관리-도구

편집 파일: autogenerate.html

<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN"
 "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">

<html xmlns="http://www.w3.org/1999/xhtml">
 <head>
 <meta http-equiv="Content-Type" content="text/html; charset=utf-8" />
 
 <title>Auto Generating Migrations &mdash; Alembic 0.8.3 documentation</title>
 
 <link rel="stylesheet" href="_static/nature_override.css" type="text/css" />
 <link rel="stylesheet" href="_static/pygments.css" type="text/css" />
 <link rel="stylesheet" href="_static/changelog.css" type="text/css" />
 <link rel="stylesheet" href="_static/sphinx_paramlinks.css" type="text/css" />
 
 <script type="text/javascript">
 var DOCUMENTATION_OPTIONS = {
 URL_ROOT: './',
 VERSION: '0.8.3',
 COLLAPSE_INDEX: false,
 FILE_SUFFIX: '.html',
 HAS_SOURCE: true
 };
 </script>
 <script type="text/javascript" src="_static/jquery.js"></script>
 <script type="text/javascript" src="_static/underscore.js"></script>
 <script type="text/javascript" src="_static/doctools.js"></script>
 <link rel="top" title="Alembic 0.8.3 documentation" href="index.html" />
 <link rel="next" title="Generating SQL Scripts (a.k.a. “Offline Mode”)" href="offline.html" />
 <link rel="prev" title="Tutorial" href="tutorial.html" /> 
 </head>
 <body role="document">
 <div class="related" role="navigation" aria-label="related navigation">
 <h3>Navigation</h3>
 <ul>
 <li class="right" style="margin-right: 10px">
 <a href="genindex.html" title="General Index"
 accesskey="I">index</a></li>
 <li class="right" >
 <a href="py-modindex.html" title="Python Module Index"
 >modules</a> |</li>
 <li class="right" >
 <a href="offline.html" title="Generating SQL Scripts (a.k.a. “Offline Mode”)"
 accesskey="N">next</a> |</li>
 <li class="right" >
 <a href="tutorial.html" title="Tutorial"
 accesskey="P">previous</a> |</li>
 <li class="nav-item nav-item-0"><a href="index.html">Alembic 0.8.3 documentation</a> &raquo;</li> 
 </ul>
 </div>

<div class="document">
 <div class="documentwrapper">
 <div class="bodywrapper">
 <div class="body" role="main">
 
