Name

sqitchtutorial - A tutorial introduction to Sqitch change management on PostgreSQL

Synopsis

  sqitch *

Description

This tutorial explains how to create a sqitch-enabled PostgreSQL project, use a VCS for deployment planning, and work with other developers to make sure changes remain in sync and in the proper order.

We'll start by creating a new project from scratch, a fictional antisocial networking site called Flipr. All examples use Git <http://git-scm.com/> as the VCS and PostgreSQL <http://www.postgresql.org/> as the storage engine, but for the most part you can substitute other VCSes and database engines in the examples as appropriate.

If you'd like to manage an SQLite database, see sqitchtutorial-sqlite.

If you'd like to manage an Oracle database, see sqitchtutorial-oracle.

If you'd like to manage an MySQL database, see sqitchtutorial-mysql.

If you'd like to manage an Firebird database, see sqitchtutorial-firebird.

If you'd like to manage an Vertica database, see sqitchtutorial-vertica.

Starting a New Project

Usually the first thing to do when starting a new project is to create a source code repository. So let's do that with Git:

  > mkdir flipr
  > cd flipr
  > git init .
  Initialized empty Git repository in /flipr/.git/
  > touch README.md
  > git add .
  > git commit -am 'Initialize project, add README.'

If you're a Git user and want to follow along the history, the repository used in these examples is on GitHub <https://github.com/theory/sqitch-intro>.

Now that we have a repository, let's get started with Sqitch. Every Sqitch project must have a name associated with it, and, optionally, a unique URI. We recommend including the URI, as it increases the uniqueness of object identifiers internally, so let's specify one when we initialize Sqitch:

  > sqitch init flipr --uri https://github.com/theory/sqitch-intro/ --engine pg
  Created sqitch.conf
  Created sqitch.plan
  Created deploy/
  Created revert/
  Created verify/

Let's have a look at sqitch.conf:

  > cat sqitch.conf
  [core]
        engine = pg
        # plan_file = sqitch.plan
        # top_dir = .
  # [engine "pg"]
        # target = db:pg:
        # registry = sqitch
        # client = /usr/local/pgsql/bin/psql

Good, it picked up on the fact that we're creating changes for the PostgreSQL engine, thanks to the "-engine pg" option, and saved it to the file. Furthermore, it wrote a commented-out "[engine "pg"]" section with all the available PostgreSQL engine-specific settings commented out and ready to be edited as appropriate.

By default, Sqitch will read sqitch.conf in the current directory for settings. But it will also read ~/.sqitch/sqitch.conf for user-specific settings. Since PostgreSQL's "psql" client is not in the path on my system, let's go ahead an tell it where to find the client on our computer:

  > sqitch config --user engine.pg.client /opt/local/pgsql/bin/psql

And let's also tell it who we are, since this data will be used in all of our projects:

  > sqitch config --user user.name 'Marge N. OXVera'
  > sqitch config --user user.email 'marge@example.com'

Have a look at ~/.sqitch/sqitch.conf and you'll see this:

  > cat ~/.sqitch/sqitch.conf
  [engine "pg"]
        client = /opt/local/pgsql/bin/psql
  [user]
        name = Marge N. OXVera
        email = marge@example.com

Which means that Sqitch should be able to find "psql" for any project, and that it will always properly identify us when planning and committing changes.

Back to the repository. Have a look at the plan file, sqitch.plan:

  > cat sqitch.plan
  %syntax-version=1.0.0
  %project=flipr
  %uri=https://github.com/theory/sqitch-intro/

Note that it has picked up on the name and URI of the app we're building. Sqitch uses this data to manage cross-project dependencies. The "%syntax-version" pragma is always set by Sqitch, so that it always knows how to parse the plan, even if the format changes in the future.

Let's commit these changes and start creating the database changes.

  > git add .
  > git commit -am 'Initialize Sqitch configuration.'
  [master 85e8d7c] Initialize Sqitch configuration.
   2 files changed, 19 insertions(+)
   create mode 100644 sqitch.conf
   create mode 100644 sqitch.plan

Our First Change

First, our project will need a schema. This creates a nice namespace for all of the objects that will be part of the flipr app. Run this command:

  > sqitch add appschema -n 'Add schema for all flipr objects.'
  Created deploy/appschema.sql
  Created revert/appschema.sql
  Created verify/appschema.sql
  Added "appschema" to sqitch.plan

The "add" command adds a database change to the plan and writes deploy, revert, and verify scripts that represent the change. Now we edit these files. The "deploy" script's job is to create the schema. So we add this to deploy/appschema.sql:

  CREATE SCHEMA flipr;

The "revert" script's job is to precisely revert the change to the deploy script, so we add this to revert/appschema.sql:

  DROP SCHEMA flipr;

Now we can try deploying this change. We tell Sqitch where to send the change via a database URI <https://github.com/theory/uri-db/>:

  > createdb flipr_test
  > sqitch deploy db:pg:flipr_test
  Adding registry tables to db:pg:flipr_test
  Deploying to db:pg:flipr_test
    + appschema .. ok

First Sqitch created registry tables used to track database changes. The structure and name of the registry varies between databases (PostgreSQL uses a schema to namespace its registry, while SQLite and MySQL use separate databases). Next, Sqitch deploys changes. We only have one so far; the "+" reinforces the idea that the change is being "added" to the database.

With this change deployed, if you connect to the database, you'll be able to see the schema:

  > psql -d flipr_test -c '\dn flipr'
  List of schemas
   Name  | Owner 
  -------+-------
   flipr | marge

Trust, But Verify

But that's too much work. Do you really want to do something like that after every deploy?

Here's where the "verify" script comes in. Its job is to test that the deploy did was it was supposed to. It should do so without regard to any data that might be in the database, and should throw an error if the deploy was not successful. In PostgreSQL, the simplest way to do so for non-queryable objects such as schemas is to take advantage the access privilege inquiry functions <http://www.postgresql.org/docs/current/static/functions-info.html#FUNCTIONS-INFO-ACCESS-TABLE>. These functions conveniently throw exceptions if the object being inquired does not exist. For our new schema, "has_schema_privilege()" will do very nicely. Put this query into verify/appschema.sql:

  SELECT pg_catalog.has_schema_privilege('flipr', 'usage');

Such functionality may not be available to other databases, but you can use any query that will throw an exception if the schema doesn't exist. One handy way to do that is to divide by zero if an object doesn't exist. So for other databases, assuming division by zero is fatal, you could do something like this:

  SELECT 1/COUNT(*) FROM information_schema.schemata WHERE schema_name = 'flipr';

Either way, run the "verify" script with the "verify" command:

  > sqitch verify db:pg:flipr_test
  Verifying db:pg:flipr_test
    * appschema .. ok
  Verify successful

Looks good! If you want to make sure that the verify script correctly dies if the schema doesn't exist, temporarily change the schema name in the script to something that doesn't exist, something like:

