First time flow (#439)

* clone sample repo, better output after dbt init

* add default target to initial profiles file

* change starter project to fishtown repo

* pep8

* add `dbt debug --config-dir` cmd

* pep8

* sample file tweaks

* add overwrite check + unit test

dbt/clients/git.py

@@ -18,8 +18,13 @@ def run_cmd(cwd, cmd):
     return proc.communicate()
 
 
-def clone(repo, cwd):
-    return run_cmd(cwd, ['git', 'clone', '--depth', '1', repo])
+def clone(repo, cwd, dirname=None):
+    clone_cmd = ['git', 'clone', '--depth', '1', repo]
+
+    if dirname is not None:
+        clone_cmd.append(dirname)
+
+    return run_cmd(cwd, clone_cmd)
 
 
 def checkout(cwd, branch=None):
@@ -39,3 +44,7 @@ def get_current_sha(cwd):
     out, err = run_cmd(cwd, ['git', 'rev-parse', 'HEAD'])
 
     return out.decode('utf-8')
+
+
+def remove_remote(cwd):
+    return run_cmd(cwd, ['git', 'remote', 'rm', 'origin'])

dbt/clients/system.py

@@ -2,6 +2,7 @@
 import fnmatch
 import os
 import os.path
+import sys
 
 
 def find_matching(root_path,
@@ -70,3 +71,28 @@ def make_directory(path):
                 pass
             else:
                 raise e
+
+
+def make_file(path, contents='', overwrite=False):
+    """
+    Make a file at `path` assuming that the directory it resides in already
+    exists. The file is saved with contents `contents`
+    """
+    if overwrite or not os.path.exists(path):
+        with open(path, 'w') as fh:
+            fh.write(contents)
+        return True
+
+    return False
+
+
+def open_dir_cmd():
+    # https://docs.python.org/2/library/sys.html#sys.platform
+    if sys.platform == 'win32':
+        return 'start'
+
+    elif sys.platform == 'darwin':
+        return 'open'
+
+    else:
+        return 'xdg-open'

dbt/main.py

@@ -291,6 +291,13 @@ def parse_args(args):
     )
 
     sub = subs.add_parser('debug', parents=[base_subparser])
+    sub.add_argument(
+        '--config-dir',
+        action='store_true',
+        help="""
+        If specified, DBT will show path information for this project
+        """
+    )
     sub.set_defaults(cls=debug_task.DebugTask, which='debug')
 
     sub = subs.add_parser('deps', parents=[base_subparser])

dbt/task/debug.py

@@ -1,15 +1,38 @@
 import pprint
+
 from dbt.logger import GLOBAL_LOGGER as logger
+import dbt.clients.system
+import dbt.project
+
+PROFILE_DIR_MESSAGE = """To view your profiles.yml file, run:
+
+{open_cmd} {profiles_dir}"""
 
 
 class DebugTask:
     def __init__(self, args, project):
         self.args = args
         self.project = project
 
-    def run(self):
+    def path_info(self):
+        open_cmd = dbt.clients.system.open_dir_cmd()
+        profiles_dir = dbt.project.default_profiles_dir
+
+        message = PROFILE_DIR_MESSAGE.format(
+            open_cmd=open_cmd,
+            profiles_dir=profiles_dir
+        )
+
+        logger.info(message)
+
+    def diag(self):
         logger.info("args: {}".format(self.args))
         logger.info("project: ")
-
-        # TODO: switch this out for a log statement
         pprint.pprint(self.project)
+
+    def run(self):
+
+        if self.args.config_dir:
+            self.path_info()
+        else:
+            self.diag()

dbt/task/init.py

@@ -1,66 +1,111 @@
-import pprint
 import os
 
-SAMPLE_CONFIG = """
-name: 'package_name'
-version: '1.0'
-
-source-paths: ["models"]   # paths with source code to compile
-analysis-paths: ["analysis"] # path with analysis files which are compiled, but
-                             # not run
-target-path: "target"      # path for compiled code
-clean-targets: ["target"]  # directories removed by the clean task
-test-paths: ["test"]       # where to store test results
-data-paths: ["data"]       # load CSVs from this directory with `dbt seed`
-
-# specify per-model configs
-#models:
-#  package_name:             # define configs for this package (called
-#                            # "package_name" above)
-#    pardot:                 # assuming pardot is listed in models/
-#      enabled: false        # disable all pardot models except where overriden
-#      pardot_emails:        # override the configs for the pardot_emails model
-#        enabled: true       # enable this specific model
-#        materialized: true  # create a table instead of a view
-
-# uncomment below and add real repositories to add dependencies to this project
-#repositories:
-#  - "git@github.com:[your-org]/[some-repo-1]"
-#  - "git@github.com:[your-org]/[some-repo-2]"
-"""
+import dbt.project
+import dbt.clients.git
+import dbt.clients.system
+
+from dbt.logger import GLOBAL_LOGGER as logger
+
+STARTER_REPO = 'https://github.com/fishtown-analytics/dbt-starter-project.git'
+DOCS_URL = 'https://dbt.readme.io/docs/configure-your-profile'
+SAMPLE_PROFILES_YML_FILE = 'https://github.com/fishtown-analytics/dbt/blob/master/sample.profiles.yml'  # noqa
+
+ON_COMPLETE_MESSAGE = """
+Your new dbt project "{project_name}" was created! If this is your first time
+using dbt, you'll need to set up your profiles.yml file -- this file will
+tell dbt how to connect to your database. You can find this file by running:
+
+  {open_cmd} {profiles_path}
 
