Annotations
Add annotation support to the viewer
Annotations in Doconut work in two directions: users draw them in the browser widget and the server keeps them per page, or your code builds them and loads them into an open session. Either way they render on the pages and can be burned into PDF or PNG exports.
Annotation support is gated by the Annotation license capability (granted automatically under an active Temporary license).
Enable the annotation UI
Annotation is a Viewer module, not a standalone toolbar. The page must include the Viewer resources, the Viewer toolbar, the Viewer mount, and an initialized objViewer; the Annotation Ribbon is then mounted and attached to that same instance.
Emit the annotation files with the viewer files. They are license-gated, so their tags appear only when the capability is granted. In MVC 5, build the tags in the controller (views have no @inject):
var viewer = Doconut.DoconutHost.Services.GetRequiredService<Viewer>();
ViewBag.ViewerCss = viewer.ReferenceCss(new CssConfig
{
IncludeViewerCss = true,
IncludeAnnotationCss = true // jquery-ui.min.css + annotationBar.css
});
ViewBag.ViewerScripts = viewer.ReferenceScripts(new ScriptConfig
{
IncludeJQuery = true,
IncludeViewerScripts = true,
IncludeAnnotationScripts = true, // jquery-ui, raphael.js, annotation.js
IncludeAnnotationBar = true // the embedded annotation ribbon
});
ViewBag.IsAnnotationEnabled = viewer.IsAnnotationEnabled;Keep the complete Viewer composition in the markup:
<nav id="toolbar" aria-label="Document viewer controls">
<!-- Viewer controls, including the button that opens Annotation -->
</nav>
<div id="annBarMount"></div>
<div id="divDocViewer"><div id="div_ctlDoc"></div></div>The Annotation files generate the Ribbon inside annBarMount; you do not copy its buttons or dialogs. Initialize docViewer first, then create the Ribbon only when the server confirms that Annotation is licensed:
@Html.Raw(ViewBag.ViewerScripts)
<script>
var annBar = null;
var currentToken = '';
var objViewer = $('#div_ctlDoc').docViewer({
BasePath: '/',
ResPath: 'doconut-res/images',
onAnnLoaded: function () { if (annBar) annBar.handleAnnLoaded(); },
onAnnSaved: function () { if (annBar) annBar.handleAnnSaved(); },
onAnnSaveError: function () { if (annBar) annBar.handleAnnSaveError(); },
onAnnClosed: function () { if (annBar) annBar.handleAnnClosed(); },
onError: function (message) { console.error('Viewer error:', message); }
});
@if ((bool)ViewBag.IsAnnotationEnabled)
{
<text>
annBar = $('#annBarMount').doconutAnnotationBar({
docId: 'ctlDoc',
getRequestParams: function () { return { token: currentToken }; },
onLayout: function () { requestAnimationFrame(function () { objViewer.Refit(); }); }
});
annBar.attach(objViewer);
</text>
}
</script>Without the capability, ReferenceScripts leaves out the Ribbon's script, so the server-side if keeps an unguarded doconutAnnotationBar(...) call from breaking the page. In Web Forms, write the same guard as <% if (IsAnnotationEnabled) { %> … <% } %>.
Saving from the Ribbon posts the data to DocImage.axd (AnnSave), which stores it in the document session per page. Loading (AnnLoad) happens automatically when a page with annotations renders.
Open and close the Ribbon from your own toolbar with annBar.open() and annBar.close(). Its public API:
| Method | Purpose |
|---|---|
attach(objViewer) | Connect the Ribbon to the initialized viewer; required once |
open() / close() | Enter or leave annotation editing |
reset() | Return the Ribbon to its closed, non-editing state |
isOpen() / annotating() | Read the Ribbon state / the viewer's annotation-editing state |
reopenEditable() | Reload the current page's annotations as editable objects |
updateActionState() | Refresh the save and delete controls after host changes |
headerSlot() | The optional header slot for host-owned controls |
onStatus, onToast, onLayout, onEditStart, and onEditEnd are optional host callbacks. The endpoints object can also provide exportPdf, exportPng, imageUpload, and imageList; controls without a configured endpoint stay hidden.
Two kinds of callback, wired two ways
The annotation code uses two callback styles, and each fails silently when wired the other way:
onAnnLoaded,onAnnSaved,onAnnSaveError,onAnnClosedare options ofdocViewer({...}), like every viewer callback.Created,Changed,Deleted, andPropertiesare global functions found by name: the viewer id prefixed to the event. For the viewer in#div_ctlDocthe id isctlDoc, so the functions arectlDoc_Created,ctlDoc_Changed,ctlDoc_Deleted, andctlDoc_Properties. A misspelled or renamed function is simply never called.
SaveAnnotations() does not return whether it saved: the result arrives through onAnnSaved or onAnnSaveError. Code that reads its return value never clears its "unsaved changes" flag.
Build annotations in C#
Get a manager bound to the open session, add annotations, and load them (with using Doconut.Annotations; for the types and using System.Drawing; for Rectangle and Color):
// POST /Annotations/LoadSample?token=...
