spark-nlp 6.0.5__py2.py3-none-any.whl → 6.1.1__py2.py3-none-any.whl

sparknlp/common/properties.py

@@ -628,7 +628,6 @@ class HasGeneratorProperties:
                              "The number of sequences to return from the beam search.",
                              typeConverter=TypeConverters.toInt)
 
-
     def setTask(self, value):
         """Sets the transformer's task, e.g. ``summarize:``.
 
@@ -639,7 +638,6 @@ class HasGeneratorProperties:
         """
         return self._set(task=value)
 
-
     def setMinOutputLength(self, value):
         """Sets minimum length of the sequence to be generated.
 
@@ -650,7 +648,6 @@ class HasGeneratorProperties:
         """
         return self._set(minOutputLength=value)
 
-
     def setMaxOutputLength(self, value):
         """Sets maximum length of output text.
 
@@ -661,7 +658,6 @@ class HasGeneratorProperties:
         """
         return self._set(maxOutputLength=value)
 
-
     def setDoSample(self, value):
         """Sets whether or not to use sampling, use greedy decoding otherwise.
 
@@ -672,7 +668,6 @@ class HasGeneratorProperties:
         """
         return self._set(doSample=value)
 
-
     def setTemperature(self, value):
         """Sets the value used to module the next token probabilities.
 
@@ -683,7 +678,6 @@ class HasGeneratorProperties:
         """
         return self._set(temperature=value)
 
-
     def setTopK(self, value):
         """Sets the number of highest probability vocabulary tokens to keep for
         top-k-filtering.
@@ -695,7 +689,6 @@ class HasGeneratorProperties:
         """
         return self._set(topK=value)
 
-
     def setTopP(self, value):
         """Sets the top cumulative probability for vocabulary tokens.
 
@@ -709,7 +702,6 @@ class HasGeneratorProperties:
         """
         return self._set(topP=value)
 
-
     def setRepetitionPenalty(self, value):
         """Sets the parameter for repetition penalty. 1.0 means no penalty.
 
@@ -725,7 +717,6 @@ class HasGeneratorProperties:
         """
         return self._set(repetitionPenalty=value)
 
-
     def setNoRepeatNgramSize(self, value):
         """Sets size of n-grams that can only occur once.
 
@@ -738,7 +729,6 @@ class HasGeneratorProperties:
         """
         return self._set(noRepeatNgramSize=value)
 
-
     def setBeamSize(self, value):
         """Sets the number of beam size for beam search.
 
@@ -749,7 +739,6 @@ class HasGeneratorProperties:
         """
         return self._set(beamSize=value)
 
-
     def setNReturnSequences(self, value):
         """Sets the number of sequences to return from the beam search.
 
