"jjSSKJr SSKrSSKrSSKJr SSKJr SSKJr SSK J r SSK J r J r Jr SSKJr SS KJrJrJrJrJr SS KJrJr S S KJr S S KJr S SKJr S SK J!r! Sr"\#"SS15r$\#"1Sk5r%Sr&\ S'Sj5r'"SS5r(S(Sjr)\"SSSS9"SS55r*\"SSSS9"SS\*55r+\"SSSS9"SS \*55r,S)S!jr-S)S"jr.\"SS#S$9"S%S&55r/g)*) annotationsN)copy) dataclass)Decimal)cache) AnnotatedAnycast) get_snapshot) AliasChoicesBeforeValidatorFieldGetCoreSchemaHandler TypeAdapter)SchemaSerializer core_schema)_utils)iter_provider_references)CostNotFoundWarning)UsageLimitExceeded) RequestUsageRunUsage UsageLimits input_tokens output_tokens>requests total_tokensrequest_tokensresponse_tokens))rr)rr c,[U5R$N)r serializer) usage_types N/home/mande/repo/quber/.venv/lib/python3.13/site-packages/pydantic_ai/usage.py_usage_serializerr&#s z " - --c\rSrSrSSjrSrg)_UsageSerializerDescriptor(c[U5$r")r&)selfinstanceowners r%__get__"_UsageSerializerDescriptor.__get__)s  ''r'N)r-objectr. type[object]returnr)__name__ __module__ __qualname____firstlineno__r/__static_attributes__r1r'r%r)r)(s(r'r) UsageBasecU"U5n[U[5(de[[[[4U5R 5nUR R5VVs0sH$upxXs;dM UcUR(aM"Xx_M& n nn[[[[4URU S[[UR5[[UR5URURURURURUR UR"UR$S9 5n UR'U 5 U$s snnf)Npython) modeincludeexcludeby_alias exclude_unsetexclude_defaults exclude_noneexclude_computed_fields round_tripserialize_as_anycontext) isinstancedictr strr r__dict__itemsrC to_pythonr>r?r@rArBrDrErFrGupdate) valueinnerinforeserved_namesextra_serializer serializedresultkeyitemextras r%_serialize_usagerY-s1uJ j$ ' '' ' $sCx.* - 2 2 4F--/ /IC  $ *.*:$BSBS  /    S#X"" dll+dll+]],,!22**$($@$@!22LL #  E$ MM% M1 s( E7E EF)repriniteqcV\rSrSr%\"5rSrS\S'SrS\S'Sr S\S'Sr S\S 'Sr S\S 'Sr S\S 'Sr S\S '\R"\\\4S 9rS\S'SrS\S'SS.SSjjr\SSj5rSSjr\S Sj5r\S!Sj5rS"SjrSrS#SjrS$SjrSr g)%r:SrzVAnnotated[int, Field(validation_alias=AliasChoices('input_tokens', 'request_tokens'))]rintcache_write_tokenscache_read_tokenszXAnnotated[int, Field(validation_alias=AliasChoices('output_tokens', 'response_tokens'))]rinput_audio_tokenscache_audio_read_tokensoutput_audio_tokensdefault_factoryz=Annotated[dict[str, int], BeforeValidator(lambda d: d or {})]detailsNDecimal | Nonecostrgc nU=(d 0UlUR5Hup4[XU5 M gr")rgrLsetattr)r,rgkwargskvs r%__init__UsageBase.__init__s)}" LLNDA DQ #r'c `^^U"U5n[S[R"U555nU[[U55-[-m[ [ R"55mSU4SjjnSUU4Sjjn[ R"UU[ R"USUS9S9$) z>Preserve arbitrary usage fields across Pydantic serialization.c38# UHoRv M g7fr")name).0fields r% 9UsageBase.__get_pydantic_core_schema__..sX8Wu 8Wsc>[U[5(aq[[[[4U5nUR 5nUR S5(d0US'[HupEXB;dM XR;dMX%bMSX5'M OSn[[U5nU"U5n[U[5(deUb-UR5HupxUT ;dM [XgU5 M U$)Nrgr) rHrIr rJr rget_LEGACY_TOKEN_ALIASESr2r:rLrl) rOrP value_dict input_value field_name legacy_namerUrVrWrRs r%validate8UsageBase.__get_pydantic_core_schema__..validates%&&!$sCx.%8 (oo/ !~~i00-/K */D+J!3 8QV`VmVu34 00E" "651 ;'Ffi00 00%!+!1!1!3IC.0T2"4Mr'c>[UUUTTS9$)N)rRrS)rY)rOrPrQrSrRs r% serialize9UsageBase.