  SELECT pg_catalog.has_schema_privilege('nonesuch', 'usage');

Then "verify" again:

  > sqitch verify db:pg:flipr_test
  Verifying db:pg:flipr_test
    * appschema .. psql:verify/appschema.sql:5: ERROR:  schema "nonesuch" does not exist
  # Verify script "verify/appschema.sql" failed.
  not ok
  
  Verify Summary Report
  ---------------------
  Changes: 1
  Errors:  1
  Verify failed

It's even nice enough to tell us what the problem is. Or, for the divide-by-zero example, change the schema name:

  SELECT 1/COUNT(*) FROM information_schema.schemata WHERE schema_name = 'nonesuch';

Then the verify will look something like:

  > sqitch verify db:pg:flipr_test
  Verifying db:pg:flipr_test
    * appschema .. psql:verify/appschema.sql:5: ERROR:  division by zero
  # Verify script "verify/appschema.sql" failed.
  not ok

  Verify Summary Report
  ---------------------
  Changes: 1
  Errors:  1
  Verify failed

Less useful error output, but enough to alert us that something has gone wrong.

Don't forget to change the schema name back before continuing!

Status, Revert, Log, Repeat

For purely informational purposes, we can always see how a deployment was recorded via the "status" command, which reads the registry tables from the database:

  > sqitch status db:pg:flipr_test
  # On database db:pg:flipr_test
  # Project:  flipr
  # Change:   c7981df861183412b01be706889e508a63d445ca
  # Name:     appschema
  # Deployed: 2013-12-30 15:27:15 -0800
  # By:       Marge N. OXVera <marge@example.com>
  # 
  Nothing to deploy (up-to-date)

Let's make sure that we can revert the change:

  > sqitch revert db:pg:flipr_test
  Revert all changes from db:pg:flipr_test? [Yes]
    - appschema .. ok

The "revert" command first prompts to make sure that we really do want to revert. This is to prevent unnecessary accidents. You can pass the "-y" option to disable the prompt. Also, notice the "-" before the change name in the output, which reinforces that the change is being removed from the database. And now the schema should be gone:

  > psql -d flipr_test -c '\dn flipr'
  List of schemas
   Name | Owner 
  ------+-------

And the status message should reflect as much:

  > sqitch status db:pg:flipr_test
  # On database db:pg:flipr_test
  No changes deployed

Of course, since nothing is deployed, the "verify" command has nothing to verify:

  > sqitch verify db:pg:flipr_test
  Verifying db:pg:flipr_test
  No changes deployed

However, we still have a record that the change happened, visible via the "log" command:

  > sqitch log db:pg:flipr_test
  On database db:pg:flipr_test
  Revert c7981df861183412b01be706889e508a63d445ca
  Name:      appschema
  Committer: Marge N. OXVera <marge@example.com>
  Date:      2013-12-30 15:38:17 -0800

      Add schema for all flipr objects.

  Deploy c7981df861183412b01be706889e508a63d445ca
  Name:      appschema
  Committer: Marge N. OXVera <marge@example.com>
  Date:      2013-12-30 15:27:15 -0800

      Add schema for all flipr objects.

Note that the actions we took are shown in reverse chronological order, with the revert first and then the deploy.

Cool. Now let's commit it.

  > git add .
  > git commit -m 'Add flipr schema.'
  [master d812132] Add flipr schema.
   4 files changed, 22 insertions(+)
   create mode 100644 deploy/appschema.sql
   create mode 100644 revert/appschema.sql
   create mode 100644 verify/appschema.sql

And then deploy again. This time, let's use the "--verify" option, so that the "verify" script is applied when the change is deployed:

  > sqitch deploy --verify db:pg:flipr_test
  Deploying changes to db:pg:flipr_test
    + appschema .. ok

And now the schema should be back:

  > psql -d flipr_test -c '\dn flipr'
  List of schemas
   Name  | Owner 
  -------+-------
   flipr | marge

When we look at the status, the deployment will be there:

  > sqitch status db:pg:flipr_test
  # On database db:pg:flipr_test
  # Project:  flipr
  # Change:   c7981df861183412b01be706889e508a63d445ca
  # Name:     appschema
  # Deployed: 2013-12-30 15:40:53 -0800
  # By:       Marge N. OXVera <marge@example.com>
  # 
  Nothing to deploy (up-to-date)

On Target

I'm getting a little tired of always having to type "db:pg:flipr_test", aren't you? This database connection URI <https://github.com/theory/uri-db/> tells Sqitch how to connect to the deployment target, but we don't have to keep using the URI. We can name the target:

  > sqitch target add flipr_test db:pg:flipr_test

The "target" command, inspired by "git-remote" <http://git-scm.com/docs/git-remote>, allows management of one or more named deployment targets. We've just added a target named "flipr_test", which means we can use the string "flipr_test" for the target, rather than the URI. But since we're doing so much testing, we can also use the The "engine" command to tell Sqitch to deploy to the "flipr_test" target by default:

  > sqitch engine add pg flipr_test

Now we can omit the target argument altogether, unless we need to deploy to another database. Which we will, eventually, but at least our examples will be simpler from here on in, e.g.:

  > sqitch status
  # On database flipr_test
  # Project:  flipr
  # Change:   c7981df861183412b01be706889e508a63d445ca
  # Name:     appschema
  # Deployed: 2013-12-30 15:40:53 -0800
  # By:       Marge N. OXVera <marge@example.com>
  # 
  Nothing to deploy (up-to-date)

Yay, that allows things to be a little more concise. Let's also make sure that changes are verified after deploying them:

  > sqitch config --bool deploy.verify true
  > sqitch config --bool rebase.verify true

We'll see the "rebase" command a bit later. In the meantime, let's commit the new configuration and and make some more changes!

  > git commit -am 'Set default deployment target and always verify.'
  [master a6267d3] Set default deployment target and always verify.
   1 file changed, 8 insertions(+)

Deploy with Dependency

Let's add another change, this time to create a table. Our app will need users, of course, so we'll create a table for them. First, add the new change:

  > sqitch add users --requires appschema -n 'Creates table to track our users.'
  Created deploy/users.sql
  Created revert/users.sql
  Created verify/users.sql
  Added "users [appschema]" to sqitch.plan

Note that we're requiring the "appschema" change as a dependency of the new "users" change. Although that change has already been added to the plan and therefore should always be applied before the "users" change, it's a good idea to be explicit about dependencies.

Now edit the scripts. When you're done, deploy/users.sql should look like this:

  -- Deploy flipr:users to pg
  -- requires: appschema

  BEGIN;

  SET client_min_messages = 'warning';

  CREATE TABLE flipr.users (
      nickname  TEXT        PRIMARY KEY,
      password  TEXT        NOT NULL,
      timestamp TIMESTAMPTZ NOT NULL DEFAULT NOW()
  );

  COMMIT;

A few things to notice here. On the second line, the dependence on the "appschema" change has been listed. This doesn't do anything, but the default "deploy" PostgreSQL template lists it here for your reference while editing the file. Useful, right?

