"""Tests looking for TODOs and prints."""
import json
import os
import re
try:
from urllib.request import urlopen
except ImportError:
from urllib import urlopen
import pytest
from pygments import lex
from pygments.lexers import get_lexer_by_name
@pytest.mark.skipif('"TRAVIS_REPO_SLUG" not in os.environ')
def test_todo_issue_validator():
"""Verify that each T.O.D.O is associated with an open GitHub issue."""
root_directory = os.path.abspath(os.path.join(os.path.dirname(__file__), '..'))
assert 'tests' in os.listdir(root_directory)
generator = (os.path.join(r, s) for r, d, f in os.walk(root_directory)
if '.tox' not in r
for s in f
if s.endswith('.py') and not s.startswith('example_'))
regex_todo = re.compile(r'^(.*)(?<!\w)(TODO|FIXME)(?!\w)(.*)$', re.IGNORECASE | re.MULTILINE)
# Find all potential TODOs in Python files. May or may not be in comments/docstrings.
potential_todos = set()
for file_path in generator:
with open(file_path) as f:
for line in f:
if regex_todo.search(line):
potential_todos.add(file_path)
break
if not potential_todos:
return
# Get all open issues.
repo_slug = os.environ['TRAVIS_REPO_SLUG']
assert re.match(r'^[a-zA-Z0-9_-]+/[a-zA-Z0-9_-]+$', repo_slug)
response = urlopen('https://api.github.com/repos/{0}/issues'.format(repo_slug))
raw_data = response.read().decode('utf-8')
parsed_data = json.loads(raw_data)
open_issues = set(['issues/{0:d}'.format(int(i.get('number'))) for i in parsed_data if i.get('state') == 'open'])
# Perform lexical analysis on the source code and find all docstrings and comments with TODOs.
todos_with_no_issues = dict()
for file_path in potential_todos:
with open(file_path) as f:
code = f.read(52428800) # Up to 50 MiB.
for token, code_piece in lex(code, get_lexer_by_name('Python')):
if str(token) not in ('Token.Comment', 'Token.Literal.String.Doc'):
continue
if not regex_todo.search(code_piece):
continue
code_line = ''.join(b for a in regex_todo.findall(code_piece) for b in a)
has_issue = bool([i for i in open_issues if i in code_line])
if has_issue:
continue # This t.o.d.o has an open issue, skipping.
# If this is reached, there is a t.o.d.o without an open issue!
if file_path not in todos_with_no_issues:
todos_with_no_issues[file_path] = list()
todos_with_no_issues[file_path].append(code_line)
assert not todos_with_no_issues
def test_print_hunter():
"""Verify that there are no print statements in the codebase."""
root_directory = os.path.abspath(os.path.join(os.path.dirname(__file__), '..'))
assert 'tests' in os.listdir(root_directory)
generator = (os.path.join(r, s) for r, d, f in os.walk(root_directory) if '.egg/' not in r and '/.tox/' not in r
for s in f if s.endswith('.py') and not s.startswith('example_'))
regex_print = re.compile(r'^(.*)(?<!\w)print(\(|\s)(.*)$', re.MULTILINE)
# Find all potential prints in Python files. May or may not be in strings.
potential_prints = set()
for file_path in generator:
with open(file_path) as f:
for line in f:
if regex_print.search(line):
potential_prints.add(file_path)
break
if not potential_prints:
return
# Perform lexical analysis on the source code and find all valid print statements/function calls.
current_line = list()
actual_prints = dict()
for file_path in potential_prints:
with open(file_path) as f:
code = f.read(52428800) # Up to 50 MiB.
for token, code_piece in lex(code, get_lexer_by_name('Python')):
if code_piece == '\n':
current_line = list() # References new list, doesn't necessarily remove old list.
continue
current_line.append(code_piece)
if (str(token), code_piece) != ('Token.Keyword', 'print'):
continue
# If this is reached, there is a print statement in the library!
if file_path not in actual_prints:
actual_prints[file_path] = list()
actual_prints[file_path].append(current_line) # Keeps reference to current list() alive.
actual_prints = dict((f, [''.join(l) for l in lst]) for f, lst in actual_prints.items())
assert not actual_prints
