Skip to main content

Validation and secure evaluation of untrusted python expressions

Project description

Evalidate

Evalidate is simple python module for safe eval()'uating user-supplied (possible malicious) logical expressions in python syntax.

Purpose

Originally it's developed for filtering complex data structures e.g.

Find cheap smartphones available for sale:

category="smartphones" and price<300 and stock>0

But also, it can be used for other expressions, e.g. arithmetical, like

a+b-100

Evalidate tries to be both secure and fast (when properly used).

Install

pip3 install evalidate

Security

Built-in python features such as compile() or eval() are quite powerful to run any kind of user-supplied code, but could be insecure if used code is malicious like os.system("rm -rf /"). Evalidate works on whitelist principle, allowing code only if it consist only of safe operations (based on authors views about what is safe and what is not, your mileage may vary - but you can supply your list of safe operations)

TL;DR. Just give me safe eval!

from evalidate import safeeval, EvalException

src="a+b" # source code
# src="__import__('os').system('clear')"
c={'a': 1, 'b': 2} # context, variables which will be available for code

try:
    result = safeeval(src,c)
    print(result)
except EvalException as e:
    print("ERR:", e)

Gives output:

3

In case of dangerous code:

src="__import__('os').system('clear')"

output will be: ERR: Operation type Call is not allowed

Exceptions

Evalidate throws exceptions CompilationException, ValidationException, ExecutionException. All of them inherit from base exception class EvalException.

Configure validation

Evalidate is very flexible, depending on parameters, same code can either pass validation or raise exception.

Safenodes and addnodes

Evalidate has built-in set of python operations, which are considered 'safe' (from author point of view). Code is considered valid only if all of it's operations are in this list. You can override this list by adding argument safenodes like:

result = evalidate.safeeval(src, context, safenodes=['Expression','BinOp','Num','Add'])

this will be enough for 1+1 expression (in src argument), but not for 1-1. If you will try '1-1', it will report error: ERROR: Validation error: Operaton type Sub is not allowed

This way you can start from scratch and allow only required operations. As an alternative, you can use built-in list of allowed operations and extend it if needed, using addnodes argument.

For example, "1*1" will give error:

ERROR: Validation error: Operaton type Mult is not allowed

But it will work with addnodes:

result = evalidate.safeeval(src,c, addnodes=['Mult'])

Please note, using 'Mult' operation isn't very secure, because for strings it can lead to Out-of-memory:

src='"a"*1000000*1000000*1000000*1000000'

and will raise runtime exception: ERROR: Runtime error (OverflowError): repeated string is too long

Allowing function calls

Evalidate does not allow any function calls by default:

>>> from evalidate import safeeval, EvalException
>>> try:
...   safeeval('int(1)')
... except EvalException as e:
...   print(e)
... 
Operation type Call is not allowed

To enable int() function, need to allow 'Call' node and add this function to list of allowed function:

>>> evalidate.safeeval('int(1)', addnodes=['Call'], funcs=['int'])
1

Attempt to call other functions will fail (because it's not in funcs list):

evalidate.safeeval('1+round(2)', addnodes=['Call'], funcs=['int'])

This will throw ValidationException.

Attributes calls ("aaa".startswith("a")) could be allowed (with proper addnodes and attrs) but other indirect function calls (like: __builtins__['eval']("print(1)")) are not allowed,

Accessing attributes (attrs parameter); data as classes

If data represented as object with attributes (not as dictionary) we have to add 'Attribute' to safe nodes. Increase salary for person for 200, and additionaly 25 for each year (s)he works in company.

from evalidate import safeeval, EvalException

class Person:
    pass

p = Person()
p.salary=1000
p.age=5

data = {'p':p}
src = 'p.salary+200+p.age*25'
try:                        
    result = safeeval(src,data,addnodes=['Attribute','Mult'], attrs=['salary', 'age'])                        
    print("result", result)
except EvalException as e:
    print("ERR:",e)

Calling attributes

This code will not work:

safeeval('"abc".startswith("a")')

Because: evalidate.ValidationException: Operation type Call is not allowed

To make it working:

print(safeeval('"abc".startswith("a")', addnodes=['Call', 'Attribute'], attrs=['startswith']))

Functions

safeeval() is simplest possible replacement to eval(). It is good to evaluate something once or few times, where speed is not an issue. If you need to eval same code 2nd time, it will take same 'long' time to parse/validate code.

evalidate() is just little more complex, but returns validated safe python AST node, which can be compiled to python bytecode, and executed at full speed. (And this code is safe after evalidate)

security.test_security() checks configuration(nodes, funcs, attrs) against set of attacks.

evalidate.safeeval()

result = safeeval(expression, context={}, safenodes=None, addnodes=None, funcs=None, attrs=None)

safeeval is higher-level wrapper of evalidate(), which validates code and runs it (if validation is successful). Throws exception if compilation(parsing), validation or execution fails.

expression - python expression like salary+100 or category="smartphones" and price<300 and stock>0.

context - dictionary of variables, available for evaluated code.

safenodes, addnodes, funcs and attrs are same as in evalidate()

returns result of evaluation of expression.

evalidate.evalidate()

node = evalidate(expression, safenodes=None, addnodes=None, funcs=None, attrs=None)

evalidate() is main (and recommended to use) method, performs parsing of python expession, validates it, and returns python AST (Abstract Syntax Tree) structure, which can be later compiled and executed. Evalidate does not evaluates code, use compile() and eval() after evalidate().

