refactor: Make the project importable as a module

README.md

@@ -6,12 +6,12 @@ You can use the scripts in this repository to analyse another git repository for
 
 Using python 3.7 or greater:
 
-    pip install -r requirements.txt
+    pip install .
 
 There are two scripts for calculating metrics (which may be combined into one soon). To find out how to use them:
 
-    python calculate_four_metrics.py --help
-    python git-metrics.py --help
+    calculate-four-metrics --help
+    git-metrics --help
 
 ## Throughput metrics: Deployment Lead Time and Deploy Interval
 
@@ -37,17 +37,17 @@ Once you have your repository annotated with tags for each deployment, you can u
 
 Use a command like this to produce the raw data for this graph and store it in a csv file:
 
-    python git_metrics.py release-lead-time [--tag-pattern=<fn_match>] [--earliest-date=<timestamp>] <path_to_git_repo> > repo_data.csv
-        
+    git-metrics release-lead-time [--tag-pattern=<fn_match>] [--earliest-date=<timestamp>] <path_to_git_repo> > repo_data.csv
+
 
 You can then plot this data with this command:
 
-    python git_metrics.py plot --release-lead-time <csv_file>
+    git-metrics plot --release-lead-time <csv_file>
 
 You can also print out average values for both release lead time and release interval like this:
 
-    calculate_four_metrics.py lead-time [--deploy-tag-pattern=<fn_match>] [--start-date=<timestamp>] <path_to_git_repo>
-    calculate_four_metrics.py deploy-interval [--deploy-tag-pattern=<fn_match>] [--start-date=<timestamp>] <path_to_git_repo>
+    calculate-four-metrics lead-time [--deploy-tag-pattern=<fn_match>] [--start-date=<timestamp>] <path_to_git_repo>
+    calculate-four-metrics deploy-interval [--deploy-tag-pattern=<fn_match>] [--start-date=<timestamp>] <path_to_git_repo>
 
 For more information about how to install these scripts, see below under 'Installation'
 
@@ -66,8 +66,8 @@ If you want to add a patch tag to a commit that has been deployed, and already h
 
 You can also print out average values for both mean time to recover and change failure rate like this:
 
-    calculate_four_metrics.py recovery-time [--deploy-tag-pattern=<fn_match>] [--patch-tag-pattern=<fn_match>] [--start-date=<timestamp>] <path_to_git_repo>
-    calculate_four_metrics.py change-fail-rate [--deploy-tag-pattern=<fn_match>] [--patch-tag-pattern=<fn_match>] [--start-date=<timestamp>] <path_to_git_repo>
+    calculate-four-metrics recovery-time [--deploy-tag-pattern=<fn_match>] [--patch-tag-pattern=<fn_match>] [--start-date=<timestamp>] <path_to_git_repo>
+    calculate-four-metrics change-fail-rate [--deploy-tag-pattern=<fn_match>] [--patch-tag-pattern=<fn_match>] [--start-date=<timestamp>] <path_to_git_repo>
 
 ## Open branches
 
@@ -77,71 +77,52 @@ Open branches gathers information about the age of all commits in a branch that
 
 Use these script commands to produce this kind of graph:
 
-    git_metrics.py open-branches [--master-branch=<branch>] <path_to_git_repo> > my_repo.csv
-    git_metrics.py plot --open-branches my_repo.csv
+    git-metrics open-branches [--master-branch=<branch>] <path_to_git_repo> > my_repo.csv
+    git-metrics plot --open-branches my_repo.csv
 
 
 ## Installation
 
 **Requirement:** Python 3.7 or later
 
 To install dependencies:
-    `pip install -r requirements.txt`
+    `python pip install .`
 
 for more help run:
-    `python git_metrics.py --help`
-
-### Python 2 systems
-
-Those OS with Python 2 as default, e.g. Ubuntu 17/16, you need to use pip and python explicitly if you didn't change the default:
-
-Call the script with:
-
-    python3 git_metrics.py --help
-
-Install pip3 if not already installed and configure dependencies with pip3, e.g. Ubuntu 16/17:
-
-    sudo apt-get install python3-pip
-    pip3 install -r requirements.txt
-
-On Ubuntu you might also miss the `python3-tk` package if doing plots
-
-    sudo apt install python3-tk
+    `git-metrics --help`
 
 ## Usage
 
 Call script with
 
-    python3 git_metrics.py --help
+    git-metrics --help
 
 which will show basic usage information like:
 
