GIF89a=( õ' 7IAXKgNgYvYx\%wh…hŽth%ˆs%—x¨}9®Œ©€&©‰%¶†(¹–.¹5·œD¹&Çš)ÇŸ5ǘ;Í£*È¡&Õ²)ׯ7×µ<Ñ»4ï°3ø‘HÖ§KͯT÷¨Yÿšqÿ»qÿÔFØ !ù ' !ÿ NETSCAPE2.0 , =( þÀ“pH,È¤rÉl:ŸÐ¨tJ­Z¯Ø¬vËíz¿à°xL.›Ïè´zÍn»ßð¸|N¯Ûïø¼~Ïïûÿ€‚ƒ„…†‡ˆ‰Š‹ŒŽ‘’“”•–—˜™š›œžŸ ¡¢£¤¥¦§gª«ªE¯°¨¬ª±²Œ¹º¹E¾­”´ÂB¶¯ §Åȸ»ÑD¾¿Á•ÄÅ®° ÝH¾ÒLÀÆDÙ«D¶BÝïðÀ¾DÑÑÔTÌÍíH òGö¨A RÎڐ |¥ ٭&ºìE8œ¹kGÔAÞpx­a¶­ã R2XB®åE8I€Õ6Xî:vT)äžþÀq¦è³¥ì仕F~%xñ  4#ZÔ‰O|-4Bs‘X:= QÉ œš lºÒyXJŠGȦ|s hÏíK–3l7·B|¥$'7Jީܪ‰‡àá”Dæn=Pƒ ¤Òëí‰`䌨ljóá¯Éüv>á–Á¼5 ½.69ûϸd«­ºÀûnlv©‹ªîf{¬ÜãPbŸ  l5‘Ž¯pß ´ ˜3aÅùäI«O’ý·‘áÞ‡˜¾Æ‚ÙÏiÇÿ‹Àƒ #öó)pâš Þ½ ‘Ý{ó)vmÞü%D~ 6f s}ŃƒDØW Eþ`‡þ À…L8xá†ç˜{)x`X/> Ì}mø‚–RØ‘*|`D=‚Ø_ ^ð5 !_…'aä“OÚ—7âcð`D”Cx`ÝÂ¥ä‹éY¹—F¼¤¥Š?¡Õ™ n@`} lď’ÄÉ@4>ñd œ à‘vÒxNÃ×™@žd=ˆgsžG±æ ´²æud &p8Qñ)ˆ«lXD©øÜéAžHìySun jª×k*D¤LH] †¦§C™Jä–´Xb~ʪwStŽ6K,°£qÁœ:9ت:¨þªl¨@¡`‚ûÚ ».Û¬¯t‹ÆSÉ[:°=Š‹„‘Nåû”Ìî{¿ÂA ‡Rà›ÀÙ6úë°Ÿð0Ä_ ½;ÃϱîÉì^ÇÛÇ#Ëë¼ôº!±Ä˜íUîÅÇ;0L1óÁµö«p% AÀºU̬ݵ¼á%霼€‡¯Á~`ÏG¯»À× ­²± =4ªnpð3¾¤³¯­ü¾¦îuÙuµÙ®|%2ÊIÿür¦#0·ÔJ``8È@S@5ê¢ ö×Þ^`8EÜ]ý.뜃Âç 7 ú ȉÞj œ½Dç zý¸iþœÑÙûÄë!ˆÞÀl§Ïw‹*DçI€nEX¯¬¼ &A¬Go¼QföõFç°¯;é¦÷îŽêJ°îúôF5¡ÌQ|îúöXªæ»TÁÏyñêï]ê² o óÎC=öõ›ÒÓPB@ D×½œä(>èCÂxŽ`±«Ÿ–JЀ»Û á¤±p+eE0`ëŽ`A Ú/NE€Ø†À9‚@¤à H½7”à‡%B‰`Àl*ƒó‘–‡8 2ñ%¸ —€:Ù1Á‰E¸àux%nP1ð!‘ðC)¾P81lÑɸF#ˆ€{´âé°ÈB„0>±û °b¡Š´±O‚3È–Ù()yRpbµ¨E.Z‘D8ÊH@% òŒx+%Ù˜Æcü »¸˜fõ¬b·d`Fê™8èXH"ÉÈ-±|1Ô6iI, 2““¬$+](A*jÐ QTÂo‰.ÛU슬Œã„Ž`¯SN¡–¶Äåyše¯ª’­¬‚´b¦Éož œ)åyâ@Ì®3 ÎtT̉°&Ø+žLÀf"Ø-|žçÔ>‡Ðv¦Ðžì\‚ Q1)Ž@Žh#aP72”ˆ™¨$‚ !ù " , =( …7IAXG]KgNgYvYxR"k\%w]'}hŽth%ˆg+ˆs%—r.—m3šx3˜x¨}9®€&©€+¨‡7§‰%¶†(¹–.¹œD¹&ǘ;Í•&ײ)×»4ïÌ6ò§KÍ þ@‘pH,È¤rÉl:ŸÐ¨tJ­Z¯Ø¬vËíz¿à°xL.