check_quotes.py

import sys
import tokenize


class DoubleQuotesChecker:
    msg = "single quotes detected, use double quotes instead"
    def __init__(self):
        self.violations = []

    def find_violations(self, filename, tokens):
        for token_type, token, (line, col), _, _ in tokens:
            if (
                token_type == tokenize.STRING
                and not token.startswith("'''") and
                token.startswith("'")
            ):
                self.violations.append((filename, line, col))

    def check(self, files):
        for filename in files:
            with tokenize.open(filename) as fd:
                tokens = tokenize.generate_tokens(fd.readline)
                self.find_violations(filename, tokens)

    def report(self):
        for violation in self.violations:
            filename, line, col = violation
            print(f"{filename}:{line}:{col}: {self.msg}")


if __name__ == '__main__':
    files = sys.argv[1:]
    checker = DoubleQuotesChecker()
    checker.check(files)
    checker.report()

example_quote.py

def simulate_quote_warnin():
    '''
    The docstring intentionally uses single quotes.
    '''
    if isinstance(shawn, 'sheep'):
        print('Shawn the sheep!')

import_check.py

import ast
from collections import defaultdict
import sys
import tokenize


def read_file(filename):
    with tokenize.open(filename) as fd:
        return fd.read()

class BaseChecker(ast.NodeVisitor):
    def __init__(self):
        self.violations = []

    def check(self, paths):
        for filepath in paths:
            self.filename = filepath
            tree = ast.parse(read_file(filepath))
            self.visit(tree)

    def report(self):
        for violation in self.violations:
            filename, lineno, msg = violation
            print(f"{filename}:{lineno}: {msg}")

class UnusedImportChecker(BaseChecker):
    def __init__(self):
        self.import_map = defaultdict(set)
        self.name_map = defaultdict(set)

    def _add_imports(self, node):
        for import_name in node.names:
            # Store only top-level module name ("os.path" -> "os").
            # We can't easily detect when "os.path" is used.
            name = import_name.name.partition(".")[0]
            self.import_map[self.filename].add((name, node.lineno))

    def visit_Import(self, node):
        self._add_imports(node)

    def visit_ImportFrom(self, node):
        self._add_imports(node)

    def visit_Name(self, node):
        # We only add those nodes for which a value is being read from.
        if isinstance(node.ctx, ast.Load):
            self.name_map[self.filename].add(node.id)

    def report(self):
        for path, imports in self.import_map.items():
            for name, line in imports:
                if name not in self.name_map[path]:
                    print(f"{path}:{line}: unused import '{name}'")


if __name__ == '__main__':
    files = sys.argv[1:]
    checker = UnusedImportChecker()
    checker.check(files)
    checker.report()

import_example.py

import antigravity
import os.path.join
import sys
import this

tmpdir = os.path.join(sys.path[0], 'tmp')

list_check.py

import ast
from collections import defaultdict
import sys
import tokenize


def read_file(filename):
    with tokenize.open(filename) as fd:
        return fd.read()

class BaseChecker(ast.NodeVisitor):
    def __init__(self):
        self.violations = []

    def check(self, paths):
        for filepath in paths:
            self.filename = filepath
            tree = ast.parse(read_file(filepath))
            self.visit(tree)

    def report(self):
        for violation in self.violations:
            filename, lineno, msg = violation
            print(f"{filename}:{lineno}: {msg}")


class ListDefinitionChecker(BaseChecker):
    msg = "usage of 'list()' detected, use '[]' instead"

    def visit_Call(self, node):
        name = getattr(node.func, "id", None)
        if name and name == list.__name__ and not node.args:
            self.violations.append((self.filename, node.lineno, self.msg))

if __name__ == '__main__':
    files = sys.argv[1:]
    checker = ListDefinitionChecker()
    checker.check(files)
    checker.report()

list_example.py

def build_herd():
    herd = list()
    for a_sheep in sheep:
        herd.append(a_sheep)
    return Herd(herd)

loop_check.py

import ast
from collections import defaultdict
import sys
import tokenize


def read_file(filename):
    with tokenize.open(filename) as fd:
        return fd.read()


class BaseChecker(ast.NodeVisitor):
    def __init__(self):
        self.violations = []

    def check(self, paths):
        for filepath in paths:
            self.filename = filepath
            tree = ast.parse(read_file(filepath))
            self.visit(tree)

    def report(self):
        for violation in self.violations:
            filename, lineno, msg = violation
            print(f"{filename}:{lineno}: {msg}")


class TooManyForLoopChecker(BaseChecker):
    msg = "too many nested for loops"

    def visit_For(self, node, parent=True):
        if parent:
            self.current_loop_depth = 1
        else:
            self.current_loop_depth += 1

        for child in node.body:
            if type(child) == ast.For:
                self.visit_For(child, parent=False)

        if parent and self.current_loop_depth > 3:
            self.violations.append((self.filename, node.lineno, self.msg))
            self.current_loop_depth = 0


if __name__ == '__main__':
    files = sys.argv[1:]
    checker = TooManyForLoopChecker()
    checker.check(files)
    checker.report()

loop_example.py

for _ in range(10):
    for _ in range(5):
        for _ in range(3):
            for _ in range(1):
                print("Baa, Baa, black sheep")

for _ in range(4):
    for _ in range(3):
        print("Have you any wool?")

for _ in range(10):
    for _ in range(5):
        for _ in range(3):
            if True:
                for _ in range(3):
                    print("Yes, sir, yes, sir!")

Cool blog. Thanks a lot for writing it.

plus one