Entry
How can I force a File Download of a .pdf file with a hyperlink (bypassing Acrobat Reader)?
Jan 26th, 2006 01:44
Andre White, Ian Vink, Justin Pasher, Arjan Kemeling, Darton Williams, Joe Lipman, Eilon Reshef, Martin Honnen, Kim Denz,
Client side JavaScript can't achieve that. It is best to teach your
users to use the existing browser functionality, that is right click-
>save as in the context menu.
[Added here]
You can use the content-disposition HTTP field on the server side.
[another added]
Sorry to contradict the above but late versions of Internet Explorer
allow a client-side save via the dHTML method
document.execCommand("SaveAs")
Although this only saves the current webpage, you could open a
seperate
window (via javascript) which temporarily opens the file, executes the
save function, then closes again. so...
myTempWindow = window.open("thePdfFile.pdf","SaysaMe");
myTempWindow.document.execCommand("SaveAs",null,"SaveAsName.pdf");
myTempWindow.close();
To be honest, I don't know what the 'null' does as it seems to work
the
same if you replace with 'true' or 'false'. If you get an error, try
replacing the 'null' with these until it works.
[Still more added]
After going around and around with a client over this issue, I finally
told the client to zip all of his pdf's (and pdf's compress very well,
by the way) and no more problems. I tried everything under the sun,
including ColdFusion CFCONTENT, Javascript, content-disposition, etc.,
and still couldn't get pdf's to reliably save to disk cross-browser.
If
you absolutely MUST use bare pdf files, then I agree with the first
answer here. In fact, you could place some instructions right on the
page telling the user how to right-click and save as.
[The latest addition so far]
I have taken the above script and tweaked it a bit.
Now it doesn't show you the window that opened (it's position is off
screen) and you can use the same script for multiple downloads.
Notice that I replaced the double quotes with single quotes.
I have posted it on my own site where you can try it out:
http://www.whitewater.be/resources/javascript/view.asp?
action=view&id=48
And this is the code :
<html>
<body>
<script>
function downloadme(x){
myTempWindow = window.open(x,’’,’left=10000,screenX=10000’);
myTempWindow.document.execCommand(’SaveAs’,’null’,x);
myTempWindow.close();
}
</script>
<a href=javascript:downloadme("/test.pdf");>Download this pdf</a>
</body>
</html>
Arjan Kemeling
[Life CAN be simple]
After having spent some time finding a way to by-pass the file
association problem for PDF's I just renamed the PDF file !
By removing the file extention '.pdf' you cause your browser to ask
you
what you want to do with the unknown file.
When saving it, it does appearently read the file and recognize it as
a
PDF. This will cause the Save As dialog box to save it with PDF file
extention and so you just need to hit the OK button !
You can test it on my website :
http://www.whitewater.be/resources/javascript/view.asp?
action=view&id=78
Arjan Kemeling
[New Comment - 6-18-2003]
Using the Content-Disposition header correctly should force IE and
other
browsers to treat the file as an attachment/download. I do it all the
time. Make sure you specify attachment as the type. ex:
Content-Disposition: attachment; filename="file.pdf"
Obviously this is easiest if you are dynamically outputing the contents
of the PDF from a server script script (like Perl, PHP, ASP, etc).
IAN VINK
--------------------
in C# with .NET you can do it like this:
string downloadPath ="c:\blah blah blah";
Response.ContentType = "APPLICATION/OCTET-STREAM";
Response.AddHeader( "Content-Disposition", "attachment;filename=\"" +
downloadPath + "\"" );
StreamWriter w = new StreamWriter(Response.OutputStream);
w.Write(xml);
[New Comment - 01-26-2006]
Andre White
Content-Disposition seems to work well (i.e. C# example above)
However, MSIE seems to ignore this information and sniff to guess the
true filetype. Although I've found no official documentation from MS,
it appears that MSIE fails to 'guess' the correct type when running IIS
6.0 with compression enabled. Mozilla / Opera properly use the Header
information supplied. I have yet to find a work around.
