How to render a python dict as an html table with vertical columns

I've got a python dict with where each key corresponds to a heading, and the list associated with each heading contains an arbitrary number of values:

data = { 
    "heading1": ['h1-val1', 'h1-val2', 'h1-val3', ],
    "heading2": ['h2-val1', ],
    "heading3": ['h3-val1', 'h3-val2', 'h3-val3', 'h3-val4', ],
} 

I need to render this in a Django template as a table, where the values are listed vertically beneath each heading, with any missing values rendered as an empty table cell:

<table>
<thead>
    <tr>
    <th>heading1</th>
    <th>heading2</th>
    <th>heading3</th>
    </tr>
</thead>
<tbody>
    <tr>
    <td>h1-val1</td>
    <td>h2-val1</td>
    <td>h3-val1</td>
    </tr>
    <tr>
    <td>h1-val2</td>
    <td></td>
    <td>h3-val2</td>
    </tr>
    <tr>
    <td>h1-val3</td>
    <td></td>
    <td>h3-val3</td>
    </tr>
    <tr>
    <td></td>
    <td></td>
    <td>h3-val4</td>
    </tr>
</tbody>
</table>

What's the best way to achieve this?

My first inclination is to rearrange the original dict into a 2D matrix, and just pass that into the template. I'm sure I'm not the first to run into this kind of problem, though, and I'm curious how others have solved this problem.

UPDATE: Just for reference, here's my original solution to this problem (which I'm not very happy with).

# Using the data dict from the question:
size = max(len(data['heading1']), len(data['heading2']), len(data['heading3']))
matrix = [[None, None, None] for i in r开发者_JAVA百科ange(size)] # initialize an empty matrix

# manually copy the data into the appropriate column :(
i = 0
for item in data['heading1']:
    matrix[i][0] = item
    i += 1
i = 0
for item in data['heading2']:
    matrix[i][1] = item
    i += 1
i = 0
for item in data['heading3']:
    matrix[i][2] = item
    i += 1

I then passed the matrix into the template which looked like this:

<table>
<thead><tr>
    <th>heading1</th>
    <th>heading2</th>
    <th>heading3</th>
</tr></thead>
<tbody>
{% for row in matrix %}
    <tr>
    {% for col in row %}
        <td>{% if col %}{{ col }}{% else %}&nbsp;{% endif %}</td>
    {% endfor %}
    </tr>
{% endfor %}
</tbody>
</table>

If we change the game a little bit, it's actually a snap to turn this around (so long as your lists are None filled...)

from django.template import Context, Template

data = {
    "heading1": ['h1-val1', 'h1-val2', 'h1-val3', ],
    "heading2": ['h2-val1', ],
    "heading3": ['h3-val1', 'h3-val2', 'h3-val3', 'h3-val4', ],
}

# we'll need to split the headings from the data
# rather than using keys() I'm just hard coding so I can control the order
headings = ["heading1", "heading2", "heading3"]

columns = [data[heading] for heading in headings]

# get the length of the longest column
max_len = len(max(columns, key=len))

for col in columns:
    # padding the short columns with None
    col += [None,] * (max_len - len(col))

# Then rotate the structure...
rows = [[col[i] for col in columns] for i in range(max_len)]


dj_template ="""
<table>
{# headings #}
    <tr>
    {% for heading in headings %}
        <th>{{ heading }}</th>
    {% endfor %}
    </tr>
{# data #}
{% for row in data %}
    <tr>
        {% for val in row %}
        <td>{{ val|default:'' }}</td>
        {% endfor %}
    </tr>
{% endfor %}
</table>
"""

# finally, the code I used to render the template:
tmpl = Template(dj_template)
tmpl.render(Context(dict(data=rows, headings=headings)))

For me, this produces the following (blank lines stripped):

<table>
    <tr>
        <th>heading1</th>
        <th>heading2</th>
        <th>heading3</th>
    </tr>
    <tr>
        <td>h1-val1</td>
        <td>h2-val1</td>
        <td>h3-val1</td>
    </tr>
    <tr>
        <td>h1-val2</td>
        <td></td>
        <td>h3-val2</td>
    </tr>
    <tr>
        <td>h1-val3</td>
        <td></td>
        <td>h3-val3</td>
    </tr>
    <tr>
        <td></td>
        <td></td>
        <td>h3-val4</td>
    </tr>
</table>

Kenneth Reitz suggested that you could also solve this problem using tablib, so I thought I'd include that here as well:

import tablib

d = tablib.Dataset()
d.append_col(['h1-val1', 'h1-val2', 'h1-val3', ''], header="heading1")
d.append_col(['h2-val1', 'h2-val2', '', ''], header="heading2")  
d.append_col(['h3-val1', 'h3-val2', 'h3-val3', 'h3-val4', ], header="heading3") 
d.headers = ['heading1', 'heading2', 'heading3']

This dumps all the pertinent data in a Dataset, which you could then render in a template with the following:

{{ d.html }}

Which produces html that looks like this:

<table>
<thead>
<tr><th>heading1</th>
<th>heading2</th>
<th>heading3</th></tr>
</thead>
<tr><td>h1-val1</td>
<td>h2-val1</td>
<td>h3-val1</td></tr>
<tr><td>h1-val2</td>
<td>h2-val2</td>
<td>h3-val2</td></tr>
<tr><td>h1-val3</td>
<td></td>
<td>h3-val3</td></tr>
<tr><td></td>
<td></td>
<td>h3-val4</td></tr>
</table>

Try the code below. Note that since you are using associative array for data, the order in which the headings appears cannot be guaranteed.

print "<table>"

print "<thead><tr>"
order = []
for k in data.keys():
    print "<td>" + k + "</td>"
    order.append(k)
print "</tr></thead>"

print "<tbody>"
for k in order:
    print "<tr>"
    for v in data[k]:
        print "<td>" + v + "</td>"
    print "</tr>"
print "</tbody>"
print "</table>"