Add component to read PDF files into assistants (#439)

app/Build/Commands/CheckRidsCommand.cs

@@ -17,5 +17,10 @@ public sealed class CheckRidsCommand
         {
             Console.WriteLine($"- {rid}");
         }
+
+        Console.WriteLine();
+        Console.WriteLine("The RID for the current OS and CPU is:");
+        var currentRid = Environment.GetCurrentRid();
+        Console.WriteLine($"- {currentRid}");
     }
 }

app/Build/Commands/Library.cs

@@ -0,0 +1,3 @@
+namespace Build.Commands;
+
+public record Library(string Path, string Filename);

app/Build/Commands/Pdfium.cs

@@ -0,0 +1,107 @@
+using System.Formats.Tar;
+using System.IO.Compression;
+
+using SharedTools;
+
+namespace Build.Commands;
+
+public static class Pdfium
+{
+    public static async Task InstallAsync(RID rid, string version)
+    {
+        Console.Write($"- Installing Pdfium {version} for {rid.ToUserFriendlyName()} ...");
+
+        var cwd = Environment.GetRustRuntimeDirectory();
+        var pdfiumTmpDownloadPath = Path.GetTempFileName();
+        var pdfiumTmpExtractPath = Directory.CreateTempSubdirectory();
+        var pdfiumUrl = GetPdfiumDownloadUrl(rid, version);
+
+        //
+        // Download the file:
+        //
+        Console.Write(" downloading ...");
+        using (var client = new HttpClient())
+        {
+            var response = await client.GetAsync(pdfiumUrl);
+            if (!response.IsSuccessStatusCode)
+            {
+                Console.WriteLine($" failed to download Pdfium {version} for {rid.ToUserFriendlyName()}");
+                return;
+            }
+
+            await using var fileStream = File.Create(pdfiumTmpDownloadPath);
+            await response.Content.CopyToAsync(fileStream);
+        }
+        
+        //
+        // Extract the downloaded file:
+        //
+        Console.Write(" extracting ...");
+        await using(var tgzStream = File.Open(pdfiumTmpDownloadPath, FileMode.Open, FileAccess.Read, FileShare.Read))
+        {
+            await using var uncompressedStream = new GZipStream(tgzStream, CompressionMode.Decompress);
+            await TarFile.ExtractToDirectoryAsync(uncompressedStream, pdfiumTmpExtractPath.FullName, true);
+        }
+        
+        //
+        // Copy the library to the target directory:
+        //
+        Console.Write(" deploying ...");
+        var library = GetLibraryPath(rid);
+        if (string.IsNullOrEmpty(library.Path))
+        {
+            Console.WriteLine($" failed to find the library path for {rid.ToUserFriendlyName()}");
+            return;
+        }
+        
+        var pdfiumLibSourcePath = Path.Join(pdfiumTmpExtractPath.FullName, library.Path);
+        var pdfiumLibTargetPath = Path.Join(cwd, "resources", "libraries", library.Filename);
+        if (!File.Exists(pdfiumLibSourcePath))
+        {
+            Console.WriteLine($" failed to find the library file '{pdfiumLibSourcePath}'");
+            return;
+        }
+        
+        Directory.CreateDirectory(Path.Join(cwd, "resources", "libraries"));
+        if (File.Exists(pdfiumLibTargetPath))
+            File.Delete(pdfiumLibTargetPath);
+        
+        File.Copy(pdfiumLibSourcePath, pdfiumLibTargetPath);
+        
+        //
+        // Cleanup:
+        //
+        Console.Write(" cleaning up ...");
+        File.Delete(pdfiumTmpDownloadPath);
+        Directory.Delete(pdfiumTmpExtractPath.FullName, true);
+        
+        Console.WriteLine(" done.");
+    }
+    
+    private static Library GetLibraryPath(RID rid) => rid switch
+    {
+        RID.LINUX_ARM64 or RID.LINUX_X64 => new(Path.Join("lib", "libpdfium.so"), "libpdfium.so"),
+        RID.OSX_ARM64 or RID.OSX_X64 => new(Path.Join("lib", "libpdfium.dylib"), "libpdfium.dylib"),
+        RID.WIN_ARM64 or RID.WIN_X64 => new(Path.Join("bin", "pdfium.dll"), "pdfium.dll"),
+        
+        _ => new(string.Empty, string.Empty),
+    };
+
+    private static string GetPdfiumDownloadUrl(RID rid, string version)
+    {
+        var baseUrl = $"https://github.com/bblanchon/pdfium-binaries/releases/download/chromium%2F{version}/pdfium-";
+        return rid switch
+        {
+            RID.LINUX_ARM64 => $"{baseUrl}linux-arm64.tgz",
+            RID.LINUX_X64 => $"{baseUrl}linux-x64.tgz",
+            
+            RID.OSX_ARM64 => $"{baseUrl}mac-arm64.tgz",
+            RID.OSX_X64 => $"{baseUrl}mac-x64.tgz",
+            
+            RID.WIN_ARM64 => $"{baseUrl}win-arm64.zip",
+            RID.WIN_X64 => $"{baseUrl}win-x64.zip",
+            
+            _ => string.Empty,
+        };
+    }
+}

