Edit Word Files With GPT

import azure.functions as func
import logging
import json
import base64
import io
import zipfile

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


@app.route(route="extract_main_word_xml")
def extract_main_word_xml(req: func.HttpRequest) -> func.HttpResponse:
    
    try:
        # Parse request JSON
        req_body = req.get_json()
        file_content_base64 = req_body.get("file_content")["$content"]
        
        if not file_content_base64:
            return func.HttpResponse("Missing 'file_content' in request.", status_code=400)
        
        # Decode base64 content
        file_bytes = base64.b64decode(file_content_base64)
        
        # Load Word file as ZIP
        zip_buffer = io.BytesIO(file_bytes)
        extracted_data = {}
        
        with zipfile.ZipFile(zip_buffer, 'r') as docx_zip:
            for file_name in docx_zip.namelist():
                # Filter to relevant XML files
                if (file_name.startswith("word/document") or file_name=="word/comments.xml" or file_name.startswith("word/_rels/document") or file_name.startswith("word/header") or file_name.startswith("word/footer") or file_name.startswith("word/endnotes") or file_name.startswith("word/footnotes") or file_name.startswith("word/numbering")) and (file_name.endswith(".xml") or file_name.endswith(".rels")):
                    with docx_zip.open(file_name) as file:
                        extracted_data[file_name] = file.read().decode("utf-8")
        
        return func.HttpResponse(json.dumps(extracted_data, indent=2), mimetype="application/json")
    
    except Exception as e:
        logging.error(f"Error processing file: {e}")
        return func.HttpResponse(f"Error: {str(e)}", status_code=500)
    


@app.route(route="update_word_xml")
def update_word_xml(req: func.HttpRequest) -> func.HttpResponse:
    
    try:
        # Parse request JSON
        req_body = req.get_json()
        file_content_base64 = req_body.get("file_content", {}).get("$content")
        updated_xml_data = req_body.get("updated_xml")
        
        if not file_content_base64 or not updated_xml_data:
            return func.HttpResponse("Missing required inputs.", status_code=400)
        
        # Decode base64 original Word file
        file_bytes = base64.b64decode(file_content_base64)
        zip_buffer = io.BytesIO(file_bytes)
        output_buffer = io.BytesIO()
        
        with zipfile.ZipFile(zip_buffer, 'r') as docx_zip:
            with zipfile.ZipFile(output_buffer, 'w', zipfile.ZIP_DEFLATED) as new_docx_zip:
                for file_name in docx_zip.namelist():
                    with docx_zip.open(file_name) as file:
                        # Replace XML content if it's updated
                        if file_name in updated_xml_data:
                            new_content = updated_xml_data[file_name].encode("utf-8")
                        else:
                            new_content = file.read()
                        new_docx_zip.writestr(file_name, new_content)
                        
                # Add new XML files that are in updated_xml_data but not in the original document
                for new_file_name, new_content in updated_xml_data.items():
                    if new_file_name not in docx_zip.namelist():
                        new_docx_zip.writestr(new_file_name, new_content.encode("utf-8"))
        
        # Get base64-encoded updated Word file
        output_buffer.seek(0)
        base64_encoded_docx = base64.b64encode(output_buffer.getvalue()).decode("utf-8")
        
        response_data = {
            "$content-type": "application/vnd.openxmlformats-officedocument.wordprocessingml.document",
            "$content": base64_encoded_docx
        }
        
        return func.HttpResponse(json.dumps(response_data), mimetype="application/json")
    
    except Exception as e:
        logging.error(f"Error updating file: {e}")
        return func.HttpResponse(f"Error: {str(e)}", status_code=500)

I want to edit a Word (.docx) file. I have un-zipped the file into its component xml parts & extracted some of the xml files, like word/document.xml, to a JSON object where the key labels are the xml file paths/names & the values are strings of the xml content from each file. The JSON is provided below between [Start JSON] & [End JSON]. Please edit the xml string value(s) in the JSON to accomplish my requested changes.
Requested changes:
-Translate all existing text & new text in the document to spanish.
-Change any links directed to www.google.com to direct to www.bing.com. (Changing links requires changing them in the word/document.xml.rels parameter/component.)
-Redact personally identifying information like names.
-Add a summary of the document's contents & seeming purpose to the end of the document. Do not duplicate the summary text. And apply yellow highlighting to that summary text. (Highlighting likely requires adding some xml like "<w:rPr><w:highlight w:val="yellow"/></w:rPr>" before the text xml).


[Start JSON]
@{body('HTTP_Extract_main_Word_xml')}
[End JSON]

Please output a JSON object that maintains key parameter labels from the above JSON for existing parameters like "word/document.xml", but now with updated xml strings. And if the xml string for a given parameter/component file doesn't have any updates, then you can exclude that parameter from the output JSON. For parameters/components with updates. do not abbreviate the xml string & do not provide only xml snippets with changes, I need the whole xml strings for the Word file components that have changes.