jorddlZddlZddlZddlZddlZddlmZmZmZm Z m Z m Z m Z m Z ddlmZddlmZddlmZddlmZmZmZmZddlmZddlm Z ddl!m"Z"ddl#m$Z$d d l%m&Z&d d l%m'Z'm(Z(m)Z)m*Z*ejVd k\rddlm,Z,nddl-m,Z,e rddl.Z/ej`e1Z2GddejfZ4Gdde4Z5Gdde4Z6y)N)OptionalAnyListDictSequenceSet TYPE_CHECKINGIterator) rest_helpers)client)ApiError)dataset_item_writedataset_version_publicevaluator_item_writeexecution_policy_write)sequence_splitter) id_helpers) constants) dataset_item convertersrest_operationsexecution_policy) )overridec eZdZdZej d deedeedeee dee de e jf dZ dd Zde fd Z ddeedee deee effd Zej deej(fd Zy)DatasetExportOperationsz Abstract base class providing export operations for dataset items. This class defines the common interface for exporting dataset items, shared by both Dataset (current state) and DatasetVersion (specific version). N nb_samples batch_sizedataset_item_ids filter_stringreturnct)a Stream dataset items as DatasetItem objects. Args: nb_samples: Maximum number of items to retrieve. batch_size: Maximum number of items to fetch per batch. dataset_item_ids: Optional list of specific item IDs to retrieve. filter_string: Optional OQL filter string to filter dataset items. Yields: DatasetItem objects one at a time. NotImplementedErrorselfr r!r"r#s y/Users/manta/Documents/Projects/TheRoad-I1/backend/.venv/lib/python3.12/site-packages/opik/api_objects/dataset/dataset.py-__internal_api__stream_items_as_dataclasses__zEDatasetExportOperations.__internal_api__stream_items_as_dataclasses__4s ("!cbt|j}tj|iS)z Convert the dataset items to a pandas DataFrame. Requires the `pandas` library to be installed. Returns: A pandas DataFrame containing all items.  keys_mapping)listr+r to_pandasr) dataset_itemss r*r1z!DatasetExportOperations.to_pandasJs*TOOQR ##MCCr,cbt|j}tj|iS)z Convert the dataset items to a JSON string. Returns: A JSON string representation of all items. r.)r0r+rto_jsonr2s r*r5zDatasetExportOperations.to_jsonVs*TOOQR !!-bAAr,c|j||Dcgc] }d|ji|j"}}|Scc}w)aB Retrieve dataset items as a list of dictionaries. Args: nb_samples: Maximum number of items to retrieve. If not set, all items are returned. filter_string: Optional OQL filter string to filter dataset items. Supports filtering by tags, data fields, metadata, etc. Supported columns include: - `id`, `source`, `trace_id`, `span_id`: String fields - `data`: Dictionary field (use dot notation, e.g., "data.category") - `tags`: List field (use "contains" operator) - `created_at`, `last_updated_at`: DateTime fields (ISO 8601 format) - `created_by`, `last_updated_by`: String fields Examples: - `tags contains "failed"` - Items with 'failed' tag - `data.category = "test"` - Items with specific data field value - `created_at >= "2024-01-01T00:00:00Z"` - Items created after date Returns: A list of dictionaries representing the dataset items. )r r#id)r+r7 get_content)r)r r#itemdataset_items_as_dictss r* get_itemsz!DatasetExportOperations.get_items`sb<JJ%]K" 477 1d..0 1 " &% " s%?ct)a^ Get version information for experiment association. Returns: DatasetVersionPublic containing version metadata (id, version_name, etc.). For Dataset, returns info about the current/latest version, or None if no version exists. For DatasetVersion, returns info about this specific version. r&r)s r*get_version_infoz(DatasetExportOperations.get_version_infos "!r,NNNN)r$ pd.DataFrameNN)__name__ __module__ __qualname____doc__abcabstractmethodrintrstrr r DatasetItemr+r1r5rrr;rDatasetVersionPublicr>r,r*rr,s %)$(04'+ "SM"SM"#49- "  } " ,** + ""* DBB%)'+"&SM"& }"& d38n  "&H  " (== > " "r,rc eZdZdZdededej dejddf dZ e defd Z e defd Z e defd Z e defd Ze deefd Ze deefdZe deefdZe deefdZe deeefdZe deefdZe deefdZe deefdZe deefdZe deefdZe deefdZe deej<fdZe deefdZ e! d"deedeedeeedeede"e#jHf dZ%e!deejfdZ&dee'fd Z(de)jTfd!Z+y)#DatasetVersionas A read-only view of a specific dataset version. This class provides access to dataset items at a specific version point in time. It supports reading version metadata and retrieving items, but does not allow mutations to the dataset. This object should not be created directly. Use :meth:`Dataset.get_dataset_version` to obtain an instance. dataset_name dataset_id rest_client version_infor$Nc<||_||_||_||_yN) _dataset_name _dataset_id _rest_client _version_info)r)rOrPrQrRs r*__init__zDatasetVersion.__init__s$*%')r,c|jS)z0The name of the dataset this version belongs to.rUr=s r*rOzDatasetVersion.dataset_name!!!r,c|jS)zIThe name of the dataset this version belongs to (alias for dataset_name).r[r=s r*namezDatasetVersion.namer\r,c|jS)z=The unique identifier of the dataset this version belongs to.rVr=s r*rPzDatasetVersion.dataset_idr,c|jS)zTThe unique identifier of the dataset this version belongs to (alias for dataset_id).r`r=s r*r7zDatasetVersion.idrar,c.|jjS)z/The unique identifier of this specific version.)rXr7r=s r* version_idzDatasetVersion.version_ids!!$$$r,c.|jjS)z>Total number of items in this version (alias for items_total).rX items_totalr=s r*dataset_items_countz"DatasetVersion.dataset_items_count!!---r,c.|jjS)z+The unique hash identifier of this version.)rX version_hashr=s r*rkzDatasetVersion.version_hash!!...r,c.|jjS)z/The sequential version name (e.g., 'v1', 'v2').)rX version_namer=s r*rnzDatasetVersion.version_namerlr,c.|jjS)z"Tags associated with this version.)rXtagsr=s r*rpzDatasetVersion.tagss!!&&&r,c.|jjS)z2Whether this is the latest version of the dataset.)rX is_latestr=s r*rrzDatasetVersion.is_latests!!+++r,c.|jjS)z&Total number of items in this version.rfr=s r*rgzDatasetVersion.items_totalrir,c.|jjS)z1Number of items added since the previous version.)rX items_addedr=s r*ruzDatasetVersion.items_addedrir,c.|jjS)z4Number of items modified since the previous version.)rXitems_modifiedr=s r*rwzDatasetVersion.items_modifieds!!000r,c.|jjS)z3Number of items deleted since the previous version.)rX items_deletedr=s r*ryzDatasetVersion.items_deleteds!!///r,c.|jjS)z'Description of changes in this version.)rXchange_descriptionr=s r*r{z!DatasetVersion.change_descriptions!!444r,c.|jjS)z(Timestamp when this version was created.)rX created_atr=s r*r}zDatasetVersion.created_at!!,,,r,c.|jjS)zUser who created this version.)rX created_byr=s r*rzDatasetVersion.created_byr~r,r r!r"r#c tj|j|j|||||jj S)NrQrOr r!r"r#dataset_version)rstream_dataset_itemsrWrUrXrkr(s r*r+zzDatasetVersion.get_version_infos!!!r,cgS)z Get suite-level evaluators for this dataset version. DatasetVersion does not support suite-level evaluators, so this always returns an empty list. Returns: Empty list. rLr=s r*get_evaluatorszDatasetVersion.get_evaluatorss  r,c>tjjS)z Get the execution policy for this dataset version. DatasetVersion does not support suite-level execution policy, so this returns the default execution