app/Build/Commands/UpdateMetadataCommands.cs

@@ -105,105 +105,102 @@ public sealed partial class UpdateMetadataCommands
         // Build the .NET project:
         //
         var pathApp = Environment.GetAIStudioDirectory();
-        var rids = Environment.GetRidsForCurrentOS();
-        foreach (var rid in rids)
-        {
-            Console.WriteLine("==============================");
-            await this.UpdateArchitecture(rid);
-            
-            Console.Write($"- Start .NET build for '{rid.AsMicrosoftRid()}' ...");
-            await this.ReadCommandOutput(pathApp, "dotnet", $"clean --configuration release --runtime {rid.AsMicrosoftRid()}");
-            var dotnetBuildOutput = await this.ReadCommandOutput(pathApp, "dotnet", $"publish --configuration release --runtime {rid.AsMicrosoftRid()} --disable-build-servers --force");
-            var dotnetBuildOutputLines = dotnetBuildOutput.Split([global::System.Environment.NewLine], StringSplitOptions.RemoveEmptyEntries);
-            var foundIssue = false;
-            foreach (var buildOutputLine in dotnetBuildOutputLines)
-            {
-                if(buildOutputLine.Contains(" error ") || buildOutputLine.Contains("#warning"))
-                {
-                    if(!foundIssue)
-                    {
-                        foundIssue = true;
-                        Console.WriteLine();
-                        Console.WriteLine("- Build has issues:");
-                    }
-
-                    Console.Write("   - ");
-                    Console.WriteLine(buildOutputLine);
-                }
-            }
-            
-            if(foundIssue)
-                Console.WriteLine();
-            else
-            {
-                Console.WriteLine(" completed successfully.");
-            }
-            
-            //
-            // Prepare the .NET artifact to be used by Tauri as sidecar:
-            //
-            var os = Environment.GetOS();
-            var tauriSidecarArtifactName = rid switch
-            {
-                RID.WIN_X64 => "mindworkAIStudioServer-x86_64-pc-windows-msvc.exe",
-                RID.WIN_ARM64 => "mindworkAIStudioServer-aarch64-pc-windows-msvc.exe",
-                
-                RID.LINUX_X64 => "mindworkAIStudioServer-x86_64-unknown-linux-gnu",
-                RID.LINUX_ARM64 => "mindworkAIStudioServer-aarch64-unknown-linux-gnu",
-                
-                RID.OSX_ARM64 => "mindworkAIStudioServer-aarch64-apple-darwin",
-                RID.OSX_X64 => "mindworkAIStudioServer-x86_64-apple-darwin",
-                
-                _ => string.Empty,
-            };
-
-            if (string.IsNullOrWhiteSpace(tauriSidecarArtifactName))
-            {
-                Console.WriteLine($"- Error: Unsupported rid '{rid.AsMicrosoftRid()}'.");
-                return;
-            }
+        var rid = Environment.GetCurrentRid();
         
-            var dotnetArtifactPath = Path.Combine(pathApp, "bin", "dist");
-            if(!Directory.Exists(dotnetArtifactPath))
-                Directory.CreateDirectory(dotnetArtifactPath);
-            
-            var dotnetArtifactFilename = os switch
+        Console.WriteLine("==============================");
+        await this.UpdateArchitecture(rid);
+        
+        var pdfiumVersion = await this.ReadPdfiumVersion();
+        await Pdfium.InstallAsync(rid, pdfiumVersion);
+        
+        Console.Write($"- Start .NET build for {rid.ToUserFriendlyName()} ...");
+        await this.ReadCommandOutput(pathApp, "dotnet", $"clean --configuration release --runtime {rid.AsMicrosoftRid()}");
+        var dotnetBuildOutput = await this.ReadCommandOutput(pathApp, "dotnet", $"publish --configuration release --runtime {rid.AsMicrosoftRid()} --disable-build-servers --force");
+        var dotnetBuildOutputLines = dotnetBuildOutput.Split([global::System.Environment.NewLine], StringSplitOptions.RemoveEmptyEntries);
+        var foundIssue = false;
+        foreach (var buildOutputLine in dotnetBuildOutputLines)
+        {
+            if(buildOutputLine.Contains(" error ") || buildOutputLine.Contains("#warning"))
             {
-                "windows" => "mindworkAIStudio.exe",
-                _ => "mindworkAIStudio",
-            };
-            
-            var dotnetPublishedPath = Path.Combine(pathApp, "bin", "release", Environment.DOTNET_VERSION, rid.AsMicrosoftRid(), "publish", dotnetArtifactFilename);
-            var finalDestination = Path.Combine(dotnetArtifactPath, tauriSidecarArtifactName);
-            
-            if(File.Exists(dotnetPublishedPath))
-                Console.WriteLine("- Published .NET artifact found.");
-            else
-            {
-                Console.WriteLine($"- Error: Published .NET artifact not found: '{dotnetPublishedPath}'.");
-                return;
-            }
+                if(!foundIssue)
+                {
+                    foundIssue = true;
+                    Console.WriteLine();
+                    Console.WriteLine("- Build has issues:");
+                }
 