Notice that all of the SQL code is wrapped in a transaction. This is handy for PostgreSQL deployments, because PostgreSQL DDLs are transactional. The upshot is that if any part of this deploy script fails, the whole change fails. Such may work less-well for database engines that don't support transactional DDLs.

The table itself will be created in the "flipr" schema. This is why we need to require the "appschema" change.

Now for the verify script. The simplest way to check that the table was created and has the expected columns without touching the data? Just select from the table with a false "WHERE" clause. Add this to verify/users.sql:

  SELECT nickname, password, timestamp
    FROM flipr.users
   WHERE FALSE;

Now for the revert script: all we have to do is drop the table. Add this to revert/users.sql:

  DROP TABLE flipr.users;

Couldn't be much simpler, right? Let's deploy this bad boy:

  > sqitch deploy
  Deploying changes to flipr_test
    + users .. ok

We know, since verification is enabled, that the table must have been created. But for the purposes of visibility, let's have a quick look:

  > psql -d flipr_test -c '\d flipr.users'
                        Table "flipr.users"
    Column   |           Type           |       Modifiers        
  -----------+--------------------------+------------------------
   nickname  | text                     | not null
   password  | text                     | not null
   timestamp | timestamp with time zone | not null default now()
  Indexes:
      "users_pkey" PRIMARY KEY, btree (nickname)

We can also verify all currently deployed changes with the "verify" command:

  > sqitch verify
  Verifying flipr_test
    * appschema .. ok
    * users ...... ok
  Verify successful

Now have a look at the status:

  > sqitch status
  # On database flipr_test
  # Project:  flipr
  # Change:   77398e1dbc5fbce58b05eb67d201f15774718727
  # Name:     users
  # Deployed: 2013-12-30 15:51:09 -0800
  # By:       Marge N. OXVera <marge@example.com>
  # 
  Nothing to deploy (up-to-date)

Success! Let's make sure we can revert the change, as well:

  > sqitch revert --to @HEAD^ -y
  Reverting changes to appschema from flipr_test
    - users .. ok

Note that we've used the "--to" option to specify the change to revert to. And what do we revert to? The symbolic tag @HEAD, when passed to "revert", always refers to the last change deployed to the database. (For other commands, it refers to the last change in the plan.) Appending the caret ("^") tells Sqitch to select the change prior to the last deployed change. So we revert to "appschema", the penultimate change. The other potentially useful symbolic tag is @ROOT, which refers to the first change deployed to the database (or in the plan, depending on the command).

Back to the database. The "users" table should be gone but the "flipr" schema should still be around:

  > psql -d flipr_test -c '\d flipr.users'
  Did not find any relation named "flipr.users".

The "status" command politely informs us that we have undeployed changes:

  > sqitch status
  # On database flipr_test
  # Project:  flipr
  # Change:   c7981df861183412b01be706889e508a63d445ca
  # Name:     appschema
  # Deployed: 2013-12-30 15:40:53 -0800
  # By:       Marge N. OXVera <marge@example.com>
  # 
  Undeployed change:
    * users

As does the "verify" command:

  > sqitch verify
  Verifying flipr_test
    * appschema .. ok
  Undeployed change:
    * users
  Verify successful

Note that the verify is successful, because all currently-deployed changes are verified. The list of undeployed changes (just "users" here) reminds us about the current state.

Okay, let's commit and deploy again:

  > git add .
  > git commit -am 'Add users table.'
  [master d58ea2f] Add users table.
   4 files changed, 31 insertions(+)
   create mode 100644 deploy/users.sql
   create mode 100644 revert/users.sql
   create mode 100644 verify/users.sql
  > sqitch deploy
  Deploying changes to flipr_test
    + users .. ok

Looks good. Check the status:

  > sqitch status
  # On database flipr_test
  # Project:  flipr
  # Change:   77398e1dbc5fbce58b05eb67d201f15774718727
  # Name:     users
  # Deployed: 2013-12-30 15:57:14 -0800
  # By:       Marge N. OXVera <marge@example.com>
  # 
  Nothing to deploy (up-to-date)

Excellent. Let's do some more!

Add Two at Once

Let's add a couple more changes to add functions for managing users.

  > sqitch add insert_user --requires users --requires appschema \
    -n 'Creates a function to insert a user.'
  Created deploy/insert_user.sql
  Created revert/insert_user.sql
  Created verify/insert_user.sql
  Added "insert_user [users appschema]" to sqitch.plan

  > sqitch add change_pass --requires users --requires appschema \
    -n 'Creates a function to change a user password.'
  Created deploy/change_pass.sql
  Created revert/change_pass.sql
  Created verify/change_pass.sql
  Added "change_pass [users appschema]" to sqitch.plan

Now might be a good time to have a look at the deployment plan:

  > cat sqitch.plan
  %syntax-version=1.0.0
  %project=flipr
  %uri=https://github.com/theory/sqitch-intro/

  appschema 2013-12-30T23:19:45Z Marge N. OXVera <marge@example.com> # Add schema for all flipr objects.
  users [appschema] 2013-12-30T23:49:00Z Marge N. OXVera <marge@example.com> # Creates table to track our users.
  insert_user [users appschema] 2013-12-30T23:57:36Z Marge N. OXVera <marge@example.com> # Creates a function to insert a user.
  change_pass [users appschema] 2013-12-30T23:57:45Z Marge N. OXVera <marge@example.com> # Creates a function to change a user password.

Each change appears on a single line with the name of the change, a bracketed list of dependencies, a timestamp, the name and email address of the user who planned the change, and a note.

Let's write the code for the new changes. Here's what deploy/insert_user.sql should look like:

  -- Deploy flipr:insert_user to pg
  -- requires: users
  -- requires: appschema
  
  BEGIN;

  CREATE OR REPLACE FUNCTION flipr.insert_user(
      nickname TEXT,
      password TEXT
  ) RETURNS VOID LANGUAGE SQL SECURITY DEFINER AS $$
      INSERT INTO flipr.users VALUES($1, md5($2));
  $$;

  COMMIT;

Here's what verify/insert_user.sql might look like:

  BEGIN;
  SELECT has_function_privilege('flipr.insert_user(text, text)', 'execute');
  COMMIT;

We simply take advantage of the fact that "has_function_privilege()" throws an exception if the specified function does not exist.

