lib/HTML/Mason
subdirectory of the distribution. After that, you can edit it by hand,
following the comments inside.
``make install'' copies Config.pm to your Perl library directory (e.g. /usr/lib/perl5/site_perl/HTML/Mason
) along with the other module files. This allows Mason internally to grab
the configuration data with ``use HTML::Mason::Config
''.
When upgrading from a previous version, ``make install'' will maintain the previous Config.pm values.
Currently this file controls:
o Whether or not certain optional modules, such as Time::HiRes, should be loaded for enhanced features.
o The type of DBM and the serialization method used for Mason's data caching. If you plan to use data caching, make sure that the DBM package is a good-quality one (DB_File or GDBM_File).
DocumentRoot /usr/local/www/htdocs PerlRequire /usr/local/mason/handler.pl DefaultType text/html <Location /> SetHandler perl-script PerlHandler HTML::Mason </Location>
handler.pl
creates three Mason objects: the Parser, Interpreter, and ApacheHandler.
The Parser compiles components into Perl subroutines; the Interpreter
executes those compiled components; and the Apache handler routes mod_perl
requests to Mason. These objects are created once in the parent httpd and
then copied to each child process.
These objects have a fair number of possible parameters. Only two of them are required, comp_root and data_dir; these are discussed in the next two subsections. The various parameters are documented in the individual reference manuals for each object: HTML::Mason::Parser, HTML::Mason::Interp, and HTML::Mason::ApacheHandler.
The advantage of embedding these parameters in objects is that advanced
configurations can create more than one set of objects, choosing which set
to use at request time. For example, suppose you have a staging site and a
production site running on the same web server, distinguishing between them
with a configuration variable called version
:
# Create Mason objects for staging site my $parser1 = new HTML::Mason::Parser; my $interp1 = new HTML::Mason::Interp (parser=>$parser1, ...); my $ah1 = new HTML::Mason::ApacheHandler (interp=>$interp1);
# Create Mason objects for production site my $parser2 = new HTML::Mason::Parser; my $interp2 = new HTML::Mason::Interp (parser=>$parser2, ...); my $ah2 = new HTML::Mason::ApacheHandler (interp=>$interp2);
sub handler { ...
# Choose the right ApacheHandler if ($r->dir_config('version') eq ' staging') { $ah1->handle_request($r); } else { $ah2->handle_request($r); } }
When Mason handles a request, the request filename ($r->filename
) must be underneath your component root -- that way Mason has a legitimate
component to start with. If the filename is not under the component root,
Mason will place a warning in the error logs and return a 404.
Unfortunately if your component root or document root goes through a soft
link, Mason will have trouble comparing the paths and will return 404. To
fix this, set your document root to the true path.
Starting in Mason 0.8 it is now possible to specify multiple component
roots to be searched in the spirit of Perl's @INC
. To do so you must specify a list of lists:
comp_root => [[key1, root1], [key2, root2], ...]
Each pair consists of a key and root. The key is a string that identifies the root mnemonically to a component developer. Keys are case-insensitive and must be distinct.
For example:
comp_root => [[private=>'/usr/home/joe/comps'], [main=>'/usr/local/www/htdocs']]
This specifies two component roots, a main component tree and a private
tree which overrides certain components. The order is respected ala @INC
, so private is searched first and main second. (I chose the =>
notation here because it looks cleaner, but note that this is a list of
lists, not a hash.)
The key has several purposes. Object and data cache filenames use the (uppercased) key to make sure different components sharing the same path have different cache and object files. For example, if a component /foo/bar is found in 'private', then the object file will be
<data_dir>/obj/PRIVATE/foo/bar
and the cache file
<data_dir>/cache/PRIVATE+2ffoo+2fbar
The key is also included whenever Mason prints the component title, as in an error message:
error while executing /foo/bar [private]: ...
This lets you know which version of the component was running.
cache: data cache files debug: debug files etc: miscellaneous files obj: compiled components preview: preview settings files
These directories will be discussed in appropriate sections throughout this manual.