-            Console.Write($"- Move the .NET artifact to the Tauri sidecar destination ...");
-            try
-            {
-                File.Move(dotnetPublishedPath, finalDestination, true);
-                Console.WriteLine(" done.");
+                Console.Write("   - ");
+                Console.WriteLine(buildOutputLine);
             }
-            catch (Exception e)
-            {
-                Console.WriteLine(" failed.");
-                Console.WriteLine($"   - Error: {e.Message}");
-            }
-            
+        }
+        
+        if(foundIssue)
             Console.WriteLine();
+        else
+        {
+            Console.WriteLine(" completed successfully.");
+        }
+        
+        //
+        // Prepare the .NET artifact to be used by Tauri as sidecar:
+        //
+        var os = Environment.GetOS();
+        var tauriSidecarArtifactName = rid switch
+        {
+            RID.WIN_X64 => "mindworkAIStudioServer-x86_64-pc-windows-msvc.exe",
+            RID.WIN_ARM64 => "mindworkAIStudioServer-aarch64-pc-windows-msvc.exe",
+            
+            RID.LINUX_X64 => "mindworkAIStudioServer-x86_64-unknown-linux-gnu",
+            RID.LINUX_ARM64 => "mindworkAIStudioServer-aarch64-unknown-linux-gnu",
+            
+            RID.OSX_ARM64 => "mindworkAIStudioServer-aarch64-apple-darwin",
+            RID.OSX_X64 => "mindworkAIStudioServer-x86_64-apple-darwin",
+            
+            _ => string.Empty,
+        };
+
+        if (string.IsNullOrWhiteSpace(tauriSidecarArtifactName))
+        {
+            Console.WriteLine($"- Error: Unsupported rid '{rid.AsMicrosoftRid()}'.");
+            return;
+        }
+    
+        var dotnetArtifactPath = Path.Combine(pathApp, "bin", "dist");
+        if(!Directory.Exists(dotnetArtifactPath))
+            Directory.CreateDirectory(dotnetArtifactPath);
+        
+        var dotnetArtifactFilename = os switch
+        {
+            "windows" => "mindworkAIStudio.exe",
+            _ => "mindworkAIStudio",
+        };
+        
+        var dotnetPublishedPath = Path.Combine(pathApp, "bin", "release", Environment.DOTNET_VERSION, rid.AsMicrosoftRid(), "publish", dotnetArtifactFilename);
+        var finalDestination = Path.Combine(dotnetArtifactPath, tauriSidecarArtifactName);
+        
+        if(File.Exists(dotnetPublishedPath))
+            Console.WriteLine("- Published .NET artifact found.");
+        else
+        {
+            Console.WriteLine($"- Error: Published .NET artifact not found: '{dotnetPublishedPath}'.");
+            return;
         }
 
+        Console.Write($"- Move the .NET artifact to the Tauri sidecar destination ...");
+        try
+        {
+            File.Move(dotnetPublishedPath, finalDestination, true);
+            Console.WriteLine(" done.");
+        }
+        catch (Exception e)
+        {
+            Console.WriteLine(" failed.");
+            Console.WriteLine($"   - Error: {e.Message}");
+        }
+        
         //
         // Build the Rust project / runtime:
         //
-        
-        Console.WriteLine("==============================");
         Console.WriteLine("- Start building the Rust runtime ...");
         
         var pathRuntime = Environment.GetRustRuntimeDirectory();
@@ -314,6 +311,18 @@ public sealed partial class UpdateMetadataCommands
         await File.WriteAllTextAsync(changelogCodePath, changelogCode, Environment.UTF8_NO_BOM);
         Console.WriteLine(" done.");
     }
+    
+    private async Task<string> ReadPdfiumVersion()
+    {
+        const int PDFIUM_VERSION_INDEX = 10;
+        
+        var pathMetadata = Environment.GetMetadataPath();
+        var lines = await File.ReadAllLinesAsync(pathMetadata, Encoding.UTF8);
+        var currentPdfiumVersion = lines[PDFIUM_VERSION_INDEX].Trim();
+        var shortVersion = currentPdfiumVersion.Split('.')[2];
+        
+        return shortVersion;
+    }
 
     private async Task UpdateArchitecture(RID rid)
     {
@@ -321,7 +330,7 @@ public sealed partial class UpdateMetadataCommands
         
         var pathMetadata = Environment.GetMetadataPath();
         var lines = await File.ReadAllLinesAsync(pathMetadata, Encoding.UTF8);
-        Console.Write("- Updating architecture ...");
+        Console.Write($"- Updating architecture to {rid.ToUserFriendlyName()} ...");
         lines[ARCHITECTURE_INDEX] = rid.AsMicrosoftRid();
         
         await File.WriteAllLinesAsync(pathMetadata, lines, Environment.UTF8_NO_BOM);

app/Build/Commands/UpdateWebAssetsCommand.cs

@@ -17,7 +17,7 @@ public sealed class UpdateWebAssetsCommand
         Console.WriteLine("=========================");
         Console.Write("- Updating web assets ...");
         
-        var rid = Environment.GetRidsForCurrentOS().First();
+        var rid = Environment.GetCurrentRid();
         var cwd = Environment.GetAIStudioDirectory();
         var contentPath = Path.Join(cwd, "bin", "release", Environment.DOTNET_VERSION, rid.AsMicrosoftRid(), "publish", "wwwroot", "_content");
         var isMudBlazorDirectoryPresent = Directory.Exists(Path.Join(contentPath, "MudBlazor"));

app/Build/Tools/Environment.cs

@@ -76,4 +76,38 @@ public static class Environment
         Console.WriteLine($"Error: Unsupported OS '{RuntimeInformation.OSDescription}'");
         return [];
     }
+    
+    public static RID GetCurrentRid()
+    {
+        var arch = RuntimeInformation.ProcessArchitecture;
+        if (RuntimeInformation.IsOSPlatform(OSPlatform.Windows))
+            return arch switch
+            {
+                Architecture.X64 => RID.WIN_X64,
+                Architecture.Arm64 => RID.WIN_ARM64,
+                
+                _ => RID.NONE,
+            };
+        
+        if (RuntimeInformation.IsOSPlatform(OSPlatform.OSX))
+            return arch switch
+            {
+                Architecture.X64 => RID.OSX_X64,
+                Architecture.Arm64 => RID.OSX_ARM64,
+                
+                _ => RID.NONE,
+            };
+        
+        if (RuntimeInformation.IsOSPlatform(OSPlatform.Linux))
+            return arch switch
+            {
+                Architecture.X64 => RID.LINUX_X64,
+                Architecture.Arm64 => RID.LINUX_ARM64,
+                
+                _ => RID.NONE,
+            };
+        
+        Console.WriteLine($"Error: Unsupported OS '{RuntimeInformation.OSDescription}'");
+        return RID.NONE;
+    }
 }