And revert/insert_user.sql should look something like this:

  -- Revert flipr:insert_user from pg
  BEGIN;
  DROP FUNCTION flipr.insert_user(TEXT, TEXT);
  COMMIT;

Now for "change_pass"; deploy/change_pass.sql might look like this:

  -- Deploy flipr:change_pass to pg
  -- requires: users
  -- requires: appschema

  BEGIN;

  CREATE OR REPLACE FUNCTION flipr.change_pass(
      nick    TEXT,
      oldpass TEXT,
      newpass TEXT
  ) RETURNS BOOLEAN LANGUAGE plpgsql SECURITY DEFINER AS $$
  BEGIN
      UPDATE flipr.users
         SET password = md5($3)
       WHERE nickname = $1
         AND password = md5($2);
      RETURN FOUND;
  END;
  $$;

  COMMIT;

Use "has_function_privilege()" in verify/change_pass.sql again:

  BEGIN;
  SELECT has_function_privilege('flipr.change_pass(text, text, text)', 'execute');
  COMMIT;

And of course, its "revert" script, revert/change_pass.sql, should look something like:

  -- Revert flipr:change_pass from pg
  BEGIN;
  DROP FUNCTION flipr.change_pass(TEXT, TEXT, TEXT);
  COMMIT;

Try em out!

  > sqitch deploy
  Deploying changes to flipr_test
    + insert_user .. ok
    + change_pass .. ok

Do we have the functions? Of course we do, they were verified. Still, have a look:

  > psql -d flipr_test -c '\df flipr.*'
                                      List of functions
   Schema |    Name     | Result data type |          Argument data types          |  Type  
  --------+-------------+------------------+---------------------------------------+--------
   flipr  | change_pass | boolean          | nick text, oldpass text, newpass text | normal
   flipr  | insert_user | void             | nickname text, password text          | normal

And what's the status?

  > sqitch status
  # On database flipr_test
  # Project:  flipr
  # Change:   01a4f6964b89284525cb5877d222df8be70d1647
  # Name:     change_pass
  # Deployed: 2013-12-30 15:59:44 -0800
  # By:       Marge N. OXVera <marge@example.com>
  # 
  Nothing to deploy (up-to-date)

Looks good. Let's make sure revert works:

  > sqitch revert -y --to @HEAD^^
  Reverting changes to users from flipr_test
    - change_pass .. ok
    - insert_user .. ok
  > psql -d flipr_test -c '\df flipr.*'
                         List of functions
   Schema | Name | Result data type | Argument data types | Type 
  --------+------+------------------+---------------------+------

Note the use of "@HEAD^^" to specify that the revert be to two changes prior the last deployed change. Looks good. Let's do the commit and re-deploy dance:

  > git add .
  > git commit -m 'Add `insert_user()` and `change_pass()`.'
  [master c9b4d68] Add `insert_user()` and `change_pass()`.
   7 files changed, 65 insertions(+)
   create mode 100644 deploy/change_pass.sql
   create mode 100644 deploy/insert_user.sql
   create mode 100644 revert/change_pass.sql
   create mode 100644 revert/insert_user.sql
   create mode 100644 verify/change_pass.sql
   create mode 100644 verify/insert_user.sql
 
  > sqitch deploy
  Deploying changes to flipr_test
    + insert_user .. ok
    + change_pass .. ok

  > sqitch status
  # On database flipr_test
  # Project:  flipr
  # Change:   01a4f6964b89284525cb5877d222df8be70d1647
  # Name:     change_pass
  # Deployed: 2013-12-30 16:00:50 -0800
  # By:       Marge N. OXVera <marge@example.com>
  # 
  Nothing to deploy (up-to-date)
  
  > sqitch verify
  Verifying flipr_test
    * appschema .... ok
    * users ........ ok
    * insert_user .. ok
    * change_pass .. ok
  Verify successful

Great, we're fully up-to-date!

Ship It!

Let's do a first release of our app. Let's call it "1.0.0-dev1" Since we want to have it go out with deployments tied to the release, let's tag it:

  > sqitch tag v1.0.0-dev1 -n 'Tag v1.0.0-dev1.'
  Tagged "change_pass" with @v1.0.0-dev1
  > git commit -am 'Tag the database with v1.0.0-dev1.'
  [master 0acef3e] Tag the database with v1.0.0-dev1.
   1 file changed, 1 insertion(+)
  > git tag v1.0.0-dev1 -am 'Tag v1.0.0-dev1'

We can try deploying to make sure the tag gets picked up like so:

  > createdb flipr_dev
  > sqitch deploy db:pg:flipr_dev
  Adding registry tables to db:pg:flipr_dev
  Deploying changes to db:pg:flipr_dev
    + appschema ................. ok
    + users ..................... ok
    + insert_user ............... ok
    + change_pass @v1.0.0-dev1 .. ok

Great, all four changes were deployed and "change_pass" was tagged with "@v1.0.0-dev1". Let's have a look at the status:

  > sqitch status db:pg:flipr_dev
  # On database db:pg:flipr_dev
  # Project:  flipr
  # Change:   01a4f6964b89284525cb5877d222df8be70d1647
  # Name:     change_pass
  # Tag:      @v1.0.0-dev1
  # Deployed: 2013-12-30 16:02:19 -0800
  # By:       Marge N. OXVera <marge@example.com>
  # 
  Nothing to deploy (up-to-date)

Note the listing of the tag as part of the status message. Now let's bundle everything up for release:

  > sqitch bundle
  Bundling into bundle/
  Writing config
  Writing plan
  Writing scripts
    + appschema
    + users
    + insert_user
    + change_pass @v1.0.0-dev1

Now we can package the bundle directory and distribute it. When it gets installed somewhere, users can use Sqitch to deploy to the database. Let's try deploying it:

  > cd bundle
  > createdb flipr_prod
  > sqitch deploy db:pg:flipr_prod
  Adding registry tables to db:pg:flipr_prod
  Deploying changes to db:pg:flipr_prod
    + appschema ................. ok
    + users ..................... ok
    + insert_user ............... ok
    + change_pass @v1.0.0-dev1 .. ok

Looks much the same as before, eh? Package it up and ship it!

Flip Out

Now that we've got the basics of user management done, let's get to work on the core of our product, the "flip." Since other folks are working on other tasks in the repository, we'll work on a branch, so we can all stay out of each other's way. So let's branch:

  > git checkout -b flips
  Switched to a new branch 'flips'

Now we can add a new change to create a table for our flips.

  > sqitch add flips -r appschema -r users -n 'Adds table for storing flips.'
  Created deploy/flips.sql
  Created revert/flips.sql
  Created verify/flips.sql
  Added "flips [appschema users]" to sqitch.plan

You know the drill by now. Edit deploy/flips.sql:

  -- Deploy flipr:flips to pg
  -- requires: appschema
  -- requires: users

  BEGIN;

  SET client_min_messages = 'warning';

  CREATE TABLE flipr.flips (
      id        BIGSERIAL   PRIMARY KEY,
      nickname  TEXT        NOT NULL REFERENCES flipr.users(nickname),
      body      TEXT        NOT NULL DEFAULT '' CHECK ( length(body) <= 180 ),
      timestamp TIMESTAMPTZ NOT NULL DEFAULT clock_timestamp()
  );

  COMMIT;

Edit verify/flips.sql:

  -- Verify flipr:flips on pg

  BEGIN;

  SELECT id
       , nickname
       , body
       , timestamp
    FROM flipr.flips
   WHERE FALSE;

  COMMIT;

And edit revert/flips.sql:

  -- Revert flipr:flips from pg

  BEGIN;

  DROP TABLE flipr.flips;

  COMMIT;

And give it a whirl:

  > sqitch deploy
  Deploying changes to flipr_test
    + flips .. ok

Look good?