 <div class="section" id="auto-generating-migrations">
<h1>Auto Generating Migrations<a class="headerlink" href="#auto-generating-migrations" title="Permalink to this headline">¶</a></h1>
Alembic can view the status of the database and compare against the table metadata
in the application, generating the &#8220;obvious&#8221; migrations based on a comparison. This
is achieved using the <code class="docutils literal">--autogenerate</code> option to the <code class="docutils literal">alembic revision</code> command,
which places so-called candidate migrations into our new migrations file. We
review and modify these by hand as needed, then proceed normally.
To use autogenerate, we first need to modify our <code class="docutils literal">env.py</code> so that it gets access
to a table metadata object that contains the target. Suppose our application
has a <a class="reference external" href="http://www.sqlalchemy.org/docs/orm/extensions/declarative.html#synopsis">declarative base</a>
in <code class="docutils literal">myapp.mymodel</code>. This base contains a <a class="reference external" href="http://www.sqlalchemy.org/docs/core/metadata.html#sqlalchemy.schema.MetaData" title="(in SQLAlchemy v1.0)"><code class="xref py py-class docutils literal">MetaData</code></a> object which
contains <a class="reference external" href="http://www.sqlalchemy.org/docs/core/metadata.html#sqlalchemy.schema.Table" title="(in SQLAlchemy v1.0)"><code class="xref py py-class docutils literal">Table</code></a> objects defining our database. We make sure this
is loaded in <code class="docutils literal">env.py</code> and then passed to <a class="reference internal" href="api/runtime.html#alembic.runtime.environment.EnvironmentContext.configure" title="alembic.runtime.environment.EnvironmentContext.configure"><code class="xref py py-meth docutils literal">EnvironmentContext.configure()</code></a> via the
<code class="docutils literal">target_metadata</code> argument. The <code class="docutils literal">env.py</code> sample script used in the
generic template already has a
variable declaration near the top for our convenience, where we replace <code class="docutils literal">None</code>
with our <a class="reference external" href="http://www.sqlalchemy.org/docs/core/metadata.html#sqlalchemy.schema.MetaData" title="(in SQLAlchemy v1.0)"><code class="xref py py-class docutils literal">MetaData</code></a>. Starting with:
<div class="highlight-python"><div class="highlight"><pre># add your model&#39;s MetaData object here
# for &#39;autogenerate&#39; support
# from myapp import mymodel
# target_metadata = mymodel.Base.metadata
target_metadata = None
</pre></div>
</div>
we change to:
<div class="highlight-python"><div class="highlight"><pre>from myapp.mymodel import Base
target_metadata = Base.metadata
</pre></div>
</div>
<div class="admonition note">
Note
The above example refers to the generic alembic env.py template, e.g.
the one created by default when calling upon <code class="docutils literal">alembic init</code>, and not
the special-use templates such as <code class="docutils literal">multidb</code>. Please consult the source
code and comments within the <code class="docutils literal">env.py</code> script directly for specific
guidance on where and how the autogenerate metadata is established.
</div>
If we look later in the script, down in <code class="docutils literal">run_migrations_online()</code>,
we can see the directive passed to <a class="reference internal" href="api/runtime.html#alembic.runtime.environment.EnvironmentContext.configure" title="alembic.runtime.environment.EnvironmentContext.configure"><code class="xref py py-meth docutils literal">EnvironmentContext.configure()</code></a>:
<div class="highlight-python"><div class="highlight"><pre>def run_migrations_online():
 engine = engine_from_config(
 config.get_section(config.config_ini_section), prefix=&#39;sqlalchemy.&#39;)

with engine.connect() as connection:
 context.configure(
 connection=connection,
 target_metadata=target_metadata
 )

with context.begin_transaction():
 context.run_migrations()
</pre></div>
</div>
We can then use the <code class="docutils literal">alembic revision</code> command in conjunction with the
<code class="docutils literal">--autogenerate</code> option. Suppose
our <a class="reference external" href="http://www.sqlalchemy.org/docs/core/metadata.html#sqlalchemy.schema.MetaData" title="(in SQLAlchemy v1.0)"><code class="xref py py-class docutils literal">MetaData</code></a> contained a definition for the <code class="docutils literal">account</code> table,
and the database did not. We&#8217;d get output like:
<div class="highlight-python"><div class="highlight"><pre>$ alembic revision --autogenerate -m &quot;Added account table&quot;
INFO [alembic.context] Detected added table &#39;account&#39;
Generating /path/to/foo/alembic/versions/27c6a30d7c24.py...done
</pre></div>
</div>
We can then view our file <code class="docutils literal">27c6a30d7c24.py</code> and see that a rudimentary migration
is already present:
<div class="highlight-python"><div class="highlight"><pre>&quot;&quot;&quot;empty message

Revision ID: 27c6a30d7c24
Revises: None
Create Date: 2011-11-08 11:40:27.089406

&quot;&quot;&quot;

# revision identifiers, used by Alembic.
revision = &#39;27c6a30d7c24&#39;
down_revision = None

from alembic import op
import sqlalchemy as sa

def upgrade():
 ### commands auto generated by Alembic - please adjust! ###
 op.create_table(
 &#39;account&#39;,
 sa.Column(&#39;id&#39;, sa.Integer()),
 sa.Column(&#39;name&#39;, sa.String(length=50), nullable=False),
 sa.Column(&#39;description&#39;, sa.VARCHAR(200)),
 sa.Column(&#39;last_transaction_date&#39;, sa.DateTime()),
 sa.PrimaryKeyConstraint(&#39;id&#39;)
 )
 ### end Alembic commands ###