app/MindWork AI Studio.sln.DotSettings

@@ -9,6 +9,7 @@
 	<s:String x:Key="/Default/CodeStyle/Naming/CSharpNaming/Abbreviations/=LM/@EntryIndexedValue">LM</s:String>
 	<s:String x:Key="/Default/CodeStyle/Naming/CSharpNaming/Abbreviations/=MSG/@EntryIndexedValue">MSG</s:String>
 	<s:String x:Key="/Default/CodeStyle/Naming/CSharpNaming/Abbreviations/=OS/@EntryIndexedValue">OS</s:String>
+	<s:String x:Key="/Default/CodeStyle/Naming/CSharpNaming/Abbreviations/=PDF/@EntryIndexedValue">PDF</s:String>
 	<s:String x:Key="/Default/CodeStyle/Naming/CSharpNaming/Abbreviations/=RAG/@EntryIndexedValue">RAG</s:String>
 	<s:String x:Key="/Default/CodeStyle/Naming/CSharpNaming/Abbreviations/=RID/@EntryIndexedValue">RID</s:String>
 	<s:String x:Key="/Default/CodeStyle/Naming/CSharpNaming/Abbreviations/=TB/@EntryIndexedValue">TB</s:String>

app/MindWork AI Studio/Assistants/I18N/allTexts.lua

@@ -631,6 +631,12 @@ UI_TEXT_CONTENT["AISTUDIO::COMPONENTS::PROFILESELECTION::T918741365"] = "You can
 -- Provider
 UI_TEXT_CONTENT["AISTUDIO::COMPONENTS::PROVIDERSELECTION::T900237532"] = "Provider"
 
+-- Use PDF content as input
+UI_TEXT_CONTENT["AISTUDIO::COMPONENTS::READPDFCONTENT::T2849276709"] = "Use PDF content as input"
+
+-- Select PDF file
+UI_TEXT_CONTENT["AISTUDIO::COMPONENTS::READPDFCONTENT::T63272795"] = "Select PDF file"
+
 -- The content is cleaned using an LLM agent: the main content is extracted, advertisements and other irrelevant things are attempted to be removed; relative links are attempted to be converted into absolute links so that they can be used.
 UI_TEXT_CONTENT["AISTUDIO::COMPONENTS::READWEBCONTENT::T1164201762"] = "The content is cleaned using an LLM agent: the main content is extracted, advertisements and other irrelevant things are attempted to be removed; relative links are attempted to be converted into absolute links so that they can be used."
 
@@ -2116,6 +2122,9 @@ UI_TEXT_CONTENT["AISTUDIO::PAGES::ABOUT::T228561878"] = "In order to use any LLM
 -- The C# language is used for the implementation of the user interface and the backend. To implement the user interface with C#, the Blazor technology from ASP.NET Core is used. All these technologies are integrated into the .NET SDK.
 UI_TEXT_CONTENT["AISTUDIO::PAGES::ABOUT::T2329884315"] = "The C# language is used for the implementation of the user interface and the backend. To implement the user interface with C#, the Blazor technology from ASP.NET Core is used. All these technologies are integrated into the .NET SDK."
 
+-- Used PDFium version
+UI_TEXT_CONTENT["AISTUDIO::PAGES::ABOUT::T2368247719"] = "Used PDFium version"
+
 -- This library is used to determine the language of the operating system. This is necessary to set the language of the user interface.
 UI_TEXT_CONTENT["AISTUDIO::PAGES::ABOUT::T2557014401"] = "This library is used to determine the language of the operating system. This is necessary to set the language of the user interface."
 

app/MindWork AI Studio/Assistants/LegalCheck/AssistantLegalCheck.razor

@@ -1,4 +1,5 @@
 @attribute [Route(Routes.ASSISTANT_LEGAL_CHECK)]
+@using AIStudio.Settings.DataModel
 @inherits AssistantBaseCore<AIStudio.Dialogs.Settings.SettingsDialogLegalCheck>
 
 @if (!this.SettingsManager.ConfigurationData.LegalCheck.HideWebContentReader)
@@ -6,6 +7,11 @@
     <ReadWebContent @bind-Content="@this.inputLegalDocument" ProviderSettings="@this.providerSettings" @bind-AgentIsRunning="@this.isAgentRunning" Preselect="@(this.SettingsManager.ConfigurationData.LegalCheck.PreselectOptions && this.SettingsManager.ConfigurationData.LegalCheck.PreselectWebContentReader)" PreselectContentCleanerAgent="@(this.SettingsManager.ConfigurationData.LegalCheck.PreselectOptions && this.SettingsManager.ConfigurationData.LegalCheck.PreselectContentCleanerAgent)"/>
 }
 
+@if (PreviewFeatures.PRE_READ_PDF_2025.IsEnabled(this.SettingsManager))
+{
+    <ReadPDFContent @bind-PDFContent="@this.inputLegalDocument"/>
+}
+
 <MudTextField T="string" Disabled="@this.isAgentRunning" @bind-Text="@this.inputLegalDocument" Validation="@this.ValidatingLegalDocument" AdornmentIcon="@Icons.Material.Filled.DocumentScanner" Adornment="Adornment.Start" Label="@T("Legal document")" Variant="Variant.Outlined" Lines="12" AutoGrow="@true" MaxLines="24" Class="mb-3" UserAttributes="@USER_INPUT_ATTRIBUTES"/>
 <MudTextField T="string" Disabled="@this.isAgentRunning" @bind-Text="@this.inputQuestions" Validation="@this.ValidatingQuestions" AdornmentIcon="@Icons.Material.Filled.QuestionAnswer" Adornment="Adornment.Start" Label="@T("Your questions")" Variant="Variant.Outlined" Lines="6" AutoGrow="@true" MaxLines="12" Class="mb-3" UserAttributes="@USER_INPUT_ATTRIBUTES"/>
 <ProviderSelection @bind-ProviderSettings="@this.providerSettings" ValidateProvider="@this.ValidatingProvider"/>

app/MindWork AI Studio/Assistants/TextSummarizer/AssistantTextSummarizer.razor

@@ -1,4 +1,5 @@
 @attribute [Route(Routes.ASSISTANT_SUMMARIZER)]
+@using AIStudio.Settings.DataModel
 @inherits AssistantBaseCore<AIStudio.Dialogs.Settings.SettingsDialogTextSummarizer>
 
