RESTful Web Services Example 2.1 - Yahoo! Web Search

Example 2.1 of the RESTful Web Services book shows a simple REST Web Service client, for the Yahoo! Web Search service.

Since the book has a detailed description of the client, I'll just show the code:

Chapter2 class>>yahooWebSearch: aString
 | baseURI document term xml |
 "Code taken from RESTful Web Services. Copyright © 2007 O'Reilly Media, Inc.
 All rights reserved. Used with permission."
 baseURI := 'http://api.search.yahoo.com/WebSearchService/V1/webSearch'.
 term := aString encodeForHTTP.
 xml := HTTPSocket httpGet: baseURI , '?appid=restbook&query=', term .
 document := XMLDOMParser parseDocumentFrom: xml.
 document ResultSet Result Title
  do: [:element |
   Transcript
    show: (element contentString);
    cr]

The code may be found on the RWS repository on SqueakSource and is included in the RWS-gc.2 package.

If you confront the Squeak code with the original Ruby code from the book, you'll notice a couple of differences. First of all, this method takes a single String argument, which can have either one search key or more than one, separated by spaces. So the #yahooWebSearch: method actually corresponds to the #print_page_titles method in the Ruby source.

Another difference may be seen on lines 5-6 of this code snippet. Where the Ruby example uses open-uri and #open():

require 'open-uri'
#...
xml = open(BASE_URI + "?appid=restbook&query=#{term}").read

# Parse the XML document into a data structure.
document = REXML::Document.new(xml)

I use the HTTPSocket class and its #httpGet: message. This message returns a stream that may be passed to XMLDOMParser>>parseDocumentFrom:. But the most glaring difference is on line 7: instead of the XPath expression used in the Ruby code, I use

document ResultSet Result Title do: [...]

that's because Squeak's XPath library, being just a 0.1.1 release, has some problems with all but the most simple XPath expressions. For this reason I concocted a small hack that allows me to navigate a DOM structure by sending messages whose selectors are the same as the XML nodes I'm interested into. In the future, either I'll build a plugin to connect to Libxml2, or I'll build a full-blown, pure-Squeak, higher-order-messaging-based system.

For the next post in this series, I'll show a Squeak version of the del.icio.us client described in chapter 2.