@@ -765,14 +754,14 @@ class HasLlamaCppProperties:
     # -------- MODEl PARAMETERS --------
     nThreads = Param(Params._dummy(), "nThreads", "Set the number of threads to use during generation",
                      typeConverter=TypeConverters.toInt)
-    nThreadsDraft = Param(Params._dummy(), "nThreadsDraft", "Set the number of threads to use during draft generation",
-                          typeConverter=TypeConverters.toInt)
+    # nThreadsDraft = Param(Params._dummy(), "nThreadsDraft", "Set the number of threads to use during draft generation",
+    #                       typeConverter=TypeConverters.toInt)
     nThreadsBatch = Param(Params._dummy(), "nThreadsBatch",
                           "Set the number of threads to use during batch and prompt processing",
                           typeConverter=TypeConverters.toInt)
-    nThreadsBatchDraft = Param(Params._dummy(), "nThreadsBatchDraft",
-                               "Set the number of threads to use during batch and prompt processing",
-                               typeConverter=TypeConverters.toInt)
+    # nThreadsBatchDraft = Param(Params._dummy(), "nThreadsBatchDraft",
+    #                            "Set the number of threads to use during batch and prompt processing",
+    #                            typeConverter=TypeConverters.toInt)
     nCtx = Param(Params._dummy(), "nCtx", "Set the size of the prompt context", typeConverter=TypeConverters.toInt)
     nBatch = Param(Params._dummy(), "nBatch",
                    "Set the logical batch size for prompt processing (must be >=32 to use BLAS)",
@@ -782,12 +771,12 @@ class HasLlamaCppProperties:
                     typeConverter=TypeConverters.toInt)
     nDraft = Param(Params._dummy(), "nDraft", "Set the number of tokens to draft for speculative decoding",
                    typeConverter=TypeConverters.toInt)
-    nChunks = Param(Params._dummy(), "nChunks", "Set the maximal number of chunks to process",
-                    typeConverter=TypeConverters.toInt)
-    nSequences = Param(Params._dummy(), "nSequences", "Set the number of sequences to decode",
-                       typeConverter=TypeConverters.toInt)
-    pSplit = Param(Params._dummy(), "pSplit", "Set the speculative decoding split probability",
-                   typeConverter=TypeConverters.toFloat)
+    # nChunks = Param(Params._dummy(), "nChunks", "Set the maximal number of chunks to process",
+    #                 typeConverter=TypeConverters.toInt)
+    # nSequences = Param(Params._dummy(), "nSequences", "Set the number of sequences to decode",
+    #                    typeConverter=TypeConverters.toInt)
+    # pSplit = Param(Params._dummy(), "pSplit", "Set the speculative decoding split probability",
+    #                typeConverter=TypeConverters.toFloat)
     nGpuLayers = Param(Params._dummy(), "nGpuLayers", "Set the number of layers to store in VRAM (-1 - use default)",
                        typeConverter=TypeConverters.toInt)
     nGpuLayersDraft = Param(Params._dummy(), "nGpuLayersDraft",
@@ -802,10 +791,10 @@ class HasLlamaCppProperties:
                          typeConverter=TypeConverters.toString)
     mainGpu = Param(Params._dummy(), "mainGpu", "Set the main GPU that is used for scratch and small tensors.",
                     typeConverter=TypeConverters.toInt)
-    tensorSplit = Param(Params._dummy(), "tensorSplit", "Set how split tensors should be distributed across GPUs",
-                        typeConverter=TypeConverters.toListFloat)
-    grpAttnN = Param(Params._dummy(), "grpAttnN", "Set the group-attention factor", typeConverter=TypeConverters.toInt)
-    grpAttnW = Param(Params._dummy(), "grpAttnW", "Set the group-attention width", typeConverter=TypeConverters.toInt)
+    # tensorSplit = Param(Params._dummy(), "tensorSplit", "Set how split tensors should be distributed across GPUs",
+    #                     typeConverter=TypeConverters.toListFloat)
+    # grpAttnN = Param(Params._dummy(), "grpAttnN", "Set the group-attention factor", typeConverter=TypeConverters.toInt)
+    # grpAttnW = Param(Params._dummy(), "grpAttnW", "Set the group-attention width", typeConverter=TypeConverters.toInt)
     ropeFreqBase = Param(Params._dummy(), "ropeFreqBase", "Set the RoPE base frequency, used by NTK-aware scaling",