policy. Returns: Default execution policy. )rDEFAULT_EXECUTION_POLICYcopyr=s r*get_execution_policyz#DatasetVersion.get_execution_policy*s 88==??r,r?),rBrCrDrErIrest_api_clientOpikApirrKrYpropertyrOr^rPr7rrdrHrhrkrnrrpboolrrrgrurwryr{datetimer}rrr rrJr+r>rrrExecutionPolicyrrLr,r*rNrNs  * * *%,, * -AA *  *"c"""c"" C   C  %HSM%%.Xc]../hsm///hsm//'htCy)'',8D>,,.Xc]...Xc]..1 110x}005HSM55-HX%6%67---HSM--%)$(04'+  SM SM #49-   }  ,** +   " " (== > " " S  @&6&F&F @r,rNc eZdZ d0dedeedej deeddf dZe jdefdZ e defd Z e deefd Ze deefd Zdeefd Zedeej&fd Z d0deedeefdZdej2fdZdeefdZdej:dej>fdZ deej>deddfdZ!deej:ddfdZ"de#e$eefddfdZ%d1dZ&dee$eefddfdZ'deededdfdZ(deeddfdZ)d1dZ*e d2d eed!eed"eeed#eede+ej:f d$Z, d3d%ed&ee$eefd'eeeddfd(Z- d3d)ed&ee$eefd'eeeddfd*Z. d3d+d,d&ee$eefd'eeeddfd-Z/d.ede0fd/Z1y)4DatasetNr^ descriptionrQrhr$ch||_||_||_||_i|_t |_y)z A Dataset object. This object should not be created directly, instead use :meth:`opik.Opik.create_dataset` or :meth:`opik.Opik.get_dataset`. N)_name _descriptionrW_dataset_items_count _id_to_hashset_hashes)r)r^rrQrhs r*rYzDataset.__init__8s4 ''$7!+-!$ r,cv|jjj|jjS)zThe id of the datasetrO)rWdatasetsget_dataset_by_identifierrr7r=s r*r7z Dataset.idJs5  ))CCD " r,c|jS)zThe name of the dataset.)rr=s r*r^z Dataset.nameQszzr,c|jS)zThe description of the dataset.)rr=s r*rzDataset.descriptionVs   r,c|jA|jjj|j}|j |_|jS)z The total number of items in the dataset. If the count is not cached locally, it will be fetched from the backend. r)rrWrrrrh)r) dataset_infos r*rhzDataset.dataset_items_count[sU  $ $ ,,,55OO!ZZPL)5(H(HD %(((r,cB|j}|r |jSdS)aG Get the current version name of the dataset. The version name is fetched from the backend and reflects the latest committed version after any mutation operations (insert, update, delete). Returns: The current version name (e.g., 'v1', 'v2'), or None if no version exists. N)r>rn)r)rRs r*get_current_version_namez Dataset.get_current_version_nameis%,,. ,8|((BdBr,c d} |jjj|jdd}|r |jsy|jdS#t$r0}|j dk(rt jdnYd}~Rd}~wwxYw)z Get version information for the current (latest) dataset version. Returns: DatasetVersionPublic containing the current version's metadata, or None if no version exists yet. Nr)r7pagesizeizFVersioning is not enabled for datasets get version info returning Noner) rWrlist_dataset_versionsr7r status_codeLOGGERdebugcontent)r)versions_responsees r*r>zDataset.get_version_infovs!  $ 1 1 : : P P77!Q! !(9(A(A ((++ }}# \  s2A B &BB evaluator_modelcddlm}ddlm}|j }| |j sgSg}|j D]} |j dk(rM|jd i|j}|jj|d|i}|j|n tjd|j |S#t$r$tjd|jd wxYw) aV Get suite-level evaluators from the current dataset version. Converts EvaluatorItemPublic objects from the BE into LLMJudge instances. Args: evaluator_model: Optional model name to use for LLMJudge evaluators. Returns: List of LLMJudge instances extracted from the version. r) llm_judge)configrmodel) init_kwargszIUnsupported evaluator type in version: %s. Only 'llm_judge' is supported.z7Failed to instantiate evaluator from version config: %sT)exc_inforL) opik.evaluation.suite_evaluatorsr*opik.evaluation.suite_evaluators.llm_judgerr> evaluatorstypeLLMJudgeConfigLLMJudge from_configappendrwarning Exceptionerror) r)rrllm_judge_configrRrevaluator_itemcfg evaluators r*rzDataset.get_evaluatorss ? ,,.  |'>'>I " *55N !&&+59*99RN >'?)C!?!I%%i0NNc&++6*  M"))!   s Arrrrr)r)rReps r*rzDataset.get_execution_policys,,.  # (E(E(Q..B##/"$!1!1$$0#%"3"3      88==??r,c|jjj|j}|jxsgS)zc Get the tags for this dataset. Returns: List of tag strings. r)rWrrrrp)r) dataset_ferns r*get_tagszDataset.get_tagss@((11KKL   &B&r,r9c d}|jrM|jDcgc]8}tj|j|j|j :}}d}|j r?tj|j j|j j}tj|j|j|j|j |j#|j$||Scc}w)zConvert a DatasetItem to REST API format. Args: item: The DatasetItem to convert. Returns: DatasetItemWrite object ready for REST API. N)r^rrr)r7trace_idspan_idsourcedatarrr)rrest_evaluator_itemEvaluatorItemWriter^rrrrest_execution_policyExecutionPolicyWriterrrest_dataset_itemDatasetItemWriter7rrrr8r)r)r9rrrs r*_convert_to_rest_itemzDataset._convert_to_rest_items ??  )A $6688 )    4II"33AA#44CC   !11ww]]LL;;!!#((!-  !s=Dbatchbatch_group_idc|tjfdtjdt y)auInsert a batch of dataset items with automatic retry on rate limit errors. Args: batch: List of dataset items to insert. batch_group_id: UUIDv7 identifier that groups all batches from a single user operation together. All batches sent as part of one insert/update call share the same batch_group_id. chjjjjS)N)rOitemsr)rWrcreate_or_update_dataset_itemsrrrr)sr*z2Dataset._insert_batch_with_retry..s-D%%..MM!ZZu^Nr,z0Successfully sent dataset items batch of size %dNr *ensure_rest_api_call_respecting_rate_limitrrlenr)rrs```r*_insert_batch_with_retryz Dataset._insert_batch_with_retrys- ??   GUTr,rc`g}|D]|}|j}||jvrtjd|8|j ||jj |||j |j<~|Dcgc]}|j|}}tj|tjtj}tj }|D]4}tjdt#||j%||6d|_ycc}w)Nz6Duplicate item found with hash: %s - ignored the event)max_payload_size_MB max_lengthz&Sending dataset items batch of size %dr) content_hashrrrraddrr7rrsplit_into_batchesrMAX_BATCH_SIZE_MBrDATASET_ITEMS_MAX_BATCH_SIZEr generate_idrrr) r)rdeduplicated_itemsr9 item_hash rest_itemsbatchesrrs r*-__internal_api__insert_items_as_dataclasses__z5Dataset.__internal_api__insert_items_as_dataclasses__"s>@D))+IDLL( L  % %d + LL  Y '(1D  TWW %DVVCU4d006CU V#66  & 8 8 ==  $//1E LLA3u: N  ) )% ) O %)!WsD+c|Dcgc])}t|trtjdi|n|+}}|j |ycc}w)z Insert new items into the dataset. A new dataset version will be created. Args: items: List of dicts (which will be converted to dataset items) to add to the dataset. NrL) isinstancedictrrJr)r)rr9r3s r*insertzDataset.insertFsU9 2.s'D%%..CC~Dr,z3Successfully deleted dataset items batch of size %dNrrs```r*_delete_batch_with_retryz Dataset._delete_batch_with_retryts. ??   JCPUJWr, items_idsc~tj|tj}t j }|D]x}t jd||j|||D]H}||jvs|j|}|jj||j|=Jzd|_ y)z Delete items from the dataset. A new dataset version will be created. Args: items_ids: List of item ids to delete. )rz Deleting dataset items batch: %srN) rrrrrrrrrrrdiscardr)r)rrrritem_idhashs r*deletezDataset.deletes$66 )"H"H $//1E LL;U C  ) )% ) O d...++G4DLL((.((1 ! %)!r,c|jDcgc]}|j |j}}|j|ycc}w)za Delete all items from the given dataset. A new dataset version will be created. N)r+r7r )r)r9rs r*clearz Dataset.clearsN JJL Lww" GGL  H  sAr r!r"r#c btj|j|j||||dS)a Stream dataset items as a generator instead of loading all at once. This method yields dataset items one at a time, enabling evaluation to start processing items before the entire dataset is downloaded. This is particularly useful for large datasets with heavy payloads (images, videos, audio). Args: nb_samples: Maximum number of items to retrieve. If None, all items are streamed. batch_size: Maximum number of items to fetch per batch from the backend. If None, uses the default value from constants.DATASET_STREAM_BATCH_SIZE. dataset_item_ids: Optional list of specific item IDs to retrieve. If provided, only items with matching IDs will be yielded. filter_string: Optional OQL filter string to filter dataset items. Yields: DatasetItem objects one at a time Nr)rrrWrr(s r*r+z5Dataset.