>>> import evalidate
>>> node = evalidate.evalidate('1+2')
>>> code = compile(node,'<usercode>','eval')
>>> eval(code)
3
  • expression - python expression salary+100 or category="smartphones" and price<300 and stock>0.
  • safenodes - list of allowed nodes. This will override built-in list of allowed nodes. e.g. safenodes=['Expression','BinOp','Num','Add'])
  • addnodes - list of allowed nodes. This will extend built-in lsit of allowed nodes. e.g. addnodes=['Mult']
  • funcs - list of allowed function calls. You need to add 'Call' to safe nodes. e.g. funcs=['int']
  • attrs - list of allowed attributes. You need to add 'Attribute' to attrs. e.g. attrs=['salary'].

evalidate() throws CompilationException if cannot parse source code and ValidationException if it doesn't like source code (if code has unsafe operations).

Even if evalidate is successful, this doesn't guarantees that code will run well, For example, code still can have NameError (if tries to access undefined variable) or ZeroDivisionError.

evalidate uses ast.parse() and returns AST node.

Warning

It is possible to crash the Python interpreter with a sufficiently large/complex string due to stack depth limitations in Python’s AST compiler.

In my test, works well with 200 nested int(): int(int(.... int(1)...)) but not with 201. Source code is 1000+ characters. But even if evalidate will get such code, it will just raise CompilationException.

evalidate.security.test_security()

Evalidate is very flexible and it's possible to shoot yourself in foot if you will try hard. test_security() checks your configuration (addnodes/safenodes, funcs, attrs) against given list of possible attack code or against built-in list of attacks. test_security() returns True if everything is OK (all attacks raised ValidationException) or False if something passed.

This code will never print (I hope).

from evalidate.security import test_security

test_security() or print("default rules are vulnerable!")

But this will fail because nodes/funcs leads to successful validation for attack (suppose you do not want anyone to call int())

from evalidate.security import test_security

attacks = ['int(1)']

test_security(attacks, addnodes=['Call'], funcs=['int'], verbose=True)

It will print:

Testing attack code:
int(1)
Problem! Attack passed validation without exception!
Code:
int(1)

Example

Filtering by user-supplied condition

This is code of examples/products.py. Expression is validated and compiled once and executed (as byte-code, very fast) many times, so filtering is both fast and secure.

#!/usr/bin/env python3

import requests
from evalidate import evalidate, ValidationException, CompilationException
import json
import sys

data = requests.get('https://dummyjson.com/products?limit=100').json()

try:
    src = sys.argv[1]
except IndexError:
    src = 'True'

try:
    node = evalidate(src)
except (ValidationException, CompilationException) as e:
    print(e)
    sys.exit(1)


code = compile(node, '<user filter>', 'eval')

c=0
for p in data['products']:
    # print(p)
    try:
        r = eval(code, p.copy())
        if r:
            print(json.dumps(p, indent=2))
            c+=1
    except Exception as e:
        print("Runtime exception:", e)
print("# {} products matches".format(c))
# print all 100 products
./products.py

# Only cheap products, 8 matches
./products.py 'price<20'

# smartphones (5)
./products.py 'category=="smartphones"'

# good smartphones
./products.py 'category=="smartphones" and rating>4.5'

# cheap smartphones
./products.py 'category=="smartphones" and price<300'

Similar projects and benchmark

asteval

While asteval can compute much more complex code (define functions, use python math libraries) it has drawbacks:

  • asteval is much slower (evalidate can be used at speed of eval() python bytecode)
  • user can provide source code which runs very long time and consumes many resources

simpleeval Very similar project, using AST approach too and optimized to re-evaluate pre-parsed expressions. But parsed expressions are stored as more high-level ast.Expr type and this approach is ~10 times slower, while evalidate uses python native code type and evaluation itself goes at speed of python eval()

evalidate is good to run short same code against different data.

Benchmarking

We use evalidate-vs-asteval.py which is in benchmark/ directory of repository. We prepare list of 1 million of products (actually, we take just 100 products sample, but repeat it 10 000 times to get 1 million), and then filter it, finding only specific products on "untrusted user-supplied expression" (price < 20 in this case)

Products: 1000000 items
test_asteval_products(): 25.920s
test_simpleeval_products(): 1.779s
test_evalidate_products(): 0.160s

As you see, evalidate is almost 10 times faster then simpleeval and both are much faster then asteval.

Maybe my test is not perfectly optimized (I'm not expert with simpleeval/asteval), if you can suggest better filtering sample code (which produces faster result), I will include it. But benchmark code must assume expression as unknown and untrusted.

Read about eval() risks

Note: realpython article shows example with nice short method of validation source (using code.co_names), but it's vulnerable, it passes "bomb" from Ned Batchelder article (bomb has empty co_names tuple) and crash interpreter. Evalidate can block this code and similar bombs (unless you will intentionally configure evalidate to pass specific bomb code. Yes, with evalidate it is hard to shoot yourself in the foot, but it is possible if you will try hard).

More info

Want more info? Check source code of module, it's very short and simple, easy to modify

Contact

Write me: yaroslaff at gmail.com

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

evalidate-1.0.3.tar.gz (10.8 kB view hashes)

Uploaded Source

Built Distribution

evalidate-1.0.3-py3-none-any.whl (9.7 kB view hashes)

Uploaded Python 3

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page