[HttpPost]
public ActionResult LoadSample(string token)
{
var viewer = Doconut.DoconutHost.Services.GetRequiredService<Viewer>();
// Bound to the open session's page dimensions
var manager = viewer.GetAnnotationManager(token);
var pageCount = viewer.GetPageCount(token);
// One stamp per page
for (int page = 1; page <= pageCount; page++)
{
manager.Add(new StampAnnotation(page, new Rectangle(30, 20, 240, 90),
$"PAGE {page}", 28, 4, Color.Maroon)
{
Opacity = 60,
Rotate = -8
});
}
manager.Add(new NoteAnnotation(1, new Rectangle(420, 150, 220, 120),
"Loaded from C# code.", Color.FromArgb(255, 255, 255, 170), 14));
// Load into the session - the viewer fetches them, and the renderer burns them into
// image/PDF exports.
viewer.LoadAnnotationData(token, manager);
return new HttpStatusCodeResult(200);
}Annotation types
All types live in Doconut.Annotations and inherit from BaseAnnotation (page number plus bounding Rectangle):
| Type | Notes |
|---|---|
StampAnnotation | Text stamp with font size, border, color; supports Opacity, Rotate |
NoteAnnotation | Sticky note with text, background color, font size, TitleColor |
RectangleAnnotation | Border and fill colors, Title/ShowTitle |
CircleAnnotation | Border and fill, ShowBorder |
EllipseAnnotation | Border and fill, ShowBorder |
TriangleAnnotation | Border color, BackColor, ShowBorder |
LineAnnotation | Straight line with width and color |
ArrowAnnotation | Line with arrowhead; settable Direction (ArrowDirection, compass points, default E) |
FreehandAnnotation | Free stroke from encoded FreehandData points |
ImageAnnotation | Image from a URL. A relative URL is resolved against the current request's host when the annotation is added — see the export note below |
The AnnotationManager API
| Member | Purpose |
|---|---|
Add(BaseAnnotation) | Queue an annotation |
GetAnnotations() / GetAnnotations(int page) | Inspect what the manager holds |
ClearAnnotations() / ClearAnnotations(int page) | Remove all / per page |
GetAnnotationData() / GetAnnotationData(int page) | Encoded annotation data — the envelope the widget consumes |
GetAnnotationXml() | XML form |
Viewer mirrors the load and read operations against a session: LoadAnnotationData(token, manager) or LoadAnnotationData(token, encodedData), LoadAnnotationXML(token, xml), and GetAnnotationXML(token).
Export with annotations burned in
// PDF of all pages with annotations rendered onto them
public async Task<ActionResult> ExportPdf(string token)
{
var viewer = Doconut.DoconutHost.Services.GetRequiredService<Viewer>();
byte[] pdf = await viewer.ExportAnnotationsToPdfAsync(token, zoom: 100, ct: Response.ClientDisconnectedToken);
return File(pdf, "application/pdf", "export.pdf");
}
// Or a ZIP of per-page PNGs
public async Task<ActionResult> ExportPngZip(string token)
{
var viewer = Doconut.DoconutHost.Services.GetRequiredService<Viewer>();
byte[] zip = await viewer.ExportAnnotationsToPngZipAsync(token, zoom: 100, ct: Response.ClientDisconnectedToken);
return File(zip, "application/zip", "annotations-png.zip");
}Exports use the same burner as on-screen rendering, and the same license and custom-watermark decision.
Image annotations on an intranet host export as "Image blocked"
The viewer and the export fetch an ImageAnnotation's image from different places:
| Surface | Who downloads the image | On an intranet or development host |
|---|---|---|
| The viewer in the browser | The user's browser | The image appears |
ExportAnnotationsToPdfAsync / …PngZipAsync | The server | A red "Image blocked" placeholder |
When the server fetches an image to burn it in, it refuses addresses that point back into private networks: loopback (localhost, 127.0.0.1), 0.0.0.0, link-local (169.254.x.x), the private ranges 10.x, 172.16–31.x, and 192.168.x, carrier-grade NAT (100.64.0.0/10), and IPv6 link-local and unique-local addresses. This protects the server from being used to reach internal systems through a crafted image URL, and there is no switch to turn it off.
A relative image URL resolves against the host of the current request, so on localhost, on an intranet server, or on any site reached through a private address, the image shows in the viewer and the export contains the placeholder instead. That is the guard working, not a rendering failure. If exports must contain the image, serve it from a host with a public address and use an absolute URL. An image URL that cannot be resolved at all makes that one annotation disappear without an error — check the rendered pixels or the network log, not a status code.
Persistence workflow
- Open the document and obtain its token.
- Load previously stored XML or encoded data into that token.
- Let the widget read and edit the session annotations.
- Retrieve XML with
GetAnnotationXML(token)when your application decides to persist. - Export PDF or PNG when a flattened deliverable is required.
- Close the document session.
Do not use the viewer token as a permanent annotation identifier — it dies with the session, and a recycle of the application pool ends every session on a single server. Store annotation data against your own document and version identifiers.
Security and rendering notes
- Annotation requests use the same session and token security as page requests.
- Validate and control any user-supplied image URL.
- Large freehand payloads and high-resolution exports increase memory use; test realistic documents and zoom values.
Troubleshooting
| Symptom | Check |
|---|---|
| The Annotation Ribbon is missing | The Annotation capability and the four annotation CSS/script flags |
| Saving reports an error | Token or session expiration, and the widget's BasePath |
ctlDoc_Created (or another event function) never runs | The function name: viewer id + _ + event, as a global function |
| C# annotations do not appear | Page numbers are one-based, and the data was loaded into the active token |
| An image annotation shows on screen but not in the export | The image's host has a private or loopback address — see above |
| A reopened document has no annotations | Persist XML or data outside the viewer session, then load it into the new token |
Questa pagina è stata utile?