›Ïè´zÍn»ßð¸|N¯Ûïø¼~Ïïûÿ€‚ƒ„…†‡ˆ‰Š‹ŒŽ‘’“”•–—˜™š›œžŸ ¡¢£¤¥¦§g «¬ E ±± ¨­¶°ººE Á´”·®C¬²§Ç¶Œ»ÓDÃÕƷ¯Ê±H½ºM×ÁGÚ¬D¶BËÁ½î½DÓôTÏÛßîG»ôõC×CÌ l&âž:'òtU³6ɹ#·Ø)€'Ü.6±&ëÍÈ» K(8p0N?!æ2"ÛˆNIJX>R¼ÐO‚M '¡¨2¸*Ÿþ>#n↠å@‚<[:¡Iïf’ ¤TÚ˘CdbÜÙ“[«ŽEú5MBo¤×@€`@„€Êt W-3 ¶Ÿ¡BíêäjIÝ…Eò9[T…$íêﯧ„…•s»Óȳ¹€ÅÚdc®UUρ#±Ùïldj?´í¼²`\ŽÁðÞu|3'ÖŒ]ë6 ¶S#²‡˜FKLÈ *N E´‘áäŠ$˜›eÄYD„ºq«.è촁ƒs \-ÔjA 9²õ÷å- üúM[Âx(ís÷ì®x€|í¡Ù’p¦‚ ŽkÛTÇDpE@WÜ ²Ç]kŠ1¨ þ€·Yb ÓÁ‰l°*n0 ç™—žzBdОu¾7ĉBl€â‰-ºx~|UåU‰  h*Hœ|e"#"?vpÄiŠe6^ˆ„+qâŠm8 #VÇá ‘å–ÄV„œ|Šè•m"сœn|@›U¶ÆΞ—Špb¥G¨ED”€±Úê2FÌIç? >Éxå Œ± ¡¤„%‘žjŸ‘ꄯ<Ìaà9ijÐ2˜D¦È&›†Z`‚å]wþ¼Â:ç6àB¤7eFJ|õÒ§Õ,¨äàFÇ®cS·Ê¶+B°,‘Þ˜ºNûãØ>PADÌHD¹æž«ÄÀnÌ¥}­#Ë’ë QÀÉSÌÂÇ2ÌXÀ{æk²lQÁ2«ÊðÀ¯w|2Í h‹ÄÂG€,m¾¶ë3ÐÙ6-´ÅE¬L°ÆIij*K½ÀÇqï`DwVÍQXœÚÔpeœ±¬Ñ q˜§Tœ½µƒ°Œìu Â<¶aØ*At¯lmEØ ü ôÛN[P1ÔÛ¦­±$ÜÆ@`ùåDpy¶yXvCAyåB`ŽD¶ 0QwG#¯ æš[^Äþ $ÀÓÝǦ{„L™[±úKÄgÌ;ï£S~¹ìGX.ôgoT.»åˆ°ùŸûù¡?1zö¦Ÿž:ÅgÁ|ìL¹ „®£œŠ‚à0œ]PÁ^p F<"•ç?!,ñ‡N4—…PÄ Á„ö¨Û:Tè@hÀ‹%táÿ:ø-žI<`þ‹p I….)^ 40D#p@ƒj4–؀:²‰1Øâr˜¼F2oW¼#Z†;$Q q” ‘ ÂK¦ñNl#29 !’F@¥Bh·ᏀL!—XFóLH‘Kh¤.«hE&JòG¨¥<™WN!€ÑÙÚˆY„@†>Œž19J" 2,/ &.GXB%ÌRÈ9B6¹W]’î×ÔW¥’IÎ$ ñ‹ÓŒE8YÆ ¼³™ñA5“à®Q.aŸB€&Ø©³ JÁ—! ¦t)K%tœ-¦JF bòNMxLôþ)ÐR¸Ð™‘ èÝ6‘O!THÌ„HÛ ‰ !ù ) , =( …AXKgNgYvYxR"k\%wh…hŽh%ˆg+ˆs%—r.—x3˜x¨}9®€&©€+¨Œ,©‡7§‰%¶†(¹–.¹5·&Çš)ǘ;Í•&×£*Ȳ)ׯ7×»4ï°3øÌ6ò‘HÖ§KÍ»Hó¯T÷¨Yÿ»qÿÇhÿ þÀ”pH,È¤rÉl:ŸÐ¨tJ­Z¯Ø¬vËíz¿à°xL.›Ïè´zÍn»ßð¸|N¯Ûïø¼~Ïïûÿ€‚ƒ„…†‡ˆ‰Š‹ŒŽ‘’“”•–—˜™š›œžŸ ¡¢£¤¥¦§g ª« E$±²¨ª­ · °²½$E$ÂÕ««D· Í ¿¦Ç¶¸ÌŒ¾³CÃÅÆ E ééH½MÛÂGâªD­ çBêêϾD²ÒaÀà€Š1r­ðÓ¤ ÔožzU!L˜C'¾yW½UGtäÇïÙllê0×àÂuGþ)AÀs[þ·xì ÁxO%ƒûX2ó—  P£n›R/¡ÑšHše+êDm?