kevin-brown

Nested ViewSets for DRF

DRF doesn't allow nested ViewSets natively, so I figured out the least amount of tweaks required to get it working.

The fields that are provided allow you to override the reverse() call with a specific mapping. These cannot be used across urls yet, as it pulls straight from the kwargs.

In your field, you must define map, which is a dictionary of destinaton => previous kwargs keys. These keys are relative to the current URL.

kevin-brown

• MyFitnessPal
• MtGox - Offline
• SpokenVote - Group or location selector
• Ckan - Language selector (and all other searchable dropdowns)
• Khan Academy
• Cargo Media
• CartoDB
• Clinical Software Solutions
• Contatta - Rebranded
• Copy - Discontinued

kevin-brown

class DisablePaginationMixin(object):
    def get_paginate_by(self, queryset=None):
        if self.paginate_by_param in self.request.QUERY_PARAMS and \
           self.request.QUERY_PARAMS[self.paginate_by_param] == '0':
            return None

        return super(DisablePaginationMixin, self).get_paginate_by(queryset)

kevin-brown

class CompatibleDispatchInitialization(object):
    """
    Allow data to be initialized before a view is fully dispatched in a way
    that is consistent across both standard Django views and Django REST
    Framework views.
    
    While initializing data would typically be done within `dispatch`, the
    request is not fully initialized at that point during DRF views which
    causes issues with authentication. For this reason, it is recommended to
    do data initialization within the `initial` method for DRF views. The

kevin-brown

# Do not include the browsable api renderer in the initial list of classes
if not DEBUG:
    REST_FRAMEWORK["DEFAULT_RENDERER_CLASSES"] += [
        "rest_framework.renderers.BrowsableAPIRenderer",
    ]

kevin-brown

import fileinput
import re
import sys

file_name = sys.argv[1]
new_file = sys.argv[2]

rh = open(file_name, "r")
lines = rh.readlines()
rh.close()

kevin-brown

Development workflow

Over the summer we switched our workflow to include GitHub in order to help up facilitate the rewrite of MouseTrap. Now that the summer is over, we are left wondering how we should handle contributions to the project.

MouseTrap workflows

MouseTrap has had two main workflows during the time that I have been on the project.

kevin-brown

Assuming bare, untouched Windows 7 Ultimate x64

Install git

http://git-scm.com/downloads

Clone repository

git clone https://github.com/GNOME-MouseTrap/mousetrap.git

kevin-brown

1. Install Python 3 and the development packages

• sudo yum install python3 python3-devel

2. Install the package manager for Python 3

• sudo yum install python3-setuptools python3-pip

3. Modify setup.py to work for us.

• Change os.walk(LOCALE_PATH).next() to next(os.walk(LOCALE_PATH)) on line 123

4. Install MouseTrap using pip

kevin-brown

#0  0x00007fffe1c8230e in int cv::predictOrderedStump<cv::HaarEvaluator>(cv::CascadeClassifierImpl&, cv::Ptr<cv::FeatureEvaluator>&, double&) () from /usr/local/lib/libopencv_objdetect.so.3.0
#1  0x00007fffe1c82e67 in cv::CascadeClassifierImpl::runAt(cv::Ptr<cv::FeatureEvaluator>&, cv::Point_<int>, int, double&) () from /usr/local/lib/libopencv_objdetect.so.3.0
#2  0x00007fffe1c85639 in cv::CascadeClassifierInvoker::operator()(cv::Range const&) const () from /usr/local/lib/libopencv_objdetect.so.3.0
#3  0x00007fffe1c863d1 in cv::CascadeClassifierImpl::detectMultiScaleNoGrouping(cv::_InputArray const&, std::vector<cv::Rect_<int>, std::allocator<cv::Rect_<int> > >&, std::vector<int, std::allocator<int> >&, std::vector<double, std::allocator<double> >&, double, cv::Size_<int>, cv::Size_<int>, bool) () from /usr/local/lib/libopencv_objdetect.so.3.0
#4  0x00007fffe1c8e42d in cv::CascadeClassifierImpl::detectMultiScale(cv::_InputArray const&, std::vector<cv::Rect_<int>, std::allocator<cv::Rect_<int> > >&, std::vect