     Calculate age of commits in open remote branches
 
     Usage:
-        git_metrics.py open-branches [--master-branch=<branch>] <path_to_git_repo>
-        git_metrics.py open-branches [--master-branch=<branch>] --plot <path_to_git_repo>
-        git_metrics.py release-lead-time [--tag-pattern=<fn_match>] [--earliest-date=<timestamp>] <path_to_git_repo>
-        git_metrics.py release-lead-time --plot [--tag-pattern=<fn_match>] <path_to_git_repo>
-        git_metrics.py plot --open-branches <csv_file>
-        git_metrics.py plot --release-lead-time <csv_file>
-        git_metrics.py batch --open-branches <path_to_git_repos>...
-        git_metrics.py batch --release-lead-time [--earliest-date=<timestamp>] <path_to_git_repos>...
-        git_metrics.py (-h | --help)
+        git-metrics open-branches [--master-branch=<branch>] <path_to_git_repo>
+        git-metrics open-branches [--master-branch=<branch>] --plot <path_to_git_repo>
+        git-metrics release-lead-time [--tag-pattern=<fn_match>] [--earliest-date=<timestamp>] <path_to_git_repo>
+        git-metrics release-lead-time --plot [--tag-pattern=<fn_match>] <path_to_git_repo>
+        git-metrics plot --open-branches <csv_file>
+        git-metrics plot --release-lead-time <csv_file>
+        git-metrics batch --open-branches <path_to_git_repos>...
+        git-metrics batch --release-lead-time [--earliest-date=<timestamp>] <path_to_git_repos>...
+        git-metrics (-h | --help)
 
         Options:
             --master-branch=<branch>    example: origin/gh-pages
 
 
 * **`--plot`** parameter will open a GnuPlot plot, and will not will not save your data.
-* To _save data_ use without plot and pipe to file for csv format: `git_metrics.py release-lead-time <path_to_git-repo> > my-csv-file.csv`
-* Use plot command to plot existing csv files, e.g. `git_metrics.py plot --release-lead-time my-csv-file.csv`
+* To _save data_ use without plot and pipe to file for csv format: `git-metrics release-lead-time <path_to_git-repo> > my-csv-file.csv`
+* Use plot command to plot existing csv files, e.g. `git-metrics plot --release-lead-time my-csv-file.csv`
 * `batch` (undocumented)
 
 ## Developer information
 
 Run the self-tests using pytest:
 
     pytest -v
-
-

calculate_four_metrics.py → git_metrics/calculate_four_metrics.py

@@ -5,12 +5,12 @@
     * Mean time to recover
 
 Usage:
-    calculate_four_metrics.py lead-time [--deploy-tag-pattern=<fn_match>] [--start-date=<timestamp>] <path_to_git_repo>
-    calculate_four_metrics.py deploy-interval [--deploy-tag-pattern=<fn_match>] [--start-date=<timestamp>] <path_to_git_repo>
-    calculate_four_metrics.py change-fail-rate [--deploy-tag-pattern=<fn_match>] [--patch-tag-pattern=<fn_match>] [--start-date=<timestamp>] <path_to_git_repo>
-    calculate_four_metrics.py recovery-time [--deploy-tag-pattern=<fn_match>] [--patch-tag-pattern=<fn_match>] [--start-date=<timestamp>] <path_to_git_repo>
-    calculate_four_metrics.py metrics-all [--deploy-tag-pattern=<fn_match>] [--patch-tag-pattern=<fn_match>] [--start-date=<timestamp>] <path_to_git_repo>
-    calculate_four_metrics.py (-h | --help)
+    calculate-four-metrics lead-time [--deploy-tag-pattern=<fn_match>] [--start-date=<timestamp>] <path_to_git_repo>
+    calculate-four-metrics deploy-interval [--deploy-tag-pattern=<fn_match>] [--start-date=<timestamp>] <path_to_git_repo>
+    calculate-four-metrics change-fail-rate [--deploy-tag-pattern=<fn_match>] [--patch-tag-pattern=<fn_match>] [--start-date=<timestamp>] <path_to_git_repo>
+    calculate-four-metrics recovery-time [--deploy-tag-pattern=<fn_match>] [--patch-tag-pattern=<fn_match>] [--start-date=<timestamp>] <path_to_git_repo>
+    calculate-four-metrics metrics-all [--deploy-tag-pattern=<fn_match>] [--patch-tag-pattern=<fn_match>] [--start-date=<timestamp>] <path_to_git_repo>
+    calculate-four-metrics (-h | --help)
 
