http://mw.hh.se/wg211/index.php?action=history&feed=atom&title=WG211%2FM10Siek 2026-04-05T21:09:07Z Revision history for this page on the wiki MediaWiki 1.43.5 http://mw.hh.se/wg211/index.php?title=WG211/M10Siek&diff=61&oldid=prev 2011-12-12T10:06:24Z

<p>1 revision</p> <p><b>New page</b></p><div>[[Category:WG211]]<br /> <br /> <br /> <br /> &#039;&#039;&#039;Composing DSLs&#039;&#039;&#039; by Jeremy Siek<br /> <br /> Domain-specific languages offer domain-appropriate syntax that lowers&lt;br /&gt;the learning curve for domain experts. However, large software&lt;br /&gt;applications often use of several domain-specific languages and suffer&lt;br /&gt;from the language interoperability problem. The alternative approach&lt;br /&gt;of embedding domain-specific languages in a general purpose language&lt;br /&gt;solves the interoperability problem, but compromises the&lt;br /&gt;domain-appropriate syntax.&lt;br /&gt;&lt;br /&gt;In this talk I propose an approach to parsing a program that consists&lt;br /&gt;of intertwined uses of multiple DSLs. This parsing problem is&lt;br /&gt;difficult because we do not assume that the DSLs have the same author,&lt;br /&gt;and as a result, the composition of their grammars is likely to be&lt;br /&gt;highly ambiguous. However, we observe that DSLs usually introduce new,&lt;br /&gt;unique types, and if those types are incorporated into each DSL&#039;s&lt;br /&gt;grammar rules, then most of the ambiguity goes away. (Our approach is&lt;br /&gt;inspired by the analogous notion of selection restriction in natural&lt;br /&gt;language processing.) We have explored type-based disambiguation&lt;br /&gt;within the context of an island parsing algorithm, a particular style&lt;br /&gt;of chart parser, and I present experimental results demonstrating that&lt;br /&gt;parse times are resilient to the number of DSLs being composed.</div>

Admin