Sharepoint as datasource in Azure AI Search, error at reading .docx

Hi One_employee,

• This problem occurs because by default, the raw file material (/document/content) obtained from Sharepoint is in binary format - not plain text. When working with file types such as .docx, .pdf, or images, Azure AI search requires the discovery and processing material using skills, especially underlying text skills such as cognitive skills.

Without a skill, the indexed draws only in binary material, which resembles the guard or base 64-encoded data. This is why you are looking at the until the delayed material in the /document /material area.

• To properly act and find .docx files, you need to define and attach a skill that includes. This will process binary material and extract readable texts that can be indexed under a searching area (eg, text).
• Create a skill with extract text skills.In the definition of your index, make sure you are doing mapping /document /content for the input of this skill.Map the skill output (eg, text) for a field in your index that is marked as searched. https://learn.microsoft.com/en-us/azure/search/search-howto-index-sharepoint-online
  https://learn.microsoft.com/en-us/azure/search/cognitive-search-defining-skillset if you have any further concerns or queries, please feel free to reach out to us.

Thanks for your answer but that's what i did. I tried with with skillset :

{
    "name": "large3-embedder",
    "description": "Skillset to chunk documents and generate embeddings",
    "skills": [
      {
        "@odata.type": "#Microsoft.Skills.Text.SplitSkill",
        "name": "#1",
        "description": "Split skill to chunk documents",
        "context": "/document",
        "defaultLanguageCode": "fr",
        "textSplitMode": "pages",
        "maximumPageLength": 10000,
        "pageOverlapLength": 700,
        "maximumPagesToTake": 0,
        "unit": "characters",
        "inputs": [
          {
            "name": "text",
            "source": "/document/content",
            "inputs": []
          }
        ],
        "outputs": [
          {
            "name": "textItems",
            "targetName": "pages"
          }
        ]
      },
      {
        "@odata.type": "#Microsoft.Skills.Text.AzureOpenAIEmbeddingSkill",
        "name": "#2",
        "context": "/document/pages/*",
        "resourceUri": "MyURI",
        "apiKey": "MyKey",
        "deploymentId": "text-embedding-3-large",
        "dimensions": 3072,
        "modelName": "text-embedding-3-large",
        "inputs": [
          {
            "name": "text",
            "source": "/document/pages/*",
            "inputs": []
          }
        ],
        "outputs": [
          {
            "name": "embedding",
            "targetName": "text_vector"
          }
        ]
      }
    ],
    "indexProjections": {
      "selectors": [
        {
          "targetIndexName": "new-index",
          "parentKeyFieldName": "parent_id",
          "sourceContext": "/document/pages/*",
          "mappings": [
            {
              "name": "text_vector",
              "source": "/document/pages/*/text_vector",
              "inputs": []
            },
            {
              "name": "chunk",
              "source": "/document/pages/*",
              "inputs": []
            },
            {
              "name": "title",
              "source": "/document/title",
              "sourceContext": null,
              "inputs": []
          },
          {
            "name": "metadata_spo_item_path",
            "source": "/document/metadata_spo_item_path",
            "sourceContext": null,
            "inputs": []
        },{
          "name": "content",
          "source": "/document/content",
          "sourceContext": null,
          "inputs": []
      },
        {
          "name": "metadata_spo_item_name",
          "source": "/document/metadata_spo_item_name",
          "sourceContext": null,
          "inputs": []
      }
          
          ]
          
          
        }
      ], 
      "parameters": {
        "projectionMode": "skipIndexingParentDocuments"
    }
    }
  }

It does not work.

I also tried with a custom skill using a Azure function (in order to read the Word binary with python-docx) instead of the TextSplitSkill : it does not work, because the binary content in /document/content is corrupted such as I've shown you the screnshot (It's a screenshot from my Azure function log).
Here is the code of my AZure function :

app = func.FunctionApp(http_auth_level=func.AuthLevel.FUNCTION)

@app.route(route="process_documents_docx", methods=[func.HttpMethod.POST])

def process_documents_docx(req: func.HttpRequest) -> func.HttpResponse:
    
    logging.info('Python HTTP trigger function processed a request.')

    # Récupération des données de l'input
    req_body = req.get_json()
    values = req_body.get('values')    
 
    # Instanciation d'une variable dans laquelle sera bâtie la réponse
    response_values = []
    
    # Traîtement des fichiers un par un
    for record in values:
        logging.info(f"values type : {type(record)}")
        logging.info(f"values  : {record}")
        
        # Récupération du record ID
        record_id = record.get("recordId")
        
        # Récupération de l'URI du fichier
        input_data = record.get("data").get("content")

        doc = Document(BytesIO(input_data)

Returns the error :
Result: Failure Exception: BadZipFile: File is not a zip file Stack: File "/azure-functions-host/workers/python/3.12/LINUX/X64/azure_functions_worker/dispatcher.py", line 676, in _handle__invocation_request call_result = await self._loop.run_in_executor( ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/opt/python/3.12/lib/python3.12/concurrent/futures/thread.py", line 58, in run result = self.fn(*self.args, **self.kwargs) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/azure-functions-host/workers/python/3.12/LINUX/X64/azure_functions_worker/dispatcher.py", line 1006, in _run_sync_func return ExtensionManager.get_sync_invocation_wrapper(context, ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/azure-functions-host/workers/python/3.12/LINUX/X64/azure_functions_worker/extension.py", line 211, in _raw_invocation_wrapper result = function(**args) ^^^^^^^^^^^^^^^^ File "/home/site/wwwroot/function_app.py", line 37, in process_documents_docx doc = Document(BytesIO(input_data)) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/home/site/wwwroot/.python_packages/lib/site-packages/docx/api.py", line 27, in Document document_part = cast("DocumentPart", Package.open(docx).main_document_part) ^^^^^^^^^^^^^^^^^^ File "/home/site/wwwroot/.python_packages/lib/site-packages/docx/opc/package.py", line 127, in open pkg_reader = PackageReader.from_file(pkg_file) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/home/site/wwwroot/.python_packages/lib/site-packages/docx/opc/pkgreader.py", line 22, in from_file phys_reader = PhysPkgReader(pkg_file) ^^^^^^^^^^^^^^^^^^^^^^^ File "/home/site/wwwroot/.python_packages/lib/site-packages/docx/opc/phys_pkg.py", line 76, in init self._zipf = ZipFile(pkg_file, "r") ^^^^^^^^^^^^^^^^^^^^^^ File "/opt/python/3.12/lib/python3.12/zipfile/init.py", line 1341, in init self._RealGetContents() File "/opt/python/3.12/lib/python3.12/zipfile/init.py", line 1408, in _RealGetContents raise BadZipFile("File is not a zip file")

This is why I can say it is corrrupted, because of "BadZipFile"