class
from invenio_workflows.models import ObjectStatus
from flask import current_app
from flask_sqlalchemy import SQLAlchemy


def update_workflow():
    query_ses = SQLAlchemy(current_app).session
    start_time = time.time()
    print 'Start of update'
    for index, workflow_model in enumerate(query_ses.query(workflow_object_class.dbmodel).filter_by(status=ObjectStatus.COMPLETED).yield_per(100)):
        batch_start_time = time.time()
        workflow = workflow_object_class(model=workflow_model)
        if not workflow.extra_data.get('is-update'):
            workflow_model_to_be_modified = db.session.query(workflow_object_class.dbmodel).filter_by(id=workflow.id).one()
            workflow_obj_to_be_modified = workflow_object_class(model=workflow_model_to_be_modified)
            workflow_obj_to_be_modified.extra_data['is-update'] = False
            workflow_obj_to_be_modified.save()
        if index % 100 == 0:
            print 'commit batch changes for {} records'.format(index)
            db.session.commit()
            batch_end_time = time.time()
            print 'Elapsed batch time {}'.format(batch_end_time - batch_start_time)
            print 'Elapsed overall time so far {}'.format(batch_end_time - start_time)
    db.session.commit()
    end_time = time.time()
    print 'End of update'
    print 'Elapsed overall time {}'.format(end_time - start_time)

This works inside an inspirehep app context (e.g after calling inspirehep shell). It's meant to be used in https://github.com/inspirehep/inspire-next .