                          typeConverter=TypeConverters.toFloat)
     ropeFreqScale = Param(Params._dummy(), "ropeFreqScale",
@@ -837,7 +826,7 @@ class HasLlamaCppProperties:
                          typeConverter=TypeConverters.toString)
     # Set the RoPE frequency scaling method, defaults to linear unless specified by the model.
     #
-    #   - UNSPECIFIED: Don't use any scaling
+    #   - NONE: Don't use any scaling
     #   - LINEAR: Linear scaling
     #   - YARN: YaRN RoPE scaling
     ropeScalingType = Param(Params._dummy(), "ropeScalingType",
@@ -845,29 +834,30 @@ class HasLlamaCppProperties:
                             typeConverter=TypeConverters.toString)
     # Set the pooling type for embeddings, use model default if unspecified
     #
-    #   - 0 NONE: Don't use any pooling
-    #   - 1 MEAN: Mean Pooling
-    #   - 2 CLS: CLS Pooling
+    #   - MEAN: Mean Pooling
+    #   - CLS: CLS Pooling
+    #   - LAST: Last token pooling
+    #   - RANK: For reranked models
     poolingType = Param(Params._dummy(), "poolingType",
                         "Set the pooling type for embeddings, use model default if unspecified",
                         typeConverter=TypeConverters.toString)
     modelDraft = Param(Params._dummy(), "modelDraft", "Set the draft model for speculative decoding",
                        typeConverter=TypeConverters.toString)
     modelAlias = Param(Params._dummy(), "modelAlias", "Set a model alias", typeConverter=TypeConverters.toString)
-    lookupCacheStaticFilePath = Param(Params._dummy(), "lookupCacheStaticFilePath",
-                                      "Set path to static lookup cache to use for lookup decoding (not updated by generation)",
-                                      typeConverter=TypeConverters.toString)
-    lookupCacheDynamicFilePath = Param(Params._dummy(), "lookupCacheDynamicFilePath",
-                                       "Set path to dynamic lookup cache to use for lookup decoding (updated by generation)",
-                                       typeConverter=TypeConverters.toString)
+    # lookupCacheStaticFilePath = Param(Params._dummy(), "lookupCacheStaticFilePath",
+    #                                   "Set path to static lookup cache to use for lookup decoding (not updated by generation)",
+    #                                   typeConverter=TypeConverters.toString)
+    # lookupCacheDynamicFilePath = Param(Params._dummy(), "lookupCacheDynamicFilePath",
+    #                                    "Set path to dynamic lookup cache to use for lookup decoding (updated by generation)",
+    #                                    typeConverter=TypeConverters.toString)
     # loraAdapters = new StructFeature[Map[String, Float]](this, "loraAdapters")
     embedding = Param(Params._dummy(), "embedding", "Whether to load model with embedding support",
                       typeConverter=TypeConverters.toBoolean)
     flashAttention = Param(Params._dummy(), "flashAttention", "Whether to enable Flash Attention",
                            typeConverter=TypeConverters.toBoolean)
-    inputPrefixBos = Param(Params._dummy(), "inputPrefixBos",
-                           "Whether to add prefix BOS to user inputs, preceding the `--in-prefix` string",
-                           typeConverter=TypeConverters.toBoolean)
+    # inputPrefixBos = Param(Params._dummy(), "inputPrefixBos",
+    #                        "Whether to add prefix BOS to user inputs, preceding the `--in-prefix` string",
+    #                        typeConverter=TypeConverters.toBoolean)
     useMmap = Param(Params._dummy(), "useMmap",
                     "Whether to use memory-map model (faster load but may increase pageouts if not using mlock)",
                     typeConverter=TypeConverters.toBoolean)
@@ -880,6 +870,10 @@ class HasLlamaCppProperties:
                          typeConverter=TypeConverters.toString)
     chatTemplate = Param(Params._dummy(), "chatTemplate", "The chat template to use",
                          typeConverter=TypeConverters.toString)
+    logVerbosity = Param(Params._dummy(), "logVerbosity", "Set the log verbosity level",
+                         typeConverter=TypeConverters.toInt)
+    disableLog = Param(Params._dummy(), "disableLog", "Whether to disable logging",
+                       typeConverter=TypeConverters.toBoolean)
 