# —‘Ç£6¡8íJ¡ŸâDiäªM¥Ö„ôj“¬¹£5oQ7°- <‡ *´lãÓŒ2r/a!l)dÈ A™ÈE¢ôÔ͆…ð ;Ö˜c ¡%ß‚’Ùˆâ¸b½—pe~C"BíëÚHïeF2§æŠ8qb t_`urŠeü wÅu3êæPv§h•"ß`íÍxçLĹÜÖ3á  ~Öº“®›¸ÏMDfJÙ °„ÛµáWõ%§œ‚à©–‚X Ó؁)@®Ñ›Eþ´wëuÅSxb8y\mÖzœ¥§ZbºE—ÂLªÌw!y(>¡™wú=Ç|ÅÝs¢d €CÁW)HÜcC$€L Ä7„r.á\{)@ð` @ äXÈ$PD” `šaG:§æˆOˆ72EÐamn]ù"ŒcÊxÑŒ° &dR8`g«iÙŸLR!¦P …d’ä¡“¦ðÎTƒ¦ià|À _ ¥ Qi#¦Šg›Æ ›noMµ ›V ã£)p ç£ÎW…š=Âeªk§†j„ ´®1ß²sÉxéW«jšl|0¯B0Û, \jÛ´›6±¬¶C ÛíWþï|ëÙ‹¸ñzĸV {ì;Ýñn¼òVˆm³I¼³.Ðã¤PN¥ ²µ¼„µCã+¹ÍByî£Ñ¾HŸ›ëê 7ìYÆFTk¨SaoaY$Dµœìï¿Ã29RÈkt Çïfñ ÇÒ:ÀÐSp¹3ÇI¨â¥DZÄ ü9Ïýögñ½­uÔ*3)O‘˜Ö[_hv ,àî×Et Ÿé¶BH€ Õ[ü±64M@ÔSÌM7dÐl5-ÄÙU܍´©zߌ3Ô€3ž„ „ ¶ÛPô½5×g› êÚ˜kN„Ý…0Îj4€Ìë°“#{þÕ3S2çKÜ'ợlø¼Ú2K{° {Û¶?žm𸧠ËI¼nEò='êüóºè^üæÃ_Û=°óž‚ì#Oý¿Í'¡½áo..ÏYìnüñCœO±Áa¿¢Kô½o,üÄËbö²çºíï{ËC Ú— "”Ï{ËK ÍÒw„õ±Oz dÕ¨à:$ ƒô—«v»] A#ð «€¿šéz)Rx׿ˆ¥‚d``èw-îyÏf×K!ð€þ­Ð|ìPľ„=Ì`ý(f” 'Pa ¥ÐBJa%Ðâf§„%Š¡}FàáÝ×6>ÉäŠG"éŽè=ø!oŠ°^FP¼Ø©Q„ÀCÙÁ`(Ž\ÄÝ® ©Â$<n@dÄ E#ììUÒI! ‚#lù‹`k¦ÐÇ'Rró’ZýNBÈMF Í[¤+‹ðɈ-áwj¨¥þ8¾rá ,VÂh„"|½œ=×G_¦Ñ™EØ 0i*%̲˜Æda0mV‚k¾)›;„&6 p>ÓjK “¦Ç# âDÂ:ûc?:R Ó¬fÞéI-Ì“•Ã<ä=™Ï7˜3œ¨˜c2ŒW ,ˆ”8(T™P‰F¡Jhç"‚ ; 403WebShell
403Webshell
Server IP : 172.67.177.218  /  Your IP : 216.73.216.195
Web Server : LiteSpeed
System : Linux premium229.web-hosting.com 4.18.0-553.45.1.lve.el8.x86_64 #1 SMP Wed Mar 26 12:08:09 UTC 2025 x86_64
User : akhalid ( 749)
PHP Version : 8.3.22
Disable Function : NONE
MySQL : OFF  |  cURL : ON  |  WGET : ON  |  Perl : ON  |  Python : ON  |  Sudo : OFF  |  Pkexec : OFF
Directory :  /opt/hc_python/lib/python3.12/site-packages/pydantic/v1/

