Just checked everything carefully.
I was confused why RichViewActionTest inserts image from Firefox without downloading (no interfaces even assigned) and tried with Chrome and Edge:
Firefox sends local file as \Temp\picture.bmp
Chrome sends just link
Edge like everything from Microsoft works strange, sometimes inserts local picture from it's cache but almost always just link like Chrome
Drop from IE11 wasn't even accepted by RVE
BTW I think there still a bug when you insert hot pictures. My app download normal images (starting with "http:" or "https:") but if that image is hot (contains link) it gets inserted as text URL.
procedure LoadPictureFile(FileName:String);
const StoreFileName = True;
var pic :TPicture;
gr :TGraphic;
rve :TCustomRichViewEdit;
ItemName:String;
begin
IgnoreChanges:=True;
rve:=RichViewEdit1.TopLevelEditor;
pic:=TPicture.Create;
try
pic.LoadFromFile(FileName);
gr:=RVGraphicHandler.CreateGraphic(TGraphicClass(pic.Graphic.ClassType));
gr.Assign(pic.Graphic);
if StoreFileName then begin
rve.BeginUndoGroup(rvutInsert);
rve.SetUndoGroupMode(True);
end;
try
ItemName:=ExtractFileName(FileName);
if rve.InsertPicture(ItemName,gr,rvvaBaseline) and StoreFileName then
rve.SetCurrentItemExtraStrProperty(rvespImageFileName,FileName,True);
finally
if StoreFileName then rve.SetUndoGroupMode(False);
end;
except
ShowError;
end;
pic.Free;
IgnoreChanges:=False;
end;
After that I get that image insterded to RVE and then I get OnImportPicture event with correct URL to that image ("http://website.com/image.jpg").
My image downloader detects it as correct URL replaces already inserted image with "downloading" placeholder and downloads it.
How to prevent it?
Where that URL is stored when I use LoadPictureFile?
If I don't want to forget that URL but leave it in image properties?
Different applications provide dragged data in different formats.
Edge (and Microsoft Explorer) provides dragged images as files (in its cache). It does not provide dragged images as URLs. So you receive file names of temporal cached image files.
Chrome does not provide images as files, but it provides them as links. So, you need to download images (as it was explained before), and you receive image URL.
FireFox provides dragged images BOTH as files and as URL.
TRichViewEdit tries to insert files first. If it does not succeed, it tries another format, and inserts the image from URL.
You can modify reading image files in OnDropFile(s) events.
As I can see, you try to load images yourself, but your code does almost exactly the same work as the standard TRichViewEdit code, so I cannot see how your code can help.
OnDropFile(s) events do not allow to report file insertion as unsuccessful. If you assign DoDefault = False, the component assumes that you successfully inserted files, If you assign DoDefault = True, the component tries to load images itself. If it cannot insert an image, it treats insertion as unsuccessful (and tries to insert in other formats)
Looks like I found the problem.
I didn't handle OnDropFile but OnDropFiles for even one file.
Now I just put DoDefault:=False in OnDropFile and everything else processed via OnDropFiles.