Resolved issue related to convertPdfToText

src/Infrastructure/BotSharp.OpenAPI/Controllers/KnowledgeBaseController.cs

@@ -1,5 +1,6 @@
 using BotSharp.Abstraction.Knowledges.Models;
 using BotSharp.Abstraction.Knowledges.Settings;
+using Microsoft.AspNetCore.Http;
 
 namespace BotSharp.OpenAPI.Controllers;
 
@@ -16,13 +17,6 @@ public KnowledgeBaseController(IKnowledgeService knowledgeService, IServiceProvi
         _services = services;
     }
 
-    [HttpPost("/knowledge-base/embed")]
-    public async Task EmbedKnowledge()
-    {
-        var chunks = await _knowledgeService.CollectChunkedKnowledge();
-        await _knowledgeService.EmbedKnowledge(chunks);
-    }
-
     [HttpGet("/knowledge/{agentId}")]
     public async Task<List<RetrievedResult>> RetrieveKnowledge([FromRoute] string agentId, [FromQuery(Name = "q")] string question)
     {
@@ -70,8 +64,13 @@ public async Task<IActionResult> FeedKnowledge([FromRoute] string agentId, List<
         foreach (var formFile in files)
         {
             var filePath = Path.GetTempFileName();
-            using var stream = System.IO.File.Create(filePath);
-            await formFile.CopyToAsync(stream);
+
+
+            using (var stream = new FileStream(filePath, FileMode.Create, FileAccess.Write, FileShare.None))
+            {
+                await formFile.CopyToAsync(stream);
+                await stream.FlushAsync(); // Ensure all data is written to the file
+            }
 
             var content = await textConverter.ConvertPdfToText(filePath, startPageNum, endPageNum);
 
@@ -84,6 +83,9 @@ await _knowledgeService.Feed(new KnowledgeFeedModel
                 AgentId = agentId,
                 Content = content
             });
+
+            // Delete the temp file after processing to clean up
+            System.IO.File.Delete(filePath);
         }
 
         return Ok(new { count = files.Count, size });

src/Plugins/BotSharp.Plugin.KnowledgeBase/Services/PigPdf2TextConverter.cs

@@ -5,28 +5,24 @@ namespace BotSharp.Plugin.KnowledgeBase.Services;
 
 public class PigPdf2TextConverter : IPdf2TextConverter
 {
-    public async Task<string> ConvertPdfToText(string filePath, int? startPageNum, int? endPageNum)
+    public Task<string> ConvertPdfToText(string filePath, int? startPageNum, int? endPageNum)
     {
-        return await OpenPdfDocumentAsync(filePath, startPageNum, endPageNum);
+        // since PdfDocument.Open is not async, we dont need to make this method async
+        // if you need this method to be async, consider wrapping the call in Task.Run for CPU-bound work
+        return Task.FromResult(OpenPdfDocument(filePath, startPageNum, endPageNum));
     }
 
-    private async Task<string> OpenPdfDocumentAsync(string filePath, int? startPageNum, int? endPageNum)
+    private string OpenPdfDocument(string filePath, int? startPageNum, int? endPageNum)
     {
-        var document = PdfDocument.Open(filePath);
-        var content = "";
+        using var fileStream = new FileStream(filePath, FileMode.Open, FileAccess.Read, FileShare.Read);
+        using var document = PdfDocument.Open(fileStream);
+        var content = new StringBuilder();
         foreach (Page page in document.GetPages())
         {
-            if (startPageNum.HasValue && page.Number < startPageNum.Value)
-            {
-                continue;
-            }
-
-            if (endPageNum.HasValue && page.Number > endPageNum.Value)
-            {
-                continue;
-            }
-            content += page.Text;
+            if (startPageNum.HasValue && page.Number < startPageNum.Value) continue;
+            if (endPageNum.HasValue && page.Number > endPageNum.Value) continue;
+            content.Append(page.Text);
         }
-        return content;
+        return content.ToString();
     }
 }