GF pdfs by TheJoeFin · Pull Request #644 · TheJoeFin/Text-Grab

TheJoeFin · 2026-04-27T03:50:48Z

Closes #641

- Introduce OpenContentKind.PdfDocument and PDF extension helpers - Support PDFs in context menu and "Open With" registration - Add FileUtilities.GetVisualDocumentFilter for images/PDFs - Route PDF files to new PdfDocumentRenderer for native text and OCR - Refactor to treat images and PDFs as visual documents - No breaking changes to existing image workflows

- Introduce PdfTextLineOverlay and PdfTextCanvas for PDF text selection - Add PDF page navigation controls to GrabFrame UI - Support loading, rendering, and extracting text from PDFs - Enable selection, search, and copy of PDF text lines - Update pan/zoom logic for PDF overlays and spacebar-panning - Refactor event handlers and utilities for PDF support - Update dialogs, menus, and help text to include PDFs - Improve XAML formatting and documentation consistency

Updated GrabFrame logic to accept both image and PDF files by using IoUtilities.IsVisualDocumentFile. Added PdfPig NuGet package for PDF handling. Improved debug message to reflect support for PDFs.

Expanded FilesIoTests to cover file type and filter logic. Added PdfDocumentRendererTests for rendering, coordinate mapping, line grouping, and OCR overlap handling.

Added a new "Open File..." menu item with a document icon to the NotifyIconWindow context menu. Implemented its click handler to asynchronously open the file picker using App.OpenFileWithPickerAsync().

Introduce FileUtilities.GetOpenDocumentFilter() for unified open file dialog filters, replacing hardcoded strings and supporting images, PDFs, spreadsheets, markdown, and text files. Refactor GrabFrame to generalize and enhance spacebar-based pan/zoom logic, including new state tracking, improved event handling, and better user experience for both images and PDFs.

Refactored ZoomBorder to use PreviewMouseDown/Up/Move for panning, improving modifier key support and event robustness. Added isPanning state, IsSpacePanModifierPressed, and RequireSpaceToPan for flexible pan activation. Middle mouse now resets zoom/pan. Improved mouse capture/release and removed obsolete handlers.

Added static helpers for file picker and drag-and-drop file handling in App.xaml.cs. Renamed TryToOpenFile to TryToOpenFilePathAsync and updated usages. Ensured EditTextWindow is activated after opening a file.

Expanded FilesIoTests to cover GetOpenDocumentFilter and drag-and-drop file handling. Added tests for document type filters, dropped file path extraction, and drag-drop effects. Included necessary using directives for System.IO and System.Windows.

Added a 300ms grace period after releasing Space before disabling pan mode, making panning with Space+mouse smoother. Ensured pan/zoom is always enabled for PDFs and moved focus away from buttons to prevent accidental activation during panning. Refactored event handling and focus logic for more robust and user-friendly pan/zoom interactions. Also improved mouse capture logic and code clarity in ZoomBorder.

Replaces individual Markdig extension calls with UseAdvancedExtensions for a cleaner and more maintainable pipeline configuration. This enables a broad set of advanced features in one step.

Copilot

Pull request overview

This PR extends Text Grab’s document-handling flow so Grab Frame and related open/OCR paths can work with PDFs in addition to images, while also adding spreadsheet paste/find-replace improvements and a few supporting infrastructure updates.

Changes:

Adds PDF loading/rendering support to Grab Frame, including page navigation, PDF text overlays, and OCR/native-text extraction paths.
Expands shared file/open/context-menu utilities so PDFs are treated as “visual documents” alongside images across the app.
Improves Edit Text spreadsheet workflows with table-aware paste and spreadsheet find/replace support, plus adds unit tests for the new utility logic.

Reviewed changes

Copilot reviewed 34 out of 34 changed files in this pull request and generated 10 comments.

