Example Python Code showing cURLing data and POSTing data to Elastic Search, but fails with escaped speech marks

pleaseEscape.py

############# My Clusters Health

curl -XGET 'http://127.0.0.1:9200/_cluster/health?pretty=true'
{
  "cluster_name" : "TweetHadoop",
  "status" : "yellow",
  "timed_out" : false,
  "number_of_nodes" : 1,
  "number_of_data_nodes" : 1,
  "active_primary_shards" : 15,
  "active_shards" : 15,
  "relocating_shards" : 0,
  "initializing_shards" : 0,
  "unassigned_shards" : 15
}

############# Works fine for posting data

import urllib
import urllib2

url = 'http://localhost:9200/twitter/tweet/22499999'
data = '{"user" : "helenax33","message" : "LIVE: http://www.justin.tv/xxhelenaxx i look like a can read+sound like a man.","pubDate" : "20090705T21:46:34","isAQuestion" : "0" }'

#data = urllib.urlencode(values)
req = urllib2.Request(url, data)
response = urllib2.urlopen(req)
the_page = response.read()


##### Also this works Fine for posting data.

import pycurl
apiURL = 'http://localhost:9200/twitter/tweet/22499999'
c = pycurl.Curl()
c.setopt(c.URL, apiURL)
c.setopt(c.POSTFIELDS, '{ "user" : "helenax33", "message" : "LIVE: http://www.justin.tv/xxhelenaxx i look like a can read+sound like a man.", "pubDate" : "20090705T21:46:34", "isAQuestion" : "0" }')
c.setopt(c.VERBOSE, True)
c.perform()


# However, if I want to put a speech mark (") in the message part, then it fails e.g.

import pycurl
apiURL = 'http://localhost:9200/twitter/tweet/22499999'
c = pycurl.Curl()
c.setopt(c.URL, apiURL)
c.setopt(c.POSTFIELDS, '{ "user" : "helenax33", "message" : "LIVE: http://www.justin.tv/xxhelenaxx i " look like a can read+sound like a man.", "pubDate" : "20090705T21:46:34", "isAQuestion" : "0" }')
c.setopt(c.VERBOSE, True)
c.perform()

# So I tried to escape the speech mark e.g.

import pycurl
apiURL = 'http://localhost:9200/twitter/tweet/22499999'
c = pycurl.Curl()
c.setopt(c.URL, apiURL)
c.setopt(c.POSTFIELDS, '{ "user" : "helenax33", "message" : "LIVE: http://www.justin.tv/xxhelenaxx i \" look like a can read+sound like a man.", "pubDate" : "20090705T21:46:34", "isAQuestion" : "0" }')
c.setopt(c.VERBOSE, True)
c.perform()

# However, this still fails. Please help :(

Ah. Looks from that that you're escaping the whole JSON string. And that's confusing it. The JSON string's quote marks should remain unescaped - it's the quote marks INSIDE the message you want to sort out. So re.escape(message) and then interpolate that into your JSON string. Hopefully you'll have a winner :-)

i.e.

{ "user" : "someusername", "message" : "something they are tweeting about that contains \"speech marks\".", "pubDate" : "20090705T21:46:34"}

Hey Shaun, Yup tried it, doesn't work. :( Going to replace " and 's with some 'unused' unicode character then replace them back to speech marks and apostrophes when displayed to users. Not ideal, but should work. Cheers for your input though. Jon

…

On 21 June 2012 11:18, Sean Handley < ***@***.*** > wrote: i.e. { "user" : "someusername", "message" : "something they are tweeting about that contains \"speech marks\".", "pubDate" : "20090705T21:46:34"} --- Reply to this email directly or view it on GitHub: https://gist.github.com/2960359

No, not ideal. Surely there's a python json lib that will parse/encode for you? In ruby you can just say .to_json and it works (escaping all dodgy chars also).