Imports Google.Cloud.Vision.V1

Public Class GoogleVisionService
    Private ReadOnly _client As ImageAnnotatorClient

    Public Sub New()
        ' Requires GOOGLE_APPLICATION_CREDENTIALS env var
        ' pointing to a service account JSON key file
        _client = ImageAnnotatorClient.Create()
    End Sub

    Public Function DetectText(imagePath As String) As String
        ' WARNING: Image uploaded to Google Cloud
        Dim image = Image.FromFile(imagePath)
        Dim response = _client.DetectText(image)

        If response.Count > 0 Then
            Return response(0).Description
        End If

        Return String.Empty
    End Function
End Class

Imports Google.Cloud.Vision.V1

Public Class GoogleVisionService
    Private ReadOnly _client As ImageAnnotatorClient

    Public Sub New()
        ' Requires GOOGLE_APPLICATION_CREDENTIALS env var
        ' pointing to a service account JSON key file
        _client = ImageAnnotatorClient.Create()
    End Sub

    Public Function DetectText(imagePath As String) As String
        ' WARNING: Image uploaded to Google Cloud
        Dim image = Image.FromFile(imagePath)
        Dim response = _client.DetectText(image)

        If response.Count > 0 Then
            Return response(0).Description
        End If

        Return String.Empty
    End Function
End Class

Public Sub New()
    ' BEFORE: Requires GOOGLE_APPLICATION_CREDENTIALS environment variable
    ' pointing to a service account JSON key file:
    ' export GOOGLE_APPLICATION_CREDENTIALS="/path/to/service-account.json"
    '
    ' The JSON file contains sensitive data:
    ' - private_key: RSA private key
    ' - client_email: service account email
    ' - project_id: GCP project identifier
    '
    ' Security concerns:
    ' - Key file must never be committed to source control
    ' - Key file must be rotated periodically
    ' - Key file must be protected with file system permissions
    ' - Key file compromise grants API access until revoked

    _client = ImageAnnotatorClient.Create()
End Sub

Public Sub New()
    ' BEFORE: Requires GOOGLE_APPLICATION_CREDENTIALS environment variable
    ' pointing to a service account JSON key file:
    ' export GOOGLE_APPLICATION_CREDENTIALS="/path/to/service-account.json"
    '
    ' The JSON file contains sensitive data:
    ' - private_key: RSA private key
    ' - client_email: service account email
    ' - project_id: GCP project identifier
    '
    ' Security concerns:
    ' - Key file must never be committed to source control
    ' - Key file must be rotated periodically
    ' - Key file must be protected with file system permissions
    ' - Key file compromise grants API access until revoked

    _client = ImageAnnotatorClient.Create()
End Sub

Public Sub New()
    ' Simple license key - no key files, no environment variables required
    IronOcr.License.LicenseKey = If(Environment.GetEnvironmentVariable("IRONOCR_LICENSE"), "YOUR-LICENSE-KEY")

    ' なし service accounts, no key rotation, no IAM configuration
    _ocr = New IronTesseract()
End Sub

Public Sub New()
    ' Simple license key - no key files, no environment variables required
    IronOcr.License.LicenseKey = If(Environment.GetEnvironmentVariable("IRONOCR_LICENSE"), "YOUR-LICENSE-KEY")

    ' なし service accounts, no key rotation, no IAM configuration
    _ocr = New IronTesseract()
End Sub

Public Async Function ProcessPdfAsync(ByVal pdfPath As String) As Task(Of String)
    ' Step 1: Create storage client
    Dim storageClient = StorageClient.Create()
    Dim objectName = $"ocr-input/{Guid.NewGuid()}.pdf"

    ' Step 2: Upload PDF to GCS (document leaves your infrastructure)
    Using stream = File.OpenRead(pdfPath)
        Await storageClient.UploadObjectAsync(_bucketName, objectName, "application/pdf", stream)
    End Using

    ' Step 3: Build async annotation request
    Dim asyncRequest = New AsyncAnnotateFileRequest With {
        .InputConfig = New InputConfig With {
            .GcsSource = New GcsSource With {.Uri = $"gs://{_bucketName}/{objectName}"},
            .MimeType = "application/pdf"
        },
        .Features = {New フィーチャー With {.Type = Feature.Types.Type.DocumentTextDetection}},
        .OutputConfig = New OutputConfig With {
            .GcsDestination = New GcsDestination With {.Uri = $"gs://{_bucketName}/ocr-output/"},
            .BatchSize = 1
        }
    }

