I have a Firefox component for PDF signing that I invoke via Javascript. The problem is that this component outputs the signed PDF as a file on the user's filesystem - there is no way to get a byte[], stream or similar.
I need to post this signed PDF file back to the server. Is it possible to do this in plain Javascript, without additional Firefox components?
You can't do this without under intervention, this would he a huge security hole. Think about visiting a webpage and it being able to grab and upload any of your files without you doing a thing...you can see how this would be abused really fast.
You might be able to do this with a Firefox extension, I'm not sure of the security limitations it imposes (though I wouldn't be surprised if it disallowed this as well), but JavaScript would not be an option here.
There's no direct way to access data on a user's file system within a browser unless you're doing it through an extension. This would be a huge security risk.
If the file can be put into an <input type="file" /> element in Firefox, you can use JavaScript to automatically upload the data to the server without user intervention, but you would need something to initially get the file into the page's memory, not the file system.
Hell no - And for very good reason. You wouldn't want every Tom, Dick and Harry automatically grabbing all your files.
I believe it's the file input that needs user interaction. You can post the form automatically but you can't fill it in.
IMO your best bet is replacing the whole Firefox component with a single Java applet - I'd all but guarantee there's a Java project for PDF signing. Perhaps even Flash or Silverlight as long as you can do what you need with your PDF.
Related
I want to allow a web site users to be able to download files from my site, but with the help of a client-side downloader with an ability to continue an interrupted download.
For example, I want to sent a person a file with a size of 30+ Meg. I want the user to have the best downloading experience, so I can't afford him downloading 25 Meg and then getting the download dropped due to the network problems on his side.
Therefore, I want to have a javascript downloader rendered on a download page, that will show the actual client-side file delivery, and when it is downloaded, to give an ability to a user to save the file.
Or is it not possible due to the fact that javascript won't be able to open a save file dialog and save to a file system?
I'm afraid that is not possible with JavaScript and that's why:
To continue downloading from the certain point you should send to the server the position number to start downloading from. And as JavaScript has no access to local file system, you can't get that position.
UPD: it seems that I was too hurrying with the reply.
The file size can be gotten using the HTML5 File API and after getting the file size you can pass it to the server which should support the partial downloading.
But anyway, after downloading another part of the file you should sew two pieces together in some way; standard web browser dialog will only suggest to overwrite the file.
UPD2: to work with files in some Internet Explorers you can use FileSystemObject:
var fso;
fso = new ActiveXObject("Scripting.FileSystemObject");
I'd look into making a plugin or extension. Like those DownloadThemAll extensions for firefox and Google chrome. Another alternative would be to use Flash, either alone or integrating it with javascript like hinted here: http://www.communitymx.com/content/article.cfm?cid=0922A
I have a web page which asks users to select a zip file on their machine and upload it. It also asks for a destination on the server where the zip file should go. I want it to work in such a way that when they tell me where the file is on their machine and where it belongs on my server, then click "send," it should be sent to my server, placed in the designated directory, and unzipped.
I want to do this in just HTML and JavaScript. Is that possible? If so, how would I go about doing it?
You have to do this serverside; the client's browser cannot write to the file system of the server directly as it's a huge security risk.
In addition to that, not even HTML5 with it's File API can extract ZIP files. You need to do it on the server with PHP, or whatever language you're using.
If you just want to do things locally on the client, you could consider a Flash/Java/Browser extension, but I wouldn't recommend it for compatibility and performance reasons. Your best bet is to send a request to the server for it to process and send back. You're already serving the HTML page, so you can use the same server to process the ZIP file.
If, on the other hand, you want to write the ZIP file to the server, you have to do it server side for reasons stated in my first paragraph.
You can unzip/zip content within JavaScript using rawdeflate.
As other posters suggest, you cannot save directly to the local filesystem using a plain browser. You either need to echo the result back off the server (in which case it may as well unzip), or you can use Flash, for example, Downloadify, if it is installed.
Most modern browsers support FileAPI for loading from the local filesystem, certainly, Chrome, Firefox and Opera do. I haven't tested IE 9.
I am trying to do a simple thing:
Let the user choose a txt file, and save its context to be used on the client side only.
no server side needed.
Is it possible ?
Thanks.
