I'm using Asp.Net Core 3.0 and I find myself in a situation where the client will pass text file(s) to my API, the API will then parse the text files into a data model using a function that I have created called ParseDataToModel(), and then store that data model into a database using Entity Framework. Since my code is parsing the files into a data model, I really don't need to copy it to the hard disk if it isn't necessary. I don't have a ton of knowledge when it comes to Streams, and I've googled quite a bit, but I was wondering if there is a way to retrieve the string data of the uploaded files without actually copying them to the hard drive? It seems like a needless extra step.... Below is my code for the file Upload and insertion into the database:
[HttpPost("upload"), DisableRequestSizeLimit]
public IActionResult Upload()
{
var filePaths = new List<string>();
foreach(var formFile in Request.Form.Files)
{
if(formFile.Length > 0)
{
var filePath = Path.GetTempFileName();
filePaths.Add(filePath);
using(var stream = new FileStream(filePath, FileMode.Create))
{
formFile.CopyTo(stream);
}
}
}
BaiFiles lastFile = null;
foreach(string s in filePaths)
{
string contents = System.IO.File.ReadAllText(s);
BaiFiles fileToCreate = ParseFileToModel(contents);
if (fileToCreate == null)
return BadRequest(ModelState);
var file = _fileRepository.GetFiles().Where(t => t.FileId == fileToCreate.FileId).FirstOrDefault();
if (file != null)
{
ModelState.AddModelError("", $"File with id {fileToCreate.FileId} already exists");
return StatusCode(422, ModelState);
}
if (!ModelState.IsValid)
return BadRequest();
if (!_fileRepository.CreateFile(fileToCreate))
{
ModelState.AddModelError("", $"Something went wrong saving file with id {fileToCreate.FileId}");
return StatusCode(500, ModelState);
}
lastFile = fileToCreate;
}
return CreatedAtRoute("GetFile", new { fileId = lastFile.FileId }, lastFile);
}
It would be nice to just hold all of the data in memory instead of copying them to the hard drive, just to turn around and open it to read the text.... I apologize if this isn't possible, or if this question has been asked before. I'm sure it has, and I just wasn't googling the correct keywords. Otherwise, I could be wrong and it is already doing exactly what I want - but System.IO.File.ReadAllText() makes me feel it's being copied to a temp directory somewhere.
After using John's answer below, here is the revised code for anyone interested:
[HttpPost("upload"), DisableRequestSizeLimit]
public IActionResult Upload()
{
var filePaths = new List<string>();
BaiFiles lastFile = null;
foreach (var formFile in Request.Form.Files)
{
if (formFile.Length > 0)
{
using (var stream = formFile.OpenReadStream())
{
using (var sr = new StreamReader(stream))
{
string contents = sr.ReadToEnd();
BaiFiles fileToCreate = ParseFileToModel(contents);
if (fileToCreate == null)
return BadRequest(ModelState);
var file = _fileRepository.GetFiles().Where(t => t.FileId == fileToCreate.FileId).FirstOrDefault();
if (file != null)
{
ModelState.AddModelError("", $"File with id {fileToCreate.FileId} already exists");
return StatusCode(422, ModelState);
}
if (!ModelState.IsValid)
return BadRequest();
if (!_fileRepository.CreateFile(fileToCreate))
{
ModelState.AddModelError("", $"Something went wrong saving file with id {fileToCreate.FileId}");
return StatusCode(500, ModelState);
}
lastFile = fileToCreate;
}
}
}
}
if(lastFile == null)
return NoContent();
else
return CreatedAtRoute("GetFile", new { fileId = lastFile.FileId }, lastFile);
}
System.IO.File.ReadAllText(filePath) is a convenience method. It essentially does this:
string text = null;
using (var stream = FileStream.OpenRead(filePath))
using (var reader = new StreamReader(stream))
{
text = reader.ReadToEnd();
}
FormFile implements an OpenReadStream method, so you can simply use this in place of stream in the above:
string text = null;
using (var stream = formFile.OpenReadStream())
using (var reader = new StreamReader(stream))
{
text = reader.ReadToEnd();
}
Related
I have this code to download a file in the database, but it only downloads the first user upload.
