Skip to main content
This is a pre-production deployment of Warehouse. Changes made here affect the production instance of PyPI (
Help us improve Python packaging - Donate today!

An un-opinionated template abstraction layer

Project Description


2013/07/29: TemplateAlchemy is in its very early stages - you should come back later.

TemplateAlchemy aims to be to the fragmented world of templates what SQLAlchemy is to the world of databases: a generic abstraction for systems that need templated data, but don’t care about what language or implementation is used to render that data.

In essence, the primary purpose of this package is to allow other packages that need templated data (such as the email generation package genemail) to remain un-opinionated about template format and source, while still providing useful higher-level functionality.

By default, TemplateAlchemy uses pkgutil for template location (i.e. files from a package, either on the filesystem or in a zip archive) and Mako for rendering, but these settings are trivial to configure to use something else. For example, you may want to store templates in a database (with built-in support) and use the Jinja2 rendering engine (supported via the TemplateAlchemy-Jinja2 package).



$ pip install templatealchemy


import templatealchemy as ta

# create a top-level template manager
root = ta.Manager(
  source = 'pkg:mypackagename:lib/templates',
  renderer = 'mako',

# load the sub-template 'foo'
foo = root.getTemplate('foo')

# render the 'text' version with some parameters; the
# actual template is then in 'mypackagename:lib/templates/foo.text'
params = dict(value='bar', zig='zog')
text = foo.render('text', params)

# get meta information about the template
if 'attachments' in foo.meta:
   for attachment in foo.meta.attachments:
     # ... do something with each attachment

# supported formats are stored in meta.formats
assert(foo.meta.formats == ['text', 'html'])


The primary API exposed by TemplateAlchemy is the templatealchemy.Template. A template has two functions:

  • Generate content: given a requested format and parameters, a template renders serialized output data.
  • Expose meta-information: certain constructs, such as email generation and generated data previews, require that the template disclose some kinds of information. Basically, templates can provide additional information to give context to the template.

In order to achieve these functions, templates use sources and renderers. Sources provide read access to persistent storage and renderers convert source data into output form.


The special format spec is reserved and used to parameterize template meta information; it must not be used as a format.


Although TemplateAlchemy comes with several built-in template sources, it exposes a generic API that can be extended to support any template storage mechanism. The following built-in sources exist:

  • file:

    Fetches templates from the filesystem. See API Details for more information.

  • pkg:

    The pkg source extracts sources from a python package. This means that it can stream the data directly from the filesystem or from a zip-archive (depending on how the package was installed). See API Details for more information.

  • sqlalchemy:

    The sqlalchemy source fetches sources from a database using the SQLAlchemy database abstraction library. See API Details for more information.

  • stream:

    The stream source reads a template from a file-like object; because it does not buffer any data, the template is single-use. This is typically used in programs that know that the template will only be used once, such as command-line programs.

  • string:

    The string source allows a simple way to provide templates inline. Generally not very useful beyond that – serious re-evaluation is recommended if this is used frequently in an application.


Once a template has been loaded from a source, it is rendered to serialized form by a renderer. Just like sources, TemplateAlchemy uses an abstract interface for this function, and therefore can support any rendering engine. TemplateAlchemy has support for the following engines built-in:

  • mako:

    Probably the most efficient and most advanced python templating engine, mako is the recommended engine. However, it does allow arbitrary python to be executed, so the input data must be trusted. See API Details for more information.

  • mustache:

    A logic-less templating engine that is very simple and effective. Since it does not allow arbitrary python to be executed, this is a better choice of renderer if the input data is not trusted. See API Details for more information.

API Details

This section provides in-depth API information. Both sources and renderers can be passed to TemplateAlchemy either as an implementation of the respective API objects or as string specifications. In the latter case, the string must be in the format TYPE:SPEC, for example "mako:default_filters=[h]". The :SPEC can be left off to use default values, for example "mako".


