U ~fhd(@sUdZddlmZddlZddlZddlZddlZddlZddlZddl m Z m Z m Z m Z mZmZmZmZmZmZmZddlmZmZmZmZddlmZddlmZmZddlm Z m!Z!m"Z"m#Z#m$Z$dd l%m&Z&dd l'm(Z(dd l)m*Z*dd l+m,Z,dd l-m.Z.ddl/m0Z0ddl1m2Z2ddl3m4Z4ddl5m6Z6ddl7m8Z8ej9ej:ej;ejdZ?GdddZ@GdddZAe reeeBe fZCneZCdZDGdddeCZEeEeAjFdZGdeHd<eEeAjIdZJdeHd<eEeAjKdZLdeHd<eLZMdeHd<d d d d!d"d#d$ZNd%d d d d&d'd(ZOeMfd dd d)d*d+ZPeMfd,dd d-d.d/ZQeMfd0dd d1d2d3ZRd d d d4d5d6ZSd dd7d8d9d:ZTd d;dd7d<d=d>ZUd dd7d8d?d@ZVd dd7d8dAdBZWd ddCd8dDdEZXd d dFd4dGdHZYd d d!d4dIdJZZd d dKd4dLdMZ[d d dNd4dOdPZ\d d d d4dQdRZ]d d d d4dSdTZ^d d d;d4dUdVZ_d d dWd4dXdYZ`d d dZd4d[d\Zad d d]d4d^d_Zbd d d`d4dadbZcd d dcd4dddeZdd dd7d8dfdgZed d dhd4didjZfeYe]eXeSecedeee[eTdkdle`efebe^e\eZe_eadmZgdneHdo<ehegZidpd;dd d<dqdrZjd ddsd)dtduZkdKddsd)dvdwZldWdd d)dxdyZmd d d dzd{d|Znd ddsd)d}d~Zod;dd d)ddZpdZdd d)ddZqdddsd)ddZrdpddsd)ddZsdddsd)ddZtdddsd)ddZudFd dsdzddZvdhd dsdzddZwdhd dsdzddZxdddsd)ddZyd d dsdddZzd d dsdddZ{e|ene}esejere"eke~eqeepeBenedenejeueete*emeele&eye,e{e.eze0eve2eoe4eoe6ewe(exiZdeHd<iZdeHd<eD] Zeedreeeej<qeddeDZeMfd dd d)ddZd d;dddZdd;dddZdd;dddZdd;dddZe0dede*de(de6de.de,diZdeHd<eBee}eejee2ee&eiZdeHd<dd d;d;d;dddZd d;ddddZd d;ddddZdS)aOTools for using Python's :mod:`json` module with BSON documents. This module provides two helper methods `dumps` and `loads` that wrap the native :mod:`json` methods and provide explicit BSON conversion to and from JSON. :class:`~bson.json_util.JSONOptions` provides a way to control how JSON is emitted and parsed, with the default being the Relaxed Extended JSON format. :mod:`~bson.json_util` can also generate Canonical or legacy `Extended JSON`_ when :const:`CANONICAL_JSON_OPTIONS` or :const:`LEGACY_JSON_OPTIONS` is provided, respectively. .. _Extended JSON: https://github.com/mongodb/specifications/blob/master/source/extended-json.rst Example usage (deserialization): .. doctest:: >>> from bson.json_util import loads >>> loads( ... '[{"foo": [1, 2]}, {"bar": {"hello": "world"}}, {"code": {"$scope": {}, "$code": "function x() { return 1; }"}}, {"bin": {"$type": "80", "$binary": "AQIDBA=="}}]' ... ) [{'foo': [1, 2]}, {'bar': {'hello': 'world'}}, {'code': Code('function x() { return 1; }', {})}, {'bin': Binary(b'...', 128)}] Example usage with :const:`RELAXED_JSON_OPTIONS` (the default): .. doctest:: >>> from bson import Binary, Code >>> from bson.json_util import dumps >>> dumps( ... [ ... {"foo": [1, 2]}, ... {"bar": {"hello": "world"}}, ... {"code": Code("function x() { return 1; }")}, ... {"bin": Binary(b"")}, ... ] ... ) '[{"foo": [1, 2]}, {"bar": {"hello": "world"}}, {"code": {"$code": "function x() { return 1; }"}}, {"bin": {"$binary": {"base64": "AQIDBA==", "subType": "00"}}}]' Example usage (with :const:`CANONICAL_JSON_OPTIONS`): .. doctest:: >>> from bson import Binary, Code >>> from bson.json_util import