__get_pydantic_core_schema__..serializes! $-!1  r'T)info_argschema) serialization)rOr rPz(core_schema.ValidatorFunctionWrapHandlerr4r:)rOr:rP)core_schema.SerializerFunctionWrapHandlerrQcore_schema.SerializationInfor4r ) frozenset dataclassesfieldsdir_LEGACY_USAGE_KEYSrr any_schemano_info_wrap_validator_function#wrap_serializer_function_ser_schema) cls source_typehandlerr field_namesrrrSrRs @@r%__get_pydantic_core_schema__&UsageBase.__get_pydantic_core_schema__s%X 8J8J;8WXX $y[1A'BBEWW+K,B,B,DE *  < 0   ::  %II)^bkqr  r'c[U5nURU5nURRUR5 URR 5UlU$)zB=N=Nt%%(9(99WTWWr'c0nUR(aURUS'UR(aURUS'URR5nUR(aURUS'URUS'UR (aUR US'UR US'UR (aUR US'UR(aURUS'UR(aURUS 'U(a2S nUR5HupEU[;aMUcMXQX4-'M U$) z7Get the token usage values as OpenTelemetry attributes.zgen_ai.usage.input_tokenszgen_ai.usage.output_tokensz(gen_ai.usage.cache_creation.input_tokensr`z$gen_ai.usage.cache_read.input_tokensrarbrcrdzgen_ai.usage.details.) rrrgrr`rarbrcrdrL_FIRST_CLASS_TOKEN_DETAIL_KEYS)r,rUrgprefixrVrOs r%opentelemetry_attributes"UsageBase.opentelemetry_attributess2!#   262C2CF. /   373E3EF/ 0,,##%  " "AEAXAXF= >,0,C,CG( )  ! !=A=S=SF9 :+/+A+AG' (  " ",0,C,CG( )  ' '151M1MG- .  # #-1-E-EG) * ,F%mmo 88$+06<(. r'cS[URR555nURRSSR U5S3$)Nc3H# UHupU(dMUSU<3v M g7f)=Nr1)rurtrOs r%rw%UsageBase.__repr__..s%b;XKD\a'tfAeY';Xs ""(z, ))sortedrKrL __class__r7join)r,kv_pairss r%__repr__UsageBase.__repr__sFb6$--BUBUBW;Xb..--.a (0C/DAFFr'c^^^[T5[T5LaV[5mTRR5TRR5-n[ UUU4SjU55$[ $)Nc3Z># UH n[TUT5[TUT5:Hv M" g7fr")getattr)rurVmissingr,rOs r%rw#UsageBase.__eq__.. s+c^bWZwtS'2geS'6RR^bs(+)rr2rKkeysallNotImplemented)r,rOrrs`` @r%__eq__UsageBase.__eq__sV :e $hG==%%'%..*=*=*??Dc^bcc cr'c[URR55=(d* [SURR 555$)z(Whether any values are set and non-zero.c3:# UHupUS:wdM Uv M g7f)rgNr1)rurnros r%rw'UsageBase.has_values..s0g?TtqXY]fXf?Ts  )anyrgvaluesrKrLrs r% has_valuesUsageBase.has_values s74<<&&()gS0gt}}?R?R?T0g-ggr')rgzdict[str, int] | Nonermr )rr rrr4zcore_schema.CoreSchema)r4r:)r4r_)r4float)r4dict[str, int])rOr2r4boolr4r)!r5r6r7r8r)__pydantic_serializer__r__annotations__r`rarrbrcrdrrvrIrJr_rgrirp classmethodrrpropertyrrrrrrr9r1r'r%r:r:Ss*9:   Ls  .C#$S$o  E $sCx.9  : 3D.<@ - - ^66 X X#JGhr'ct\rSrSrSr\S5rS SjrS Sjr\ SSS.S S jj5r S r g)rizLLM usage associated with a single request. This is an implementation of `genai_prices.types.AbstractUsage` so it can be used to calculate the price of the request using [genai-prices](https://github.com/pydantic/genai-prices). cg)Nrr1rs r%rRequestUsage.requestssr'c0[X5 [X5 gzPIncrement the usage in place. Args: incr_usage: The usage to increment by. N)_incr_usage_tokens_incr_usage_costr, incr_usages r%incrRequestUsage.incrs 4,*r'c>[U5nURU5 U$)zAdd two RequestUsages together. This is provided so it's trivial to sum usage information from multiple parts of a response. **WARNING:** this CANNOT be used to sum multiple requests without breaking some pricing calculations. rrr,other new_usages r%__add__RequestUsage.