I tried using LastOrDefault and reverse to download the most recent file but I couldn't.
Does anyone have any idea how to do it?
[HttpGet]
[Route("~/DownloadFileCNPJ/{Id}")]
public async Task<IActionResult> DownloadFileCNPJ(long Id)
{
using (var context = new MyDbContext())
{
var cartaoCnpj = (from j in context.CartaoCNPJCedentes
where j.CedenteId.Equals(Id)
select j).Reverse().FirstOrDefault();
if (cartaoCnpj != null)
{
var content = new System.IO.MemoryStream(cartaoCnpj.CartaoCNPJ);
var contentType = cartaoCnpj.TipoCartãoCNPJ;
var fileName = cartaoCnpj.NomeArquivoCNPJ;
return File(content, contentType, fileName);
}
return null;
}
}
I'm currently using tinyMCE to create some news post for a website.
I need to be able to upload images, however i've hit a stopblock.
When I hit my controller I get an
System.ObjectDisposedException: Cannot access a disposed object.
Object name: 'FileBufferingReadStream'.
at Microsoft.AspNetCore.WebUtilities.FileBufferingReadStream.ThrowIfDisposed()
at Microsoft.AspNetCore.WebUtilities.FileBufferingReadStream.set_Position(Int64 value)
at Microsoft.AspNetCore.Http.Internal.ReferenceReadStream..ctor(Stream inner, Int64 offset, Int64 length)
at Microsoft.AspNetCore.Http.Internal.FormFile.OpenReadStream()
at Microsoft.AspNetCore.Http.Internal.FormFile.CopyToAsync(Stream target, CancellationToken cancellationToken)
at HardwareOnlineDk.Web.Areas.Admin.Controllers.ImageController.Upload(IFormFile inputFile) in D:\Kode\HardwareOnlineRider\HOL\SourceCode\Main\Web\Areas\Admin\Controllers\ImageController.cs:line 139
My code looks like this:
[HttpPost, ActionName("Upload")]
public async Task<IActionResult> Upload(IFormFile inputFile)
{
try
{
var filesCount = Request.Form.Files.Count;
if (filesCount == 0)
return BadRequest("Ingen fil fundet");
// Get HTTP posted file based on the fieldname.
var file = Request.Form.Files.GetFile("file");
if (file == null)
return BadRequest("Fejlkonfiguration: Filnavn ikke korrekt");
// Check if the file is valid.
if (!Check(file.FileName, file.ContentType))
return BadRequest("Fil ikke gyldig");
var memoryStream = new MemoryStream();
await file.CopyToAsync(memoryStream);
var medie = new Medie
{
Name = file.FileName.Trim('\"'),
ParentId = _imageService.TempFolderGuid,
ContentLength = file.Length,
Content = memoryStream.ToArray()
};
try
{
var imageId = await _imageService.Medier_InsertMedie(medie);
//TODO Her skal vi gemme ImageId i Session
return Json(new
{
location = $"/api/media/{imageId.Id}.jpg"
});
}
catch
{
return BadRequest("Kunne ikke gemme billede");
}
}
catch
{
return StatusCode(500);
}
}
And the Check Method if needed
private static bool Check(string filePath, string mimeType)
{
return AllowedImageExts.Contains(GetFileExtension(filePath)) &&
AllowedImageMimetypes.Contains(mimeType.ToLower());
}
The code fails when i'm doing:
await file.CopyToAsync(memoryStream)
Can anyone help me here. I'm lost.
UPDATE 1
I just tried to fix it with the suggested answer, so my code now looks like this:
[HttpPost, ActionName("Upload")]
public async Task<IActionResult> Upload([FromForm]IFormFile file)
{
try
{
if (file == null)
return BadRequest("Fejlkonfiguration: Filnavn ikke korrekt");
var memoryStream = new MemoryStream();
await file.CopyToAsync(memoryStream);
var filesCount = Request.Form.Files.Count;
if (!Request.ContentType.StartsWith(MultipartContentType))
return BadRequest("Contenttype er ikke korrekt");
if (filesCount == 0)
return BadRequest("Ingen fil fundet");
// Get HTTP posted file based on the fieldname.