{ package HTML::Mason::Commands; use CGI ':standard'; use LWP::UserAgent; ... }
In any case, for optimal memory utilization, make sure all Perl modules are used in the parent process, and not in components. Otherwise, each child allocates its own copy and you lose the benefit of shared memory between parent processes and their children. See Vivek Khera's mod_perl tuning FAQ (perl.apache.org/tuning) for details.
To work around this conflict, Mason remembers all directories and files
created at startup, returning them in response to
interp->files_written
. This list can be fed to a chown()
at the end of the startup
code in handler.pl
:
chown (scalar(getpwnam "nobody"), scalar(getgrnam "nobody"), $interp->files_written);
The file eg/session_handler.pl in the distribution contains the lines to activate cookie-based sessions using Apache::Session and CGI::Cookie. You can use eg/session_handler.pl as your new handler.pl base, or just copy out the appropriate pieces to your existing handler.pl.
The session code is customizable; you can change the user ID location (e.g. URL instead of cookie), the user data storage mechanism (e.g. DBI database), and the name of the global hash.
my
) variables in components, there is very little need for globals at all.
That said, there are times when it is very useful to make a value available
to all Mason components: a DBI database handler, a hash of user session
information, the server root for forming absolute URLs. Usually you
initialize the global in your handler.pl, either outside the
handler()
subroutine (if you only need to set it once) or
inside (if you need to set it every request).
Mason by default parses components in strict
mode, so you can't simply start referring to a new global or you'll get a
fatal warning. The solution is to invoke use vars
inside the package that components execute in, by default HTML::Mason::Commands:
{ package HTML::Mason::Commands; use vars qw($dbh %session); }
Alternatively you can use the allow_globals parameter or method:
my $parser = new HTML::Mason::Parser (..., allow_globals => [qw($dbh %session)]); $parser->allow_globals(qw($foo @bar))
The only advantage to allow_globals is that it will do the right thing if you've chosen a different package for components to run in (via the in_package Parser parameter.)
Similarly, to initialize the variable in handler.pl, you need to set it in the component package:
$HTML::Mason::Commands::dbh = DBI->connect(...);
Alternatively you can use the set_global Interp method:
$interp->set_global(dbh => DBI->connect(...));
Again, set_global will do the right thing if you've chosen a different package for components.
Now when referring to these globals inside components, you can use the plain variable name:
$dbh->prepare...
The most important task is selecting a good DBM package. Most standard DBM packages (SDBM, ODBM, NDBM) are unsuitable for data caching due to significant limitations on the size of keys and values. Perl only comes with SDBM, so you'll need to obtain a good-quality package if you haven't already. At this time the best options are Berkeley DB (DB_File) version 2.x, available at www.sleepycat.com, and GNU's gdbm (GDBM), available at GNU mirror sites everywhere. Stay away from Berkeley DB version 1.x on Linux which has a serious memory leak (and is unfortunately pre-installed on many distributions).
As far as the serialization methods, all of them should work fine. Data::Dumper is probably simplest: it comes with the latest versions of Perl, is required by Mason anyway, and produces readable output (possibly useful for debugging cache files). On the other hand Storable is significantly faster than the other options according to the MLDBM documentation.
Data caching will not work on systems lacking flock(),
such as
Windows 95 and 98.
$m->cache
or $m->cache_self
for the first time, Mason automatically creates a new cache file under data_dir/cache
. The name of the file is determined by encoding the path as follows:
s/([^\w\.\-\~])/sprintf('+%02x', ord $1)/eg;
like URL encoding with a '+' escape character. For example, the cache file
for component /foo/bar
is data_dir/cache/foo+2fbar
.
Currently Mason never deletes cache files, not even when the associated component file is modified. (This may change in the near future.) Thus cache files hang around and grow indefinitely. You may want to use a cron job or similar mechanism to delete cache files that get too large or too old. For example:
# Shoot cache files more than 30 days old foreach (<data_dir/cache>) { # path to cache directory unlink $_ if (-M >= 30); }
In general you can feel free to delete cache files periodically and without warning, because the data cache mechanism is explicitly not guaranteed -- developers are warned that cached data may disappear anytime and components must still function.
