Added OneNote sections as a document type (#1161)

app/MindWork AI Studio/Tools/ContentStreamDocumentDetails.cs

@@ -7,6 +7,13 @@ public sealed class ContentStreamDocumentDetails
     [JsonPropertyName("page_number")]
     public int? PageNumber { get; set; }
 
+    /// <summary>
+    /// Whether the page is one the file itself knows, such as a page of a OneNote section, rather
+    /// than one the runtime estimated from the page breaks of a Word or OpenDocument file.
+    /// </summary>
+    [JsonPropertyName("page_is_exact")]
+    public bool PageIsExact { get; set; }
+
     [JsonPropertyName("image")]
     public ContentStreamPptxImageData? Image { get; set; }
 }

app/MindWork AI Studio/Tools/DocumentManager.cs

@@ -26,12 +26,13 @@ public sealed class DocumentManager
 
             //
             // A Word or OpenDocument file carries no fixed page layout, so the runtime derives these
-            // boundaries from page breaks and heuristics. We note the estimate as a comment rather
-            // than as a heading: a heading would sit on the same level as the document's own first
-            // level headings, leaving the AI unable to tell the structure of the document apart from
-            // our boundaries. The presentation reader marks its slides the same way.
+            // boundaries from page breaks and heuristics and says so. The pages of a OneNote section,
+            // on the other hand, are those of the file. We note the page as a comment rather than as
+            // a heading: a heading would sit on the same level as the document's own first level
+            // headings, leaving the AI unable to tell the structure of the document apart from our
+            // boundaries. The presentation reader marks its slides the same way.
             //
-            this.currentPageContent.AppendLine($"<!-- Estimated page {pageNumber} -->");
+            this.currentPageContent.AppendLine(metadata.Document?.PageIsExact is true ? $"<!-- Page {pageNumber} -->" : $"<!-- Estimated page {pageNumber} -->");
             this.currentPageContent.AppendLine();
             this.currentPageContent.Append(content);
 

app/MindWork AI Studio/Tools/FileExclusions.cs

@@ -95,9 +95,10 @@ internal static class FileExclusions
     private static readonly string[] ALWAYS_EXCLUDED_FILES = ["Thumbs.db", "ehthumbs.db", "ehthumbs_vista.db", "desktop.ini", ".DS_Store", "Icon\r", OFFICE_LOCK_FILE_PATTERN];
 
     /// <summary>
-    /// Folders nobody ever wanted as content: what operating systems and ZIP files on macOS leave behind.
+    /// Folders nobody ever wanted as content: what operating systems and ZIP files on macOS leave behind,
+    /// and the recycle bin of a OneNote notebook, which holds the pages the user deleted.
     /// </summary>
-    private static readonly string[] ALWAYS_EXCLUDED_FOLDERS = ["__MACOSX", "$RECYCLE.BIN", "System Volume Information"];
+    private static readonly string[] ALWAYS_EXCLUDED_FOLDERS = ["__MACOSX", "$RECYCLE.BIN", "System Volume Information", "OneNote_RecycleBin"];
 
     private static readonly char[] SEPARATORS = [';', '\r', '\n'];
 

app/MindWork AI Studio/Tools/Rust/FileTypes.cs

@@ -84,7 +84,11 @@ public static class FileTypes
     // The legacy binary ".ppt" is missing on purpose: AI Studio has no reader for it, so offering
     // it would only let users attach a file which cannot be read.
     public static readonly FileTypeFilter POWER_POINT = FileTypeFilter.Leaf("PowerPoint", "pptx", "odp");
-    public static readonly FileTypeFilter MAIL        = FileTypeFilter.Leaf(TB("Mail"), "eml", "msg", "mbox");
+
+    // Only the sections of a notebook carry its pages. The table of contents (".onetoc2") holds none,
+    // and a package (".onepkg") would have to be unpacked into memory as a whole.
+    public static readonly FileTypeFilter ONENOTE     = FileTypeFilter.Leaf("OneNote", "one");
+    public static readonly FileTypeFilter MAIL       = FileTypeFilter.Leaf(TB("Mail"), "eml", "msg", "mbox");
     public static readonly FileTypeFilter LATEX       = FileTypeFilter.Leaf("LaTeX", "tex", "bib", "sty", "cls", "log");
 
     // Only the LaTeX document itself, without the auxiliary files of the LaTeX family: this is
@@ -96,7 +100,7 @@ public static class FileTypes
     public static readonly FileTypeFilter POWER_POINT_PRESENTATION = FileTypeFilter.Leaf("PowerPoint", "pptx");
 
     public static readonly FileTypeFilter OFFICE_FILES = FileTypeFilter.Parent(TB("Office Files"),
-        WORD, SPREADSHEET, POWER_POINT, PDF);
+        WORD, SPREADSHEET, POWER_POINT, ONENOTE, PDF);
     public static readonly FileTypeFilter DOCUMENT     = FileTypeFilter.Parent(TB("Document"),
         TEXT, TABULAR, OFFICE_FILES, SOURCE_CODE, LATEX);