Abstract Interface

The abstract interface for a TemplateAlchemy source is in templatealchemy.api.Source, which has the following definition:

class templatealchemy.api.Source(object):

  def get(self, format):
    Returns the source content stream for the current template
    source for the specified `format`. The returned object must be a
    file-like object supporting read access.

  def getSource(self, name):
    Returns a subsidiary source template, relative to the current
    template, with the specified `name`. This is seen as a hierchical
    relationship, and is typically represented as a slash ('/')
    delimited path.

  def getFormats(self):
    Returns a list of all the available formats for this source.

  def getRelated(self, name):
    Returns a content stream for the related object `name` that
    is relative to the current template. Typically this is used
    for meta-information *spec* definitions using the "!include"
    or "!include-raw" directives. As with :meth:`get`, the
    returned object must be a file-like object supporting read

File Hierarchy (‘file’ and ‘pkg’ sources)

The file source expects the path to the template hierarchy as a specification, e.g. if the templates are located in /var/lib/templates, then the source spec should be file:/var/lib/templates.

The pkg source expects the package name and relative path to the template hierarchy as a specification separated by a colon (‘:’), e.g. if the templates are located in the demo package and within its templates directory, then the source spec should be pkg:demo:templates.

Template hierarchies for the file and pkg sources map directly to filesystem hierarchies. (Note that for the pkg source, these may be stored in a zip archive depending on installation method, but will be treated the same.) When rendering, the format maps directly to the file extension, adjusted for any spec rules.

For example, given the following filesystem structure:

-- /myroot/
   `-- foo/
       |-- bar.html      | content: '<html><p>{{name}}</p></html>'
       `-- bar.text      | content: 'Name is {{name}}'

The following code will pass the assert:

import templatealchemy as ta
root = ta.Manager(source='file:/myroot', renderer='mustache')
bar  = root.getTemplate('foo/bar')

assert(bar.render('text', dict(name='Joe')) == 'Name is Joe')
assert(bar.render('html', dict(name='Joe')) == '<html><p>Joe</p></html>')


The sqlalchemy source allows templates to be store in any database that the SQLAlchemy python library supports. The sqlalchemy specification is simply the database URL as you would pass it to sqlalchemy.create_engine. For example, if the templates were stored in the /var/lib/templates.db sqlite database, then the source spec would be sqlalchemy:sqlite:////var/lib/templates.db.

By default, the sqlalchemy source expects a table named template to exist in the database, with the columns name, format and content. Currently, the templatealchemy.sqlalchemy implementation does not support the use of sessions; to use them instead of the standard direct connection, use a subclass of templatealchemy.sqlalchemy.SaSource.

For example, given the following database content:

$ sqlite3 -header -column /var/lib/templates.db 'select * from template'
name        format      content
----------  ----------  ----------------------------
foo/bar     html        <html><p>{{name}}</p></html>
foo/bar     text        Name is {{name}}

The following code will pass the assert:

import templatealchemy as ta
root = ta.Manager(source='sqlalchemy:sqlite:////var/lib/templates.db',
bar  = root.getTemplate('foo/bar')

assert(bar.render('text', dict(name='Joe')) == 'Name is Joe')
assert(bar.render('html', dict(name='Joe')) == '<html><p>Joe</p></html>')


Abstract Interface

The abstract interface for a TemplateAlchemy renderer is in templatealchemy.api.Renderer, which has the following definition:

class templatealchemy.api.Renderer(object):

  def render(self, context, stream, params):
    Renders the given template data `stream` (as a read-access
    file-like object) to serialized rendered output. The given
    `params` provide variables that are typically passed to the
    template using template-specific mechanisms.

    todo: update this when the time comes:

    `context` is a reserved parameter that is intended to enable
    cross-driver optimizations, but has not been defined at this


TODO: add docs


TODO: add docs

Release History

Release History

This version
History Node


History Node


History Node


History Node


History Node


History Node


History Node


History Node


History Node


History Node


History Node


History Node


History Node


History Node


History Node


History Node


History Node


History Node


History Node


Download Files

Download Files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

File Name & Checksum SHA256 Checksum Help Version File Type Upload Date
TemplateAlchemy-0.1.22.tar.gz (20.1 kB) Copy SHA256 Checksum SHA256 Source Jul 14, 2017

Supported By

WebFaction WebFaction Technical Writing Elastic Elastic Search Pingdom Pingdom Monitoring Dyn Dyn DNS Sentry Sentry Error Logging CloudAMQP CloudAMQP RabbitMQ Heroku Heroku PaaS Kabu Creative Kabu Creative UX & Design Fastly Fastly CDN DigiCert DigiCert EV Certificate Rackspace Rackspace Cloud Servers DreamHost DreamHost Log Hosting