  > sqitch status --show-tags
  # On database flipr_test
  # Project:  flipr
  # Change:   4d164ef5986450f00a565735518b1d126f8ee69d
  # Name:     flips
  # Deployed: 2013-12-30 16:34:38 -0800
  # By:       Marge N. OXVera <marge@example.com>
  # 
  # Tag:
  #   @v1.0.0-dev1 - 2013-12-30 16:34:38 -0800 - Marge N. OXVera <marge@example.com>
  # 
  Nothing to deploy (up-to-date)

Note the use of "--show tags" to show all the deployed tags. Now make it so:

  > git add .
  [flips e8f4655] Add flips table.
  > git commit -am 'Add flips table.'
   4 files changed, 37 insertions(+)
   create mode 100644 deploy/flips.sql
   create mode 100644 revert/flips.sql
   create mode 100644 verify/flips.sql

Wash, Rinse, Repeat

Now comes the time to add functions to manage flips. I'm sure you have things nailed down now. Go ahead and add "insert_flip" and "delete_flip" changes and commit them. The "insert_flip" deploy script might look something like:

  -- Deploy flipr:insert_flip to pg
  -- requires: flips
  -- requires: appschema
  -- requires: users

  BEGIN;

  CREATE OR REPLACE FUNCTION flipr.insert_flip(
     nickname TEXT,
     body     TEXT
  ) RETURNS BIGINT LANGUAGE sql SECURITY DEFINER AS $$
      INSERT INTO flipr.flips (nickname, body)
      VALUES ($1, $2)
      RETURNING id;
  $$;

  COMMIT;

And the "delete_flip" deploy script might look something like:

  -- Deploy flipr:delete_flip to pg
  -- requires: flips
  -- requires: appschema
  -- requires: users

  BEGIN;

  CREATE OR REPLACE FUNCTION flipr.delete_flip(
     flip_id BIGINT
  ) RETURNS BOOLEAN LANGUAGE plpgsql SECURITY DEFINER AS $$
  BEGIN
      DELETE FROM flipe.flips WHERE id = flip_id;
      RETURN FOUND;
  END;
  $$;

  COMMIT;

The "verify" scripts are:

  -- Verify flipr:insert_flip on pg

  BEGIN;

  SELECT has_function_privilege('flipr.insert_flip(text, text)', 'execute');

  COMMIT;

And:

  -- Verify flipr:delete_flip on pg

  BEGIN;

  SELECT has_function_privilege('flipr.delete_flip(bigint)', 'execute');

  COMMIT;

The "revert" scripts are:

  -- Revert flipr:insert_flip from pg

  BEGIN;

  DROP FUNCTION flipr.insert_flip(TEXT, TEXT);

  COMMIT;

And:

  -- Revert flipr:delete_flip from pg

  BEGIN;

  DROP FUNCTION flipr.delete_flip(BIGINT);

  COMMIT;

Check the example git repository <https://github.com/theory/sqitch-intro> for the complete details. Test "deploy" and "revert", then commit it to the repository. The status should end up looking something like this:

  > sqitch status --show-tags
  # On database flipr_test
  # Project:  flipr
  # Change:   9a645034b35fa46df37a3725c480982628cc64ec
  # Name:     delete_flip
  # Deployed: 2013-12-30 16:37:51 -0800
  # By:       Marge N. OXVera <marge@example.com>
  # 
  # Tag:
  #   @v1.0.0-dev1 - 2013-12-30 16:34:38 -0800 - Marge N. OXVera <marge@example.com>
  # 
  Nothing to deploy (up-to-date)

Good, we've finished this feature. Time to merge back into "master".

Emergency

Let's do it:

  > git checkout master
  Switched to branch 'master'
  > git pull
  Updating 0acef3e..d4cbd7d
  Fast-forward
   deploy/delete_list.sql | 20 ++++++++++++++++++++
   deploy/insert_list.sql | 17 +++++++++++++++++
   deploy/lists.sql       | 16 ++++++++++++++++
   revert/delete_list.sql |  7 +++++++
   revert/insert_list.sql |  7 +++++++
   revert/lists.sql       |  7 +++++++
   sqitch.plan            |  4 ++++
   verify/delete_list.sql |  7 +++++++
   verify/insert_list.sql |  7 +++++++
   verify/lists.sql       |  9 +++++++++
   10 files changed, 101 insertions(+)
   create mode 100644 deploy/delete_list.sql
   create mode 100644 deploy/insert_list.sql
   create mode 100644 deploy/lists.sql
   create mode 100644 revert/delete_list.sql
   create mode 100644 revert/insert_list.sql
   create mode 100644 revert/lists.sql
   create mode 100644 verify/delete_list.sql
   create mode 100644 verify/insert_list.sql
   create mode 100644 verify/lists.sql

Hrm, that's interesting. Looks like someone made some changes to "master". They added list support. Well, let's see what happens when we merge our changes.

  > git merge --no-ff flips
  Auto-merging sqitch.plan
  CONFLICT (content): Merge conflict in sqitch.plan
  Automatic merge failed; fix conflicts and then commit the result.

Oh, a conflict in sqitch.plan. Not too surprising, since both the merged "lists" branch and our "flips" branch added changes to the plan. Let's try a different approach.

The truth is, we got lazy. Those changes when we pulled master from the origin should have raised a red flag. It's considered a bad practice not to look at what's changed in "master" before merging in a branch. What one should do is either:

Rebase the flips branch from master before merging. This "rewinds" the branch changes, pulls from "master", and then replays the changes back on top of the pulled changes.
Create a patch and apply that to master. This is the sort of thing you might have to do if you're sending changes to another user, especially if the VCS is not Git.

So let's restore things to how they were at master:

  > git reset --hard HEAD
  HEAD is now at ff60b9b Merge branch 'lists'

That throws out our botched merge. Now let's go back to our branch and rebase it on "master":

  > git checkout flips
  Switched to branch 'flips'
  > git rebase master
  First, rewinding head to replay your work on top of it...
  Applying: Add flips table.
  Using index info to reconstruct a base tree...
  M     sqitch.plan
  Falling back to patching base and 3-way merge...
  Auto-merging sqitch.plan
  CONFLICT (content): Merge conflict in sqitch.plan
  Failed to merge in the changes.
  Patch failed at 0001 Add flips table.
  The copy of the patch that failed is found in:
     .git/rebase-apply/patch

  When you have resolved this problem, run "git rebase --continue".
  If you prefer to skip this patch, run "git rebase --skip" instead.
  To check out the original branch and stop rebasing, run "git rebase --abort".

