Added DeploymentData document generation

This PS adds a DeploymentData document to sites collected by Pegleg.
This document describes the repos Pegleg collected, including their
commit SHA, tag, and whether the repo was dirty.

If the source directory is not a git repo, these values will be
None.

Change-Id: I7919b02d70c9797f689cdad85066d3953b978901

pegleg/engine/site.py

@@ -16,10 +16,12 @@ import logging
 import os
 
 import click
+import git
 import yaml
 
 from prettytable import PrettyTable
 
+from pegleg import config
 from pegleg.engine import util
 from pegleg.engine.util import files
 
@@ -48,6 +50,13 @@ def _collect_to_stdout(site_name):
             for line in _read_and_format_yaml(filename):
                 # This code is a pattern to convert \r\n to \n.
                 click.echo("\n".join(line.splitlines()))
+        res = yaml.safe_dump(_get_deployment_data_doc(),
+                             explicit_start=True,
+                             explicit_end=True,
+                             default_flow_style=False)
+        # Click isn't splitting these lines correctly, so do it manually
+        for line in res.split('\n'):
+            click.echo(line)
     except Exception as ex:
         raise click.ClickException("Error printing output: %s" % str(ex))
 
@@ -60,6 +69,8 @@ def _collect_to_file(site_name, save_location):
     files.check_file_save_location(save_location)
 
     save_files = dict()
+    curr_site_repo = files.path_leaf(config.get_site_repo())
+
     try:
         for repo_base, filename in util.definition.site_files_by_repo(
                 site_name):
@@ -69,6 +80,9 @@ def _collect_to_file(site_name, save_location):
                 save_files[repo_name] = open(save_file, "w")
             LOG.debug("Collecting file %s to file %s", filename, save_file)
             save_files[repo_name].writelines(_read_and_format_yaml(filename))
+        save_files[curr_site_repo].writelines(yaml.safe_dump(
+            _get_deployment_data_doc(), default_flow_style=False,
+            explicit_start=True, explicit_end=True))
     except Exception as ex:
         raise click.ClickException("Error saving output: %s" % str(ex))
     finally:
@@ -140,3 +154,55 @@ def show(site_name, output_stream):
                 ["", data['site_name'], data['site_type'], file])
     # Write tables to specified output_stream
     output_stream.write(site_table.get_string() + "\n")
+
+
+def _get_deployment_data_doc():
+    stanzas = {files.path_leaf(repo): _get_repo_deployment_data_stanza(repo)
+               for repo in config.all_repos()}
+    return {
+        "schema": "pegleg/DeploymentData/v1",
+        "metadata": {
+            "schema": "metadata/Document/v1",
+            "name": "deployment-version",
+        },
+        "layeringDefinition": {
+            "abstract": "false",
+            "layer": "global"
+        },
+        "storagePolicy": "cleartext",
+        "data": {
+            "documents": stanzas
+        }
+    }
+
+
+def _get_repo_deployment_data_stanza(repo_path):
+    try:
+        repo = git.Repo(repo_path)
+        commit = repo.commit()
+
+        # If we're at a particular tag, reference it
+        tag = [tag.name for tag in
+               repo.tags if tag.commit == commit]
+        if tag:
+            tag == ", ".join(tag)
+        else:
+            # Otherwise just use the branch name
+            try:
+                tag = repo.active_branch.name
+            except TypeError as e:
+                if "HEAD is a detached symbolic reference" in str(e):
+                    tag = "Detached HEAD"
+                else:
+                    raise e
+        return {
+            "commit": commit.hexsha,
+            "tag": tag,
+            "dirty": repo.is_dirty()
+        }
+    except git.InvalidGitRepositoryError:
+        return {
+            "commit": "None",
+            "tag": "None",
+            "dirty": "None"
+        }

pegleg/engine/util/files.py

@@ -394,3 +394,20 @@ def file_in_subdir(filename, _dir):
     file_path, file_name = os.path.split(
         os.path.realpath(filename))
     return _dir in file_path.split(os.path.sep)
+
+
+def path_leaf(path):
+    """
+    Return the last element in a path, UNLESS it's empty,
+    then return the second to last element (unlike os.path.split)
+
+    :param path: a path as a string
+    :return: the last non-empty element of a string
+    :rtype: str
+    """
+    split_path = [i for i in path.split(os.sep)
+                  if i]
+    if split_path:
+        return split_path[-1]
+    else:
+        return None

tests/unit/engine/test_site_collect.py

@@ -58,6 +58,7 @@ def _expected_document_names(site_name):
         _site_definition(site_name)["metadata"]["name"],
         '%s-chart' % site_name,
         '%s-passphrase' % site_name,
+        'deployment-version'
     ]
     return EXPECTED_DOCUMENT_NAMES
 
@@ -77,6 +78,7 @@ def _test_site_collect_to_file(tmpdir, site_name, collection_path):
 
         assert sorted(_expected_document_names(site_name)) == sorted(
             [x['metadata']['name'] for x in deployment_documents])
+        assert "pegleg/DeploymentData/v1" in lines
     finally:
         if os.path.exists(collection_str_path):
             shutil.rmtree(collection_str_path, ignore_errors=True)
@@ -96,8 +98,9 @@ def _test_site_collect_to_stdout(site_name):
     all_lines = [x[1][0].strip() for x in mock_echo.mock_calls]
 
     assert all_lines, "Nothing written to stdout"
+    assert any("pegleg/DeploymentData/v1" in line for line in all_lines)
     for expected in expected_names:
-        assert 'name: %s' % expected in all_lines
+        assert 'name: {}'.format(expected) in all_lines
 
 
 def test_site_collect_to_stdout(create_tmp_deployment_files):