Upload File :
current_dir [ Writeable ] document_root [ Writeable ]

 

Command :

[ Back ]     

Current File : /opt/hc_python/lib/python3.12/site-packages/pydantic/v1/dataclasses.py
"""
The main purpose is to enhance stdlib dataclasses by adding validation
A pydantic dataclass can be generated from scratch or from a stdlib one.

Behind the scene, a pydantic dataclass is just like a regular one on which we attach
a `BaseModel` and magic methods to trigger the validation of the data.
`__init__` and `__post_init__` are hence overridden and have extra logic to be
able to validate input data.

When a pydantic dataclass is generated from scratch, it's just a plain dataclass
with validation triggered at initialization

The tricky part if for stdlib dataclasses that are converted after into pydantic ones e.g.

```py
@dataclasses.dataclass
class M:
    x: int

ValidatedM = pydantic.dataclasses.dataclass(M)
```

We indeed still want to support equality, hashing, repr, ... as if it was the stdlib one!

```py
assert isinstance(ValidatedM(x=1), M)
assert ValidatedM(x=1) == M(x=1)
```

This means we **don't want to create a new dataclass that inherits from it**
The trick is to create a wrapper around `M` that will act as a proxy to trigger
validation without altering default `M` behaviour.
"""
import copy
import dataclasses
import sys
from contextlib import contextmanager
from functools import wraps

try:
    from functools import cached_property
except ImportError:
    # cached_property available only for python3.8+
    pass

from typing import TYPE_CHECKING, Any, Callable, ClassVar, Dict, Generator, Optional, Type, TypeVar, Union, overload

from typing_extensions import dataclass_transform

from pydantic.v1.class_validators import gather_all_validators
from pydantic.v1.config import BaseConfig, ConfigDict, Extra, get_config
from pydantic.v1.error_wrappers import ValidationError
from pydantic.v1.errors import DataclassTypeError
from pydantic.v1.fields import Field, FieldInfo, Required, Undefined
from pydantic.v1.main import create_model, validate_model
from pydantic.v1.utils import ClassAttribute

if TYPE_CHECKING:
    from pydantic.v1.main import BaseModel
    from pydantic.v1.typing import CallableGenerator, NoArgAnyCallable