 """
 import csv
@@ -24,10 +24,10 @@
 
 import docopt
 
-from git_metrics_release_lead_time import commit_author_time_tag_author_time_and_from_to_tag_name, \
+from .git_metrics_release_lead_time import commit_author_time_tag_author_time_and_from_to_tag_name, \
     fetch_tags_and_author_dates, fetch_tags_and_sha
-from process import mk_run
-from recovery_time import Deployment, find_is_patch, find_outages
+from .process import mk_run
+from .recovery_time import Deployment, find_is_patch, find_outages
 
 
 def configure_logging():
@@ -143,7 +143,7 @@ def calculate_lead_time(path_to_git_repo, pattern, start_date):
     log.info("calculating lead time data from deployments %s", deployment_tag_pairs)
     lead_time_data = [(tat - cat) for cat, tat, old_tag, tag in deployment_data]
     return statistics.mean(lead_time_data) if lead_time_data else "N/A"
-    
+
 
 def write_four_metrics_csv_file(data):
     writer = csv.writer(sys.stdout, delimiter=',', lineterminator='\n')

git_metrics.py → git_metrics/git_metrics.py

@@ -1,15 +1,15 @@
 """Calculate age of commits in open remote branches
 
 Usage:
-    git_metrics.py open-branches [--master-branch=<branch>] <path_to_git_repo>
-    git_metrics.py open-branches [--master-branch=<branch>] --plot <path_to_git_repo>
-    git_metrics.py release-lead-time [--tag-pattern=<fn_match>] [--earliest-date=<timestamp>] <path_to_git_repo>
-    git_metrics.py release-lead-time --plot [--tag-pattern=<fn_match>] <path_to_git_repo>
-    git_metrics.py plot --open-branches <csv_file>
-    git_metrics.py plot --release-lead-time <csv_file>
-    git_metrics.py batch --open-branches <path_to_git_repos>...
-    git_metrics.py batch --release-lead-time [--earliest-date=<timestamp>] <path_to_git_repos>...
-    git_metrics.py (-h | --help)
+    git-metrics open-branches [--master-branch=<branch>] <path_to_git_repo>
+    git-metrics open-branches [--master-branch=<branch>] --plot <path_to_git_repo>
+    git-metrics release-lead-time [--tag-pattern=<fn_match>] [--earliest-date=<timestamp>] <path_to_git_repo>
+    git-metrics release-lead-time --plot [--tag-pattern=<fn_match>] <path_to_git_repo>
+    git-metrics plot --open-branches <csv_file>
+    git-metrics plot --release-lead-time <csv_file>
+    git-metrics batch --open-branches <path_to_git_repos>...
+    git-metrics batch --release-lead-time [--earliest-date=<timestamp>] <path_to_git_repos>...
+    git-metrics (-h | --help)
 
     Options:
         --master-branch=<branch>    example: origin/gh-pages
@@ -23,12 +23,12 @@
 import docopt
 import sys
 
-from git_metrics_open_branches import plot_open_branches_metrics
-from git_metrics_open_branches import get_branches
-from git_metrics_open_branches import commit_author_time_and_branch_ref
-from git_metrics_release_lead_time import commit_author_time_tag_author_time_and_from_to_tag_name
-from git_metrics_release_lead_time import plot_release_lead_time_metrics
-from process import mk_run
+from .git_metrics_open_branches import plot_open_branches_metrics
+from .git_metrics_open_branches import get_branches
+from .git_metrics_open_branches import commit_author_time_and_branch_ref
+from .git_metrics_release_lead_time import commit_author_time_tag_author_time_and_from_to_tag_name
+from .git_metrics_release_lead_time import plot_release_lead_time_metrics
+from .process import mk_run
 
 
 def read_open_branches_csv_file(filename):
@@ -136,4 +136,3 @@ def assert_master_branch(run, master_branch):
 
 if __name__ == "__main__":
     main()
-

git_metrics_open_branches.py → git_metrics/git_metrics_open_branches.py

@@ -1,5 +1,5 @@
-from data import columns
-from custom_git import for_each_ref, log
+from .data import columns
+from .custom_git import for_each_ref, log
 
 
 def plot_open_branches_metrics(data):

git_metrics_release_lead_time.py → git_metrics/git_metrics_release_lead_time.py

@@ -3,10 +3,10 @@
 
 import matplotlib
 
-from data import columns, zip_with_tail
-from custom_git import for_each_ref, show
-from custom_git import cherry
-from process import proc_to_stdout
+from .data import columns, zip_with_tail
+from .custom_git import for_each_ref, show
+from .custom_git import cherry
+from .process import proc_to_stdout
 
