Pylance crashes when parsing large integers

[Faholan]

Environment data

• Language Server version: 2022.1.3
• OS and version: Windows 11 21H2
• Python version (& distribution if applicable, e.g. Anaconda): 3.10.1

Expected behaviour

Pylance should correctly parse large integers

Actual behaviour

Pylance crashes when parsing big integers, stating "RangeError: The number Infinity cannot be converted to a BigInt because it is not an integer"

Logs

[Error - 12:16:45] An internal error occurred while parsing file "c:\Users\Waylander\Documents\Coding\TIPE\tipe\src\core\crypto\config.py": RangeError: The number Infinity cannot be converted to a BigInt because it is not an integer
    at BigInt (<anonymous>)
    at t.Tokenizer._tryNumber (c:\Users\Waylander\.vscode\extensions\ms-python.vscode-pylance-2022.1.3\dist\pyright-internal\src\parser\tokenizer.ts:762:25)
    at t.Tokenizer._handleCharacter (c:\Users\Waylander\.vscode\extensions\ms-python.vscode-pylance-2022.1.3\dist\pyright-internal\src\parser\tokenizer.ts:461:30)
    at t.Tokenizer._addNextToken (c:\Users\Waylander\.vscode\extensions\ms-python.vscode-pylance-2022.1.3\dist\pyright-internal\src\parser\tokenizer.ts:323:19)
    at t.Tokenizer.tokenize (c:\Users\Waylander\.vscode\extensions\ms-python.vscode-pylance-2022.1.3\dist\pyright-internal\src\parser\tokenizer.ts:239:18)
    at T._startNewParse (c:\Users\Waylander\.vscode\extensions\ms-python.vscode-pylance-2022.1.3\dist\pyright-internal\src\parser\parser.ts:322:43)
    at callback (c:\Users\Waylander\.vscode\extensions\ms-python.vscode-pylance-2022.1.3\dist\pyright-internal\src\parser\parser.ts:216:18)
    at i.timeOperation (c:\Users\Waylander\.vscode\extensions\ms-python.vscode-pylance-2022.1.3\dist\pyright-internal\src\common\timing.ts:44:28)
    at T.parseSourceFile (c:\Users\Waylander\.vscode\extensions\ms-python.vscode-pylance-2022.1.3\dist\pyright-internal\src\parser\parser.ts:215:38)
    at c:\Users\Waylander\.vscode\extensions\ms-python.vscode-pylance-2022.1.3\dist\pyright-internal\src\analyzer\sourceFile.ts:643:45

Code Snippet / Additional information

"""Default arguments.

Copyright (C) 2022  Faholan <https://github.com/Faholan>
"""
import typing as t

DEFAULT_RSA = {
    "nlen": 3072,
    "hash_name": "sha3_512",
    "gen_method": "probable_aux_prob",
}

DEFAULT_ECDSA = {
    "curve_name": "frp256v1",
    "hash_name": "sha3_512",
}

FFDHE_FIELDS: t.Dict[str, t.Tuple[int, int, int, int]] = {
    # p: prime, g: generator, minimial secret key size, byte size
    "ffdhe2048": (
        0xFFFFFFFFFFFFFFFFADF85458A2BB4A9AAFDC5620273D3CF1D8B9C583CE2D3695A9E13641146433FBCC939DCE249B3EF97D2FE363630C75D8F681B202AEC4617AD3DF1ED5D5FD65612433F51F5F066ED0856365553DED1AF3B557135E7F57C935984F0C70E0E68B77E2A689DAF3EFE8721DF158A136ADE73530ACCA4F483A797ABC0AB182B324FB61D108A94BB2C8E3FBB96ADAB760D7F4681D4F42A3DE394DF4AE56EDE76372BB190B07A7C8EE0A6D709E02FCE1CDF7E2ECC03404CD28342F619172FE9CE98583FF8E4F1232EEF28183C3FE3B1B4C6FAD733BB5FCBC2EC22005C58EF1837D1683B2C6F34A26C1B2EFFA886B423861285C97FFFFFFFFFFFFFFFF,
        2,
        255,
        256,
    ),
    "ffdhe3072": (
        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
        2,
        275,
        384,
    ),
    "ffdhe4096": (
        0xFFFFFFFFFFFFFFFFADF85458A2BB4A9AAFDC5620273D3CF1D8B9C583CE2D3695A9E13641146433FBCC939DCE249B3EF97D2FE363630C75D8F681B202AEC4617AD3DF1ED5D5FD65612433F51F5F066ED0856365553DED1AF3B557135E7F57C935984F0C70E0E68B77E2A689DAF3EFE8721DF158A136ADE73530ACCA4F483A797ABC0AB182B324FB61D108A94BB2C8E3FBB96ADAB760D7F4681D4F42A3DE394DF4AE56EDE76372BB190B07A7C8EE0A6D709E02FCE1CDF7E2ECC03404CD28342F619172FE9CE98583FF8E4F1232EEF28183C3FE3B1B4C6FAD733BB5FCBC2EC22005C58EF1837D1683B2C6F34A26C1B2EFFA886B4238611FCFDCDE355B3B6519035BBC34F4DEF99C023861B46FC9D6E6C9077AD91D2691F7F7EE598CB0FAC186D91CAEFE130985139270B4130C93BC437944F4FD4452E2D74DD364F2E21E71F54BFF5CAE82AB9C9DF69EE86D2BC522363A0DABC521979B0DEADA1DBF9A42D5C4484E0ABCD06BFA53DDEF3C1B20EE3FD59D7C25E41D2B669E1EF16E6F52C3164DF4FB7930E9E4E58857B6AC7D5F42D69F6D187763CF1D5503400487F55BA57E31CC7A7135C886EFB4318AED6A1E012D9E6832A907600A918130C46DC778F971AD0038092999A333CB8B7A1A1DB93D7140003C2A4ECEA9F98D0ACC0A8291CDCEC97DCF8EC9B55A7F88A46B4DB5A851F44182E1C68A007E5E655F6AFFFFFFFFFFFFFFFF,
        2,
        325,
        512,
    ),
    "ffdhe6144": (
        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
        2,
        375,
        768,
    ),
    "ffdhe8192": (
        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
        2,
        400,
        1024,
    ),
}

[kevin-he-01]

One of the shortest valid Python source I found to reproduce this bug is just this literal (A single literal on a line is valid Python):

0xffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffff

The number is 0x followed by 256 fs. It is equal to 2 ** 1024 - 1
I found that any big integer that overflows a Python float (i.e. larger than about 1.7976931348623157e+308) would work, like this:

>>> float(2 ** 1024 - 1)
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
OverflowError: int too large to convert to float

The overflow explains why the number is being erroneously converted to the floating point infinity in the stack trace.

This bug impacts cryptographic Python scripts or modules that frequently use huge integers as constants.

[kevin-he-01]

One ugly workaround is to wrap the big number literals xxx in int('xxx', base=0). You need base=0 to accept octal or hexadecimal numbers. Of course eval('xxx') also works but the mere use of eval makes the code look like obfuscated malware in my opinion.

[Faholan]

This bug should've been introduced very recently, because when I wrote this code, a few weeks ago, it didn't fail. I guess it must've been introduced in the past week or so

[erictraut]

Thanks for reporting, and sorry for the regression. In the past week, I added support for large integers. But it appears that there are limits beyond which the implementation fails. My test cases tested very large numbers but not very very very large numbers. I'm working on a fix now.

[erictraut]

This will be fixed in the next release.

[debonte]

This issue was fixed in version 2022.1.5. You can find the changelog here: CHANGELOG.md