 @if (!this.SettingsManager.ConfigurationData.TextSummarizer.HideWebContentReader)
@@ -6,6 +7,11 @@
     <ReadWebContent @bind-Content="@this.inputText" ProviderSettings="@this.providerSettings" @bind-AgentIsRunning="@this.isAgentRunning" Preselect="@(this.SettingsManager.ConfigurationData.TextSummarizer.PreselectOptions && this.SettingsManager.ConfigurationData.TextSummarizer.PreselectWebContentReader)" PreselectContentCleanerAgent="@(this.SettingsManager.ConfigurationData.TextSummarizer.PreselectOptions && this.SettingsManager.ConfigurationData.TextSummarizer.PreselectContentCleanerAgent)"/>
 }
 
+@if (PreviewFeatures.PRE_READ_PDF_2025.IsEnabled(this.SettingsManager))
+{
+    <ReadPDFContent @bind-PDFContent="@this.inputText"/>
+}
+
 <MudTextField T="string" Disabled="@this.isAgentRunning" @bind-Text="@this.inputText" Validation="@this.ValidatingText" AdornmentIcon="@Icons.Material.Filled.DocumentScanner" Adornment="Adornment.Start" Label="@T("Your input")" Variant="Variant.Outlined" Lines="6" AutoGrow="@true" MaxLines="12" Class="mb-3" UserAttributes="@USER_INPUT_ATTRIBUTES"/>
 <EnumSelection T="CommonLanguages" NameFunc="@(language => language.Name())" @bind-Value="@this.selectedTargetLanguage" Icon="@Icons.Material.Filled.Translate" Label="@T("Target language")" AllowOther="@true" @bind-OtherInput="@this.customTargetLanguage" OtherValue="CommonLanguages.OTHER" LabelOther="@T("Custom target language")" ValidateOther="@this.ValidateCustomLanguage" />
 <EnumSelection T="Complexity" NameFunc="@(complexity => complexity.Name())" @bind-Value="@this.selectedComplexity" Icon="@Icons.Material.Filled.Layers" Label="@T("Target complexity")" AllowOther="@true" @bind-OtherInput="@this.expertInField" OtherValue="Complexity.SCIENTIFIC_LANGUAGE_OTHER_EXPERTS" LabelOther="@T("Your expertise")" ValidateOther="@this.ValidateExpertInField" />

app/MindWork AI Studio/Assistants/Translation/AssistantTranslation.razor

@@ -1,4 +1,5 @@
 @attribute [Route(Routes.ASSISTANT_TRANSLATION)]
+@using AIStudio.Settings.DataModel
 @inherits AssistantBaseCore<AIStudio.Dialogs.Settings.SettingsDialogTranslation>
 
 @if (!this.SettingsManager.ConfigurationData.Translation.HideWebContentReader)
@@ -6,6 +7,11 @@
     <ReadWebContent @bind-Content="@this.inputText" ProviderSettings="@this.providerSettings" @bind-AgentIsRunning="@this.isAgentRunning" Preselect="@(this.SettingsManager.ConfigurationData.Translation.PreselectOptions && this.SettingsManager.ConfigurationData.Translation.PreselectWebContentReader)" PreselectContentCleanerAgent="@(this.SettingsManager.ConfigurationData.Translation.PreselectOptions && this.SettingsManager.ConfigurationData.Translation.PreselectContentCleanerAgent)"/>
 }
 
+@if (PreviewFeatures.PRE_READ_PDF_2025.IsEnabled(this.SettingsManager))
+{
+    <ReadPDFContent @bind-PDFContent="@this.inputText"/>
+}
+
 <MudTextSwitch Label="@T("Live translation")" @bind-Value="@this.liveTranslation" LabelOn="@T("Live translation")" LabelOff="@T("No live translation")"/>
 @if (this.liveTranslation)
 {

app/MindWork AI Studio/Components/ReadPDFContent.razor

@@ -0,0 +1,4 @@
+@inherits MSGComponentBase
+<MudButton StartIcon="@Icons.Material.Filled.Description" OnClick="async () => await this.SelectFile()" Variant="Variant.Filled" Class="mb-3">
+    @T("Use PDF content as input")
+</MudButton>

app/MindWork AI Studio/Components/ReadPDFContent.razor.cs

@@ -0,0 +1,33 @@
+using AIStudio.Tools.Services;
+
+using Microsoft.AspNetCore.Components;
+
+namespace AIStudio.Components;
+
+public partial class ReadPDFContent : MSGComponentBase
+{
+    [Parameter]
+    public string PDFContent { get; set; } = string.Empty;
+    
+    [Parameter]
+    public EventCallback<string> PDFContentChanged { get; set; }
+    
+    [Inject]
+    private RustService RustService { get; init; } = null!;
+    
+    private async Task SelectFile()
+    {
+        var pdfFile = await this.RustService.SelectFile(T("Select PDF file"));
+        if (pdfFile.UserCancelled)
+            return;
+
+        if (!pdfFile.SelectedFilePath.EndsWith(".pdf", StringComparison.OrdinalIgnoreCase))
+            return;
+        
+        if(!File.Exists(pdfFile.SelectedFilePath))
+            return;
+        
+        var pdfText = await this.RustService.GetPDFText(pdfFile.SelectedFilePath);
+        await this.PDFContentChanged.InvokeAsync(pdfText);
+    }
+}