def downgrade():
 ### commands auto generated by Alembic - please adjust! ###
 op.drop_table(&quot;account&quot;)
 ### end Alembic commands ###
</pre></div>
</div>
The migration hasn&#8217;t actually run yet, of course. We do that via the usual <code class="docutils literal">upgrade</code>
command. We should also go into our migration file and alter it as needed, including
adjustments to the directives as well as the addition of other directives which these may
be dependent on - specifically data changes in between creates/alters/drops.
<div class="section" id="what-does-autogenerate-detect-and-what-does-it-not-detect">
<h2>What does Autogenerate Detect (and what does it not detect?)<a class="headerlink" href="#what-does-autogenerate-detect-and-what-does-it-not-detect" title="Permalink to this headline">¶</a></h2>
The vast majority of user issues with Alembic centers on the topic of what
kinds of changes autogenerate can and cannot detect reliably, as well as
how it renders Python code for what it does detect. it is critical to
note that autogenerate is not intended to be perfect. It is always
necessary to manually review and correct the candidate migrations
that autogenererate produces. The feature is getting more and more
comprehensive and error-free as releases continue, but one should take
note of the current limitations.
Autogenerate will detect:
<ul class="simple">
<li>Table additions, removals.</li>
<li>Column additions, removals.</li>
<li>Change of nullable status on columns.</li>
<li>Basic changes in indexes and explcitly-named unique constraints</li>
</ul>
<div class="versionadded">
New in version 0.6.1: Support for autogenerate of indexes and unique constraints.
</div>
<ul class="simple">
<li>Basic changes in foreign key constraints</li>
</ul>
<div class="versionadded">
New in version 0.7.1: Support for autogenerate of foreign key constraints.
</div>
Autogenerate can optionally detect:
<ul class="simple">
<li>Change of column type. This will occur if you set
the <a class="reference internal" href="api/runtime.html#alembic.runtime.environment.EnvironmentContext.configure.params.compare_type" title="alembic.runtime.environment.EnvironmentContext.configure"><code class="xref py py-paramref docutils literal">EnvironmentContext.configure.compare_type</code></a> parameter
to <code class="docutils literal">True</code>, or to a custom callable function.
The feature works well in most cases,
but is off by default so that it can be tested on the target schema
first. It can also be customized by passing a callable here; see the
section <a class="reference internal" href="#compare-types">Comparing Types</a> for details.</li>
<li>Change of server default. This will occur if you set
the <a class="reference internal" href="api/runtime.html#alembic.runtime.environment.EnvironmentContext.configure.params.compare_server_default" title="alembic.runtime.environment.EnvironmentContext.configure"><code class="xref py py-paramref docutils literal">EnvironmentContext.configure.compare_server_default</code></a>
parameter to <code class="docutils literal">True</code>, or to a custom callable function.
This feature works well for simple cases but cannot always produce
accurate results. The Postgresql backend will actually invoke
the &#8220;detected&#8221; and &#8220;metadata&#8221; values against the database to
determine equivalence. The feature is off by default so that
it can be tested on the target schema first. Like type comparison,
it can also be customized by passing a callable; see the
function&#8217;s documentation for details.</li>
</ul>
Autogenerate can not detect:
<ul class="simple">
<li>Changes of table name. These will come out as an add/drop of two different
tables, and should be hand-edited into a name change instead.</li>
<li>Changes of column name. Like table name changes, these are detected as
a column add/drop pair, which is not at all the same as a name change.</li>