It is possible to do so with HTML5 Files API as explained in these resources:
https://developer.mozilla.org/en-US/docs/Using_files_from_web_applications
http://www.html5rocks.com/en/tutorials/file/dndfiles/
I guess you mean "save its content" and conclude you want to do anything with this content on the client side, e.g. extract some parts to fill a form. Anyway saving the whole file unchanged, on the same machine where it comes from, does not make sense.
So the problem is not how to upload, but how to open/read a file. You can do this with a Java Applet, Flash, Silverlight, ActiveX ... just to name a few.
JavaScript is not an option. It cannot access the file system.
If the html page, that is hosting your javascript, is from a remote server. This script is not trusted to do actions on your local filesystem.
<Obscure solution mode level = 1>
You can give more trust to a page, but this is something your user has to do. If this is an app/web only for use within an enterprise, you can probably do this centrally. And every browser handles this differently. So it is not something you can rely on, when you do not have a limited userbase.
<Obscure solution mode level = high>
If your (enterprise) users are using Internet Explorer, you could also create a HTML Application (simply give your html page an hta extension). These pages have full trust, but can only be started from a trusted location, or require confirmation from the user.
The only way you can acheive this successfully is to build an ActiveX type plugin/component (or java applet) you will have much more control of the client machine.
No. JavaScript cannot access the local filesystem.
However, you could install a webserver on your machine and e.g. run PHP on that one. Then you could do it without ever sending your data over a network connection. That would require you to do your data processing in PHP though.. probably not what you want. Or you could simply send back the data to your javascript.. but that'd be pretty awful to run an upload just to make the data available to JavaScript.
Is it possible to read the contents of an uploaded file (through <form><input type="file">) from inside javascript with no backend? I suppose it's possible with HTML5, but what about HTML4?
You can read (not upload - well, you can upload too, but that's like sending fan-mail to Edgar Allen Poe; nothing stops you but he won't be able to read it) from files if your script runs in a very trusted context.
If anyone on your network is browsing the web in a browser set to give that much trust, disconnect their machine from the network first, make their settings tighter second.
I don't know if I understand correctly. How about using something like uploadify together with jQuery. I guess you'll have to grant the webserver the privilege to write in the specified folders.
The security model will not let you unless you change the security to allow this.
http://www.mozilla.org/projects/security/components/signed-scripts.html
Can you elaborate on what you want to do?
I read the question as "How do I grab the contents from a file chosen using a file upload field and process it on the client without involving a backend/webserver".
I am thinking UniversalBrowserRead access for FF and HTA filesystem for IE for example
I need a Javascript sample which reads contents from a url and creates a file in the client with same contents when some button is clicked. I also understand that Javascript cannot be allowed to access the local file system (Unless you use ActiveX objects). I do not want to rely on ActiveX objects (since my client is not IE always).
So here is what I did. I used the standard XmlHttpRequest made a request and got my bytes. Now I thought I could kinda stream this contents to the user, first by opening a standard dialog box (the box that pops up when you attempt download something from internet with options like Open/Save/Cancel) and then asking the user to save it somewhere.
I know how to do the read part, can someone show some javascript/html sample on "How to stream open a confirm dialog box to the user and stream some contents?"
PS: Not too sure if this could be done in javascript at all but with Javascript you never know what is possible and what is not :)
Rather than using Javascript to stream the content, which is not possible for obvious security reasons, you need to point the browser at a URL that will return a 'Content-Type' header of 'application/octet-stream'. In most cases this will force the browser to initiate a 'save as' operation and ask the user what to do with it.
I believe it is possible to do this using an iframe in the same page, such that the user will not have to navigate away from the page or open a new tab/window.
See it this way: if you could read/write to users' computers using JavaScript, then no computer would be safe from browsing the web.
Having said so, you cannot read/write client-side files using JavaScript.
But you could if you use Flash / a Java applet; in those cases the embedded objects asks for your permission before doing such actions.
You may want to have a look at TiddlyWiki that claims it can write itself to the disk...
If I understand what your question is, then you want to use JavaScript to write data like a server-side script would (PHP, Python, Java, etc..) but in browser?
If so, then what your asking isn't possible with in browser JavaScript.
However if for some reason you wanted to do this with server-side JavaScript, then yes the "streaming" part is possible.