Show a summary per file

File	Description
`Text-Grab/Views/GrabFrame.xaml.cs`	Adds PDF state, loading, page switching, overlay selection, and zoom/pan behavior in Grab Frame.
`Text-Grab/Views/GrabFrame.xaml`	Adds PDF page controls and a dedicated PDF text overlay canvas.
`Text-Grab/Views/FirstRunWindow.xaml`	Updates first-run help text to mention PDFs in Grab Frame.
`Text-Grab/Views/EditTextWindow.xaml.cs`	Adds spreadsheet paste/search/replace helpers and shared open-filter usage.
`Text-Grab/Views/EditTextWindow.xaml`	Wires spreadsheet key handling and minor XAML cleanup.
`Text-Grab/Utilities/PdfDocumentRenderer.cs`	New PDF rendering/text extraction utility for native text and OCR fallback.
`Text-Grab/Utilities/OcrUtilities.cs`	Routes PDF files through the new PDF extraction path.
`Text-Grab/Utilities/MarkdownDocumentUtilities.cs`	Refactors regex declarations and markdown pipeline setup.
`Text-Grab/Utilities/IoUtilities.cs`	Classifies PDFs as a supported open-content kind and visual document type.
`Text-Grab/Utilities/ImplementAppOptions.cs`	Extends “Open with” registration to include PDF support.
`Text-Grab/Utilities/FileUtilities.cs`	Adds reusable open-file filters for visual documents and broader document opening.
`Text-Grab/Utilities/ContextMenuUtilities.cs`	Extends Explorer context-menu registration from images to visual documents including PDFs.
`Text-Grab/Utilities/ClipboardUtilities.cs`	Adds HTML-table-to-tab-separated clipboard conversion for spreadsheet pasting.
`Text-Grab/Text-Grab.csproj`	Adds the PDF parsing dependency.
`Text-Grab/Styles/TextBoxStyles.xaml`	Formatting-only XAML cleanup.
`Text-Grab/Styles/ListViewScrollFix.xaml`	Formatting-only XAML cleanup.
`Text-Grab/Styles/DataGridStyles.xaml`	Formatting-only XAML cleanup.
`Text-Grab/Styles/ButtonStyles.xaml`	Formatting-only XAML cleanup.
`Text-Grab/Pages/GeneralSettings.xaml`	Updates settings text to mention PDF support.
`Text-Grab/Models/FindResult.cs`	Adds spreadsheet cell location metadata for find results.
`Text-Grab/Enums.cs`	Adds `PdfDocument` as an open-content enum value.
`Text-Grab/Controls/ZoomBorder.cs`	Reworks pan handling for space-to-pan and PDF overlay interaction.
`Text-Grab/Controls/PdfTextLineOverlay.cs`	New selectable overlay control for PDF text lines.
`Text-Grab/Controls/NotifyIconWindow.xaml.cs`	Adds tray-menu file-open command.
`Text-Grab/Controls/NotifyIconWindow.xaml`	Adds tray-menu “Open File…” UI.
`Text-Grab/Controls/FindAndReplaceWindow.xaml.cs`	Adds spreadsheet-aware search, navigation, and replace/delete operations.
`Text-Grab/Controls/FindAndReplaceWindow.xaml`	Updates result list layout to show spreadsheet locations.
`Text-Grab/App.xaml.cs`	Adds picker-based file opening and shared dropped-file helpers; routes PDFs to Grab Frame.
`Tests/PdfDocumentRendererTests.cs`	Adds unit tests for PDF geometry and line-grouping helpers.
`Tests/FilesIoTests.cs`	Adds tests for content-kind classification, filters, and dropped-file helpers.
`Tests/ClipboardUtilitiesTests.cs`	Adds tests for HTML table clipboard conversion.
`.github/workflows/buildDev.yml`	Updates build workflow action versions.
`.github/workflows/Release.yml`	Updates release workflow action versions.
`.codetesting/AnalysisReport_20260125_220624_678.md`	Removes a stale analysis report artifact.