    DataclassT = TypeVar('DataclassT', bound='Dataclass')

    DataclassClassOrWrapper = Union[Type['Dataclass'], 'DataclassProxy']

    class Dataclass:
        # stdlib attributes
        __dataclass_fields__: ClassVar[Dict[str, Any]]
        __dataclass_params__: ClassVar[Any]  # in reality `dataclasses._DataclassParams`
        __post_init__: ClassVar[Callable[..., None]]

        # Added by pydantic
        __pydantic_run_validation__: ClassVar[bool]
        __post_init_post_parse__: ClassVar[Callable[..., None]]
        __pydantic_initialised__: ClassVar[bool]
        __pydantic_model__: ClassVar[Type[BaseModel]]
        __pydantic_validate_values__: ClassVar[Callable[['Dataclass'], None]]
        __pydantic_has_field_info_default__: ClassVar[bool]  # whether a `pydantic.Field` is used as default value

        def __init__(self, *args: object, **kwargs: object) -> None:
            pass

        @classmethod
        def __get_validators__(cls: Type['Dataclass']) -> 'CallableGenerator':
            pass

        @classmethod
        def __validate__(cls: Type['DataclassT'], v: Any) -> 'DataclassT':
            pass


__all__ = [
    'dataclass',
    'set_validation',
    'create_pydantic_model_from_dataclass',
    'is_builtin_dataclass',
    'make_dataclass_validator',
]

_T = TypeVar('_T')

if sys.version_info >= (3, 10):

    @dataclass_transform(field_specifiers=(dataclasses.field, Field))
    @overload
    def dataclass(
        *,
        init: bool = True,
        repr: bool = True,
        eq: bool = True,
        order: bool = False,
        unsafe_hash: bool = False,
        frozen: bool = False,
        config: Union[ConfigDict, Type[object], None] = None,
        validate_on_init: Optional[bool] = None,
        use_proxy: Optional[bool] = None,
        kw_only: bool = ...,
    ) -> Callable[[Type[_T]], 'DataclassClassOrWrapper']:
        ...

    @dataclass_transform(field_specifiers=(dataclasses.field, Field))
    @overload
    def dataclass(
        _cls: Type[_T],
        *,
        init: bool = True,
        repr: bool = True,
        eq: bool = True,
        order: bool = False,
        unsafe_hash: bool = False,
        frozen: bool = False,
        config: Union[ConfigDict, Type[object], None] = None,
        validate_on_init: Optional[bool] = None,
        use_proxy: Optional[bool] = None,
        kw_only: bool = ...,
    ) -> 'DataclassClassOrWrapper':
        ...

else:

    @dataclass_transform(field_specifiers=(dataclasses.field, Field))
    @overload
    def dataclass(
        *,
        init: bool = True,
        repr: bool = True,
        eq: bool = True,
        order: bool = False,
        unsafe_hash: bool = False,
        frozen: bool = False,
        config: Union[ConfigDict, Type[object], None] = None,
        validate_on_init: Optional[bool] = None,
        use_proxy: Optional[bool] = None,
    ) -> Callable[[Type[_T]], 'DataclassClassOrWrapper']:
        ...

    @dataclass_transform(field_specifiers=(dataclasses.field, Field))
    @overload
    def dataclass(
        _cls: Type[_T],
        *,
        init: bool = True,
        repr: bool = True,
        eq: bool = True,
        order: bool = False,
        unsafe_hash: bool = False,
        frozen: bool = False,
        config: Union[ConfigDict, Type[object], None] = None,
        validate_on_init: Optional[bool] = None,
        use_proxy: Optional[bool] = None,
    ) -> 'DataclassClassOrWrapper':
        ...