Oy, that's kind of a pain. It seems like no matter what we do, we'll need to resolve conflicts in that file. Except in Git. Fortunately for us, we can tell Git to resolve conflicts in sqitch.plan differently. Because we only ever append lines to the file, we can have it use the "union" merge driver, which, according to its docs <http://git-scm.com/docs/gitattributes#_built-in_merge_drivers>:

Run 3-way file level merge for text files, but take lines from both versions, instead of leaving conflict markers. This tends to leave the added lines in the resulting file in random order and the user should verify the result. Do not use this if you do not understand the implications.

This has the effect of appending lines from all the merging files, which is exactly what we need. So let's give it a try. First, back out the botched rebase:

  > git rebase --abort

Now add the union merge driver to .gitattributes for sqitch.plan and rebase again:

  > echo sqitch.plan merge=union > .gitattributes
  > git rebase master                            
  First, rewinding head to replay your work on top of it...
  Applying: Add flips table.
  Using index info to reconstruct a base tree...
  M     sqitch.plan
  Falling back to patching base and 3-way merge...
  Auto-merging sqitch.plan
  Applying: Add functions to insert and delete flips.
  Using index info to reconstruct a base tree...
  M     sqitch.plan
  Falling back to patching base and 3-way merge...
  Auto-merging sqitch.plan

Ah, that looks a bit better. Let's have a look at the plan:

  > cat sqitch.plan
  %syntax-version=1.0.0
  %project=flipr
  %uri=https://github.com/theory/sqitch-intro/

  %syntax-version=1.0.0
  %project=flipr
  %uri=https://github.com/theory/sqitch-intro/

  appschema 2013-12-30T23:19:45Z Marge N. OXVera <marge@example.com> # Add schema for all flipr objects.
  users [appschema] 2013-12-30T23:49:00Z Marge N. OXVera <marge@example.com> # Creates table to track our users.
  insert_user [users appschema] 2013-12-30T23:57:36Z Marge N. OXVera <marge@example.com> # Creates a function to insert a user.
  change_pass [users appschema] 2013-12-30T23:57:45Z Marge N. OXVera <marge@example.com> # Creates a function to change a user password.
  @v1.0.0-dev1 2013-12-31T00:01:22Z Marge N. OXVera <marge@example.com> # Tag v1.0.0-dev1.

  lists [appschema users] 2013-12-31T00:39:40Z Marge N. OXVera <marge@example.com> # Adds table for storing lists.
  insert_list [lists appschema users] 2013-12-31T00:41:29Z Marge N. OXVera <marge@example.com> # Creates a function to insert a list.
  delete_list [lists appschema users] 2013-12-31T00:41:37Z Marge N. OXVera <marge@example.com> # Creates a function to delete a list.
  flips [appschema users] 2013-12-31T00:32:39Z Marge N. OXVera <marge@example.com> # Adds table for storing flips.
  insert_flip [flips appschema users] 2013-12-31T00:35:59Z Marge N. OXVera <marge@example.com> # Creates a function to insert a flip.
  delete_flip [flips appschema users] 2013-12-31T00:36:34Z Marge N. OXVera <marge@example.com> # Creates a function to delete a flip.

Note that it has appended the changes from the merged "lists" branch, and then merged the changes from our "flips" branch. Test it to make sure it works as expected:

  > sqitch rebase -y
  Reverting all changes from flipr_test
    - delete_flip ............... ok
    - insert_flip ............... ok
    - flips ..................... ok
    - change_pass @v1.0.0-dev1 .. ok
    - insert_user ............... ok
    - users ..................... ok
    - appschema ................. ok
  Deploying changes to flipr_test
    + appschema ................. ok
    + users ..................... ok
    + insert_user ............... ok
    + change_pass @v1.0.0-dev1 .. ok
    + lists ..................... ok
    + insert_list ............... ok
    + delete_list ............... ok
    + flips ..................... ok
    + insert_flip ............... ok
    + delete_flip ............... ok

Note the use of "rebase", which combines a "revert" and a "deploy" into a single command. Handy, right? It correctly reverted our changes, and then deployed them all again in the proper order. So let's commit .gitattributes; seems worthwhile to keep that change:

  > git add .
  > git commit -m 'Add `.gitattributes` with union merge for `sqitch.plan`.'
  [flips f5ad242] Add `.gitattributes` with union merge for `sqitch.plan`.
   1 file changed, 1 insertion(+)
   create mode 100644 .gitattributes

Merges Mastered

And now, finally, we can merge into "master":

  > git checkout master
  Switched to branch 'master'
  > git merge --no-ff flips
  Merge made by the 'recursive' strategy.
   .gitattributes         |  1 +
   deploy/delete_flip.sql | 17 +++++++++++++++++
   deploy/flips.sql       | 16 ++++++++++++++++
   deploy/insert_flip.sql | 17 +++++++++++++++++
   revert/delete_flip.sql |  7 +++++++
   revert/flips.sql       |  7 +++++++
   revert/insert_flip.sql |  7 +++++++
   sqitch.plan            |  3 +++
   verify/delete_flip.sql |  7 +++++++
   verify/flips.sql       | 12 ++++++++++++
   verify/insert_flip.sql |  7 +++++++
   11 files changed, 101 insertions(+)
   create mode 100644 .gitattributes
   create mode 100644 deploy/delete_flip.sql
   create mode 100644 deploy/flips.sql
   create mode 100644 deploy/insert_flip.sql
   create mode 100644 revert/delete_flip.sql
   create mode 100644 revert/flips.sql
   create mode 100644 revert/insert_flip.sql
   create mode 100644 verify/delete_flip.sql
   create mode 100644 verify/flips.sql
   create mode 100644 verify/insert_flip.sql

And double-check our work:

  > cat sqitch.plan
  %syntax-version=1.0.0
  %project=flipr
  %uri=https://github.com/theory/sqitch-intro/

  appschema 2013-12-30T23:19:45Z Marge N. OXVera <marge@example.com> # Add schema for all flipr objects.
  users [appschema] 2013-12-30T23:49:00Z Marge N. OXVera <marge@example.com> # Creates table to track our users.
  insert_user [users appschema] 2013-12-30T23:57:36Z Marge N. OXVera <marge@example.com> # Creates a function to insert a user.
  change_pass [users appschema] 2013-12-30T23:57:45Z Marge N. OXVera <marge@example.com> # Creates a function to change a user password.
  @v1.0.0-dev1 2013-12-31T00:01:22Z Marge N. OXVera <marge@example.com> # Tag v1.0.0-dev1.

  lists [appschema users] 2013-12-31T00:39:40Z Marge N. OXVera <marge@example.com> # Adds table for storing lists.
  insert_list [lists appschema users] 2013-12-31T00:41:29Z Marge N. OXVera <marge@example.com> # Creates a function to insert a list.
  delete_list [lists appschema users] 2013-12-31T00:41:37Z Marge N. OXVera <marge@example.com> # Creates a function to delete a list.
  flips [appschema users] 2013-12-31T00:32:39Z Marge N. OXVera <marge@example.com> # Adds table for storing flips.
  insert_flip [flips appschema users] 2013-12-31T00:35:59Z Marge N. OXVera <marge@example.com> # Creates a function to insert a flip.
  delete_flip [flips appschema users] 2013-12-31T00:36:34Z Marge N. OXVera <marge@example.com> # Creates a function to delete a flip.

