pySCG: Numbers related landing page to explain generic issues, such as array with ctypes and numpy

Offen
  • Stimmt etwas nicht?
#13 0 Kommentare 0 Reaktionen 0 zugewiesene Personen Auf GitHub ansehen

Dieses Issue hat noch niemand übernommen.

Bewertung

Schwierigkeit
4/5
Geschätzter Aufwand
3-5 Tage
Anfängerfreundlichkeit
42/100
Issue-Typ
Dokumentation
Klarheit
Größtenteils klar
Aktivitätsstatus
Veraltet
Tech-Stack
numpy, python
Bereich
documentation, security

Rechercherichtung

Start by reviewing the repository’s existing CWE-191, CWE-1335, and CWE-197 pages and the linked material, then check how numeric issue pages are organized. Draft a shared landing page covering Python integers, array, numpy, ctypes, conversions, numeric bounds, and the bytes examples. Done when it explains these topics accurately and links to the related numeric issues.

Vom Indexierungsmodell aus dem Issue-Text verfasst.

Beschreibung

Numbers landing page to explain generic issues around:

  • Hardware limitations
  • A Python int can be much bigger than what is known as 64 bit int and why NUM13 does not apply
  • Python relation to primitive int numpy ctypes
  • implicit conversion can come unexpected
  • explicit conversion can have unexpected results.
  • Upper boundaries
  • Lower boundaries.

Content that is helpful for understanding numbers or above points and applicable for all numeric related issues such as
Avoid loss of precision when converting primitive integers to floating-point

Pythons array module is the closest thing to primitive types mentioned in the Java or C rules such as unsigned or signed int, float, long or char. The array module should be at least mentioned together with numpy and ctypes.

CWE-191: Detect or prevent integer overflow

"by classes and are protected against overflows.  Primitive types such as numpy , time  or ctype share issues known from C , or C++"

CWE-1335: Incorrect Bitwise Shift of Integer

CWE-1335: Promote readability and compatibility by using mathematical written code with arithmetic operations instead of bit-wise operations

"outcomes. Python's ctypes  module allows integration of C based system libraries into Python and direct access to C-type variables that can have"

CWE-197: Control rounding when converting to less precise numbers

"Python has ctypes module allowing users to work with many different data types. "

Additional notes from previous conversations:
A build in type byte does not exist in Python 3, try byte(0) vs bytes(0).

The bytes and bytearray classes are both designed for binary data as a sequence of integers capped to byte size range of decimal 0 to 255. The use of bytes for memory optimization is unlikely of help as it requires the extra handling and storage for the sequence part; compare sys.getsizeof(bytes(10)) vs sys.getsizeof(int(10)).

String presentations such as decimal 10 displayed by the programming IDE as a newline \n character can be confusing.

Storing anything bigger than an decimal 255 in bytes requires to spread the number across a sequence of byte sized objects.

Designing with bytes and bytearray classes requires a good understanding of example01.py code. Python objects can be stored directly via the pickle and unpickle modules. Memory optimization is better addressed by understanding the difference between mutable and nonmutable objects in Python.

example01.py:

# byte object does not exist:
byte(10)     {}# NameError: name 'byte' is not defined
 
# bytes(10) creates a sequence with 10 items
bytes(10)    {}# b'\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00'
 
# string representation can be confusing
bytes([{}}}\{{{}10{}}}\{{{}])  {}# b'\n' ;IDE printing hex 0a as newline \n
bytes([{}}}\{{{}10{}}}\{{{}]).hex()     {}# '0a'
b'10' == b'\x31\x30'  # True ; decimal string 10 to hex string \x31\x30
 
# storing byte sized decimal in bytes compared to build in int:
print(sys.getsizeof(bytes([{}}}\{{{}10{}}}\{{{}])))  {}# 34
print(sys.getsizeof(10))           {}{ {# 28 }
}
 
# different platform store number invers to each other:
int(1).to_bytes(2, {}'little')  {}# b'\x01\x00'
int(1).to_bytes(2, {}'big')     {}# b'\x00\x01'
 
# bytes comes with Overflow protection
int(256).to_bytes(1, {}'big')   {}# OverflowError: int too big to convert
Vorherrschende Sprache
Python
Sterne
2
Forks
2
Ø Merge
11 Min.
Gemergte PRs (30 T.)
2

Entwicklungsumgebung

Erste Schritte

  1. Lesen Sie das ganze Issue und danach den Beitragsleitfaden des Projekts.
  2. Schreiben Sie ins Issue, dass Sie es übernehmen — das erspart doppelte Arbeit.
  3. Forken Sie das Repository und arbeiten Sie in einem Branch.
  4. Öffnen Sie einen Pull Request, der die Issue-Nummer nennt.

Mehr aus ossf/python-secure-coding-guide

Alle Issues in ossf/python-secure-coding-guide

Ähnliche Issues

Weitere Issues zu Python

Neue Issues direkt in Ihr Postfach

Eine kurze Übersicht über anfängerfreundliche GitHub-Issues.