     # -------- INFERENCE PARAMETERS --------
     inputPrefix = Param(Params._dummy(), "inputPrefix", "Set the prompt to start generation with",
@@ -948,17 +942,17 @@ class HasLlamaCppProperties:
         """Set the number of threads to use during generation"""
         return self._set(nThreads=nThreads)
 
-    def setNThreadsDraft(self, nThreadsDraft: int):
-        """Set the number of threads to use during draft generation"""
-        return self._set(nThreadsDraft=nThreadsDraft)
+    # def setNThreadsDraft(self, nThreadsDraft: int):
+    #     """Set the number of threads to use during draft generation"""
+    #     return self._set(nThreadsDraft=nThreadsDraft)
 
     def setNThreadsBatch(self, nThreadsBatch: int):
         """Set the number of threads to use during batch and prompt processing"""
         return self._set(nThreadsBatch=nThreadsBatch)
 
-    def setNThreadsBatchDraft(self, nThreadsBatchDraft: int):
-        """Set the number of threads to use during batch and prompt processing"""
-        return self._set(nThreadsBatchDraft=nThreadsBatchDraft)
+    # def setNThreadsBatchDraft(self, nThreadsBatchDraft: int):
+    #     """Set the number of threads to use during batch and prompt processing"""
+    #     return self._set(nThreadsBatchDraft=nThreadsBatchDraft)
 
     def setNCtx(self, nCtx: int):
         """Set the size of the prompt context"""
@@ -976,17 +970,17 @@ class HasLlamaCppProperties:
         """Set the number of tokens to draft for speculative decoding"""
         return self._set(nDraft=nDraft)
 
-    def setNChunks(self, nChunks: int):
-        """Set the maximal number of chunks to process"""
-        return self._set(nChunks=nChunks)
+    # def setNChunks(self, nChunks: int):
+    #     """Set the maximal number of chunks to process"""
+    #     return self._set(nChunks=nChunks)
 
-    def setNSequences(self, nSequences: int):
-        """Set the number of sequences to decode"""
-        return self._set(nSequences=nSequences)
+    # def setNSequences(self, nSequences: int):
+    #     """Set the number of sequences to decode"""
+    #     return self._set(nSequences=nSequences)
 
-    def setPSplit(self, pSplit: float):
-        """Set the speculative decoding split probability"""
-        return self._set(pSplit=pSplit)
+    # def setPSplit(self, pSplit: float):
+    #     """Set the speculative decoding split probability"""
+    #     return self._set(pSplit=pSplit)
 
     def setNGpuLayers(self, nGpuLayers: int):
         """Set the number of layers to store in VRAM (-1 - use default)"""
@@ -1004,17 +998,17 @@ class HasLlamaCppProperties:
         """Set the main GPU that is used for scratch and small tensors."""
         return self._set(mainGpu=mainGpu)
 
-    def setTensorSplit(self, tensorSplit: List[float]):
-        """Set how split tensors should be distributed across GPUs"""
-        return self._set(tensorSplit=tensorSplit)
+    # def setTensorSplit(self, tensorSplit: List[float]):
+    #     """Set how split tensors should be distributed across GPUs"""
+    #     return self._set(tensorSplit=tensorSplit)
 
-    def setGrpAttnN(self, grpAttnN: int):
-        """Set the group-attention factor"""
-        return self._set(grpAttnN=grpAttnN)
+    # def setGrpAttnN(self, grpAttnN: int):
+    #     """Set the group-attention factor"""
+    #     return self._set(grpAttnN=grpAttnN)
 
-    def setGrpAttnW(self, grpAttnW: int):
-        """Set the group-attention width"""
-        return self._set(grpAttnW=grpAttnW)
+    # def setGrpAttnW(self, grpAttnW: int):
+    #     """Set the group-attention width"""
+    #     return self._set(grpAttnW=grpAttnW)
 
     def setRopeFreqBase(self, ropeFreqBase: float):
         """Set the RoPE base frequency, used by NTK-aware scaling"""
@@ -1049,7 +1043,16 @@ class HasLlamaCppProperties:
         return self._set(defragmentationThreshold=defragmentationThreshold)
 
     def setNumaStrategy(self, numaStrategy: str):
-        """Set optimization strategies that help on some NUMA systems (if available)"""
+        """Set optimization strategies that help on some NUMA systems (if available)
+
+        Possible values:
+
+        - DISABLED: No NUMA optimizations
+        - DISTRIBUTE: spread execution evenly over all
+        - ISOLATE: only spawn threads on CPUs on the node that execution started on
+        - NUMA_CTL: use the CPU map provided by numactl
+        - MIRROR: Mirrors the model across NUMA nodes
+        """
         numaUpper = numaStrategy.upper()
         numaStrategies = ["DISABLED", "DISTRIBUTE", "ISOLATE", "NUMA_CTL", "MIRROR"]
         if numaUpper not in numaStrategies:
@@ -1060,13 +1063,35 @@ class HasLlamaCppProperties:
         return self._set(numaStrategy=numaStrategy)
 
