File: json.rst

package info (click to toggle)
python-pyramid 1.6%2Bdfsg-1
  • links: PTS, VCS
  • area: main
  • in suites: stretch
  • size: 9,112 kB
  • ctags: 8,169
  • sloc: python: 41,764; makefile: 111; sh: 17
file content (103 lines) | stat: -rw-r--r-- 3,285 bytes parent folder | download
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
.. _qtut_json:

========================================
14: Ajax Development With JSON Renderers
========================================

Modern web apps are more than rendered HTML. Dynamic pages now use
JavaScript to update the UI in the browser by requesting server data as
JSON. Pyramid supports this with a *JSON renderer*.

Background
==========

As we saw in :doc:`templating`, view declarations can specify a
renderer. Output from the view is then run through the renderer,
which generates and returns the ``Response``. We first used a Chameleon
renderer, then a Jinja2 renderer.

Renderers aren't limited, however, to templates that generate HTML.
Pyramid supplies a JSON renderer which takes Python data,
serializes it to JSON, and performs some other functions such as
setting the content type. In fact, you can write your own renderer (or
extend a built-in renderer) containing custom logic for your unique
application.

Steps
=====

#. First we copy the results of the ``view_classes`` step:

   .. code-block:: bash

    $ cd ..; cp -r view_classes json; cd json
    $ $VENV/bin/python setup.py develop

#. We add a new route for ``hello_json`` in
   ``json/tutorial/__init__.py``:

   .. literalinclude:: json/tutorial/__init__.py
    :linenos:

#. Rather than implement a new view, we will "stack" another decorator
   on the ``hello`` view in ``views.py``:

   .. literalinclude:: json/tutorial/views.py
    :linenos:

#. We need a new functional test at the end of
   ``json/tutorial/tests.py``:

   .. literalinclude:: json/tutorial/tests.py
    :linenos:

#. Run the tests:

   .. code-block:: bash

    $ $VENV/bin/nosetests tutorial

#. Run your Pyramid application with:

   .. code-block:: bash

    $ $VENV/bin/pserve development.ini --reload

#. Open http://localhost:6543/howdy.json in your browser and you
   will see the resulting JSON response.

Analysis
========

Earlier we changed our view functions and methods to return Python
data. This change to a data-oriented view layer made test writing
easier, decoupling the templating from the view logic.

Since Pyramid has a JSON renderer as well as the templating renderers,
it is an easy step to return JSON. In this case we kept the exact same
view and arranged to return a JSON encoding of the view data. We did
this by:

- Adding a route to map ``/howdy.json`` to a route name

- Providing a ``@view_config`` that associated that route name with an
  existing view

- *overriding* the view defaults in the view config that mentions the 
  ``hello_json`` route, so that when the route is matched, we use the JSON 
  renderer rather than the ``home.pt`` template renderer that would otherwise 
  be used.

In fact, for pure Ajax-style web applications, we could re-use the existing
route by using Pyramid's view predicates to match on the
``Accepts:`` header sent by modern Ajax implementation.

Pyramid's JSON renderer uses the base Python JSON encoder,
thus inheriting its strengths and weaknesses. For example,
Python can't natively JSON encode DateTime objects. There are a number
of solutions for this in Pyramid, including extending the JSON renderer
with a custom renderer.

.. seealso:: :ref:`views_which_use_a_renderer`,
   :ref:`json_renderer`, and
   :ref:`adding_and_overriding_renderers`