Diesel-Engine Documentation |
|
Introduction back to top -
When Diesel-Engine executes a Diesel file, it needs to parse the
Diesel-Script commands and the HTML. When the source file is parsed,
Diesel-Engine creates a "parse tree" of the logic in the page. This parse
tree is taken and executed apon to produce the dynamical page content.
The parsing of Diesel files is very quick, but is often the slowest step
in the generation of the content. If the Diesel pages are not changing
very often, it might seem a waste of resources to have to parse the same
Diesel files for every request. Diesel addresses this with "Object
Caches". Diesel can write out its "parse tree" into byte code. The byte
code of all the pages in a site are then stored in a single flat file
read-only database. Essentially, the entire site is parsed and the byte
code is compiled into a single application file used to run the web
application. When Diesel-Engine processes a request, the page is looked up
in the application database and the Diesel byte code is loaded and
executed.
Diesel Object Caches back to top -
A Diesel Object Cache is created by the 'dsl2ops' command line utility
described in detail below. The utility creates a single file which
contains an entire directory structure of Diesel pages, parsed into byte
code and indexed in a read-only CDB database.
To configure Diesel-Engine to use a Diesel Object Cache, simply create a
.htaccess file in the directory containing the 'diesel-ops.cdb' file. The
contents of .htaccess is:
DieselOpsCache diesel-ops.cdb
The directory containing the 'diesel-ops.cdb' file must be in the web
server document directory tree. Any Diesel file requested which resolves
to a directory in or below the directory containing the 'diesel-ops.cdb'
database file is matched against the database for a match. If found, the
byte code in the database file is used (even if the actual file exists). The dsl2ops Program back to top -
To create a 'diesel-ops.cdb' database file, use the command:
dsl2ops -r
in a directory containing Diesel files and other directories. Using the
'-r' option, directories are recursively scanned and all Diesel files are
parsed and will be saved in the database as byte code. It is not required
to have the actual pages themselves present when using a Object Cache. The
source files can be on one machine which produces a 'diesel-ops.cdb'
file. This file contains the entire application and can be put on
production web servers as the only content required to run the site
(images included).
|
|
Copyright © 2005 Blueprint Networks, Inc.
|