// Check if the file is valid.
if (!Check(file.FileName, file.ContentType))
return BadRequest("Fil ikke gyldig");
var medie = new Medie
{
Name = file.FileName.Trim('\"'),
ParentId = _imageService.TempFolderGuid,
ContentLength = file.Length,
Content = memoryStream.ToArray()
};
try
{
var imageId = await _imageService.Medier_InsertMedie(medie);
//TODO Her skal vi gemme ImageId i Session
return Json(new
{
location = $"/api/media/{imageId.Id}.jpg"
});
}
catch
{
return BadRequest("Kunne ikke gemme billede");
}
}
catch
{
return StatusCode(500);
}
}
The input parameter is no longer null, but it still throws the same exception
You indicated that inputFile is always null.
Binding matches form files by name.
Reference Upload files in ASP.NET Core
Based on
// Get HTTP posted file based on the fieldname.
var file = Request.Form.Files.GetFile("file");
the name of the posted field is "file"
rename the action parameter to match and explicitly state when to bind the data using [FromForm]
[HttpPost, ActionName("Upload")]
public async Task<IActionResult> Upload([FromForm]IFormFile file) {
try {
if (file == null)
return BadRequest("Fejlkonfiguration: Filnavn ikke korrekt");
// Check if the file is valid.
if (!Check(file.Name, file.ContentType))
return BadRequest("Fil ikke gyldig");
var medie = new Medie {
Name = file.Name.Trim('\"'),
ParentId = _imageService.TempFolderGuid
};
var fileStream = file.OpenStreamRead();
using (var memoryStream = new MemoryStream()) {
await fileStream.CopyToAsync(memoryStream);
medie.Content = memoryStream.ToArray();
medie.ContentLength = memoryStream.Length,
}
try {
var imageId = await _imageService.Medier_InsertMedie(medie);
//TODO Her skal vi gemme ImageId i Session
return Json(new {
location = $"/api/media/{imageId.Id}.jpg"
});
} catch {
return BadRequest("Kunne ikke gemme billede");
}
} catch {
return StatusCode(500);
}
}
I feel like I should post how I solved it
The issue was that I have a log in middleware, that reads the httpcontext.
Using the answer found here: How can I read http request body in netcore 3 more than once?
solved the issue for me.
you can pass additional parameters with a class with a IFormFile element. pass the byte data to a stored procure as a byte[] array. where the FileData field is a varbinary parameter in the stored procedure.
public class FormData
{
public string FileName { get; set; }
public string FileDescription { get; set; }
public IFormFile file { get; set; }
}
controller end point
public async Task<IActionResult> UploadDocument([FromForm] FormData formData)
{
if (formData.file.Length > 0)
{
using (MemoryStream ms = new MemoryStream())
{
await formData.file.CopyToAsync(ms);
data = ms.ToArray();
await _repository.spProcedure(FileData: data);
}
}
I have to create a zip file from set of urls. and it should have a proper folder structure.