    ' Step 4: Submit and wait for async operation
    Dim operation = Await _visionClient.AsyncBatchAnnotateFilesAsync({asyncRequest})
    Dim completedOperation = Await operation.PollUntilCompletedAsync()

    ' Step 5: Download and parse results from GCS output
    Dim outputUri = completedOperation.Result.Responses(0).OutputConfig.GcsDestination.Uri
    Dim text = Await DownloadAndParseResultsAsync(storageClient, outputUri)

    ' Step 6: Clean up input file from GCS
    Await storageClient.DeleteObjectAsync(_bucketName, objectName)

    Return text
End Function

Public Async Function ProcessPdfAsync(ByVal pdfPath As String) As Task(Of String)
    ' Step 1: Create storage client
    Dim storageClient = StorageClient.Create()
    Dim objectName = $"ocr-input/{Guid.NewGuid()}.pdf"

    ' Step 2: Upload PDF to GCS (document leaves your infrastructure)
    Using stream = File.OpenRead(pdfPath)
        Await storageClient.UploadObjectAsync(_bucketName, objectName, "application/pdf", stream)
    End Using

    ' Step 3: Build async annotation request
    Dim asyncRequest = New AsyncAnnotateFileRequest With {
        .InputConfig = New InputConfig With {
            .GcsSource = New GcsSource With {.Uri = $"gs://{_bucketName}/{objectName}"},
            .MimeType = "application/pdf"
        },
        .Features = {New フィーチャー With {.Type = Feature.Types.Type.DocumentTextDetection}},
        .OutputConfig = New OutputConfig With {
            .GcsDestination = New GcsDestination With {.Uri = $"gs://{_bucketName}/ocr-output/"},
            .BatchSize = 1
        }
    }

    ' Step 4: Submit and wait for async operation
    Dim operation = Await _visionClient.AsyncBatchAnnotateFilesAsync({asyncRequest})
    Dim completedOperation = Await operation.PollUntilCompletedAsync()

    ' Step 5: Download and parse results from GCS output
    Dim outputUri = completedOperation.Result.Responses(0).OutputConfig.GcsDestination.Uri
    Dim text = Await DownloadAndParseResultsAsync(storageClient, outputUri)

    ' Step 6: Clean up input file from GCS
    Await storageClient.DeleteObjectAsync(_bucketName, objectName)

    Return text
End Function

Public Function ProcessPdf(ByVal pdfPath As String) As String
    ' DirectPDF処理- no GCS, no async, no cleanup
    Using input As New OcrInput()
        input.LoadPdf(pdfPath)
        Return _ocr.Read(input).Text
    End Using
End Function

Public Function ProcessPdfPages(ByVal pdfPath As String, ByVal startPage As Integer, ByVal endPage As Integer) As String
    ' Process specific page range
    Using input As New OcrInput()
        input.LoadPdfPages(pdfPath, startPage, endPage)
        Return _ocr.Read(input).Text
    End Using
End Function

Public Function ProcessEncryptedPdf(ByVal pdfPath As String, ByVal password As String) As String
    ' Password-protected PDFs - not possible with Google Cloud Vision
    Using input As New OcrInput()
        input.LoadPdf(pdfPath, Password:=password)
        Return _ocr.Read(input).Text
    End Using
End Function

Public Function ProcessPdf(ByVal pdfPath As String) As String
    ' DirectPDF処理- no GCS, no async, no cleanup
    Using input As New OcrInput()
        input.LoadPdf(pdfPath)
        Return _ocr.Read(input).Text
    End Using
End Function

Public Function ProcessPdfPages(ByVal pdfPath As String, ByVal startPage As Integer, ByVal endPage As Integer) As String
    ' Process specific page range
    Using input As New OcrInput()
        input.LoadPdfPages(pdfPath, startPage, endPage)
        Return _ocr.Read(input).Text
    End Using
End Function

Public Function ProcessEncryptedPdf(ByVal pdfPath As String, ByVal password As String) As String
    ' Password-protected PDFs - not possible with Google Cloud Vision
    Using input As New OcrInput()
        input.LoadPdf(pdfPath, Password:=password)
        Return _ocr.Read(input).Text
    End Using
End Function

