# Changeset 901

Ignore:
Timestamp:
Feb 7, 2011, 7:19:13 PM (8 years ago)
Message:

Fixes for Tom's points.

Location:
docs/EuroPar2011
Files:
2 edited

Unmodified
Removed
• ## docs/EuroPar2011/europar-cameron.tex

 r900 Figure \ref{fig:decref} shows the parallel parsing of decimal references together with error checking. decimal references together with error checking. For clarity, the streams are now shown in left-to-right order as indicated by the $\vartriangleright$ symbol. The source data includes four instances of potential decimal references beginning with the \verb:&: character. indicating the positions of all detected errors. One question that may arise is: how are marker bitstreams initialized?   In general, this is an important problem, and dependent on the task at hand. Initialization of marker streams may be achieved in various ways, dependent on the task at hand. In the XML parsing context, we rely on an important property of well-formed error-check streams for simple post-bitstream processing or composite stream over which iterative stack-based procedures can be defined for checking recursive syntax. can be defined for checking recursive syntax.  To assess the completness of our analysis, we have confirmed that our implementations correctly handle all the well-formedness checks of the W3C XML Conformance Test Suite. \section{Compilation to Block-Based Processing} \label{sec:compile} While a Python implementation of the techniques described in the previous section works on unbounded bitstreams, a corresponding While our Python implementation of the techniques described in the previous section works on unbounded bitstreams, a corresponding C implementation needs to process an input stream in blocks of size equal to the SIMD register width of the processor it runs on.
Note: See TracChangeset for help on using the changeset viewer.