 TAGS_WITH_AUTHOR_DATE_CMD = for_each_ref(
     'refs/tags/**',

test_data.py → git_metrics/test_data.py

@@ -1,5 +1,5 @@
-from data import columns
-from data import zip_with_tail
+from .data import columns
+from .data import zip_with_tail
 
 
 def test_single_column():

test_git.py → git_metrics/test_git.py

@@ -1,5 +1,5 @@
-from custom_git import for_each_ref, cherry, show
-from custom_git import log
+from .custom_git import for_each_ref, cherry, show
+from .custom_git import log
 
 
 def test_for_each_ref_normal():

test_integration.py → git_metrics/test_integration.py

@@ -5,7 +5,7 @@
 import pytest
 from git import Repo
 
-from calculate_four_metrics import calculate_lead_time, calculate_deploy_interval, calculate_change_fail_rate, \
+from .calculate_four_metrics import calculate_lead_time, calculate_deploy_interval, calculate_change_fail_rate, \
     calculate_MTTR
 
 
@@ -52,7 +52,7 @@ def _run_git_cmd(git_repo, command, environment):
                     "GIT_AUTHOR_NAME": "Integration Test",
                     "EMAIL": "[email protected]"}
     env_variables.update(environment)
-    subprocess.run(command, cwd=git_repo.working_dir, 
+    subprocess.run(command, cwd=git_repo.working_dir,
         env=env_variables)
 
 
@@ -111,4 +111,3 @@ def test_calculate_deploy_interval_no_matching_tags(git_repo_DDDP):
 def test_lead_time_multiple_deploys(git_repo_DDDP):
     mean_lead_time = calculate_lead_time(git_repo_DDDP.working_dir, "FOO*", 1548321540)
     assert mean_lead_time == "N/A"
-

test_process.py → git_metrics/test_process.py

@@ -1,7 +1,7 @@
 import os
 from subprocess import Popen, PIPE
 
-from process import proc_to_stdout
+from .process import proc_to_stdout
 
 
 def test_proc_to_pocess():

test_recovery_time.py → git_metrics/test_recovery_time.py

@@ -1,4 +1,4 @@
-from recovery_time import find_outages, split_sequence, Deployment, find_is_patch
+from .recovery_time import find_outages, split_sequence, Deployment, find_is_patch
 
 deployment_zero = Deployment(False, 0)
 deployment_two = Deployment(False, 2)

test_release_lead_time.py → git_metrics/test_release_lead_time.py

@@ -1,10 +1,10 @@
 from collections import namedtuple
 from contextlib import contextmanager
 
-from git_metrics_release_lead_time import parse_tags_with_date
-from git_metrics_release_lead_time import tags_with_author_date
-from git_metrics_release_lead_time import fetch_tags_and_sha
-import git_metrics_release_lead_time
+from .git_metrics_release_lead_time import parse_tags_with_date
+from .git_metrics_release_lead_time import tags_with_author_date
+from .git_metrics_release_lead_time import fetch_tags_and_sha
+from . import git_metrics_release_lead_time
 
 
 def test_parse_tags_with_author_date():

pyproject.toml

@@ -0,0 +1,3 @@
+[build-system]
+requires = ["setuptools>=61.0"]
+build-backend = "setuptools.build_meta"

setup.py

@@ -0,0 +1,21 @@
+from setuptools import setup, find_packages
+
+setup(
+    name='git_metrics',
+    version='0.1.0',
+    packages=find_packages(),
+    install_requires=[
+        "docopt >= 0.6.2",
+        "GitPython >= 3.1.14",
+        "matplotlib >= 3.3.4",
+        "pandas >= 1.2.3",
+        "pytest >= 6.2.2",
+        "requests >= 2.25.1",
+    ],
+    entry_points={
+        'console_scripts': [
+            'calculate-four-metrics = git_metrics.calculate_four_metrics:main',
+            'git-metrics = git_metrics.git_metrics:main',
+        ],
+    },
+)