Welcome, guest | Sign In | My Account | Store | Cart

serving dynamically-generated PDF files via XML/RPC (Python recipe) by gian paolo ciceri
ActiveState Code (http://code.activestate.com/recipes/413404/)

It's only a matter to put together Reportlab to generate a pdf file on the fly and Cherrypy xmlrpc filter to serve it as an XML/RPC binary object (base64-encoded).

Python, 51 lines
Download
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
#################################################################
#################################################################
import xmlrpclib
from cherrypy.lib.filter.xmlrpcfilter import XmlRpcFilter
from cherrypy import cpg
from reportlab.pdfgen import canvas


class Root:
  _cpFilterList = [XmlRpcFilter()]

  def _cpOnError(self):
    import traceback, StringIO
    bodyFile = StringIO.StringIO()
    traceback.print_exc(file = bodyFile)
    errorBody = bodyFile.getvalue()
    if cpg.request.isRPC: 
      ## isRPC boolean is set on xml-rpc requests by the filter
      ## convert the traceback to a dumped Fault object: 
      ## the XML-RPC exception
      cpg.response.body = [xmlrpclib.dumps(xmlrpclib.Fault(1,errorBody))]
    else:
      ## handle regular web errors
      cpg.response.body = ['<pre>%s</pre>' % errorBody]
      
  def getPdf(self,aMsg):
    c = canvas.Canvas(None)
    c.drawString(100,100,aMsg)
    c.showPage()
    c.save()
    return xmlrpclib.Binary(c.getpdfdata())
   
  getPdf.exposed = True 

cpg.root = Root()
if __name__=='__main__':
  cpg.server.start(configMap = {'socketPort': 9001,
                                'threadPool':0,
                                'socketQueueSize':10 })
########################################################
########################################################

# a simple client ######################################
########################################################
import xmlrpclib
server = xmlrpclib.ServerProxy('http://localhost:9001')
f = open('helloWorld.pdf','w')
f.write(server.getPdf('Hello World!').data)
f.close()
########################################################
########################################################
Tags: distributed

2 comments

Graham Dumpleton 18 years, 11 months ago  # | flag

Even simpler using mod_python and Vampire. Even simpler when using mod_python and Vampire, as you don't need to add any special error handling code which determines whether it was an XML-RPC request or not.

from reportlab.pdfgen import canvas
import xmlrpclib
import vampire

class _Service:
  def getPdf(self,aMsg):
    c = canvas.Canvas(None)
    c.drawString(100,100,aMsg)
    c.showPage()
    c.save()
    return xmlrpclib.Binary(c.getpdfdata())

handler = vampire.Service(_Service())

Setting the "handler" to be an actual service object, automatically ensures that all the appropriate XML-RPC specific work is done. The object being wrapped as a service can be an existing object which knows nothing about mod_python, whereas in the original example the XML-RPC code is intertwined with the CherryPy code.

Mod_python is available from "http://www.modpython.org" and Vampire from "http://www.dscpl.com.au/projects/vampire".

Vasudev Ram 14 years, 1 month ago  # | flag

Another approach is shown below. I've not evaluated whether it is "simpler" or "better" than above ones, but on one hand it does not require mod_python or Vampire, on the other it does require xtopdf (my toolkit for PDF creation / conversion):

Just use PDFXMLRPC. It is a pair of programs, a client and a server, for PDF creation from text, over the Internet, using XML-RPC. It is both a pair of applications and a pair of libraries (one of each for the client and server sides).

Get it here:

http://www.dancingbison.com/products.html#PDFXMLRPC

xtopdf, which PDFXMLRPC requires, can be got here:

http://www.dancingbison.com/products.html#xtopdf

xtopdf in turn requires ReportLab like the recipes above.

It is also possible to do the text-to-PDF conversion over XML-RPC without using xtopdf, by just using ReportLab and XML-RPC, in a similar way to the technique used in PDFXMLRPC to wrap xtopdf's methods - just apply the same technique to Reportlab's methods.