Comments on: Pecia: Towards a Fluent Interface for Building Documents http://blog.xebia.com/2009/06/25/fluent-interface-for-documentation/ Thu, 11 Mar 2010 16:23:01 +0100 http://wordpress.org/?v=2.8.4 hourly 1 By: Wilfred springer http://blog.xebia.com/2009/06/25/fluent-interface-for-documentation/comment-page-1/#comment-92372 Wilfred springer Sun, 30 Aug 2009 15:32:16 +0000 http://blog.xebia.com/?p=2202#comment-92372 Hi Anthony, Pecia does not return Strings; if you call .para() on an object, you cat a Para implementation upon which you can call other methods, for creating paragraph content. Those objects are basically shortly-living objects. In the current implementation of Pecia (found here: http://pecia.flotsam.nl/), data will be written to an StAX type of document writer as quickly as possible. Hi Anthony,

Pecia does not return Strings; if you call .para() on an object, you cat a Para implementation upon which you can call other methods, for creating paragraph content. Those objects are basically shortly-living objects. In the current implementation of Pecia (found here: http://pecia.flotsam.nl/), data will be written to an StAX type of document writer as quickly as possible.

]]> By: Anthony Gatlin http://blog.xebia.com/2009/06/25/fluent-interface-for-documentation/comment-page-1/#comment-92366 Anthony Gatlin Sat, 29 Aug 2009 07:44:56 +0000 http://blog.xebia.com/?p=2202#comment-92366 I think this is a very interesting idea but there are a few wrinkles that must be considered. As you understand, the primary idea behind method chaining is that every method returns a type from which any available methods on that type can be called. If the output of every method was a string, this would cause incredible performance problems since strings are immutable and have to be recreated with each modification. However, if the methods which were called returned some implementation of something like StringBuilder, then the final document would not have to be rendered until it was complete. All of the method chaining could occur on the StringBuilder object or its derivatives. Your idea has great promise. I would be interested in discussing this idea further. Thanks for sharing it! I think this is a very interesting idea but there are a few wrinkles that must be considered. As you understand, the primary idea behind method chaining is that every method returns a type from which any available methods on that type can be called. If the output of every method was a string, this would cause incredible performance problems since strings are immutable and have to be recreated with each modification. However, if the methods which were called returned some implementation of something like StringBuilder, then the final document would not have to be rendered until it was complete. All of the method chaining could occur on the StringBuilder object or its derivatives. Your idea has great promise. I would be interested in discussing this idea further. Thanks for sharing it!

]]>