+                        <StackPanel
+                            x:Name="PdfPagePanel"
+                            Margin="0,0,6,0"
+                            Orientation="Horizontal"
+                            Visibility="Collapsed">
+                            <Button
+                                x:Name="PreviousPdfPageButton"
+                                Width="30"
+                                Height="30"
+                                Padding="0"
+                                Click="PreviousPdfPageButton_Click"
+                                ToolTip="Previous PDF page">
+                                <ui:SymbolIcon Symbol="ChevronLeft24" />
+                            </Button>
+                            <TextBlock
+                                x:Name="PdfPageTextBlock"
+                                Margin="6,0"
+                                VerticalAlignment="Center"
+                                Text="Page 1 / 1" />
+                            <Button
+                                x:Name="NextPdfPageButton"
+                                Width="30"
+                                Height="30"
+                                Margin="6,0,0,0"
+                                Padding="0"
+                                Click="NextPdfPageButton_Click"
+                                ToolTip="Next PDF page">
+                                <ui:SymbolIcon Symbol="ChevronRight24" />
+                            </Button>
+                        </StackPanel>


+            _loadedPdfDocument = await PdfDocumentRenderer.LoadAsync(path);
+            _currentImagePath = Path.GetFullPath(path);
+            await ShowPdfPageAsync(0);


+    private async Task ShowPdfPageAsync(int pageIndex)
+    {
+        if (_loadedPdfDocument is null)
+            return;
+
+        reDrawTimer.Stop();
+        ResetGrabFrame();
+        await Task.Delay(300);
+
+        _currentPdfPageContent = await _loadedPdfDocument.GetPageContentAsync(pageIndex);


+        List<PdfPageTextLine> combinedLines = [.. pageContent.NativeLines];
+        IReadOnlyList<PdfPageTextLine> imageOcrLines = await GetOcrLinesAsync(
+            pageContent.RenderedPage,
+            resolvedLanguage,
+            sourceRect => ShouldIncludeOcrLine(sourceRect, pageContent.ImageRegions));
+
+        combinedLines.AddRange(imageOcrLines);
+        return SortLines(combinedLines);


        FrameText = "";
        wordBorders.Clear();
+        pdfTextLineOverlays.Clear();
        UpdateFrameText();


+    private const double DefaultRenderScale = 2.0;
+    private readonly WinPdfDocument renderDocument;
+    private readonly PigPdfDocument textDocument;
+    private readonly Dictionary<int, PdfPageContent> pageCache = [];


+
+        if (IoUtilities.IsPdfFileExtension(Path.GetExtension(absolutePath)))
+        {
+            PdfDocumentRenderer pdfDocument = await PdfDocumentRenderer.LoadAsync(absolutePath);


-                ocrText = await GrabTemplateExecutor.ExecuteTemplateOnBitmapAsync(grabTemplate, bmp, selectedLanguage);
+                if (IoUtilities.IsPdfFileExtension(Path.GetExtension(path)))
+                {
+                    PdfDocumentRenderer pdfDocument = await PdfDocumentRenderer.LoadAsync(path);


+        var updates = targets
+            .Where(r => r.RowIndex.HasValue && r.ColumnIndex.HasValue)
+            .GroupBy(r => (r.RowIndex!.Value, r.ColumnIndex!.Value))
+            .Select(g =>
+            {
+                int row = g.Key.Item1, col = g.Key.Item2;
+                string oldValue = row < tableDocument.Rows.Count && col < tableDocument.Rows[row].Count
+                    ? tableDocument.Rows[row][col] ?? string.Empty : string.Empty;
+                return (RowIndex: row, ColumnIndex: col, Value: pattern.Replace(oldValue, replaceWith));


+            int cellStart;
+            string endTag;
+            if (tdPos >= 0 && (thPos < 0 || tdPos <= thPos))
+            {
+                cellStart = tdPos;
+                endTag = "</td>";
+            }
+            else
+            {
+                cellStart = thPos;
+                endTag = "</th>";
+            }
+
+            int openEnd = rowHtml.IndexOf('>', cellStart);
+            if (openEnd < 0) break;
+
+            int contentStart = openEnd + 1;
+            int contentEnd = rowHtml.IndexOf(endTag, contentStart, StringComparison.OrdinalIgnoreCase);
+            contentEnd = contentEnd >= 0 ? contentEnd : rowHtml.Length;
+
+            cells.Add(CleanHtmlCellContent(rowHtml[contentStart..contentEnd]));
+            pos = contentEnd + endTag.Length;