     def setRopeScalingType(self, ropeScalingType: str):
-        """Set the RoPE frequency scaling method, defaults to linear unless specified by the model"""
-        return self._set(ropeScalingType=ropeScalingType)
+        """Set the RoPE frequency scaling method, defaults to linear unless specified by the model.
+
+        Possible values:
+
+        - NONE: Don't use any scaling
+        - LINEAR: Linear scaling
+        - YARN: YaRN RoPE scaling
+        """
+        ropeScalingTypeUpper = ropeScalingType.upper()
+        ropeScalingTypes = ["NONE", "LINEAR", "YARN"]
+        if ropeScalingTypeUpper not in ropeScalingTypes:
+            raise ValueError(
+                f"Invalid RoPE scaling type: {ropeScalingType}. "
+                + f"Valid values are: {ropeScalingTypes}"
+            )
+        return self._set(ropeScalingType=ropeScalingTypeUpper)
 
     def setPoolingType(self, poolingType: str):
-        """Set the pooling type for embeddings, use model default if unspecified"""
+        """Set the pooling type for embeddings, use model default if unspecified
+
+        Possible values:
+
+        - MEAN: Mean Pooling
+        - CLS: CLS Pooling
+        - LAST: Last token pooling
+        - RANK: For reranked models
+        """
         poolingTypeUpper = poolingType.upper()
-        poolingTypes = ["NONE", "MEAN", "CLS", "LAST"]
+        poolingTypes = ["NONE", "MEAN", "CLS", "LAST", "RANK"]
         if poolingTypeUpper not in poolingTypes:
             raise ValueError(
                 f"Invalid pooling type: {poolingType}. "
@@ -1082,25 +1107,21 @@ class HasLlamaCppProperties:
         """Set a model alias"""
         return self._set(modelAlias=modelAlias)
 
-    def setLookupCacheStaticFilePath(self, lookupCacheStaticFilePath: str):
-        """Set path to static lookup cache to use for lookup decoding (not updated by generation)"""
-        return self._set(lookupCacheStaticFilePath=lookupCacheStaticFilePath)
+    # def setLookupCacheStaticFilePath(self, lookupCacheStaticFilePath: str):
+    #     """Set path to static lookup cache to use for lookup decoding (not updated by generation)"""
+    #     return self._set(lookupCacheStaticFilePath=lookupCacheStaticFilePath)
 
-    def setLookupCacheDynamicFilePath(self, lookupCacheDynamicFilePath: str):
-        """Set path to dynamic lookup cache to use for lookup decoding (updated by generation)"""
-        return self._set(lookupCacheDynamicFilePath=lookupCacheDynamicFilePath)
-
-    def setEmbedding(self, embedding: bool):
-        """Whether to load model with embedding support"""
-        return self._set(embedding=embedding)
+    # def setLookupCacheDynamicFilePath(self, lookupCacheDynamicFilePath: str):
+    #     """Set path to dynamic lookup cache to use for lookup decoding (updated by generation)"""
+    #     return self._set(lookupCacheDynamicFilePath=lookupCacheDynamicFilePath)
 
     def setFlashAttention(self, flashAttention: bool):
         """Whether to enable Flash Attention"""
         return self._set(flashAttention=flashAttention)
 
-    def setInputPrefixBos(self, inputPrefixBos: bool):
-        """Whether to add prefix BOS to user inputs, preceding the `--in-prefix` string"""
-        return self._set(inputPrefixBos=inputPrefixBos)
+    # def setInputPrefixBos(self, inputPrefixBos: bool):
+    #     """Whether to add prefix BOS to user inputs, preceding the `--in-prefix` string"""
+    #     return self._set(inputPrefixBos=inputPrefixBos)
 
     def setUseMmap(self, useMmap: bool):
         """Whether to use memory-map model (faster load but may increase pageouts if not using mlock)"""
@@ -1246,11 +1267,19 @@ class HasLlamaCppProperties:
     def setUseChatTemplate(self, useChatTemplate: bool):
         """Set whether generate should apply a chat template"""
         return self._set(useChatTemplate=useChatTemplate)
-    
+
     def setNParallel(self, nParallel: int):
         """Sets the number of parallel processes for decoding. This is an alias for `setBatchSize`."""
         return self.setBatchSize(nParallel)
 