app/MindWork AI Studio/Components/ReadWebContent.razor

@@ -1,5 +1,5 @@
 @inherits MSGComponentBase
-<MudPaper Class="pa-3 mb-8 border-dashed border rounded-lg">
+<MudPaper Class="pa-3 mb-3 border-dashed border rounded-lg">
     <MudTextSwitch Label="@T("Read content from web?")" Disabled="@this.AgentIsRunning" @bind-Value="@this.showWebContentReader" LabelOn="@T("Show web content options")" LabelOff="@T("Hide web content options")" />
     @if (this.showWebContentReader)
     {

app/MindWork AI Studio/MindWork AI Studio.csproj

@@ -80,6 +80,7 @@
             <MetaTauriVersion>$([System.String]::Copy( $(Metadata) ).Split( ';' )[ 7 ])</MetaTauriVersion>
             <MetaAppCommitHash>$([System.String]::Copy( $(Metadata) ).Split( ';' )[ 8 ])</MetaAppCommitHash>
             <MetaArchitecture>$([System.String]::Copy( $(Metadata) ).Split( ';' )[ 9 ])</MetaArchitecture>
+            <MetaPdfiumVersion>$([System.String]::Copy( $(Metadata) ).Split( ';' )[ 10 ])</MetaPdfiumVersion>
             
             <GenerateAssemblyInfo>true</GenerateAssemblyInfo>
             
@@ -104,6 +105,9 @@
             <AssemblyAttribute Include="AIStudio.Tools.Metadata.MetaDataArchitecture">
                 <_Parameter1>$(MetaArchitecture)</_Parameter1>
             </AssemblyAttribute>
+            <AssemblyAttribute Include="AIStudio.Tools.Metadata.MetaDataLibraries">
+                <_Parameter1>$(MetaPdfiumVersion)</_Parameter1>
+            </AssemblyAttribute>
         </ItemGroup>
         
     </Target>

app/MindWork AI Studio/Pages/About.razor

@@ -18,6 +18,7 @@
                     <MudListItem T="string" Icon="@Icons.Material.Outlined.Build" Text="@VersionDotnetSdk"/>
                     <MudListItem T="string" Icon="@Icons.Material.Outlined.Memory" Text="@VersionDotnetRuntime"/>
                     <MudListItem T="string" Icon="@Icons.Material.Outlined.Build" Text="@VersionRust"/>
+                    <MudListItem T="string" Icon="@Icons.Material.Outlined.DocumentScanner" Text="@VersionPdfium"/>
                     <MudListItem T="string" Icon="@Icons.Material.Outlined.Widgets" Text="@MudBlazorVersion"/>
                     <MudListItem T="string" Icon="@Icons.Material.Outlined.Memory" Text="@TauriVersion"/>
                     <MudListItem T="string" Icon="@Icons.Material.Outlined.Translate" Text="@this.OSLanguage"/>
@@ -111,6 +112,7 @@
                     <ThirdPartyComponent Name="rcgen" Developer="RustTLS developers, est31 & Open Source Community" LicenseName="MIT" LicenseUrl="https://github.com/rustls/rcgen/blob/main/LICENSE" RepositoryUrl="https://github.com/rustls/rcgen" UseCase="@T("For the secure communication between the user interface and the runtime, we need to create certificates. This Rust library is great for this purpose.")"/>
                     <ThirdPartyComponent Name="file-format" Developer="Mickaël Malécot & Open Source Community" LicenseName="MIT" LicenseUrl="https://github.com/mmalecot/file-format/blob/main/LICENSE-MIT" RepositoryUrl="https://github.com/mmalecot/file-format" UseCase="@T("This library is used to determine the file type of a file. This is necessary, e.g., when we want to stream a file.")"/>
                     <ThirdPartyComponent Name="calamine" Developer="Johann Tuffe, Joel Natividad, Eric Jolibois, Dmitriy & Open Source Community" LicenseName="MIT" LicenseUrl="https://github.com/tafia/calamine/blob/master/LICENSE-MIT.md" RepositoryUrl="https://github.com/tafia/calamine" UseCase="@T("This library is used to read Excel and OpenDocument spreadsheet files. This is necessary, e.g., for using spreadsheets as a data source for a chat.")"/>
+                    <ThirdPartyComponent Name="PDFium" Developer="Lei Zhang, Tom Sepez, Dan Sinclair, and Foxit, Google, Chromium, Collabora, Ada, DocsCorp, Dropbox, Microsoft, and PSPDFKit Teams & Open Source Community" LicenseName="Apache-2.0" LicenseUrl="https://pdfium.googlesource.com/pdfium/+/refs/heads/main/LICENSE" RepositoryUrl="https://pdfium.googlesource.com/pdfium" UseCase="@T("This library is used to read PDF files. This is necessary, e.g., for using PDFs as a data source for a chat.")"/>
                     <ThirdPartyComponent Name="pdfium-render" Developer="Alastair Carey, Dorian Rudolph & Open Source Community" LicenseName="MIT" LicenseUrl="https://github.com/ajrcarey/pdfium-render/blob/master/LICENSE.md" RepositoryUrl="https://github.com/ajrcarey/pdfium-render" UseCase="@T("This library is used to read PDF files. This is necessary, e.g., for using PDFs as a data source for a chat.")"/>
                     <ThirdPartyComponent Name="sys-locale" Developer="1Password Team, ComplexSpaces & Open Source Community" LicenseName="MIT" LicenseUrl="https://github.com/1Password/sys-locale/blob/main/LICENSE-MIT" RepositoryUrl="https://github.com/1Password/sys-locale" UseCase="@T("This library is used to determine the language of the operating system. This is necessary to set the language of the user interface.")"/>
                     <ThirdPartyComponent Name="Lua-CSharp" Developer="Yusuke Nakada & Open Source Community" LicenseName="MIT" LicenseUrl="https://github.com/nuskey8/Lua-CSharp/blob/main/LICENSE" RepositoryUrl="https://github.com/nuskey8/Lua-CSharp" UseCase="@T("We use Lua as the language for plugins. Lua-CSharp lets Lua scripts communicate with AI Studio and vice versa. Thank you, Yusuke Nakada, for this great library.")" />