Much much better, a nice clean master now. And because it is now identical to the "flips" branch, we can just carry on. Go ahead and tag it, bundle, and release:

  > sqitch tag v1.0.0-dev2 -n 'Tag v1.0.0-dev2.'
  Tagged "delete_flip" with @v1.0.0-dev2
  > git commit -am 'Tag the database with v1.0.0-dev2.'
  [master 230603b] Tag the database with v1.0.0-dev2.
   1 file changed, 1 insertion(+)
  > git tag v1.0.0-dev2 -am 'Tag v1.0.0-dev2'
  > sqitch bundle --dest-dir flipr-1.0.0-dev2
  Bundling into flipr-1.0.0-dev2
  Writing config
  Writing plan
  Writing scripts
    + appschema
    + users
    + insert_user
    + change_pass @v1.0.0-dev1
    + lists
    + insert_list
    + delete_list
    + flips
    + insert_flip
    + delete_flip @v1.0.0-dev2

Note the use of the "--dest-dir" option to "sqitch bundle". Just a nicer way to create the top-level directory name so we don't have to rename it from bundle.

In Place Changes

Uh-oh, someone just noticed that MD5 hashing is not particularly secure. Why? Have a look at this:

  > psql -d flipr_test -c "
      SELECT flipr.insert_user('foo', 'secr3t'), flipr.insert_user('bar', 'secr3t');
      SELECT * FROM flipr.users;
  "
   nickname |             password             |           timestamp           
  ----------+----------------------------------+-------------------------------
   foo      | 9695da4dd567a19f9b92065f240c6725 | 2013-12-31 00:56:20.240481+00
   bar      | 9695da4dd567a19f9b92065f240c6725 | 2013-12-31 00:56:20.240481+00

If user "foo" ever got access to the database, she could quickly discover that user "bar" has the same password and thus be able to exploit the account. Not a great idea. So we need to modify the "insert_user()" and "change_pass()" functions to fix that. How?

We'll use "pgcrypto" <http://www.postgresql.org/docs/current/static/pgcrypto.html>'s "crypt()" function to encrypt passwords with a salt, so that they're all unique. We just add a change to add "pgcrypto" to the database, and then we can use it. The deploy script should be:

  CREATE EXTENSION pgcrypto;

And the revert script should be:

  DROP EXTENSION pgcrypto;

If you're on PostgreSQL 9.0 or lower, you won't be able to deploy "pgcrypto" with a Sqitch change, alas. You'll have to install it manually, like so:

    psql -d flipr_test -f /path/to/pgsql/share/contrib/pgcrypto.sql

Don't forget to do this with your staging and production databases, too. Or consider upgrading to PostgreSQL 9.1 or higher; the SQL-level extension support is amazingly useful.

We're going to use the "crypt()" and "gen_salt()" functions, so in the "verify" script, let's make sure that the extension exists and that both those functions exist:

  SELECT 1/count(*) FROM pg_extension WHERE extname = 'pgcrypto';
  SELECT has_function_privilege('crypt(text, text)', 'execute');
  SELECT has_function_privilege('gen_salt(text)', 'execute');

Now we can use "pgcrypto". But how to deploy the changes to "insert_user()" and "change_pass()"?

Normally, modifying functions in database changes is a PITA <http://www.urbandictionary.com/define.php?term=pita>. You have to make changes like these:

1.: Copy deploy/insert_user.sql to deploy/insert_user_crypt.sql.
2.: Edit deploy/insert_user_crypt.sql to switch from "MD5()" to "crypt()" and to add a dependency on the "pgcrypto" change.
3.: Copy deploy/insert_user.sql to revert/insert_user_crypt.sql. Yes, copy the original change script to the new revert change.
4.: Copy verify/insert_user.sql to verify/insert_user_crypt.sql.
5.: Edit verify/insert_user_crypt.sql to test that the function now properly uses "crypt()".
6.: Test the changes to make sure you can deploy and revert the "insert_user_crypt" change.
7.: Now do the same for the "change_pass" scripts.

But you can have Sqitch do it for you. The only requirement is that a tag appear between the two instances of a change we want to modify. In general, you're going to make a change like this after a release, which you've tagged anyway, right? Well we have, with "@v1.0.0-dev2" added in the previous section. With that, we can let Sqitch do most of the hard work for us, thanks to the "rework" command, which is similar to "add", including support for the "--requires" option:

  > sqitch rework insert_user --requires pgcrypto -n 'Change insert_user to use pgcrypto.'
  Added "insert_user [insert_user@v1.0.0-dev2 pgcrypto]" to sqitch.plan.
  Modify these files as appropriate:
    * deploy/insert_user.sql
    * revert/insert_user.sql
    * verify/insert_user.sql

Oh, so we can edit those files in place. Nice! How does Sqitch do it? Well, in point of fact, it has copied the files to stand in for the previous instance of the "insert_user" change, which we can see via "git status":

  > git status
  # On branch master
  # Changes not staged for commit:
  #   (use "git add <file>..." to update what will be committed)
  #   (use "git checkout -- <file>..." to discard changes in working directory)
  #
  #     modified:   revert/insert_user.sql
  #     modified:   sqitch.plan
  #
  # Untracked files:
  #   (use "git add <file>..." to include in what will be committed)
  #
  #     deploy/insert_user@v1.0.0-dev2.sql
  #     revert/insert_user@v1.0.0-dev2.sql
  #     verify/insert_user@v1.0.0-dev2.sql
  no changes added to commit (use "git add" and/or "git commit -a")

The "untracked files" part of the output is the first thing to notice. They are all named "insert_user@v1.0.0-dev2.sql". What that means is: "the "insert_user" change as it was implemented as of the "@v1.0.0-dev2" tag." These are copies of the original scripts, and thereafter Sqitch will find them when it needs to run scripts for the first instance of the "insert_user" change. As such, it's important not to change them again. But hey, if you're reworking the change, you shouldn't need to.