+    def setLogVerbosity(self, logVerbosity: int):
+        """Set the log verbosity level"""
+        return self._set(logVerbosity=logVerbosity)
+
+    def setDisableLog(self, disableLog: bool):
+        """Whether to disable logging"""
+        return self._set(disableLog=disableLog)
+
     # -------- JAVA SETTERS --------
     def setTokenIdBias(self, tokenIdBias: Dict[int, float]):
         """Set token id bias"""
@@ -1260,9 +1289,9 @@ class HasLlamaCppProperties:
         """Set token id bias"""
         return self._call_java("setTokenBias", tokenBias)
 
-    def setLoraAdapters(self, loraAdapters: Dict[str, float]):
-        """Set LoRA adapters with their scaling factors"""
-        return self._call_java("setLoraAdapters", loraAdapters)
+    # def setLoraAdapters(self, loraAdapters: Dict[str, float]):
+    #     """Set LoRA adapters with their scaling factors"""
+    #     return self._call_java("setLoraAdapters", loraAdapters)
 
     def getMetadata(self):
         """Gets the metadata of the model"""

sparknlp/internal/__init__.py

@@ -1182,4 +1182,13 @@ class _E5VEmbeddingsLoader(ExtendedJavaWrapper):
             path,
             jspark,
             use_openvino
+        )
+
+class _Phi4Loader(ExtendedJavaWrapper):
+    def __init__(self, path, jspark, use_openvino=False):
+        super(_Phi4Loader, self).__init__(
+            "com.johnsnowlabs.nlp.annotators.seq2seq.Phi4Transformer.loadSavedModel",
+            path,
+            jspark,
+            use_openvino,
         )