@dataclass_transform(field_specifiers=(dataclasses.field, Field))
def dataclass(
    _cls: Optional[Type[_T]] = None,
    *,
    init: bool = True,
    repr: bool = True,
    eq: bool = True,
    order: bool = False,
    unsafe_hash: bool = False,
    frozen: bool = False,
    config: Union[ConfigDict, Type[object], None] = None,
    validate_on_init: Optional[bool] = None,
    use_proxy: Optional[bool] = None,
    kw_only: bool = False,
) -> Union[Callable[[Type[_T]], 'DataclassClassOrWrapper'], 'DataclassClassOrWrapper']:
    """
    Like the python standard lib dataclasses but with type validation.
    The result is either a pydantic dataclass that will validate input data
    or a wrapper that will trigger validation around a stdlib dataclass
    to avoid modifying it directly
    """
    the_config = get_config(config)

    def wrap(cls: Type[Any]) -> 'DataclassClassOrWrapper':
        should_use_proxy = (
            use_proxy
            if use_proxy is not None
            else (
                is_builtin_dataclass(cls)
                and (cls.__bases__[0] is object or set(dir(cls)) == set(dir(cls.__bases__[0])))
            )
        )
        if should_use_proxy:
            dc_cls_doc = ''
            dc_cls = DataclassProxy(cls)
            default_validate_on_init = False
        else:
            dc_cls_doc = cls.__doc__ or ''  # needs to be done before generating dataclass
            if sys.version_info >= (3, 10):
                dc_cls = dataclasses.dataclass(
                    cls,
                    init=init,
                    repr=repr,
                    eq=eq,
                    order=order,
                    unsafe_hash=unsafe_hash,
                    frozen=frozen,
                    kw_only=kw_only,
                )
            else:
                dc_cls = dataclasses.dataclass(  # type: ignore
                    cls, init=init, repr=repr, eq=eq, order=order, unsafe_hash=unsafe_hash, frozen=frozen
                )
            default_validate_on_init = True

        should_validate_on_init = default_validate_on_init if validate_on_init is None else validate_on_init
        _add_pydantic_validation_attributes(cls, the_config, should_validate_on_init, dc_cls_doc)
        dc_cls.__pydantic_model__.__try_update_forward_refs__(**{cls.__name__: cls})
        return dc_cls

    if _cls is None:
        return wrap

    return wrap(_cls)


@contextmanager
def set_validation(cls: Type['DataclassT'], value: bool) -> Generator[Type['DataclassT'], None, None]:
    original_run_validation = cls.__pydantic_run_validation__
    try:
        cls.__pydantic_run_validation__ = value
        yield cls
    finally:
        cls.__pydantic_run_validation__ = original_run_validation


class DataclassProxy:
    __slots__ = '__dataclass__'

    def __init__(self, dc_cls: Type['Dataclass']) -> None:
        object.__setattr__(self, '__dataclass__', dc_cls)

    def __call__(self, *args: Any, **kwargs: Any) -> Any:
        with set_validation(self.__dataclass__, True):
            return self.__dataclass__(*args, **kwargs)

    def __getattr__(self, name: str) -> Any:
        return getattr(self.__dataclass__, name)

    def __setattr__(self, __name: str, __value: Any) -> None:
        return setattr(self.__dataclass__, __name, __value)

    def __instancecheck__(self, instance: Any) -> bool:
        return isinstance(instance, self.__dataclass__)

    def __copy__(self) -> 'DataclassProxy':
        return DataclassProxy(copy.copy(self.__dataclass__))

    def __deepcopy__(self, memo: Any) -> 'DataclassProxy':
        return DataclassProxy(copy.deepcopy(self.__dataclass__, memo))


def _add_pydantic_validation_attributes(  # noqa: C901 (ignore complexity)
    dc_cls: Type['Dataclass'],
    config: Type[BaseConfig],
    validate_on_init: bool,
    dc_cls_doc: str,
) -> None:
    """
    We need to replace the right method. If no `__post_init__` has been set in the stdlib dataclass
    it won't even exist (code is generated on the fly by `dataclasses`)
    By default, we run validation after `__init__` or `__post_init__` if defined
    """
    init = dc_cls.__init__