So i tried like
public async Task<byte[]> CreateZip(Guid ownerId)
{
try
{
string startPath = Path.Combine(AppDomain.CurrentDomain.BaseDirectory, "zipFolder");//base folder
if (Directory.Exists(startPath))
{
DeleteAllFiles(startPath);
Directory.Delete(startPath);
}
Directory.CreateDirectory(startPath);
string zipPath = Path.Combine(AppDomain.CurrentDomain.BaseDirectory, $"{ownerId.ToString()}"); //folder based on ownerid
if (Directory.Exists(zipPath))
{
DeleteAllFiles(zipPath);
Directory.Delete(zipPath);
}
Directory.CreateDirectory(zipPath);
var attachemnts = await ReadByOwnerId(ownerId);
attachemnts.Data.ForEach(i =>
{
var fileLocalPath = $"{startPath}\\{i.Category}";
if (!Directory.Exists(fileLocalPath))
{
Directory.CreateDirectory(fileLocalPath);
}
using (var client = new WebClient())
{
client.DownloadFile(i.Url, $"{fileLocalPath}//{i.Flags ?? ""}_{i.FileName}");
}
});
var zipFilename = $"{zipPath}//result.zip";
if (File.Exists(zipFilename))
{
File.Delete(zipFilename);
}
ZipFile.CreateFromDirectory(startPath, zipFilename, CompressionLevel.Fastest, true);
var result = System.IO.File.ReadAllBytes(zipFilename);
return result;
}
catch (Exception ex)
{
var a = ex;
return null;
}
}
currently im writing all files in my base directory(may be not a good idea).corrently i have to manually delete all folders and files to avoid exception/unwanted files. Can everything be written in memory?
What changes required to write all files and folder structure in memory?
No you can't. Not with the built in Dotnet any way.
As per my comment I would recommend storing the files in a custom location based on a Guid or similar. Eg:
"/xxxx-xxxx-xxxx-xxxx/Folder-To-Zip/....".
This would ensure you could handle multiple requests with the same files or similar file / folder names.
Then you just have to cleanup and delete the folder again afterwards so you don't run out of space.
Hope the below code does the job.
public async Task<byte[]> CreateZip(Guid ownerId)
{
try
{
string startPath = Path.Combine(Path.GetTempPath(), $"{Guid.NewGuid()}_zipFolder");//folder to add
Directory.CreateDirectory(startPath);
var attachemnts = await ReadByOwnerId(ownerId);
attachemnts.Data = filterDuplicateAttachments(attachemnts.Data);
//filtering youtube urls
attachemnts.Data = attachemnts.Data.Where(i => !i.Flags.Equals("YoutubeUrl", StringComparison.OrdinalIgnoreCase)).ToList();
attachemnts.Data.ForEach(i =>
{
var fileLocalPath = $"{startPath}\\{i.Category}";
if (!Directory.Exists(fileLocalPath))
{
Directory.CreateDirectory(fileLocalPath);
}
using (var client = new WebClient())
{
client.DownloadFile(i.Url, $"{fileLocalPath}//{i.Flags ?? ""}_{i.FileName}");
}
});
using (var ms = new MemoryStream())
{
using (var zipArchive = new ZipArchive(ms, ZipArchiveMode.Create, true))
{
System.IO.DirectoryInfo di = new DirectoryInfo(startPath);
var allFiles = di.GetFiles("",SearchOption.AllDirectories);
foreach (var attachment in allFiles)
{
var file = File.OpenRead(attachment.FullName);
var type = attachemnts.Data.Where(i => $"{ i.Flags ?? ""}_{ i.FileName}".Equals(attachment.Name, StringComparison.OrdinalIgnoreCase)).FirstOrDefault();
var entry = zipArchive.CreateEntry($"{type.Category}/{attachment.Name}", CompressionLevel.Fastest);
using (var entryStream = entry.Open())
{
file.CopyTo(entryStream);
}
}
}
var result = ms.ToArray();
return result;
}
}
catch (Exception ex)
{
var a = ex;
return null;
}
}
I have a controller action which is returning a file using the following line of code. However, I need return a filestream with a model which may contain some other JSON serialized data.
I've tried a few different approaches but haven't found a solution which seems to work very well. Has anyone done this before and what is considered the best approach?
this.File(fs, "application/zip", fileName);
Note: The question answered here is referring to multiple FileStreams, whereas I need to return back a FileStream and a Model as JSON.