<li>Anonymously named constraints. Give your constraints a name,
e.g. <code class="docutils literal">UniqueConstraint('col1', 'col2', name=&quot;my_name&quot;)</code>. See the section
<a class="reference internal" href="naming.html">The Importance of Naming Constraints</a> for background on how to configure automatic naming schemes
for constraints.</li>
<li>Special SQLAlchemy types such as <a class="reference external" href="http://www.sqlalchemy.org/docs/core/type_basics.html#sqlalchemy.types.Enum" title="(in SQLAlchemy v1.0)"><code class="xref py py-class docutils literal">Enum</code></a> when generated
on a backend which doesn&#8217;t support ENUM directly - this because the
representation of such a type
in the non-supporting database, i.e. a CHAR+ CHECK constraint, could be
any kind of CHAR+CHECK. For SQLAlchemy to determine that this is actually
an ENUM would only be a guess, something that&#8217;s generally a bad idea.
To implement your own &#8220;guessing&#8221; function here, use the
<a class="reference external" href="http://www.sqlalchemy.org/docs/core/events.html#sqlalchemy.events.DDLEvents.column_reflect" title="(in SQLAlchemy v1.0)"><code class="xref py py-meth docutils literal">sqlalchemy.events.DDLEvents.column_reflect()</code></a> event
to detect when a CHAR (or whatever the target type is) is reflected,
and change it to an ENUM (or whatever type is desired) if it is known that
that&#8217;s the intent of the type. The
<a class="reference external" href="http://www.sqlalchemy.org/docs/core/events.html#sqlalchemy.events.DDLEvents.after_parent_attach" title="(in SQLAlchemy v1.0)"><code class="xref py py-meth docutils literal">sqlalchemy.events.DDLEvents.after_parent_attach()</code></a>
can be used within the autogenerate process to intercept and un-attach
unwanted CHECK constraints.</li>
</ul>
Autogenerate can&#8217;t currently, but will eventually detect:
<ul class="simple">
<li>Some free-standing constraint additions and removals,
like CHECK, PRIMARY KEY - these are not fully implemented.</li>
<li>Sequence additions, removals - not yet implemented.</li>
</ul>
</div>
<div class="section" id="comparing-and-rendering-types">
<h2>Comparing and Rendering Types<a class="headerlink" href="#comparing-and-rendering-types" title="Permalink to this headline">¶</a></h2>
The area of autogenerate&#8217;s behavior of comparing and rendering Python-based type objects
in migration scripts presents a challenge, in that there&#8217;s
a very wide variety of types to be rendered in scripts, including those
part of SQLAlchemy as well as user-defined types. A few options
are given to help out with this task.
<div class="section" id="controlling-the-module-prefix">
<h3>Controlling the Module Prefix<a class="headerlink" href="#controlling-the-module-prefix" title="Permalink to this headline">¶</a></h3>
When types are rendered, they are generated with a module prefix, so
that they are available based on a relatively small number of imports.
The rules for what the prefix is is based on the kind of datatype as well
as configurational settings. For example, when Alembic renders SQLAlchemy
types, it will by default prefix the type name with the prefix <code class="docutils literal">sa.</code>:
<div class="highlight-python"><div class="highlight"><pre>Column(&quot;my_column&quot;, sa.Integer())
</pre></div>
</div>
The use of the <code class="docutils literal">sa.</code> prefix is controllable by altering the value
of <a class="reference internal" href="api/runtime.html#alembic.runtime.environment.EnvironmentContext.configure.params.sqlalchemy_module_prefix" title="alembic.runtime.environment.EnvironmentContext.configure"><code class="xref py py-paramref docutils literal">EnvironmentContext.configure.sqlalchemy_module_prefix</code></a>:
<div class="highlight-python"><div class="highlight"><pre>def run_migrations_online():
 # ...