TheJoeFin · 2026-05-02T15:16:47Z

@copilot apply changes based on the comments in this thread

…replace, colspan/rowspan Agent-Logs-Url: https://github.com/TheJoeFin/Text-Grab/sessions/c5acdc4a-84dd-4662-b5d3-335886f0b193 Co-authored-by: TheJoeFin <7809853+TheJoeFin@users.noreply.github.com>

Agent-Logs-Url: https://github.com/TheJoeFin/Text-Grab/sessions/c5acdc4a-84dd-4662-b5d3-335886f0b193 Co-authored-by: TheJoeFin <7809853+TheJoeFin@users.noreply.github.com>

Copilot · 2026-05-02T15:28:32Z

@copilot apply changes based on the comments in this thread

Applied all changes from the review thread (commit 703f09f, optimized in 5f4772a):

PdfDocumentRenderer disposal (OcrUtilities.cs): Added using to both PDF renderer creation sites so the document is disposed after extraction.
Dispose before reassignment (GrabFrame.xaml.cs): TryLoadPdfFromPath now calls ClearLoadedPdfDocument() before loading a new PDF, releasing the old file handle first.
Dispose on window close (GrabFrame.xaml.cs): GrabFrameWindow_Closing now calls ClearLoadedPdfDocument() to release the file handle when the window closes.
Page navigation race condition (GrabFrame.xaml.cs): Added _pdfPageNavCts — each call to ShowPdfPageAsync cancels the previous in-flight navigation before starting a new one, preventing out-of-order page display.
Duplicate OCR content (PdfDocumentRenderer.cs): When a page has native text and image regions, OCR lines that already overlap with native text regions are filtered out, preventing double-reporting on scan PDFs with hidden text layers.
Page cache eviction (PdfDocumentRenderer.cs): Implemented an LRU cache capped at 10 pages — the least-recently-used page is evicted when the limit is reached.
Find/replace granularity (EditTextWindow.xaml.cs): SearchSpreadsheetCells now stores m.Index in FindResult.Index; ReplaceInSpreadsheetCells uses a MatchEvaluator keyed on those indices to replace only the selected occurrences rather than all matches in a cell.
HTML colspan/rowspan (ClipboardUtilities.cs): Rewrote ParseHtmlTableToGrid with a span-aware grid algorithm using a rowspanMap to carry spanned cells forward; added ParseSpanAttribute helper and updated ParseHtmlRowCells to return span info. Added unit tests for both colspan and rowspan.

Warning

Firewall rules blocked me from connecting to one or more addresses (expand for details)

I tried to connect to the following addresses, but was blocked by firewall rules:

az667904.vo.msecnd.net
- Triggering command: /usr/bin/dotnet dotnet build Text-Grab/Text-Grab.csproj -p:EnableWindowsTargeting=true --no-restore (dns block)
- Triggering command: /usr/bin/dotnet dotnet build Tests/Tests.csproj -p:EnableWindowsTargeting=true --no-restore (dns block)

If you need me to access, download, or install something from one of these locations, you can either:

Configure Actions setup steps to set up my environment, which run before the firewall is enabled
Add the appropriate URLs or hosts to the custom allowlist in this repository's Copilot coding agent settings (admins only)

Expanded file filtering to include PDFs for OCR, updated UI messages to reference "files" instead of just "images," clarified variable typing in spreadsheet undo logic, and set StaysOpenOnClick for relevant menu items to enhance user experience.

Replaced custom Copy with standard Cut/Copy/Paste commands in the spreadsheet context menu. Added command bindings and keyboard shortcut handling for these actions. Implemented TryCutSpreadsheetCellValues for cut logic with clipboard failure detection. Updated clipboard helper to return success status. Added unit tests for cut scenarios.