sparknlp/reader/reader2doc.py

@@ -0,0 +1,210 @@
+#  Copyright 2017-2025 John Snow Labs
+#
+#  Licensed under the Apache License, Version 2.0 (the "License");
+#  you may not use this file except in compliance with the License.
+#  You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+#  Unless required by applicable law or agreed to in writing, software
+#  distributed under the License is distributed on an "AS IS" BASIS,
+#  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+#  See the License for the specific language governing permissions and
+#  limitations under the License.
+from pyspark import keyword_only
+from pyspark.ml.param import TypeConverters, Params, Param
+
+from sparknlp.common import AnnotatorType
+from sparknlp.internal import AnnotatorTransformer
+from sparknlp.partition.partition_properties import *
+
+
+class Reader2Doc(
+    AnnotatorTransformer,
+    HasEmailReaderProperties,
+    HasExcelReaderProperties,
+    HasHTMLReaderProperties,
+    HasPowerPointProperties,
+    HasTextReaderProperties
+):
+    """
+    The Reader2Doc annotator allows you to use reading files more smoothly within existing
+    Spark NLP workflows, enabling seamless reuse of your pipelines.
+
+    Reader2Doc can be used for extracting structured content from various document types
+    using Spark NLP readers. It supports reading from many file types and returns parsed
+    output as a structured Spark DataFrame.
+
+    Supported formats include:
+
+    - Plain text
+    - HTML
+    - Word (.doc/.docx)
+    - Excel (.xls/.xlsx)
+    - PowerPoint (.ppt/.pptx)
+    - Email files (.eml, .msg)
+    - PDFs
+
+    Examples
+    --------
+    >>> from johnsnowlabs.reader import Reader2Doc
+    >>> from johnsnowlabs.nlp.base import DocumentAssembler
+    >>> from pyspark.ml import Pipeline
+    >>> # Initialize Reader2Doc for PDF files
+    >>> reader2doc = Reader2Doc() \\
+    ...     .setContentType("application/pdf") \\
+    ...     .setContentPath(f"{pdf_directory}/")
+    >>> # Build the pipeline with the Reader2Doc stage
+    >>> pipeline = Pipeline(stages=[reader2doc])
+    >>> # Fit the pipeline to an empty DataFrame
+    >>> pipeline_model = pipeline.fit(empty_data_set)
+    >>> result_df = pipeline_model.transform(empty_data_set)
+    >>> # Show the resulting DataFrame
+    >>> result_df.show()
+    +------------------------------------------------------------------------------------------------------------------------------------+
+    |document                                                                                                                            |
+    +------------------------------------------------------------------------------------------------------------------------------------+
+    |[{'document', 0, 14, 'This is a Title', {'pageNumber': 1, 'elementType': 'Title', 'fileName': 'pdf-title.pdf'}, []}]               |
+    |[{'document', 15, 38, 'This is a narrative text', {'pageNumber': 1, 'elementType': 'NarrativeText', 'fileName': 'pdf-title.pdf'}, []}]|
+    |[{'document', 39, 68, 'This is another narrative text', {'pageNumber': 1, 'elementType': 'NarrativeText', 'fileName': 'pdf-title.pdf'}, []}]|
+    +------------------------------------------------------------------------------------------------------------------------------------+
+"""
+
+    name = "Reader2Doc"
+    outputAnnotatorType = AnnotatorType.DOCUMENT
+
+    contentPath = Param(
+        Params._dummy(),
+        "contentPath",
+        "contentPath path to files to read",
+        typeConverter=TypeConverters.toString
+    )
+
+    outputCol = Param(
+        Params._dummy(),
+        "outputCol",
+        "output column name",
+        typeConverter=TypeConverters.toString
+    )
+
+    contentType = Param(
+        Params._dummy(),
+        "contentType",
+        "Set the content type to load following MIME specification",
+        typeConverter=TypeConverters.toString
+    )
+
+    explodeDocs = Param(
+        Params._dummy(),
+        "explodeDocs",
+        "whether to explode the documents into separate rows",
+        typeConverter=TypeConverters.toBoolean
+    )
+
+    flattenOutput = Param(
+        Params._dummy(),
+        "flattenOutput",
+        "If true, output is flattened to plain text with minimal metadata",
+        typeConverter=TypeConverters.toBoolean
+    )
+
+    titleThreshold = Param(
+        Params._dummy(),
+        "titleThreshold",
+        "Minimum font size threshold for title detection in PDF docs",
+        typeConverter=TypeConverters.toFloat
+    )
+
+    outputFormat = Param(
+        Params._dummy(),
+        "outputFormat",
+        "Output format for the table content. Options are 'plain-text' or 'html-table'. Default is 'json-table'.",
+        typeConverter=TypeConverters.toString
+    )
+
+    @keyword_only
+    def __init__(self):
+        super(Reader2Doc, self).__init__(classname="com.johnsnowlabs.reader.Reader2Doc")
+        self._setDefault(
+            outputCol="document",
+            explodeDocs=False,
+            contentType="",
+            flattenOutput=False,
+            titleThreshold=18
+        )
+    @keyword_only
+    def setParams(self):
+        kwargs = self._input_kwargs
+        return self._set(**kwargs)
+
+    def setContentPath(self, value):
+        """Sets content path.
+
+        Parameters
+        ----------
+        value : str
+            contentPath path to files to read
+        """
+        return self._set(contentPath=value)
+
+    def setContentType(self, value):
+        """
+        Set the content type to load following MIME specification
+
+        Parameters
+        ----------
+        value : str
+            content type to load following MIME specification
+        """
+        return self._set(contentType=value)
+
+    def setExplodeDocs(self, value):
+        """Sets whether to explode the documents into separate rows.
+
+
+        Parameters
+        ----------
+        value : boolean
+        Whether to explode the documents into separate rows
+        """
+        return self._set(explodeDocs=value)
+
+    def setOutputCol(self, value):
+        """Sets output column name.
+
+        Parameters
+        ----------
+        value : str
+            Name of the Output Column
+        """
+        return self._set(outputCol=value)
+
+    def setFlattenOutput(self, value):
+        """Sets whether to flatten the output to plain text with minimal metadata.
+
+        Parameters
+        ----------
+        value : bool
+            If true, output is flattened to plain text with minimal metadata
+        """
+        return self._set(flattenOutput=value)
+
+    def setTitleThreshold(self, value):
+        """Sets the minimum font size threshold for title detection in PDF documents.
+
+        Parameters
+        ----------
+        value : float
+            Minimum font size threshold for title detection in PDF docs
+        """
+        return self._set(titleThreshold=value)
+
+    def setOutputFormat(self, value):
+        """Sets the output format for the table content.
+
+        Parameters
+        ----------
+        value : str
+            Output format for the table content. Options are 'plain-text' or 'html-table'. Default is 'json-table'.
+        """
+        return self._set(outputFormat=value)