The other thing to notice is that revert/insert_user.sql has changed. Sqitch replaced it with the original deploy script. As of now, deploy/insert_user.sql and revert/insert_user.sql are identical. This is on the assumption that the deploy script will be changed (we're reworking it, remember?), and that the revert script should actually change things back to how they were before. Of course, the original deploy script may not be idempotent <http://en.wikipedia.org/wiki/Idempotence> -- that is, able to be applied multiple times without changing the result beyond the initial application. If it's not, you will likely need to modify it so that it properly restores things to how they were after the original deploy script was deployed. Or, more simply, it should revert changes back to how they were as-of the deployment of deploy/insert_user@v1.0.0-dev2.sql.

Fortunately, our function deploy scripts are already idempotent, thanks to the use of the "OR REPLACE" expression. No matter how many times a deployment script is run, the end result will be the same instance of the function, with no duplicates or errors.

As a result, there is no need to explicitly add changes. So go ahead. Modify the script to switch to "crypt()". Make this change to deploy/insert_user.sql:

  @@ -1,6 +1,7 @@
   -- Deploy flipr:insert_user to pg
   -- requires: users
   -- requires: appschema
  +-- requires: pgcrypto
 
   BEGIN;
 
  @@ -8,7 +9,7 @@ CREATE OR REPLACE FUNCTION flipr.insert_user(
       nickname TEXT,
       password TEXT
   ) RETURNS VOID LANGUAGE SQL SECURITY DEFINER AS $$
  -    INSERT INTO flipr.users VALUES($1, md5($2));
  +    INSERT INTO flipr.users values($1, crypt($2, gen_salt('md5')));
   $$;
 
   COMMIT;

Go ahead and rework the "change_pass" change, too:

  > sqitch rework change_pass --requires pgcrypto -n 'Change change_pass to use pgcrypto.'
  Added "change_pass [change_pass@v1.0.0-dev2 pgcrypto]" to sqitch.plan.
  Modify these files as appropriate:
    * deploy/change_pass.sql
    * revert/change_pass.sql
    * verify/change_pass.sql

And make this change to deploy/change_pass.sql:

  @@ -1,6 +1,7 @@
   -- Deploy flipr:change_pass to pg
   -- requires: users
   -- requires: appschema
  +-- requires: pgcrypto
 
   BEGIN;
 
  @@ -11,9 +12,9 @@ CREATE OR REPLACE FUNCTION flipr.change_pass(
   ) RETURNS BOOLEAN LANGUAGE plpgsql SECURITY DEFINER AS $$
   BEGIN
       UPDATE flipr.users
  -       SET password = md5($3)
  +       SET password = crypt($3, gen_salt('md5'))
        WHERE nickname = $1
  -       AND password = md5($2);
  +       AND password = crypt($2, password);
       RETURN FOUND;
   END;
   $$;

And then try a deployment:

  > sqitch deploy
  Deploying changes to flipr_test
    + insert_user .. ok
    + change_pass .. ok

So, are the changes deployed?

  > psql -d flipr_test -c "
      DELETE FROM flipr.users;
      SELECT flipr.insert_user('foo', 'secr3t'), flipr.insert_user('bar', 'secr3t');
      SELECT * FROM flipr.users;
  "
   nickname |              password              |           timestamp           
  ----------+------------------------------------+-------------------------------
   foo      | $1$pRNfJjI9$CdcEXJ9xCoJPD.R5Z/7.R1 | 2013-12-31 01:03:15.398572+00
   bar      | $1$Nf1LcU.p$B9sKzdu8vMgu5oxbimo5P1 | 2013-12-31 01:03:15.398572+00

Awesome, the stored passwords are different now. But can we revert, even though we haven't written any reversion scripts?

  > sqitch revert --to @HEAD^^ -y
  Reverting changes to pgcrypto from flipr_test
    - change_pass .. ok
    - insert_user .. ok

Did that work, are the "MD5()" passwords back?

  > psql -d flipr_test -c "
      DELETE FROM flipr.users;
      SELECT flipr.insert_user('foo', 'secr3t'), flipr.insert_user('bar', 'secr3t');
      SELECT * FROM flipr.users;
  "
   nickname |             password             |           timestamp           
  ----------+----------------------------------+-------------------------------
   foo      | 9695da4dd567a19f9b92065f240c6725 | 2013-12-31 01:03:57.263583+00
   bar      | 9695da4dd567a19f9b92065f240c6725 | 2013-12-31 01:03:57.263583+00

Yes, it works! Sqitch properly finds the original instances of these changes in the new script files that include tags.

But what about the verify script? How can we verify that the functions have been modified to use "crypt()"? I think the simplest thing to do is to examine the body of the function, using "pg_get_functiondef()" <http://www.postgresql.org/docs/9.2/static/functions-info.html#FUNCTIONS-INFO-CATALOG-TABLE>. So the "insert_user" verify script looks like this:

  -- Verify flipr:insert_user on pg

  BEGIN;

  SELECT has_function_privilege('flipr.insert_user(text, text)', 'execute');

  SELECT 1/COUNT(*)
    FROM pg_catalog.pg_proc
   WHERE proname = 'insert_user'
     AND pg_get_functiondef(oid) LIKE $$%crypt($2, gen_salt('md5'))%$$;

  COMMIT;

And the "change_pass" verify script looks like this:

  -- Verify flipr:change_pass on pg

  BEGIN;

  SELECT has_function_privilege('flipr.change_pass(text, text, text)', 'execute');

  SELECT 1/COUNT(*)
    FROM pg_catalog.pg_proc
   WHERE proname = 'change_pass'
     AND pg_get_functiondef(oid) LIKE $$%crypt($3, gen_salt('md5'))%$$;

  COMMIT;

Make sure these pass by re-deploying:

  > sqitch deploy
  Deploying changes to flipr_test
    + insert_user .. ok
    + change_pass .. ok

Excellent. Let's go ahead and commit these changes:

  > git add .
  > git commit -m 'Use pgcrypto to encrypt passwords.'
  [master 4257ae6] Use pgcrypto to encrypt passwords.
   13 files changed, 107 insertions(+), 9 deletions(-)
   create mode 100644 deploy/change_pass@v1.0.0-dev2.sql
   create mode 100644 deploy/insert_user@v1.0.0-dev2.sql
   create mode 100644 revert/change_pass@v1.0.0-dev2.sql
   create mode 100644 revert/insert_user@v1.0.0-dev2.sql
   create mode 100644 verify/change_pass@v1.0.0-dev2.sql
   create mode 100644 verify/insert_user@v1.0.0-dev2.sql

  > sqitch status
  # On database flipr_test
  # Project:  flipr
  # Change:   d3ffa30b72abaf9619ae1f0e726026667612f2b1
  # Name:     change_pass
  # Deployed: 2013-12-30 17:05:08 -0800
  # By:       Marge N. OXVera <marge@example.com>
  # 
  Nothing to deploy (up-to-date)

More to Come

Sqitch is a work in progress. Better integration with version control systems is planned to make managing idempotent reworkings even easier. Stay tuned.

Author

David E. Wheeler <david@justatheory.com>

License

Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions:

The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.