How to bulk-generate Google Docs documents

Question

I have a requirement to modify an existing .NET-based document generator, which today creates Word documents or HTML files. The App is moving into the cloud, and so have a full install of MS Office on every image is too complicated. So I'd like to generate GoogleDocs format documents instead, because now there's a great API which lets me do that. Except I can't find a way through the mass of options.

Plan A So far, I've managed to:

• Create a new (empty) document
• Re-open that document, and batch-up changes to it using the 'bachUpdate' API.

( the googledocs .net examples from https://developers.google.com/docs/api/quickstart/dotnet are helpful to get past all the authentication magic)

But my documents are usually 10s, sometimes 100s of pages long, so going back to the server, even with batches of updates, seems an unnecessary performance overhead.

Plan B For best performance, and to keep most of the existing generator code, I tried creating the whole document in the JSON style which Google requires, just as a .JSON file. I tried the example file from https://developers.google.com/docs/api/samples/output-json so I'm sure it's in a valid format.

So then I tried:

• changing the file type to gdoc then..
• copying it to a google drive - Google won't let me
• uploading it to GoogleDrive: Google didn't like that either
• I even tried uploading the HTML-format file from the exisitng generator, but it loses lots of the clever formatting which Word, and GoogleDocs, allow

So I need a way to upload a JSON file so that that Google understands its a GoogleDocs document.

How to download / upload the JSON representation of a Google doc? comes close, but doesn't quite show me the 'upload the JSON document' bit. The Google examples https://developers.google.com/docs/api/samples also only show small number of updates, not the single-file, large scale upload I need.

Am I missing something here? I seem to be so close to a working solution. Or is there a Plan C ?

Answer 1

It is not possible unfortunately

Although like you, I assumed it would be possible, upon further investigation and experiments I was re-reading the docs to see if I had missed anything and I saw this from the Docs API create page:

So unfortunately you cannot simply build a JSON representation of the document and upload it in one go. All that the API will listen to, is the title.

For example, using this code:

static void createDocFromJSON(DocsService service)
        {
            string text = File.ReadAllText("doc.json");
            Document doc = JsonSerializer.Deserialize<Document>(text);
            DocumentsResource.CreateRequest request = 
                                    service.Documents.Create(doc);
            Document createdDoc = request.Execute();
            Console.WriteLine(createdDoc.DocumentId);
        }

With this json:

{
    "Title": "Test Document",
    "Body": {
      "Content": [
        {
          "Paragraph": {
            "Elements": [
              {
                "TextRun": {
                  "Content": "Hello\n"
                }
              }
            ]
          }
        },
        {
          "Paragraph": {
            "Elements": [
              {
                "TextRun": {
                  "Content": "This is a document created from dotnet\n"
                }
              }
            ]
          }
        }
      ]
    }
  }

Will only produce a blank document with the title

:(

---

Use the Drive API

Yet there is hope:

Instead of generating your documents in JSON format I might suggest doing it in HTML. This would arguably be easier than creating the content > paragraph > element > text run > content type of structure that you would hypothetically need for the JSON to work. With HTML you can use h1 h2 p ul tags and more and the API will convert it to a google document format. To do this you just upload HTML file and set the MimeType as application/vnd.google-apps.document.

Adapted from the dot net library page:

https://developers.google.com/api-client-library/dotnet/guide/media_upload

static Google.Apis.Drive.v3.Data.File
        uploadFile(DriveService service, string fileToUpload)
{
    Google.Apis.Drive.v3.Data.File body = 
        new Google.Apis.Drive.v3.Data.File();
    body.Name = System.IO.Path.GetFileName(fileToUpload);
    // Set what the end mime type will be
    body.MimeType = "application/vnd.google-apps.document";
    byte[] byteArray = System.IO.File.ReadAllBytes(fileToUpload);
    
    System.IO.MemoryStream stream = 
        new System.IO.MemoryStream(byteArray);
    FilesResource.CreateMediaUpload request = 
        service.Files.Create(
            body,
            stream,
            "text/html"); // the upload format
    request.Upload();
    return request.ResponseBody;
}

which I called from the main function like:

uploadFile(service, "test.html");

Where test.html was a file in the project folder:

<h1>Hello World</h1>
<p>Document created with dotnet</p>

Which resulted in:

For reference, here is the repository for the dotnet library:

https://github.com/googleapis/google-api-dotnet-client

Answer 2

It is possible:

step 1:

use DocumentService.Create to create a new blank document. This Api call will generate a new document with a DocumentId.

step 2:

perform a BatchUpdateDocumentRequest using the documentId [1] using the InsertTextRequest [2]. This call will insert the new text into the document.

step 3:

format the document with additional BatchUpdateDocumentRequests.

[1] https://pkg.go.dev/google.golang.org/api/docs/v1#BatchUpdateDocumentRequest
[2] https://pkg.go.dev/google.golang.org/api/docs/v1#InsertTextRequest

Please look at https://github.com/prr123/gdoc/blob/master/CvtTxtToGdoc.go to see an example.