context.configure(
 connection=connection,
 target_metadata=target_metadata,
 sqlalchemy_module_prefix=&quot;sqla.&quot;,
 # ...
 )

# ...
</pre></div>
</div>
In either case, the <code class="docutils literal">sa.</code> prefix, or whatever prefix is desired, should
also be included in the imports section of <code class="docutils literal">script.py.mako</code>; it also
defaults to <code class="docutils literal">import sqlalchemy as sa</code>.
For user-defined types, that is, any custom type that
is not within the <code class="docutils literal">sqlalchemy.</code> module namespace, by default Alembic will
use the value of __module__ for the custom type:
<div class="highlight-python"><div class="highlight"><pre>Column(&quot;my_column&quot;, myapp.models.utils.types.MyCustomType())
</pre></div>
</div>
The imports for the above type again must be made present within the migration,
either manually, or by adding it to <code class="docutils literal">script.py.mako</code>.
<div class="versionchanged">
Changed in version 0.7.0: The default module prefix rendering for a user-defined type now makes use
of the type&#8217;s <code class="docutils literal">__module__</code> attribute to retrieve the prefix, rather than
using the value of
<a class="reference internal" href="api/runtime.html#alembic.runtime.environment.EnvironmentContext.configure.params.sqlalchemy_module_prefix" title="alembic.runtime.environment.EnvironmentContext.configure"><code class="xref py py-paramref docutils literal">sqlalchemy_module_prefix</code></a>.
</div>
The above custom type has a long and cumbersome name based on the use
of <code class="docutils literal">__module__</code> directly, which also implies that lots of imports would
be needed in order to accomodate lots of types. For this reason, it is
recommended that user-defined types used in migration scripts be made
available from a single module. Suppose we call it <code class="docutils literal">myapp.migration_types</code>:
<div class="highlight-python"><div class="highlight"><pre># myapp/migration_types.py

from myapp.models.utils.types import MyCustomType
</pre></div>
</div>
We can first add an import for <code class="docutils literal">migration_types</code> to our <code class="docutils literal">script.py.mako</code>:
<div class="highlight-python"><div class="highlight"><pre>from alembic import op
import sqlalchemy as sa
import myapp.migration_types
${imports if imports else &quot;&quot;}
</pre></div>
</div>
We then override Alembic&#8217;s use of <code class="docutils literal">__module__</code> by providing a fixed
prefix, using the <a class="reference internal" href="api/runtime.html#alembic.runtime.environment.EnvironmentContext.configure.params.user_module_prefix" title="alembic.runtime.environment.EnvironmentContext.configure"><code class="xref py py-paramref docutils literal">EnvironmentContext.configure.user_module_prefix</code></a>
option:
<div class="highlight-python"><div class="highlight"><pre>def run_migrations_online():
 # ...

context.configure(
 connection=connection,
 target_metadata=target_metadata,
 user_module_prefix=&quot;myapp.migration_types.&quot;,
 # ...
 )

# ...
</pre></div>
</div>
Above, we now would get a migration like:
<div class="highlight-python"><div class="highlight"><pre>Column(&quot;my_column&quot;, myapp.migration_types.MyCustomType())
</pre></div>
</div>
Now, when we inevitably refactor our application to move <code class="docutils literal">MyCustomType</code>
somewhere else, we only need modify the <code class="docutils literal">myapp.migration_types</code> module,
instead of searching and replacing all instances within our migration scripts.
<div class="versionadded">
New in version 0.6.3: Added <a class="reference internal" href="api/runtime.html#alembic.runtime.environment.EnvironmentContext.configure.params.user_module_prefix" title="alembic.runtime.environment.EnvironmentContext.configure"><code class="xref py py-paramref docutils literal">EnvironmentContext.configure.user_module_prefix</code></a>.
</div>
</div>
<div class="section" id="affecting-the-rendering-of-types-themselves">
<h3>Affecting the Rendering of Types Themselves<a class="headerlink" href="#affecting-the-rendering-of-types-themselves" title="Permalink to this headline">¶</a></h3>
The methodology Alembic uses to generate SQLAlchemy and user-defined type constructs
as Python code is plain old <code class="docutils literal">__repr__()</code>. SQLAlchemy&#8217;s built-in types
for the most part have a <code class="docutils literal">__repr__()</code> that faithfully renders a
Python-compatible constructor call, but there are some exceptions, particularly
in those cases when a constructor accepts arguments that aren&#8217;t compatible
with <code class="docutils literal">__repr__()</code>, such as a pickling function.
When building a custom type that will be rendered into a migration script,
it is often necessary to explicitly give the type a <code class="docutils literal">__repr__()</code> that will
faithfully reproduce the constructor for that type. This, in combination
with <a class="reference internal" href="api/runtime.html#alembic.runtime.environment.EnvironmentContext.configure.params.user_module_prefix" title="alembic.runtime.environment.EnvironmentContext.configure"><code class="xref py py-paramref docutils literal">EnvironmentContext.configure.user_module_prefix</code></a>, is usually
enough. However, if additional behaviors are needed, a more comprehensive
hook is the <a class="reference internal" href="api/runtime.html#alembic.runtime.environment.EnvironmentContext.configure.params.render_item" title="alembic.runtime.environment.EnvironmentContext.configure"><code class="xref py py-paramref docutils literal">EnvironmentContext.configure.render_item</code></a> option.
This hook allows one to provide a callable function within <code class="docutils literal">env.py</code> that will fully take
over how a type is rendered, including its module prefix:
<div class="highlight-python"><div class="highlight"><pre>def render_item(type_, obj, autogen_context):
 &quot;&quot;&quot;Apply custom rendering for selected items.&quot;&quot;&quot;