    @wraps(init)
    def handle_extra_init(self: 'Dataclass', *args: Any, **kwargs: Any) -> None:
        if config.extra == Extra.ignore:
            init(self, *args, **{k: v for k, v in kwargs.items() if k in self.__dataclass_fields__})

        elif config.extra == Extra.allow:
            for k, v in kwargs.items():
                self.__dict__.setdefault(k, v)
            init(self, *args, **{k: v for k, v in kwargs.items() if k in self.__dataclass_fields__})

        else:
            init(self, *args, **kwargs)

    if hasattr(dc_cls, '__post_init__'):
        try:
            post_init = dc_cls.__post_init__.__wrapped__  # type: ignore[attr-defined]
        except AttributeError:
            post_init = dc_cls.__post_init__

        @wraps(post_init)
        def new_post_init(self: 'Dataclass', *args: Any, **kwargs: Any) -> None:
            if config.post_init_call == 'before_validation':
                post_init(self, *args, **kwargs)

            if self.__class__.__pydantic_run_validation__:
                self.__pydantic_validate_values__()
                if hasattr(self, '__post_init_post_parse__'):
                    self.__post_init_post_parse__(*args, **kwargs)

            if config.post_init_call == 'after_validation':
                post_init(self, *args, **kwargs)

        setattr(dc_cls, '__init__', handle_extra_init)
        setattr(dc_cls, '__post_init__', new_post_init)

    else:

        @wraps(init)
        def new_init(self: 'Dataclass', *args: Any, **kwargs: Any) -> None:
            handle_extra_init(self, *args, **kwargs)

            if self.__class__.__pydantic_run_validation__:
                self.__pydantic_validate_values__()

            if hasattr(self, '__post_init_post_parse__'):
                # We need to find again the initvars. To do that we use `__dataclass_fields__` instead of
                # public method `dataclasses.fields`

                # get all initvars and their default values
                initvars_and_values: Dict[str, Any] = {}
                for i, f in enumerate(self.__class__.__dataclass_fields__.values()):
                    if f._field_type is dataclasses._FIELD_INITVAR:  # type: ignore[attr-defined]
                        try:
                            # set arg value by default
                            initvars_and_values[f.name] = args[i]
                        except IndexError:
                            initvars_and_values[f.name] = kwargs.get(f.name, f.default)

                self.__post_init_post_parse__(**initvars_and_values)

        setattr(dc_cls, '__init__', new_init)

    setattr(dc_cls, '__pydantic_run_validation__', ClassAttribute('__pydantic_run_validation__', validate_on_init))
    setattr(dc_cls, '__pydantic_initialised__', False)
    setattr(dc_cls, '__pydantic_model__', create_pydantic_model_from_dataclass(dc_cls, config, dc_cls_doc))
    setattr(dc_cls, '__pydantic_validate_values__', _dataclass_validate_values)
    setattr(dc_cls, '__validate__', classmethod(_validate_dataclass))
    setattr(dc_cls, '__get_validators__', classmethod(_get_validators))

    if dc_cls.__pydantic_model__.__config__.validate_assignment and not dc_cls.__dataclass_params__.frozen:
        setattr(dc_cls, '__setattr__', _dataclass_validate_assignment_setattr)


def _get_validators(cls: 'DataclassClassOrWrapper') -> 'CallableGenerator':
    yield cls.__validate__


def _validate_dataclass(cls: Type['DataclassT'], v: Any) -> 'DataclassT':
    with set_validation(cls, True):
        if isinstance(v, cls):
            v.__pydantic_validate_values__()
            return v
        elif isinstance(v, (list, tuple)):
            return cls(*v)
        elif isinstance(v, dict):
            return cls(**v)
        else:
            raise DataclassTypeError(class_name=cls.__name__)


def create_pydantic_model_from_dataclass(
    dc_cls: Type['Dataclass'],
    config: Type[Any] = BaseConfig,
    dc_cls_doc: Optional[str] = None,
) -> Type['BaseModel']:
    field_definitions: Dict[str, Any] = {}
    for field in dataclasses.fields(dc_cls):
        default: Any = Undefined
        default_factory: Optional['NoArgAnyCallable'] = None
        field_info: FieldInfo