Imports System.Drawing

Public Function ExtractDocumentStructure(imagePath As String) As DocumentStructure
    Dim image = Image.FromFile(imagePath)
    Dim annotation = _client.DetectDocumentText(image)

    Dim structure As New DocumentStructure With {
        .FullText = annotation.Text,
        .Pages = New List(Of PageInfo)()
    }

    ' Navigate: Pages -> Blocks -> Paragraphs -> Words -> Symbols
    For Each page In annotation.Pages
        Dim pageInfo As New PageInfo With {
            .Confidence = page.Confidence,
            .Paragraphs = New List(Of ParagraphInfo)()
        }

        For Each block In page.Blocks
            For Each paragraph In block.Paragraphs
                ' Must concatenate symbols to get paragraph text
                Dim text = String.Join("", paragraph.Words _
                    .SelectMany(Function(w) w.Symbols) _
                    .Select(Function(s) s.Text))

                pageInfo.Paragraphs.Add(New ParagraphInfo With {
                    .Text = text,
                    .Confidence = paragraph.Confidence
                })
            Next
        Next
        structure.Pages.Add(pageInfo)
    Next

    Return structure
End Function

Imports System.Drawing

Public Function ExtractDocumentStructure(imagePath As String) As DocumentStructure
    Dim image = Image.FromFile(imagePath)
    Dim annotation = _client.DetectDocumentText(image)

    Dim structure As New DocumentStructure With {
        .FullText = annotation.Text,
        .Pages = New List(Of PageInfo)()
    }

    ' Navigate: Pages -> Blocks -> Paragraphs -> Words -> Symbols
    For Each page In annotation.Pages
        Dim pageInfo As New PageInfo With {
            .Confidence = page.Confidence,
            .Paragraphs = New List(Of ParagraphInfo)()
        }

        For Each block In page.Blocks
            For Each paragraph In block.Paragraphs
                ' Must concatenate symbols to get paragraph text
                Dim text = String.Join("", paragraph.Words _
                    .SelectMany(Function(w) w.Symbols) _
                    .Select(Function(s) s.Text))

                pageInfo.Paragraphs.Add(New ParagraphInfo With {
                    .Text = text,
                    .Confidence = paragraph.Confidence
                })
            Next
        Next
        structure.Pages.Add(pageInfo)
    Next

    Return structure
End Function

Public Function ExtractDocumentStructure(imagePath As String) As DocumentStructure
    Dim result = _ocr.Read(imagePath)

    ' Direct access - no symbol concatenation, no nested loops
    Return New DocumentStructure With {
        .FullText = result.Text,
        .Confidence = result.Confidence,
        .Paragraphs = result.Paragraphs.Select(Function(p) New ParagraphInfo With {
            .Text = p.Text,
            .Confidence = p.Confidence
        }).ToList(),
        .Lines = result.Lines.Select(Function(l) New LineInfo With {
            .Text = l.Text,
            .X = l.X,
            .Y = l.Y
        }).ToList()
    }
End Function

Public Function ExtractDocumentStructure(imagePath As String) As DocumentStructure
    Dim result = _ocr.Read(imagePath)

    ' Direct access - no symbol concatenation, no nested loops
    Return New DocumentStructure With {
        .FullText = result.Text,
        .Confidence = result.Confidence,
        .Paragraphs = result.Paragraphs.Select(Function(p) New ParagraphInfo With {
            .Text = p.Text,
            .Confidence = p.Confidence
        }).ToList(),
        .Lines = result.Lines.Select(Function(l) New LineInfo With {
            .Text = l.Text,
            .X = l.X,
            .Y = l.Y
        }).ToList()
    }
End Function

Imports System.Threading.Tasks

Public Async Function ProcessPdfAsync(pdfPath As String) As Task(Of String)
    Return Await Task.Run(Function()
                              Using input As New OcrInput()
                                  input.LoadPdf(pdfPath)
                                  Return _ocr.Read(input).Text
                              End Using
                          End Function)
End Function

Imports System.Threading.Tasks

Public Async Function ProcessPdfAsync(pdfPath As String) As Task(Of String)
    Return Await Task.Run(Function()
                              Using input As New OcrInput()
                                  input.LoadPdf(pdfPath)
                                  Return _ocr.Read(input).Text
                              End Using
                          End Function)
End Function