# default rendering for other objects
 return False

def run_migrations_online():
 # ...

context.configure(
 connection=connection,
 target_metadata=target_metadata,
 render_item=render_item,
 # ...
 )

# ...
</pre></div>
</div>
In the above example, we&#8217;d ensure our <code class="docutils literal">MySpecialType</code> includes an appropriate
<code class="docutils literal">__repr__()</code> method, which is invoked when we call it against <code class="docutils literal">&quot;%r&quot;</code>.
The callable we use for <a class="reference internal" href="api/runtime.html#alembic.runtime.environment.EnvironmentContext.configure.params.render_item" title="alembic.runtime.environment.EnvironmentContext.configure"><code class="xref py py-paramref docutils literal">EnvironmentContext.configure.render_item</code></a>
can also add imports to our migration script. The <a class="reference internal" href="api/autogenerate.html#alembic.autogenerate.api.AutogenContext" title="alembic.autogenerate.api.AutogenContext"><code class="xref py py-class docutils literal">AutogenContext</code></a> passed in
contains a datamember called <a class="reference internal" href="api/autogenerate.html#alembic.autogenerate.api.AutogenContext.imports" title="alembic.autogenerate.api.AutogenContext.imports"><code class="xref py py-attr docutils literal">AutogenContext.imports</code></a>, which is a Python
<code class="docutils literal">set()</code> for which we can add new imports. For example, if <code class="docutils literal">MySpecialType</code>
were in a module called <code class="docutils literal">mymodel.types</code>, we can add the import for it
as we encounter the type:
<div class="highlight-python"><div class="highlight"><pre>def render_item(type_, obj, autogen_context):
 &quot;&quot;&quot;Apply custom rendering for selected items.&quot;&quot;&quot;

if type_ == &#39;type&#39; and isinstance(obj, MySpecialType):
 # add import for this type
 autogen_context.imports.add(&quot;from mymodel import types&quot;)
 return &quot;types.%r&quot; % obj

# default rendering for other objects
 return False
</pre></div>
</div>
<div class="versionchanged">
Changed in version 0.8: The <code class="docutils literal">autogen_context</code> data member passed to
the <code class="docutils literal">render_item</code> callable is now an instance of <a class="reference internal" href="api/autogenerate.html#alembic.autogenerate.api.AutogenContext" title="alembic.autogenerate.api.AutogenContext"><code class="xref py py-class docutils literal">AutogenContext</code></a>.
</div>
<div class="versionchanged">
Changed in version 0.8.3: The &#8220;imports&#8221; data member of the autogen context
is restored to the new <a class="reference internal" href="api/autogenerate.html#alembic.autogenerate.api.AutogenContext" title="alembic.autogenerate.api.AutogenContext"><code class="xref py py-class docutils literal">AutogenContext</code></a> object as
<a class="reference internal" href="api/autogenerate.html#alembic.autogenerate.api.AutogenContext.imports" title="alembic.autogenerate.api.AutogenContext.imports"><code class="xref py py-attr docutils literal">AutogenContext.imports</code></a>.
</div>
The finished migration script will include our imports where the
<code class="docutils literal">${imports}</code> expression is used, producing output such as:
<div class="highlight-python"><div class="highlight"><pre>from alembic import op
import sqlalchemy as sa
from mymodel import types