dumps, CANONICAL_JSON_OPTIONS >>> dumps( ... [ ... {"foo": [1, 2]}, ... {"bar": {"hello": "world"}}, ... {"code": Code("function x() { return 1; }")}, ... {"bin": Binary(b"")}, ... ], ... json_options=CANONICAL_JSON_OPTIONS, ... ) '[{"foo": [{"$numberInt": "1"}, {"$numberInt": "2"}]}, {"bar": {"hello": "world"}}, {"code": {"$code": "function x() { return 1; }"}}, {"bin": {"$binary": {"base64": "AQIDBA==", "subType": "00"}}}]' Example usage (with :const:`LEGACY_JSON_OPTIONS`): .. doctest:: >>> from bson import Binary, Code >>> from bson.json_util import dumps, LEGACY_JSON_OPTIONS >>> dumps( ... [ ... {"foo": [1, 2]}, ... {"bar": {"hello": "world"}}, ... {"code": Code("function x() { return 1; }", {})}, ... {"bin": Binary(b"")}, ... ], ... json_options=LEGACY_JSON_OPTIONS, ... ) '[{"foo": [1, 2]}, {"bar": {"hello": "world"}}, {"code": {"$code": "function x() { return 1; }", "$scope": {}}}, {"bin": {"$binary": "AQIDBA==", "$type": "00"}}]' Alternatively, you can manually pass the `default` to :func:`json.dumps`. It won't handle :class:`~bson.binary.Binary` and :class:`~bson.code.Code` instances (as they are extended strings you can't provide custom defaults), but it will be faster as there is less recursion. .. note:: If your application does not need the flexibility offered by :class:`JSONOptions` and spends a large amount of time in the `json_util` module, look to `python-bsonjs `_ for a nice performance improvement. `python-bsonjs` is a fast BSON to MongoDB Extended JSON converter for Python built on top of `libbson `_. `python-bsonjs` works best with PyMongo when using :class:`~bson.raw_bson.RawBSONDocument`. ) annotationsN) TYPE_CHECKINGAnyCallableMappingMutableMappingOptionalSequenceTupleTypeUnioncast)ALL_UUID_SUBTYPES UUID_SUBTYPEBinaryUuidRepresentation)Code) CodecOptionsDatetimeConversion) _MAX_UTC_MS EPOCH_AWARE DatetimeMS_datetime_to_millis_millis_to_datetime)DBRef) Decimal128)Int64)MaxKey)MinKey)ObjectId)Regex)RE_TYPE Timestamp)utc)ilmsuxc@seZdZdZdZdZdS)DatetimeRepresentationrN)__name__ __module__ __qualname__LEGACY NUMBERLONGISO8601r4r42/tmp/pip-unpacked-wheel-36gvocj8/bson/json_util.pyr+s  r+c@seZdZdZdZdZdS)JSONModerr,r-N)r.r/r0r1RELAXED CANONICALr4r4r4r5r6s r6lc seZdZUded<ded<ded<ded<ded<d d d fd d Zd d d ejfdddddd d ddfdd Zddfdd Zddfdd Z d ddddZ Z S) JSONOptionsint json_modeboolstrict_number_longdatetime_representation strict_uuidzType[MutableMapping[str, Any]]document_classr)argskwargscstdS)a Encapsulates JSON options for :func:`dumps` and :func:`loads`. :param strict_number_long: If ``True``, :class:`~bson.int64.Int64` objects are encoded to MongoDB Extended JSON's *Strict mode* type `NumberLong`, ie ``'{"$numberLong": "" }'``. Otherwise they will be encoded as an `int`. Defaults to ``False``. :param datetime_representation: The representation to use when encoding instances of :class:`datetime.datetime`. Defaults to :const:`~DatetimeRepresentation.LEGACY`. :param strict_uuid: If ``True``, :class:`uuid.UUID` object are encoded to MongoDB Extended JSON's *Strict mode* type `Binary`. Otherwise it will be encoded as ``'{"$uuid": "" }'``. Defaults to ``False``. :param json_mode: The :class:`JSONMode` to use when encoding BSON types to Extended JSON. Defaults to :const:`~JSONMode.LEGACY`. :param document_class: BSON documents returned by :func:`loads` will be decoded to an instance of this class. Must be a subclass of :class:`collections.MutableMapping`. Defaults to :class:`dict`. :param uuid_representation: The :class:`~bson.binary.UuidRepresentation` to use when encoding and decoding instances of :class:`uuid.UUID`. Defaults to :const:`~bson.binary.UuidRepresentation.UNSPECIFIED`. :param tz_aware: If ``True``, MongoDB Extended JSON's *Strict mode* type `Date` will be decoded to timezone aware instances of :class:`datetime.datetime`. Otherwise they will be naive. Defaults to ``False``. :param tzinfo: A :class:`datetime.tzinfo` subclass that specifies the timezone from which :class:`~datetime.datetime` objects should be decoded. Defaults to :const:`~bson.tz_util.utc`. :param datetime_conversion: Specifies how UTC datetimes should be decoded within BSON. Valid options include 'datetime_ms' to return as a DatetimeMS, 'datetime' to return as a datetime.datetime and raising a ValueError for out-of-range values, 'datetime_auto' to return DatetimeMS objects when the underlying datetime is out-of-range and 'datetime_clamp' to clamp to the minimum and maximum possible datetimes. Defaults to 'datetime'. See :ref:`handling-out-of-range-datetimes` for details. :param args: arguments to :class:`~bson.codec_options.CodecOptions` :param kwargs: arguments to :class:`~bson.codec_options.CodecOptions` .. seealso:: The specification for Relaxed and Canonical `Extended JSON`_. .. versionchanged:: 4.0 The default for `json_mode` was changed from :const:`JSONMode.LEGACY` to :const:`JSONMode.RELAXED`. The default for `uuid_representation` was changed from :const:`~bson.binary.UuidRepresentation.PYTHON_LEGACY` to :const:`~bson.binary.UuidRepresentation.UNSPECIFIED`. .. versionchanged:: 3.5 Accepts the optional parameter `json_mode`. .. versionchanged:: 4.0 Changed default value of `tz_aware` to False. N)super__init__)selfrArB __class__r4r5rDs6zJSONOptions.__init__NzType[JSONOptions]zOptional[bool]z Optional[int])clsr=r>r?r;rArBreturncs||dd|d<|dr(|dt|d<|tjtjtjdfkrFtdttt j |f||}|t jt j t j fkr|td||_|jt j kr|rtd|dtjfkrtd|dkrtd d|_tj|_d |_n|jt j kr4|dkrtd |dtjfkr td |dkrtd d |_tj|_d |_nDd|_tj|_d|_|dk rX||_|dk rh||_|dk rx||_|S) Ntz_awareFtzinfoznJSONOptions.datetime_representation must be one of LEGACY, NUMBERLONG, or ISO8601 from DatetimeRepresentation.zQJSONOptions.json_mode must be one of LEGACY, RELAXED, or CANONICAL from JSONMode.zr?)rHr=r>r?r;rArBrErFr4r5rN1sh      zJSONOptions.