app/MindWork AI Studio/Pages/About.razor.cs

@@ -21,7 +21,8 @@ public partial class About : MSGComponentBase
     
     private static readonly Assembly ASSEMBLY = Assembly.GetExecutingAssembly();
     private static readonly MetaDataAttribute META_DATA = ASSEMBLY.GetCustomAttribute<MetaDataAttribute>()!;
-    private static readonly MetaDataArchitecture META_DATA_ARCH = ASSEMBLY.GetCustomAttribute<MetaDataArchitecture>()!;
+    private static readonly MetaDataArchitectureAttribute META_DATA_ARCH = ASSEMBLY.GetCustomAttribute<MetaDataArchitectureAttribute>()!;
+    private static readonly MetaDataLibrariesAttribute META_DATA_LIBRARIES = ASSEMBLY.GetCustomAttribute<MetaDataLibrariesAttribute>()!;
 
     private string osLanguage = string.Empty;
     
@@ -40,6 +41,8 @@ public partial class About : MSGComponentBase
     private string VersionDotnetSdk => $"{T("Used .NET SDK")}: v{META_DATA.DotnetSdkVersion}";
     
     private string BuildTime => $"{T("Build time")}: {META_DATA.BuildTime}";
+    
+    private string VersionPdfium => $"{T("Used PDFium version")}: v{META_DATA_LIBRARIES.PdfiumVersion}";
 
     private GetLogPathsResponse logPaths;
     

app/MindWork AI Studio/Plugins/languages/de-de-43065dbc-78d0-45b7-92be-f14c2926e2dc/plugin.lua

@@ -633,6 +633,12 @@ UI_TEXT_CONTENT["AISTUDIO::COMPONENTS::PROFILESELECTION::T918741365"] = "Hier ka
 -- Provider
 UI_TEXT_CONTENT["AISTUDIO::COMPONENTS::PROVIDERSELECTION::T900237532"] = "Anbieter"
 
+-- Use PDF content as input
+UI_TEXT_CONTENT["AISTUDIO::COMPONENTS::READPDFCONTENT::T2849276709"] = "PDF-Inhalt als Eingabe verwenden"
+
+-- Select PDF file
+UI_TEXT_CONTENT["AISTUDIO::COMPONENTS::READPDFCONTENT::T63272795"] = "PDF-Datei auswählen"
+
 -- The content is cleaned using an LLM agent: the main content is extracted, advertisements and other irrelevant things are attempted to be removed; relative links are attempted to be converted into absolute links so that they can be used.
 UI_TEXT_CONTENT["AISTUDIO::COMPONENTS::READWEBCONTENT::T1164201762"] = "Der Inhalt wird mithilfe eines LLM-Agents bereinigt: Der Hauptinhalt wird extrahiert, Werbung und andere irrelevante Elemente werden nach Möglichkeit entfernt. Relative Links werden nach Möglichkeit in absolute Links umgewandelt, damit sie verwendet werden können."
 
@@ -2118,6 +2124,9 @@ UI_TEXT_CONTENT["AISTUDIO::PAGES::ABOUT::T228561878"] = "Um ein beliebiges LLM n
 -- The C# language is used for the implementation of the user interface and the backend. To implement the user interface with C#, the Blazor technology from ASP.NET Core is used. All these technologies are integrated into the .NET SDK.
 UI_TEXT_CONTENT["AISTUDIO::PAGES::ABOUT::T2329884315"] = "Die Programmiersprache C# wird für die Umsetzung der Benutzeroberfläche und des Backends verwendet. Für die Entwicklung der Benutzeroberfläche mit C# kommt die Blazor-Technologie aus ASP.NET Core zum Einsatz. Alle diese Technologien sind im .NET SDK integriert."
 
+-- Used PDFium version
+UI_TEXT_CONTENT["AISTUDIO::PAGES::ABOUT::T2368247719"] = "Verwendete PDFium-Version"
+
 -- This library is used to determine the language of the operating system. This is necessary to set the language of the user interface.
 UI_TEXT_CONTENT["AISTUDIO::PAGES::ABOUT::T2557014401"] = "Diese Bibliothek wird verwendet, um die Sprache des Betriebssystems zu erkennen. Dies ist notwendig, um die Sprache der Benutzeroberfläche einzustellen."
 

app/MindWork AI Studio/Plugins/languages/en-us-97dfb1ba-50c4-4440-8dfa-6575daf543c8/plugin.lua

@@ -633,6 +633,12 @@ UI_TEXT_CONTENT["AISTUDIO::COMPONENTS::PROFILESELECTION::T918741365"] = "You can
 -- Provider
 UI_TEXT_CONTENT["AISTUDIO::COMPONENTS::PROVIDERSELECTION::T900237532"] = "Provider"
 
+-- Use PDF content as input
+UI_TEXT_CONTENT["AISTUDIO::COMPONENTS::READPDFCONTENT::T2849276709"] = "Use PDF content as input"
+
+-- Select PDF file
+UI_TEXT_CONTENT["AISTUDIO::COMPONENTS::READPDFCONTENT::T63272795"] = "Select PDF file"
+
 -- The content is cleaned using an LLM agent: the main content is extracted, advertisements and other irrelevant things are attempted to be removed; relative links are attempted to be converted into absolute links so that they can be used.
 UI_TEXT_CONTENT["AISTUDIO::COMPONENTS::READWEBCONTENT::T1164201762"] = "The content is cleaned using an LLM agent: the main content is extracted, advertisements and other irrelevant things are attempted to be removed; relative links are attempted to be converted into absolute links so that they can be used."
 