def upgrade():
 op.add_column(&#39;sometable&#39;, Column(&#39;mycolumn&#39;, types.MySpecialType()))
</pre></div>
</div>
</div>
<div class="section" id="comparing-types">
<h3>Comparing Types<a class="headerlink" href="#comparing-types" title="Permalink to this headline">¶</a></h3>
The default type comparison logic will work for SQLAlchemy built in types as
well as basic user defined types. This logic is only enabled if the
<a class="reference internal" href="api/runtime.html#alembic.runtime.environment.EnvironmentContext.configure.params.compare_type" title="alembic.runtime.environment.EnvironmentContext.configure"><code class="xref py py-paramref docutils literal">EnvironmentContext.configure.compare_type</code></a> parameter
is set to True:
<div class="highlight-python"><div class="highlight"><pre>context.configure(
 # ...
 compare_type = True
)
</pre></div>
</div>
Alternatively, the <a class="reference internal" href="api/runtime.html#alembic.runtime.environment.EnvironmentContext.configure.params.compare_type" title="alembic.runtime.environment.EnvironmentContext.configure"><code class="xref py py-paramref docutils literal">EnvironmentContext.configure.compare_type</code></a>
parameter accepts a callable function which may be used to implement custom type
comparison logic, for cases such as where special user defined types
are being used:
<div class="highlight-python"><div class="highlight"><pre>def my_compare_type(context, inspected_column,
 metadata_column, inspected_type, metadata_type):
 # return True if the types are different,
 # False if not, or None to allow the default implementation
 # to compare these types
 return None

context.configure(
 # ...
 compare_type = my_compare_type
)
</pre></div>
</div>
Above, <code class="docutils literal">inspected_column</code> is a <a class="reference external" href="http://www.sqlalchemy.org/docs/core/metadata.html#sqlalchemy.schema.Column" title="(in SQLAlchemy v1.0)"><code class="xref py py-class docutils literal">sqlalchemy.schema.Column</code></a> as
returned by
<a class="reference external" href="http://www.sqlalchemy.org/docs/core/reflection.html#sqlalchemy.engine.reflection.Inspector.reflecttable" title="(in SQLAlchemy v1.0)"><code class="xref py py-meth docutils literal">sqlalchemy.engine.reflection.Inspector.reflecttable()</code></a>, whereas
<code class="docutils literal">metadata_column</code> is a <a class="reference external" href="http://www.sqlalchemy.org/docs/core/metadata.html#sqlalchemy.schema.Column" title="(in SQLAlchemy v1.0)"><code class="xref py py-class docutils literal">sqlalchemy.schema.Column</code></a> from the
local model environment. A return value of <code class="docutils literal">None</code> indicates that default
type comparison to proceed.
Additionally, custom types that are part of imported or third party
packages which have special behaviors such as per-dialect behavior
should implement a method called <code class="docutils literal">compare_against_backend()</code>
on their SQLAlchemy type. If this method is present, it will be called
where it can also return True or False to specify the types compare as
equivalent or not; if it returns None, default type comparison logic
will proceed:
<div class="highlight-python"><div class="highlight"><pre>class MySpecialType(TypeDecorator):