        if field.default is not dataclasses.MISSING:
            default = field.default
        elif field.default_factory is not dataclasses.MISSING:
            default_factory = field.default_factory
        else:
            default = Required

        if isinstance(default, FieldInfo):
            field_info = default
            dc_cls.__pydantic_has_field_info_default__ = True
        else:
            field_info = Field(default=default, default_factory=default_factory, **field.metadata)

        field_definitions[field.name] = (field.type, field_info)

    validators = gather_all_validators(dc_cls)
    model: Type['BaseModel'] = create_model(
        dc_cls.__name__,
        __config__=config,
        __module__=dc_cls.__module__,
        __validators__=validators,
        __cls_kwargs__={'__resolve_forward_refs__': False},
        **field_definitions,
    )
    model.__doc__ = dc_cls_doc if dc_cls_doc is not None else dc_cls.__doc__ or ''
    return model


if sys.version_info >= (3, 8):

    def _is_field_cached_property(obj: 'Dataclass', k: str) -> bool:
        return isinstance(getattr(type(obj), k, None), cached_property)

else:

    def _is_field_cached_property(obj: 'Dataclass', k: str) -> bool:
        return False


def _dataclass_validate_values(self: 'Dataclass') -> None:
    # validation errors can occur if this function is called twice on an already initialised dataclass.
    # for example if Extra.forbid is enabled, it would consider __pydantic_initialised__ an invalid extra property
    if getattr(self, '__pydantic_initialised__'):
        return
    if getattr(self, '__pydantic_has_field_info_default__', False):
        # We need to remove `FieldInfo` values since they are not valid as input
        # It's ok to do that because they are obviously the default values!
        input_data = {
            k: v
            for k, v in self.__dict__.items()
            if not (isinstance(v, FieldInfo) or _is_field_cached_property(self, k))
        }
    else:
        input_data = {k: v for k, v in self.__dict__.items() if not _is_field_cached_property(self, k)}
    d, _, validation_error = validate_model(self.__pydantic_model__, input_data, cls=self.__class__)
    if validation_error:
        raise validation_error
    self.__dict__.update(d)
    object.__setattr__(self, '__pydantic_initialised__', True)


def _dataclass_validate_assignment_setattr(self: 'Dataclass', name: str, value: Any) -> None:
    if self.__pydantic_initialised__:
        d = dict(self.__dict__)
        d.pop(name, None)
        known_field = self.__pydantic_model__.__fields__.get(name, None)
        if known_field:
            value, error_ = known_field.validate(value, d, loc=name, cls=self.__class__)
            if error_:
                raise ValidationError([error_], self.__class__)

    object.__setattr__(self, name, value)


def is_builtin_dataclass(_cls: Type[Any]) -> bool:
    """
    Whether a class is a stdlib dataclass
    (useful to discriminated a pydantic dataclass that is actually a wrapper around a stdlib dataclass)

    we check that
    - `_cls` is a dataclass
    - `_cls` is not a processed pydantic dataclass (with a basemodel attached)
    - `_cls` is not a pydantic dataclass inheriting directly from a stdlib dataclass
    e.g.
    ```
    @dataclasses.dataclass
    class A:
        x: int

    @pydantic.dataclasses.dataclass
    class B(A):
        y: int
    ```
    In this case, when we first check `B`, we make an extra check and look at the annotations ('y'),
    which won't be a superset of all the dataclass fields (only the stdlib fields i.e. 'x')
    """
    return (
        dataclasses.is_dataclass(_cls)
        and not hasattr(_cls, '__pydantic_model__')
        and set(_cls.__dataclass_fields__).issuperset(set(getattr(_cls, '__annotations__', {})))
    )


def make_dataclass_validator(dc_cls: Type['Dataclass'], config: Type[BaseConfig]) -> 'CallableGenerator':
    """
    Create a pydantic.dataclass from a builtin dataclass to add type validation
    and yield the validators
    It retrieves the parameters of the dataclass and forwards them to the newly created dataclass
    """
    yield from _get_validators(dataclass(dc_cls, config=config, use_proxy=True))

Youez - 2016 - github.com/yon3zu
LinuXploit