__add__&sJ ur'defaultN) api_flavorrgc LU=(d 0n[X2US9Houpx[5RSXx5n U RXS9upU"S0U RR 5V V s0sH upU cM X_M sn n DSU0D6s $ U"US9$s sn n f![ a Mf=f)anExtract usage information from the response data using genai-prices. Args: data: The response data from the model API. provider: The actual provider ID provider_url: The provider base_url provider_fallback: The fallback provider ID to use if the actual provider is not found in genai-prices. For example, an OpenAI model should set this to "openai" in case it has an obscure provider ID. api_flavor: The API flavor to use when extracting usage information, e.g. 'chat' or 'responses' for OpenAI. details: Becomes the `details` field on the returned `RequestUsage` for convenience. )provider_api_url provider_idprovider_fallbackN)rrgrjr1)rr find_provider extract_usagerKrL Exception)rdataprovider provider_urlrrrgrr provider_obj _model_refextracted_usagernros r%extractRequestUsage.extract1s.-R-E)Sd. )K +~;;D+` .:.H.H.H.e+ s/G/G/M/M/Oa/OtqSTdad/Oaskrss . 7##b  s*A B( B5B; BB B#"B#r1)rrr4None)rrr4r)rr rrJrrJrrJrrJrgzdict[str, Any] | Noner4r) r5r6r7r8__doc__rrrrrrr9r1r'r%rrs + $)- $ $ $  $  $ $' $  $ $r'rc\rSrSr%SrSrS\S'SrS\S'SrS\S'Sr S\S'Sr S\S 'Sr S\S 'Sr S\S 'Sr S\S '\R"\\\4S 9rS\S'SSjrSSjrSSjrSrg)riUzLLM usage associated with an agent run. Responsibility for calculating request usage is on the model; Pydantic AI simply sums the usage information across requests. rr_r tool_callsrr`rarbrcrrerrgc[U[5(a>U=RUR- slU=RUR- sl[ X5 [ X5 gr)rHrrrrrrs r%r RunUsage.incrwsI j( + + MMZ00 0M OOz44 4O4,*r'c>[U5nURU5 U$)zkAdd two RunUsages together. This is provided so it's trivial to sum usage information from multiple runs. rrs r%rRunUsage.__add__s J ur'c&URUR-Vs0sH<o"URRUS5URRUS5- _M> nn[URUR- URUR- UR UR - UR UR - URUR- URUR- URUR- URUR- URUR- UURb>URUR:wa$URUR=(d S- S9 $SS9 $s snf)a8Return the field-by-field usage accumulated since `other`. This is useful when a nested operation shares a run's mutable usage object and needs to report only the requests, tool calls, tokens, details, and cost added by that operation. Unknown costs remain `None`; an unchanged known cost also produces `None`. rN) rrrr`rarrbrcrdrgri) rgrzrrrrr`rarrbrcrdri)r,rrtrgs r%__sub__RunUsage.__sub__sdVZUaUadidqdqUq UqT$,,""4+emm.?.?a.HH HUq  ]]U^^3)9)99**U-?-??#669Q9QQ"44u7N7NN,,u/B/BB#669Q9QQ$($@$@5C`C`$` $ 8 85;T;T T26))2GDIIY^YcYcLcejjoA.  jn   sAFr1N)rRunUsage | RequestUsager4r)rrr4r)rrr4r)r5r6r7r8rrrrrr`rarbrcrrrvrIrJr_rgrrrr9r1r'r%rrUs Hc1JBL#.6s5-#$S$;M33)//S#XOG^O2 + r'rcnURb(UR=(d SUR-Ulgg)Nr)ri)slfrs r%rrs)"HHMZ__4#r'cURR5URR5-1Sk- Hcn[XS5n[XS5n[U[[ 45(dM8[U[[ 45(dMU[ XX4-5 Me URR5HNupV[U[[ 45(dM"URRUS5U-URU'MP g)zqIncrement the usage in place. Args: slf: The usage to increment. incr_usage: The usage to increment by. >rirgrrrN) rKrrrHr_rrlrgrLrz)rrrn slf_value incr_valuerVrOs r%rrsll!J$7$7$<$<$>>Bo oCA& ZA. i#u . .:j3PU,3W3W CI2 3 p !