-GIT_IGNORE = """
-target/
-dbt_modules/
+For more information on how to configure the profiles.yml file,
+please consult the dbt documentation here:
+
+  {docs_url}
+
+One more thing:
+
+Need help? Don't hesitate to reach out to us via GitHub issues or on Slack --
+There's a link to our Slack group in the GitHub Readme. Happy modeling!
 """
 
 
+STARTER_PROFILE = """
+# For more information on how to configure this file, please see:
+# {profiles_sample}
+
+default:
+  outputs:
+    dev:
+      type: redshift
+      threads: 1
+      host: 127.0.0.1
+      port: 5439
+      user: alice
+      pass: pa55word
+      dbname: warehouse
+      schema: dbt_alice
+    prod:
+      type: redshift
+      threads: 1
+      host: 127.0.0.1
+      port: 5439
+      user: alice
+      pass: pa55word
+      dbname: warehouse
+      schema: analytics
+  target: dev
+""".format(profiles_sample=SAMPLE_PROFILES_YML_FILE)
+
+
 class InitTask:
     def __init__(self, args, project=None):
         self.args = args
         self.project = project
 
-    def __write(self, path, filename, contents):
-        file_path = os.path.join(path, filename)
+    def clone_starter_repo(self, project_name):
+        dbt.clients.git.clone(STARTER_REPO, '.', project_name)
+        dbt.clients.git.remove_remote(project_name)
+
+    def create_profiles_dir(self, profiles_dir):
+        if not os.path.exists(profiles_dir):
+            dbt.clients.system.make_directory(profiles_dir)
+            return True
+        return False
+
+    def create_profiles_file(self, profiles_file):
+        if not os.path.exists(profiles_file):
+            dbt.clients.system.make_file(profiles_file, STARTER_PROFILE)
+            return True
+        return False
 
-        with open(file_path, 'w') as fh:
-            fh.write(contents)
+    def get_addendum(self, project_name, profiles_path):
+        open_cmd = dbt.clients.system.open_dir_cmd()
+
+        return ON_COMPLETE_MESSAGE.format(
+            open_cmd=open_cmd,
+            project_name=project_name,
+            profiles_path=profiles_path,
+            docs_url=DOCS_URL
+        )
 
     def run(self):
         project_dir = self.args.project_name
 
+        profiles_dir = dbt.project.default_profiles_dir
+        profiles_file = os.path.join(profiles_dir, 'profiles.yml')
+
+        self.create_profiles_dir(profiles_dir)
+        self.create_profiles_file(profiles_file)
+
+        msg = "Creating dbt configuration folder at {}"
+        logger.info(msg.format(profiles_dir))
+
         if os.path.exists(project_dir):
             raise RuntimeError("directory {} already exists!".format(
                 project_dir
             ))
 
-        os.mkdir(project_dir)
-
-        project_dir = self.args.project_name
-        self.__write(project_dir, 'dbt_project.yml', SAMPLE_CONFIG)
-        self.__write(project_dir, '.gitignore', GIT_IGNORE)
+        self.clone_starter_repo(project_dir)
 
-        dirs = ['models', 'analysis', 'tests', 'data']
-        for dir_name in dirs:
-            dir_path = os.path.join(project_dir, dir_name)
-            os.mkdir(dir_path)
+        addendum = self.get_addendum(project_dir, profiles_dir)
+        logger.info(addendum)

sample.dbt_project.yml

@@ -16,7 +16,7 @@ name: 'your_package_name'
 
 # version: Required. This indicates the current version of your package and
 # should conform to semantic versioning. The field is currently unused
-version: '0.0.1'
+version: '0.1.0'
 
 
 
@@ -40,7 +40,7 @@ target-path: "target"
 
 # test-paths: Optional. Specify which path(s) dbt should look in to find data
 # test definitions.
-test-paths: ["test"]
+test-paths: ["tests"]
 
 # data-paths: Optional. Specify which path(s) dbt should look in to find CSV
 # files. Running `dbt seed` will load these CSVs as tables in your warehouse
@@ -191,5 +191,18 @@ repositories:
 # Archival
 #
 
-# Data archival is a powerful feature intended for advanced dbt users. For more
-# information, consult: https://dbt.readme.io/reference#archives
+# dbt's archival process records snapshots of specified tables so that
+# you can analyze how these tables change over time. In the example below,
+# the public.users table is configured for archival. When the `updated_at`
+# value changes for a given user record (identified by the `id` field), dbt
+# will record a new record in `users_archived` table which reflects the
+# changed state of that row. For more information on this command, consult
+# the dbt documentation: https://dbt.readme.io/reference#archive
+archive:
+  - source_schema: public
+    target_schema: public
+    tables:
+    - source_table: users
+      target_table: users_archived
+      updated_at: updated_at
+      unique_key: id