__new__str)rIcs d|j|j|j|jtS)Nz[strict_number_long={!r}, datetime_representation={!r}, strict_uuid={!r}, json_mode={!r}, {})formatr=r>r?r;rC_arguments_repr)rErFr4r5rQtszJSONOptions._arguments_reprzdict[Any, Any]cs*t}||j|j|j|jd|S)Nr=r>r?r;)rC _options_dictupdater=r>r?r;)rEZ options_dictrFr4r5rSs zJSONOptions._options_dict)rBrIcKs<|}dD]}||t||||<q ||tf|S)a Make a copy of this JSONOptions, overriding some options:: >>> from bson.json_util import CANONICAL_JSON_OPTIONS >>> CANONICAL_JSON_OPTIONS.tz_aware True >>> json_options = CANONICAL_JSON_OPTIONS.with_options(tz_aware=False, tzinfo=None) >>> json_options.tz_aware False .. versionadded:: 3.12 rR)rSrLgetattrrTr9)rErBoptsoptr4r4r5 with_optionss  zJSONOptions.with_options) r.r/r0__annotations__rDr6r7rNrQrSrX __classcell__r4r4rFr5r9s :"C  r9)r;LEGACY_JSON_OPTIONSCANONICAL_JSON_OPTIONSRELAXED_JSON_OPTIONSDEFAULT_JSON_OPTIONSrrO)objrArBrIcOs$|dt}tjt||f||S)aQHelper function that wraps :func:`json.dumps`. Recursive function that handles all BSON types including :class:`~bson.binary.Binary` and :class:`~bson.code.Code`. :param json_options: A :class:`JSONOptions` instance used to modify the encoding of MongoDB Extended JSON types. Defaults to :const:`DEFAULT_JSON_OPTIONS`. .. versionchanged:: 4.0 Now outputs MongoDB Relaxed Extended JSON by default (using :const:`DEFAULT_JSON_OPTIONS`). .. versionchanged:: 3.4 Accepts optional parameter `json_options`. See :class:`JSONOptions`. json_options)popr^jsondumps _json_convert)r_rArBr`r4r4r5rcs rczUnion[str, bytes, bytearray])r(rArBrIcsJ|dtjtkr(fdd|d<nfdd|d<tj|f||S)aHelper function that wraps :func:`json.loads`. Automatically passes the object_hook for BSON type conversion. Raises ``TypeError``, ``ValueError``, ``KeyError``, or :exc:`~bson.errors.InvalidId` on invalid MongoDB Extended JSON. :param json_options: A :class:`JSONOptions` instance used to modify the decoding of MongoDB Extended JSON types. Defaults to :const:`DEFAULT_JSON_OPTIONS`. .. versionchanged:: 4.0 Now loads :class:`datetime.datetime` instances as naive by default. To load timezone aware instances utilize the `json_options` parameter. See :ref:`tz_aware_default_change` for an example. .. versionchanged:: 3.5 Parses Relaxed and Canonical Extended JSON as well as PyMongo's legacy format. Now raises ``TypeError`` or ``ValueError`` when parsing JSON type wrappers with values of the wrong type or any extra keys. .. versionchanged:: 3.4 Accepts optional parameter `json_options`. See :class:`JSONOptions`. r`cs t|SN) object_hookr_r`r4r5zloads..rfcs t|Sre)object_pairs_hook)pairsrhr4r5rirjrk)rar^r@dictrbloads)r(rArBr4rhr5rns   rn)r_r`rIcstt|dr fdd|DSt|drJt|ttfsJfdd|DSz t|WStk rn|YSXdS)z]Recursive helper method that converts BSON types so they can be converted into json. itemscsi|]\}}|t|qSr4rd).0kvrhr4r5 sz!