|
Santiago M. Pericas-Geertsen
XML-Fluent Mobile Ambients
PhD thesis, Boston University, 2001
The Internet is considered to be among the greatest inventions of the
twentieth century. Despite its immense importance we, programming language
designers, have yet to propose a programming language that can exploit a
network that spans the entire globe.
In this thesis, we introduce a language for a programming model based on
XML-speaking mobile agents. The thesis is divided into three parts: the
first part defines the mobility aspect of the language, the second
part introduces a typed calculus for manipulating XML documents, and the
third part proposes a way of integrating these two calculi. The final goal
is to design a typed language in which computation can be carried out by
mobile agents exchanging data in the form of XML documents.
The Ambient Calculus was developed by Cardelli and Gordon as a formal
framework to study issues of mobility and migrant. In Part I, we consider
an extension to the Ambient Calculus where ambients are equipped with a set
of local channels over which processes can communicate. This extension
incorporates certain features from the Pi-Calculus, such as the ability
to communicate over higher-order channels, but without violating design
principles of the Ambient Calculus (e.g., that two processes are unable to
communicate if located in different administrative domains). For the
extended calculus, we introduce an appropriate operational semantics and a
provably sound type system. We conclude this part by showing an encoding
of the Ambient Calculus into our calculus that is both semantics
preserving and type preserving.
XML is a simple and portable way of representing structured data in the
form of trees. It is an ideal language for machine-to-machine (or
peer-to-peer) exchange of data because (i) it is portable, so there is no
need to worry about things like integer representations (big endian
vs. little endian) or string representations (null terminated vs. fixed
length) and (ii) structured data is easily represented in the form of
trees. Interestingly, the ability of XML to represent almost any form of
structured data is often an important weakness at the same time: for a
data exchange to be successful, a producer and a consumer must agree on
the structure of the exchanged documents so that the latter can
find/extract what it needs.
This problem can be addressed by the use of a type system. A type
system is needed in order to keep track of the meta-data (i.e. data about
the data) which can be used by a consumer as a guideline for data
extraction. A number of type systems for XML have been proposed in recent
years. Perhaps surprisingly, specially if we look at the evolution of
programming languages, the use of a type system in XML (as well as in
languages for manipulating XML documents) is not mandatory.
Domain-specific languages for XML proposed in the last few years are all,
with the exception of XDuce, dynamically typed. Experience has showed us that
statically-typed languages are better suited for real-world applications than
dynamically-typed ones. Despite requiring more from the programmer's
perspective (at least in the absence of a type-inference engine) the
benefits greatly outcast the requirements. Specifically, the use of a
statically-typed language ensures that the final product is free of an
important number of errors (so-called type errors). In Part II, we
present the dlcalculus, a statically-typed calculus for manipulating XML
documents, and show how it can be integrated with existing XML technology
by means of two algorithms for importing and exporting documents into the
calculus.
[ bib |
.ps |
.pdf ]
Back
This file has been generated by
bibtex2html 1.61
Copyright notice: The documents contained
in these pages are included by the contributing authors as a means to
ensure timely dissemination of scholarly and technical work on a
non-commercial basis. Copyright and all rights therein are maintained
by the authors or by other copyright holders, notwithstanding that
they have offered their works here electronically. It is understood that all persons copying this information will
adhere to the terms and constraints invoked by each author's
copyright. These works
may not be reposted without the explicit permission of the copyright
holder.
If you experience problems downloading any of the files above,
it is most likely because your browser does not handle compressed
files correctly.
In particular, Netscape might save the file in the compressed
gz-format with extension .ps or
.pdf (indicating postscript or PDF, resp.). You can work around this by saving the file,
renaming it to .ps.gz or .pdf.gz, and then
uncrompress it.
|