((..0  ec5\ * *"{{sA6>CKK 1r'T)rZkw_onlyc\rSrSr%SrSrS\S'SrS\S'SrS\S 'Sr S\S 'Sr S\S 'Sr S\S 'Sr S\S 'Sr S\S'SSjrSSjrSS.SSjjrSSjrSSjrSSjrSSjr\R,rSrg) ria5Limits on model usage. The request count is tracked by pydantic_ai, and the request limit is checked before each request to the model. Token counts are provided in responses from the model, and the token limits are checked after each response. Each of the limits can be set to `None` to disable that limit. Nrh cost_limit2z int | None request_limittool_calls_limitinput_tokens_limitoutput_tokens_limittotal_tokens_limitper_request_input_tokens_limitFrcount_tokens_before_requestc~[SURURURUR455$)a~Returns `True` if this instance places any limits on token counts. If this returns `False`, the `check_tokens` and `check_per_request_input_tokens` methods will never raise an error. This is useful because if we have token limits, we need to check them after receiving each streamed message. If there are no limits, we can skip that processing in the streaming response iterator. c3*# UH nUSLv M g7fr"r1)rulimits r%rw/UsageLimits.has_token_limits..s    s)rrr r r rs r%has_token_limitsUsageLimits.has_token_limitssF ''((''33    r'cURnUbURU:a[SU35eURnURb,X0R:a[SURSU<S35eUR nUR b,X@R :a[SUR SU<S35eURnUb;URb-XPR:a[SURS U<S35eggg) z[Raises a `UsageLimitExceeded` exception if the next request would exceed any of the limits.Nz3The next request would exceed the request_limit of z8The next request would exceed the input_tokens_limit of  (input_tokens=rz8The next request would exceed the total_tokens_limit of  (total_tokens=z2The next request would exceed the `cost_limit` of z (`cost`=) rrrrrrr rir)r,usagerrrris r%check_before_request UsageLimits.check_before_requests%**  $=)H$'Z[hZi%jk k))  " " . &r'cURnURb,X R:a[SURSU<S35eURnURb,X0R:a[SURSU<S35eUR nUR b-X@R :a[SUR SU<S35egg) zURaises a `UsageLimitExceeded` exception if the usage exceeds any of the token limits.Nz#Exceeded the input_tokens_limit of rrz$Exceeded the output_tokens_limit of z (output_tokens=z#Exceeded the total_tokens_limit of r)rrrrr rr )r,rrrrs r% check_tokensUsageLimits.check_tokens6s))  " " . r'r1r)rrr4r)rrrrr4r)r&rr4r)r*r_r4r)r5r6r7r8rrrrrrr r r r rrrrr#r'r+rdataclasses_no_defaults_reprrr9r1r'r%rrs"&J%* "M:">#'j'M%) )<&**?%) )R15"J5&).-  $0OS vx  22Hr'r)r$r3r4r) rOr:rPrrQrrRzfrozenset[str]rSrr4zdict[str, Any])rrrrr4r)0 __future__r _annotationsrrrrdecimalr functoolsrtypingrr r genai_prices.data_snapshotr pydanticr r rrr pydantic_corerrr _genai_pricesr _warningsr exceptionsr__all__rrrr{r&r)rYr:rrrrrr1r'r%r;sl2!''3\\73** 3!*NO+L!Mn `akb..(( # # 4# (# # # ' ##L Ee,zhzh-zhz Ee,@$9@$-@$F Ee,L yL -L ^5 ?& t$]3]3%]3r'