_json_convert..__iter__csg|]}t|qSr4rp)rqrsrhr4r5 sz!_json_convert..N)hasattrro isinstancerObytesdefault TypeErrorr_r`r4rhr5rds  rdzSequence[Tuple[str, Any]])rlr`rIcCst|||Sre)rfr@)rlr`r4r4r5rk srkzMapping[str, Any])dctr`rIcCs4d}|D]}|tkr|}qq|r0t|||S|Sre) _PARSERS_SET_PARSERS)r}r`matchrrr4r4r5rfsrf)docdummy0rIcCsJ|d}t|ttfs|Sd}|ddD]}|t|dO}q*t||S)N$regexr$options)rxrOryrL _RE_OPT_TABLEr )rrpatternflagsrWr4r4r5_parse_legacy_regexsrzUnion[Binary, uuid.UUID])rr`rIcCsht|dkrtd|t|dts6td||jtjkrVtt |dSt |dSdS)z*Decode a JSON legacy $uuid to Python UUID.r,zBad $uuid, extra field(s): $uuidz$uuid must be a string: N) lenr{rxrOuuid_representationr UNSPECIFIEDr from_uuiduuidUUIDrr`r4r4r5_parse_legacy_uuid)s  rr:)datasubtyper`rIcCsn|tkrP|j}t||}|tjkr&|S|tkr6tj}n|tjkrFtj}||S|dkrdt t j |St||SNr) rrrrrrZSTANDARDZ PYTHON_LEGACYZas_uuidr rr)rrr`rZ binary_valuer4r4r5_binary_or_uuid5s     rcCsht|dtrd|d|d<t|dd}|dkrJt|dddd}t|d}t|||S)N$type%02xl$binary)rxr:base64 b64decodeencoder)rr`rrr4r4r5_parse_legacy_binaryJsrcCs|d}|d}|d}t|ts0td|t|trFt|dkrTtd|t|dkrntd|t|}t|t|d|S) NrrsubTypez!$binary base64 must be a string: r-z7$binary subType must be a string at most 2 characters: z=$binary must include only "base64" and "subType" components: r) rxrOr{rrrrrr:)rr`binaryZb64rrr4r4r5_parse_canonical_binaryTs  rz$Union[datetime.datetime, DatetimeMS]c Csd|d}t|dkr"td|t|trP|ddkrL|dd}d}n|ddkr~|d d kr~|dd}|dd}nT|d dkr|dd }|d d}n.|d dkr|dd }|d d}n|}d }|d }d}|dkrtt||dd}|d|}tj|dj |t d}|r|dkrt|dkrt|dd d \}} t|dt| d} nZt|dkrt|dddt|ddd} n"t|dkrt|ddd} |ddkr| d9} |tj | d}|j r*|jr||j}|jtjkr&t|S|S|j dd} |jtjkrLt| S| Stt|td|S)z3Decode a JSON datetime to python datetime.datetime.$dater,zBad $date, extra field(s): ZNi)+-:r.ri@B%Y-%m-%dT%H:%M:%S) microsecondrKri<r)secondsrKzCodecOptions[Any])rr{rxrOrfindr:floatdatetimestrptimereplacer$split timedeltarJrK astimezoneZdatetime_conversionrZ DATETIME_MSrrr ) rr`ZdtmdtoffsetZ dot_indexrZawarehoursminutesZsecsZaware_tzinfo_noner4r4r5_parse_canonical_datetimecsb            *  rrcCs&t|dkrtd|t|dS)z1Decode a JSON ObjectId to bson.objectid.ObjectId.r,zBad $oid, extra field(s): $oid)rr{rrrr4r4r5_parse_canonical_oids rcCs*|d}t|dkr"td|t|S)z&Decode a JSON symbol to Python string.$symbolr,zBad $symbol, extra field(s): )rr{rO)rrsymbolr4r4r5_parse_canonical_symbols rrcCs6|D]}|dkrtd|qt|d|ddS)z%Decode a JSON code to bson.code.Code.