However, I have modified this solution with the following. What does everyone think?
return new MultipartResult()
{
new MultipartFileContent()
{
ContentType = "application/zip",
FileName = fileName,
Stream = this.OpenFile(filePath)
},
new MultipartModelContent()
{
ContentType = "application/json",
Model = model
}
};
public async Task ExecuteResultAsync(ActionContext context)
{
foreach (var item in this.OfType<MultipartFileContent>())
{
if (item.Stream != null)
{
var c = new StreamContent(item.Stream);
if (item.ContentType != null)
{
c.Headers.ContentType = new System.Net.Http.Headers.MediaTypeHeaderValue(item.ContentType);
}
if (item.FileName != null)
{
var contentDisposition = new ContentDispositionHeaderValue("attachment");
contentDisposition.SetHttpFileName(item.FileName);
c.Headers.ContentDisposition = new System.Net.Http.Headers.ContentDispositionHeaderValue("attachment");
c.Headers.ContentDisposition.FileName = contentDisposition.FileName.Value;
c.Headers.ContentDisposition.FileNameStar = contentDisposition.FileNameStar.Value;
}
this.content.Add(c);
}
}
foreach (var item in this.OfType<MultipartModelContent>())
{
var c = new ObjectContent<object>(item, new JsonMediaTypeFormatter(), item.ContentType);
if (item.ContentType != null)
{
c.Headers.ContentType = new System.Net.Http.Headers.MediaTypeHeaderValue(item.ContentType);
}
this.content.Add(c);
}
context.HttpContext.Response.ContentLength = content.Headers.ContentLength;
context.HttpContext.Response.ContentType = content.Headers.ContentType.ToString();
await content.CopyToAsync(context.HttpContext.Response.Body);
}
In my app, I am using OneDrive to keep data in sync. I am successfully writing the file to OneDrive, but am having no luck replacing the local outdated data with the newer OneDrive data.
My current method, which completes without throwing an exception, does not return the same text data that the file on OneDrive contains.
Goal of the method is to compare the datemodified to the OneDrive file to the local file, and if OneDrive is newer, write the contents of the OndeDrive file to the local StorageFile, and then return it to be de-serialized.
private async Task<string> GetSavedDataFileAsync(string filename)
{
string filepath = _appFolder + #"\" + KOWGame + #"\" + filename;
StorageFile localread;
BasicProperties localprops = null;
string txt;
try
{
localread = await local.GetFileAsync(filepath);
localprops = await localread.GetBasicPropertiesAsync();
}
catch (FileNotFoundException)
{ localread = null; }
if (_userDrive != null)
{
if (_userDrive.IsAuthenticated)
{
try
{
Item item = await _userDrive.Drive.Special.AppRoot.ItemWithPath(filepath).Request().GetAsync();
if (item != null)
{
DateTimeOffset drivemodified = (DateTimeOffset)item.FileSystemInfo.LastModifiedDateTime;
if (localprops != null)
{
if (drivemodified > localprops.DateModified)
{
Stream stream = await localread.OpenStreamForWriteAsync();
using (stream)
{ await _userDrive.Drive.Special.AppRoot.ItemWithPath(filepath).Request().GetAsync(); }
}
}
}
}
catch (OneDriveException e)
{
if (e.IsMatch(OneDriveErrorCode.ActivityLimitReached.ToString()))
{ string stop; }
}
}
}
if (localread == null) return string.Empty;
txt = await FileIO.ReadTextAsync(localread);
return txt;
}
I tried to reverse engineer another answer I found on Stack regarding writing a StorageFile to OneDrive, in that I needed to open the stream of the local file, but I doesn't appear to be working properly.
To get the content of a OneDrive item, we need use following method:
var contentStream = await _userDrive.Drive.Special.AppRoot.ItemWithPath(filepath).Content.Request().GetAsync();
While using
await _userDrive.Drive.Special.AppRoot.ItemWithPath(filepath).Request().GetAsync();
you are getting the OneDrive Item not its content.
So you can change your code like following to write the content of a Onedrive item to a local file:
if (drivemodified > localprops.DateModified)
{
using (var stream = await localread.OpenStreamForWriteAsync())
{
using (var contentStream = await _userDrive.Drive.Special.AppRoot.ItemWithPath(filepath).Content.Request().GetAsync())
{
contentStream.CopyTo(stream);
}
}
}