__internal_api__stream_items_as_dataclasses__s6433))!!-'   r, json_arrayr/ ignore_keyscn|in|}|gn|}tj|||}|j|y)a Args: json_array: json string of format: "[{...}, {...}, {...}]" where every dictionary is to be transformed into dataset item keys_mapping: dictionary that maps json keys to item fields names Example: {'Expected output': 'expected_output'} ignore_keys: if your json dicts contain keys that are not needed for DatasetItem construction - pass them as ignore_keys argument N)r/r)r from_jsonr)r)rr/r new_itemss r*insert_from_jsonzDataset.insert_from_jsonsA*1r| '/b[ (( \{  Ir, file_pathcl|in|}|gn|}tj|||}|j|y)a Read JSONL from a file and insert it into the dataset. Args: file_path: Path to the JSONL file keys_mapping: dictionary that maps json keys to item fields names Example: {'Expected output': 'expected_output'} ignore_keys: if your json dicts contain keys that are not needed for DatasetItem construction - pass them as ignore_keys argument N)rfrom_jsonl_filer)r)rr/rrs r*read_jsonl_from_filezDataset.read_jsonl_from_files< *1r| '/b[ ..y, T  Ir, dataframer@cl|in|}|gn|}tj|||}|j|y)a Requires: `pandas` library to be installed. Args: dataframe: pandas dataframe keys_mapping: Dictionary that maps dataframe column names to dataset item field names. Example: {'Expected output': 'expected_output'} ignore_keys: if your dataframe contains columns that are not needed for DatasetItem construction - pass them as ignore_keys argument N)r from_pandasr)r)rr/rrs r*insert_from_pandaszDataset.insert_from_pandass< *1r| '/b[ **9lKP  Ir,rnctj|j|j|}|&t j d|d|j dt|j |j|j|S)a Get a read-only view of a specific dataset version. The returned DatasetVersion object allows reading version metadata and retrieving items via :meth:`DatasetVersion.get_items`, but does not support mutations. Args: version_name: The version name (e.g., 'v1', 'v2'). Returns: A read-only DatasetVersion object for accessing the specified version. Raises: opik.exceptions.DatasetVersionNotFound: If the specified version does not exist. Example: >>> dataset = client.get_dataset("my_dataset") >>> version = dataset.get_version_view("v1") >>> items = version.get_items() )rQrPrnzDataset version 'z' not found in dataset '')rOrPrQrR)rfind_version_by_namerWr7rDatasetVersionNotFoundrrN)r)rnrRs r*get_version_viewzDataset.get_version_views,';;))ww%  33#L>1I$**UVW ww))%   r,rT)r$Nr?rA)2rBrCrDrIrrrrHrY functoolscached_propertyr7rr^rrhrrrrKr>rrrrrrrrrJrrrrrrrrrrrr r r r+rrrrNr rLr,r*rr7s .2 ''c]'%,, ' &c] '  '$ C  c!Xc]!! )Xc] ) ) C(3- C, (== >,,<*..!#. c.`@  ) )@. '$s) '&  ,,&  + +& PU%667UU  U("),223") ")H JHT#s(^4 J J 4Dc3h0T$XCyXX  X()S )d)4 %)$(04'+ ! SM! SM! #49- !  } ! ,** + ! ! L26+/ tCH~.d3i(   626+/ tCH~.d3i(   026+/ !tCH~.d3i(   .& S& ^& r,r)7rFrloggingr!systypingrrrrrrr r opik.api_objectsr opik.rest_apir ropik.rest_api.core.api_errorr opik.rest_api.typesrrrrrrr opik.message_processing.batchingropikropik.exceptionsr opik.configrrrrrrrRrtyping_extensionspandaspd getLoggerrBrABCrrNrrLr,r*r4s     *31 ?$IIw*   8 $d"cggd"Na@,a@HE %E r,