# Changeset 3880 for docs

Ignore:
Timestamp:
Jun 17, 2014, 7:55:53 PM (5 years ago)
Message:

Various formatting items

Location:
docs/Working/re
Files:
7 edited

Unmodified
Removed
• ## docs/Working/re/avx2.tex

 r3875 \paragraph*{AVX2 256-Bit Addition} \begin{figure}[tbh] \end{figure} Bitstream addition at the 256-bit block size was implemented using the \paragraph*{AVX2 256-Bit Addition} Bitstream addition at the 256-bit block size was implemented using the long-stream addition technique.   The AVX2 instruction set directly supports the \verb#hsimd<64>::mask(X)# operation using
• ## docs/Working/re/compilation.tex

 r3876 \paragraph{Compilation} Using the marker stream and MatchStar concept, we now \paragraph*{Compilation} Using the marker stream and MatchStar concept, we now outline our compilation algorithm.  This is implemented in a Java program.   First the regular expression is
• ## docs/Working/re/conclusion.tex

 r3870 \section{Discussion}\label{sec:Concl} \paragraph*{Contributions} A new class of regular expression matching algorithm has been \paragraph*{Contributions} A new class of regular expression matching algorithm has been introduced based on the concept of bit-parallel data streams together with the MatchStar operation.  The algorithm is effort. \paragraph{Related Work} Much of the previous work in parallelizing of regular processing \paragraph*{Related Work} Much of the previous work in parallelizing of regular processing has dealt with the problem of using parallel resources to handle multiple instances of a matching problem in parallel.   It is thus of 5X or more. \paragraph*{Ongoing and Future Work} Based on the techniques presented here a fully integrated \paragraph*{Ongoing and Future Work} Based on the techniques presented here a fully integrated grep version with a dynamic code generator implemented with LLVM is being developed by another team working with the Parabix
• ## docs/Working/re/pact051-cameron.tex

 r3878 % % 1st. author \alignauthor Rob Cameron\\ %\alignauthor Rob Cameron, Tom Shermer, Arrvindh Shriraman, Ken Herdy, Dan Lin, Ben Hull, Meng Lin\\ \affaddr{School of Computing Science}\\ \affaddr{Simon Fraser University}\\ \affaddr{Surrey, British Columbia}\\ \email{cameron@cs.sfu.ca} % 2nd. author \alignauthor Tom Shermer\\ \affaddr{School of Computing Science}\\ \affaddr{Simon Fraser University}\\ \affaddr{Surrey, British Columbia}\\ \email{shermer@cs.sfu.ca} % 3rd. author \alignauthor Arrvindh Shriraman\\ \affaddr{School of Computing Science}\\ \affaddr{Simon Fraser University}\\ \affaddr{Surrey, British Columbia}\\ \email{ashriram@cs.sfu.ca} \and  % use '\and' if you need 'another row' of author names % 4th. author \alignauthor Ken Herdy\\ \affaddr{School of Computing Science}\\ \affaddr{Simon Fraser University}\\ \affaddr{Surrey, British Columbia}\\ \email{ksherdy@sfu.ca} % 5th. author \alignauthor Dan Lin\\ \affaddr{School of Computing Science}\\ \affaddr{Simon Fraser University}\\ \affaddr{Surrey, British Columbia}\\ \email{lindanl@sfu.ca} % 6th. author \alignauthor Ben Hull\\ \affaddr{School of Computing Science}\\ \affaddr{Simon Fraser University}\\ \affaddr{Surrey, British Columbia}\\ \email{bhull@sfu.ca} % 7th. author \alignauthor Meng Lin\\ \affaddr{School of Computing Science}\\ \affaddr{Simon Fraser University}\\ \affaddr{Surrey, British Columbia}\\ \email{linmengl@sfu.ca} \email{\{cameron,shermer,ashrira,ksherdy,lindanl,bhull,linmengl\}@sfu.ca} } % There's nothing stopping you putting the seventh, eighth, etc. % author on the opening page (as the 'third row') but we ask, % for aesthetic reasons that you place these 'additional authors' % in the \additional authors block, viz. %\additionalauthors{Additional authors: John Smith (The Th{\o}rv{\"a}ld Group, %email: {\texttt{jsmith@affiliation.org}}) and Julius P.~Kumquat %(The Kumquat Consortium, email: {\texttt{jpkumquat@consortium.net}}).} %\date{30 July 1999} % Just remember to make sure that the TOTAL number of authors % is the number that will appear on the first page PLUS the % number that will appear in the \additionalauthors section. \paragraph*{MatchStar} MatchStar takes a marker bitstream and a character class bitstream as input.  It returns all positions that can be reached by advancing the marker bitstream zero or more times through the character class bitstream. \paragraph*{MatchStar}  MatchStar takes a marker bitstream and a character class bitstream as input.  It returns all positions that can be reached by advancing the marker bitstream zero or more times through the character class bitstream. \begin{figure}[tbh] and then extracting the carry-out {\tt q} from the high bit position. $\text{\tt i} = \text{\tt MatchStar(c*2+p, b)}$ $\text{\tt q} = \text{\tt i >> f}$ $\text{\tt q} = \text{verb:i >> f:}$ As described subsequently, we use a two-level long-stream addition technique
• ## docs/Working/re/re-Unicode.tex

 r3876 \paragraph*{Unicode} The introduction of Unicode as a common encoding system including \paragraph*{Unicode}  The introduction of Unicode as a common encoding system including the characters of all the world's written languages and notation systems has introduced some complexity for regular expression matching
• ## docs/Working/re/sse2.tex

 r3876 \paragraph{Implementation Notes} Our regular expression compiler directly uses the Parabix tool chain \paragraph*{Implementation Notes} Our regular expression compiler directly uses the Parabix tool chain to compile regular expression into SSE2-based implementations. Our compiler essentially scripts three other compilers to perform as they are encountered. \paragraph{Comparative Implementations} We evaluate our bitwise data parallel implementation versus several \paragraph*{Comparative Implementations} We evaluate our bitwise data parallel implementation versus several alternatives. We report data for two of these: \paragraph{Test Expressions} Each grep implementation was evaluated \paragraph*{Test Expressions} Each grep implementation was evaluated against the five regular expressions shown in Table \ref{RegularExpressions}. in matching the simplest possible regular expression, a single character. Date, Email, and URIOrEmail provide examples of commonly used regular expression. This set of expressions were drawn from the \textit{Benchmark of Regex This set of expressions were modified from the \textit{Benchmark of Regex Libraries} (http://lh3lh3.users.sourceforge.net/reb.shtml). HexBytes matches delimited byte strings in hexadecimal notation, and expression illustrates the performance of a repetition operator implemented using a while loop. a while loop.  StarHeight is an expression designed to further stress our while loop implementation with 4 levels of Kleene closure. All tests were run on a version of a \textit{Linux 3Dfx howto} \paragraph{Results} \paragraph*{Results} \begin{figure} \begin{center}
Note: See TracChangeset for help on using the changeset viewer.