Ticket #7441: trac_7441_upload.patch

sagenb/data/sage/html/upload.html

@@ -15 +15 @@
             <input size="50" type="file" name="file" />
         </div>
         <div>
-            <label for="url">Or enter the URL of a worksheet file on the web:</label>
+            <label for="url">Or enter the URL of a worksheet file on the
+              web. This can be a .sws file, a published worksheet, or
+              any webpage with an associated worksheet:</label>
             <input size="50" type="text" name="url" />
         </div>
         <div>

sagenb/notebook/twist.py

@@ -443 +443 @@
     def render(self, ctx):
         backlinks = """ Return to <a href="/upload" title="Upload a worksheet"><strong>Upload File</strong></a>."""
 
+        print 'ctx.args:', ctx.args
         url = ctx.args['url'][0].strip()
         dir = ''  # we will delete the directory below if it is used
-        if url != '':
+        if url.endswith('.sws'):
             # downloading a file from the internet
             filename = tmp_filename()+".sws"
+        elif url != '':
+            # regular web page with associated worksheet
+            filename = tmp_filename()+".html"
         else:
             # uploading a file from the user's computer
             dir = tmp_dir()
@@ -466 +470 @@
         # We make a callback so that we can download a file remotely
         # while allowing the server to still serve requests.
         def callback(result):
+            print '_dd_: in callback, filename:', filename
 
             if ctx.args.has_key('name'):
                 new_name = ctx.args['name'][0].strip()
@@ -474 +479 @@
 
             try:
                 try:
-
                     if filename.endswith('.zip'):
                         # Extract all the .sws files from a zip file.
                         zip_file = zipfile.ZipFile(filename)
@@ -487 +491 @@
                                     W.set_name("%s - %s" % (new_name, W.name()))
                         return RedirectResponse('/')
 
+                    elif filename.endswith('html'):
+                        # look through source for appropriate link_rel
+
+                        from HTMLParser import HTMLParser
+                        class GetLinkRelWorksheets(HTMLParser):
+                            def __init__(self):
+                                HTMLParser.__init__(self)
+                                self.worksheets = []
+
+                            def handle_starttag(self, tag, attrs):
+                                if (tag == 'link' and
+                                    ('rel', 'alternate') in attrs and
+                                    ('type', 'application/sage') in attrs):
+                                    self.worksheets.append({'title': [_ for _ in attrs if _[0] == 'title'][0][1],
+                                                              'url': [_ for _ in attrs if _[0] == 'href'][0][1]})
+
+                        parser = GetLinkRelWorksheets()
+
+                        with open(filename) as f:
+                            parser.feed(f.read(50000))
+
+                        # did we find any worksheets? If so, we
+                        # (arbitrarily) only download the first one. I
+                        # want to just call self.render() with an
+                        # appropriate context, but I don't know how to
+                        # do that. :(
+                        # TODO: fix this up, so relative and absolute
+                        # urls are okay
+                        # TODO: url encode this and do joining of
+                        # relative URls properly
+                        import urllib
+                        print 'url:', url
+                        print 'from link rel:', parser.worksheets[0]['url']
+                        u = url + urllib.quote(urllib.unquote(parser.worksheets[0]['url']))
+                        print 'u:', u
+                        ctx.args = {'url': [u], 'name': ['']}
+                        print 'about to recurse...'
+                        self.render(ctx)
                     else:
+                        print '_dd_: got a filename of', filename
                         W = notebook.import_worksheet(filename, self.username)
 
                 except Exception, msg:
-                    s = 'There was an error uploading the worksheet.  It could be an old unsupported format or worse.  If you desperately need its contents contact the <a href="http://groups.google.com/group/sage-support">sage-support group</a> and post a link to your worksheet.  Alternatively, an sws file is just a bzip2 tarball; take a look inside!%s' % backlinks
+                    # TODO: put in <p> tags below
+                    s = 'There was an error uploading the worksheet.  It could be an old unsupported format or worse.  If you desperately need its contents contact the <a href="http://groups.google.com/group/sage-support">sage-support group</a> and post a link to your worksheet. Alternatively, an sws file is just a bzip2 tarball; take a look inside!<br />The error was:<br />%s<br /><br />%s' % (msg, backlinks)
                     return HTMLResponse(stream=message(s,'/'))
                 finally:
                     # Clean up the temporarily uploaded filename.
-                    os.unlink(filename)
+                    # FIXME TODO uncomment this later           os.unlink(filename)
                     # if a temp directory was created, we delete it now.
                     if dir:
                         shutil.rmtree(dir)
@@ -512 +556 @@
             return RedirectResponse('/home/'+W.filename())
 
         if url != '':
+            print '_dd_ got a url of ', url
             # We use the downloadPage function which returns a
             # deferred which we are allowed to return to the server.
             # The server waits until the download is finished and then runs
@@ -521 +566 @@
             d.addCallback(callback)
             def errback(result):
                 msg = "There was an error uploading '%s' (please recheck the URL).%s" % (url, backlinks)
+                msg += str(result)
                 return HTMLResponse(stream=message(msg,'/'))
             d.addErrback(errback)
             return d