$code$scopezBad $code, extra field(s): rr)scope)r{rrL)rrkeyr4r4r5_parse_canonical_codesrz Regex[str]cCsl|d}t|dkr"td|t|dkrr+r3r:r_encode_datetimeZ as_datetimer1rOr|r4r4r5_encode_datetimemsLs   rcCs0|jdkrdt|iSt|t|j|dSdS)Nrr)rrOrdr|r4r4r5 _encode_codeWs  rcCs|jrdt|iSt|SdS)Nr)r=rOr:r|r4r4r5 _encode_int64^s r)r_rrIcCs|Srer4r_rr4r4r5 _encode_noopesrcCsd}|jtj@r|d7}|jtj@r,|d7}|jtj@r@|d7}|jtj@rT|d7}|jtj@rh|d7}|jtj@r||d7}t|j t r|j }n |j d}|j t jkr||d Sd ||d iS) Nrr%r&r'r(r)r*zutf-8)rrr)rr)rre IGNORECASELOCALE MULTILINEDOTALLUNICODEVERBOSErxrrOrr;r6r1)r_r`rrr4r4r5 _encode_regexis&          rcCsB|jtjkr>t |kr"tkr2nn dt|iSdt|iS|S)Nrr)r;r6r8 _INT32_MAXrOr|r4r4r5 _encode_ints    rcCs`|jtjkr\t|rddiSt|r@|dkr4dnd}d|iS|jtjkr\dtt|iS|S)NrNaNrInfinityz -Infinity) r;r6r1mathisnanisinfr8rOrepr)r_r`Zrepresentationr4r4r5 _encode_floats    rzdatetime.datetimecCs|jtjkr|js,|jtd}|jdk s,t|tkr|j|}|j |j |j fdkrZd}n | d}t |jd}|rd|fnd}dd | d ||iSt|}|jtjkrd|iSdd t|iiS) Nr)rrrrz%ziz.%03drrz{}{}{}rr)r>r+r3rKrr$AssertionErrorr utcoffsetdaysr microsecondsstrftimer:rrPrr1rO)r_r`offZ tz_stringZmillisZfracsecsr4r4r5rs$     rcCs t|d|Sr)rr|r4r4r5 _encode_bytessr rcCst||j|Sre)rrr|r4r4r5_encode_binary_objsr z uuid.UUIDcCs2|jr$tj||jd}t||j|Sd|jiSdS)N)rr)r?rrrrrhex)r_r`Zbinvalr4r4r5 _encode_uuidsrcCs dt|iS)NrrOrr4r4r5_encode_objectidsrcCsd|j|jdiS)Nr)rr%)timeincrr4r4r5_encode_timestampsrcCs dt|iS)Nrrrr4r4r5_encode_decimal128srrcCst||dS)Nrh)rdrr|r4r4r5 _encode_dbrefsr)rdummy1rIcCsddiS)Nrr,r4rrr4r4r5_encode_minkeysrcCsddiS)Nrr,r4rr4r4r5_encode_maxkeysrz-dict[Type, Callable[[Any, JSONOptions], Any]] _ENCODERSz,dict[int, Callable[[Any, JSONOptions], Any]]_MARKERS _type_markerccs|] }|VqdSrer4)rqrr4r4r5 srcCsztt|||WStk r(YnXt|dr`|j}|tkr`t|}|tt|<|||StD]0}t||rdt|}|tt|<|||Sqdtd|dS)Nrz%r is not JSON serializable) rrKeyErrorrwrr_BUILT_IN_TYPESrxr{)r_r`markerfuncbaser4r4r5rzs      rz)r_rIcCst|Sre)rrgr4r4r5 _get_str_sizesr#cCsdtt|S)Nr)rrOrrgr4r4r5_get_datetime_sizesr$r cCsdt|jS)N)rrrgr4r4r5_get_regex_size sr&cCsdt|jS)N")rZ collectionrgr4r4r5_get_dbref_size$sr( zdict[Any, int]_CONSTANT_SIZE_TABLEzdict[Any, Callable[[Any], int]]_VARIABLE_SIZE_TABLE)r_max_size current_sizerIcCs,||kr |St|}z t|WStk r2YnXzt||WStk rVYnX|tkr|jr|dt|j||t|t|j7}n|dt|7}n|tkr| D]8\}}|t|||7}|t|||7}||kr|Sqn:t |dr(|D](}|t|||7}||kr|Sq|S)z!Recursively finds size of objectsrru) rr-rr.rrget_sizerrmrorw)r_r/r0Zobj_typerrrsr%r4r4r5r1;s: "    r1zTuple[Any, int])r_ max_lengthrIcCs|dkr dS|}t|dr^i}|D].\}}t||\}}|rH|||<|dkr&qVq&||fSt|drt|ttfsg}|D],}t||\}}|r|||dkr~qq~||fSt||SdS)zMRecursively truncate documents as needed to fit inside max_length characters.rrroruN)rwro_truncate_documentsrxrOryappend _truncate)r_r2 remaining truncatedrrrsZ truncated_vr4r4r5r3ds,  r3)r_r6rIcCsXt||}||kr|||fSz|d|}Wntk rF|}YnX|||fSdSre)r1r{)r_r6sizer7r4r4r5r5s   r5)r)__doc__ __future__rrrrbrrrtypingrrrrrrr r r r r Z bson.binaryrrrrZ bson.coderZbson.codec_optionsrrZbson.datetime_msrrrrrZ bson.dbrefrZbson.decimal128rZ bson.int64rZ bson.max_keyrZ bson.min_keyrZ bson.objectidrZ bson.regexr Zbson.sonr!Zbson.timestampr#Z bson.tz_utilr$ILMSUXrr+r6rOZ _BASE_CLASSrr9r1r[rYr8r\r7r]r^rcrnrdrkrfrrrrrrrrrrrrrrrrrrrrrsetr~rrrrrrrrrr r rrrrrrrr<ryrr:rrrrZ_typrwrtuplerrzr#r$r&r(r-r.r1r3r5r4r4r4r5srW 4            $+1 "    B             )