# ...

def compare_against_backend(self, dialect, conn_type):
 # return True if the types are different,
 # False if not, or None to allow the default implementation
 # to compare these types
 if dialect.name == &#39;postgresql&#39;:
 return isinstance(conn_type, postgresql.UUID)
 else:
 return isinstance(conn_type, String)
</pre></div>
</div>
The order of precedence regarding the
<a class="reference internal" href="api/runtime.html#alembic.runtime.environment.EnvironmentContext.configure.params.compare_type" title="alembic.runtime.environment.EnvironmentContext.configure"><code class="xref py py-paramref docutils literal">EnvironmentContext.configure.compare_type</code></a> callable vs. the
type itself implementing <code class="docutils literal">compare_against_backend</code> is that the
<a class="reference internal" href="api/runtime.html#alembic.runtime.environment.EnvironmentContext.configure.params.compare_type" title="alembic.runtime.environment.EnvironmentContext.configure"><code class="xref py py-paramref docutils literal">EnvironmentContext.configure.compare_type</code></a> callable is favored
first; if it returns <code class="docutils literal">None</code>, then the <code class="docutils literal">compare_against_backend</code> method
will be used, if present on the metadata type. If that returns <code class="docutils literal">None</code>,
then a basic check for type equivalence is run.
<div class="versionadded">
New in version 0.7.6: - added support for the <code class="docutils literal">compare_against_backend()</code>
method.
</div>
</div>
</div>
</div>

</div>
 </div>
 </div>
 <div class="sphinxsidebar" role="navigation" aria-label="main navigation">
 <div class="sphinxsidebarwrapper">
 <h3><a href="index.html">Table Of Contents</a></h3>
 <ul>
<li><a class="reference internal" href="#">Auto Generating Migrations</a><ul>
<li><a class="reference internal" href="#what-does-autogenerate-detect-and-what-does-it-not-detect">What does Autogenerate Detect (and what does it not detect?)</a></li>
<li><a class="reference internal" href="#comparing-and-rendering-types">Comparing and Rendering Types</a><ul>
<li><a class="reference internal" href="#controlling-the-module-prefix">Controlling the Module Prefix</a></li>
<li><a class="reference internal" href="#affecting-the-rendering-of-types-themselves">Affecting the Rendering of Types Themselves</a></li>
<li><a class="reference internal" href="#comparing-types">Comparing Types</a></li>
</ul>
</li>
</ul>
</li>
</ul>

<h4>Previous topic</h4>
 <a href="tutorial.html"
 title="previous chapter">Tutorial</a>
 <h4>Next topic</h4>
 <a href="offline.html"
 title="next chapter">Generating SQL Scripts (a.k.a. &#8220;Offline Mode&#8221;)</a>
 <div role="note" aria-label="source link">
 <h3>This Page</h3>
 <ul class="this-page-menu">
 <li><a href="_sources/autogenerate.txt"
 rel="nofollow">Show Source</a></li>
 </ul>
 </div>
<div id="searchbox" style="display: none" role="search">
 <h3>Quick search</h3>
 <form class="search" action="search.html" method="get">
 <input type="text" name="q" />
 <input type="submit" value="Go" />
 <input type="hidden" name="check_keywords" value="yes" />
 <input type="hidden" name="area" value="default" />
 </form>
 
 Enter search terms or a module, class or function name.
 
</div>
<script type="text/javascript">$('#searchbox').show(0);</script>
 </div>
 </div>
 <div class="clearer"></div>
 </div>
 <div class="related" role="navigation" aria-label="related navigation">
 <h3>Navigation</h3>
 <ul>
 <li class="right" style="margin-right: 10px">
 <a href="genindex.html" title="General Index"
 >index</a></li>
 <li class="right" >
 <a href="py-modindex.html" title="Python Module Index"
 >modules</a> |</li>
 <li class="right" >
 <a href="offline.html" title="Generating SQL Scripts (a.k.a. “Offline Mode”)"
 >next</a> |</li>
 <li class="right" >
 <a href="tutorial.html" title="Tutorial"
 >previous</a> |</li>
 <li class="nav-item nav-item-0"><a href="index.html">Alembic 0.8.3 documentation</a> &raquo;</li> 
 </ul>
 </div>
 <div class="footer" role="contentinfo">
 &copy; Copyright 2010-2015, Mike Bayer.
 Created using <a href="http://sphinx-doc.org/">Sphinx</a> 1.3.1.
 </div>
 </body>
</html>