@@ -2118,6 +2124,9 @@ UI_TEXT_CONTENT["AISTUDIO::PAGES::ABOUT::T228561878"] = "In order to use any LLM
 -- The C# language is used for the implementation of the user interface and the backend. To implement the user interface with C#, the Blazor technology from ASP.NET Core is used. All these technologies are integrated into the .NET SDK.
 UI_TEXT_CONTENT["AISTUDIO::PAGES::ABOUT::T2329884315"] = "The C# language is used for the implementation of the user interface and the backend. To implement the user interface with C#, the Blazor technology from ASP.NET Core is used. All these technologies are integrated into the .NET SDK."
 
+-- Used PDFium version
+UI_TEXT_CONTENT["AISTUDIO::PAGES::ABOUT::T2368247719"] = "Used PDFium version"
+
 -- This library is used to determine the language of the operating system. This is necessary to set the language of the user interface.
 UI_TEXT_CONTENT["AISTUDIO::PAGES::ABOUT::T2557014401"] = "This library is used to determine the language of the operating system. This is necessary to set the language of the user interface."
 
@@ -2519,3 +2528,4 @@ UI_TEXT_CONTENT["AISTUDIO::PAGES::WRITER::T3948127789"] = "Suggestion"
 
 -- Your stage directions
 UI_TEXT_CONTENT["AISTUDIO::PAGES::WRITER::T779923726"] = "Your stage directions"
+

app/MindWork AI Studio/Settings/DataModel/PreviewFeatures.cs

@@ -10,4 +10,5 @@ public enum PreviewFeatures
     PRE_RAG_2024,
     
     PRE_PLUGINS_2025,
+    PRE_READ_PDF_2025,
 }

app/MindWork AI Studio/Settings/DataModel/PreviewFeaturesExtensions.cs

@@ -6,7 +6,9 @@ public static class PreviewFeaturesExtensions
     {
         PreviewFeatures.PRE_WRITER_MODE_2024 => "Writer Mode: Experiments about how to write long texts using AI",
         PreviewFeatures.PRE_RAG_2024 => "RAG: Preview of our RAG implementation where you can refer your files or integrate enterprise data within your company",
+        
         PreviewFeatures.PRE_PLUGINS_2025 => "Plugins: Preview of our plugin system where you can extend the functionality of the app",
+        PreviewFeatures.PRE_READ_PDF_2025 => "Read PDF: Preview of our PDF reading system where you can read and extract text from PDF files",
         
         _ => "Unknown preview feature"
     };

app/MindWork AI Studio/Settings/DataModel/PreviewVisibilityExtensions.cs

@@ -19,6 +19,7 @@ public static class PreviewVisibilityExtensions
         
         if (visibility >= PreviewVisibility.PROTOTYPE)
         {
+            features.Add(PreviewFeatures.PRE_READ_PDF_2025);
             features.Add(PreviewFeatures.PRE_RAG_2024);
         }
         

app/MindWork AI Studio/Tools/Metadata/MetaDataArchitecture.cs

@@ -2,7 +2,7 @@
 namespace AIStudio.Tools.Metadata;
 
 [AttributeUsage(AttributeTargets.Assembly)]
-public class MetaDataArchitecture(string architecture) : Attribute
+public class MetaDataArchitectureAttribute(string architecture) : Attribute
 {
     public string Architecture => architecture;
 }

app/MindWork AI Studio/Tools/Metadata/MetaDataLibrariesAttribute.cs

@@ -0,0 +1,6 @@
+namespace AIStudio.Tools.Metadata;
+
+public class MetaDataLibrariesAttribute(string pdfiumVersion) : Attribute
+{
+    public string PdfiumVersion => pdfiumVersion;
+}

app/MindWork AI Studio/Tools/Services/RustService.Retrieval.cs

@@ -0,0 +1,16 @@
+namespace AIStudio.Tools.Services;
+
+public sealed partial class RustService
+{
+    public async Task<string> GetPDFText(string filePath)
+    {
+        var response = await this.http.GetAsync($"/retrieval/fs/read/pdf?file_path={filePath}");
+        if (!response.IsSuccessStatusCode)
+        {
+            this.logger!.LogError($"Failed to read the PDF file due to an network error: '{response.StatusCode}'");
+            return string.Empty;
+        }
+
+        return await response.Content.ReadAsStringAsync();
+    }
+}

app/MindWork AI Studio/wwwroot/changelog/v0.9.42.md

@@ -1,8 +1,9 @@
 # v0.9.42, build 217 (2025-05-xx xx:xx UTC)
 - Added the writing style "Changelog" to the rewrite & improve text assistant. This helps to create changelogs for your projects.
+- Added an option to load PDF files directly into the translation, text summarization, and legal check assistants as a preview prototype for testing before release.
 - Improved the model selection for OpenAI by removing all `o1-pro` models. These models cannot be used right now, since OpenAI introduced a new API, which is not yet supported by MindWork AI Studio.
 - Improved the internal plugin maintenance so that removed resources are now removed from the file system.
 - Improved the app settings to apply the chosen language immediately.
 - Fixed an issue where empty lines in source code were being ignored by the Markdown renderer. Thanks My Nihongo for fixing this bug in the `MudBlazor.Markdown` repository.
 - Fixed the localization assistant not being able to load the localization file when used in the release app.
-- Upgraded .NET dependencies.
+- Upgraded Rust & .NET dependencies.