TheJoeFin added 9 commits April 25, 2026 18:38

Add PDF support to GrabFrame and update file checks

3d49a18

Updated GrabFrame logic to accept both image and PDF files by using IoUtilities.IsVisualDocumentFile. Added PdfPig NuGet package for PDF handling. Improved debug message to reflect support for PDFs.

Add tests for file type classification and PDF rendering

02fec33

Expanded FilesIoTests to cover file type and filter logic. Added PdfDocumentRendererTests for rendering, coordinate mapping, line grouping, and OCR overlap handling.

Add "Open File..." option to NotifyIcon context menu

4a0b99a

Added a new "Open File..." menu item with a document icon to the NotifyIconWindow context menu. Implemented its click handler to asynchronously open the file picker using App.OpenFileWithPickerAsync().

Enhance file open and drag-and-drop support

c2407f7

Added static helpers for file picker and drag-and-drop file handling in App.xaml.cs. Renamed TryToOpenFile to TryToOpenFilePathAsync and updated usages. Ensured EditTextWindow is activated after opening a file.

TheJoeFin added enhancement New feature or request Grab Frame Relating to the Grab Frame experience labels Apr 27, 2026

TheJoeFin changed the base branch from main to dev April 27, 2026 03:51

TheJoeFin added 9 commits April 27, 2026 23:16

Add clipboard tests

2766e95

improve the way the clipboard can handle html table data and more

da3ee2d

send table data to etw properly

a8b61c9

improve perf and options around markdown

5721793

enhance the find and replace to be compatible with etw spreadsheet mode

9e8aeb1

update github actions

493b304

remove unused file

2091e93

Simplify Markdig pipeline with UseAdvancedExtensions

38b679f

Replaces individual Markdig extension calls with UseAdvancedExtensions for a cleaner and more maintainable pipeline configuration. This enables a broad set of advanced features in one step.

TheJoeFin requested a review from Copilot May 2, 2026 15:05

Copilot started reviewing on behalf of TheJoeFin May 2, 2026 15:06 View session

Copilot AI reviewed May 2, 2026

View reviewed changes

Copilot started work on behalf of TheJoeFin May 2, 2026 15:16 View session

Copilot AI and others added 2 commits May 2, 2026 15:25

Fix review comments: disposal, race conditions, cache eviction, find/…

703f09f

…replace, colspan/rowspan Agent-Logs-Url: https://github.com/TheJoeFin/Text-Grab/sessions/c5acdc4a-84dd-4662-b5d3-335886f0b193 Co-authored-by: TheJoeFin <7809853+TheJoeFin@users.noreply.github.com>

Optimize: materialize native rects once before OCR predicate

5f4772a

Agent-Logs-Url: https://github.com/TheJoeFin/Text-Grab/sessions/c5acdc4a-84dd-4662-b5d3-335886f0b193 Co-authored-by: TheJoeFin <7809853+TheJoeFin@users.noreply.github.com>

Copilot finished work on behalf of TheJoeFin May 2, 2026 15:29

TheJoeFin added 2 commits May 2, 2026 10:44

TheJoeFin merged commit 17e8809 into dev May 5, 2026
1 check passed

TheJoeFin deleted the gf-pdfs branch May 5, 2026 13:22

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

GF pdfs#644

GF pdfs#644
TheJoeFin merged 22 commits intodevfrom
gf-pdfs

TheJoeFin commented Apr 27, 2026

Uh oh!

Copilot AI left a comment

Uh oh!

TheJoeFin commented May 2, 2026

Uh oh!

Copilot AI commented May 2, 2026 •

edited

Loading

I tried to connect to the following addresses, but was blocked by firewall rules:

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

Uh oh!

Conversation

TheJoeFin commented Apr 27, 2026

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Reviewed changes

Uh oh!

TheJoeFin commented May 2, 2026

Uh oh!

Copilot AI commented May 2, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

I tried to connect to the following addresses, but was blocked by firewall rules:

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

Copilot AI commented May 2, 2026 •

edited

Loading