If some reason you want to disable data caching, specify
use_data_cache=>0 to the Interp object. This will cause all $m->cache
calls to return undef without doing anything.
$r
) and calls the same PerlHandler that Apache called. Debug files are
created under data_dir/debug/<username>
for authenticated users, otherwise they are placed in
data_dir/debug/anon
. Several ApacheHandler parameters are required to activate and configure
debug files:
/usr/bin/perl
. This is used in the Unix ``shebang'' line at the top of each debug file.
handler.pl
script. Debug files invoke
handler.pl
just as Apache does as startup, to load needed modules and create Mason
objects.
handler.pl
. This routine is called with the saved Apache request object.
ApacheHandler
constructor with all debug options:
my $ah = new HTML::Mason::ApacheHandler (interp=>$interp, debug_mode=>'all', debug_perl_binary=>'/usr/local/bin/perl', debug_handler_script=>'/usr/local/mason/eg/handler.pl', debug_handler_proc=>'HTML::Mason::handler');
When replaying a request through a debug file, the global variable
$HTML::Mason::IN_DEBUG_FILE
will be set to 1. This is useful if you want to omit certain flags (like
preloading) in handler.pl when running under debug. For example:
my %extra_flags = ($HTML::Mason::IN_DEBUG_FILE) ? () : (preloads=>[...]); my $interp = new HTML::Mason::Interp (..., %extra_flags);
Follow these steps to activate the Previewer:
Listen your.site.ip.address:3001 ... Listen your.site.ip.address:3005
You'll also probably want to restrict access to these ports in your access.conf. If you have multiple site developers, it is helpful to use username/password access control, since the previewer will use the username to keep configurations separate.
handler.pl
, add the line
use HTML::Mason::Preview;
somewhere underneath ``use HTML::Mason''. Then add code to your handler routine to intercept Previewer requests on the ports defined above. Your handler should end up looking like this:
sub handler { my ($r) = @_;
# Compute port number from Host header my $host = $r->header_in('Host'); my ($port) = ($host =~ /:([0-9]+)$/); $port = 80 if (!defined($port));
# Handle previewer request on special ports if ($port >= 3001 && $port <= 3005) { my $parser = new HTML::Mason::Parser(...); my $interp = new HTML::Mason::Interp(...); my $ah = new HTML::Mason::ApacheHandler (...); return HTML::Mason::Preview::handle_preview_request($r,$ah); } else { $ah->handle_request($r); # else, normal request handler } }
The three ``new'' lines inside the if block should look exactly the same as
the lines at the top of handler.pl
. Note that these separate Mason objects are created for a single request
and discarded. The reason is that the previewer may alter the objects'
settings, so it is safer to create new ones every time.
The format of the system log was designed to be easy to parse by programs, although it is not unduly hard to read for humans. Every event is logged on one line. Each line consists of multiple fields delimited by a common separator, by default ctrl-A. The first three fields are always the same: time, the name of the event, and the current pid ($$). These are followed by one or more fields specific to the event.
The events are:
EVENT NAME DESCRIPTION EXTRA FIELDS
REQ_START start of HTTP request request number, URL + query string REQ_END end of HTTP request request number, error flag (1 if error occurred, 0 otherwise) CACHE_READ attempt to read from component path, cache key, success data cache (C<$m-E<gt>cache>) flag (1 if item found, 0 otherwise) CACHE_STORE store to data cache component path, cache key COMP_LOAD component loaded into memory component path for first time
The request number is an incremental value that uniquely identifies each request for a given child process. Use it to match up REQ_START/REQ_END pairs.
To turn on logging, specify a string value to system_log_events containing one or more event names separated by '|'. In additional to individual event names, the following names can be used to specify multiple events:
REQUEST = REQ_START | REQ_END CACHE = CACHE_READ | CACHE_STORE ALL = All events
For example, to log REQ_START, REQ_END, and COMP_LOAD events, you could use system_log_events => ``REQUEST|COMP_LOAD'' Note that this is a string, not a set of constants or'd together.
Configuration Options
By default, the system log will be placed in data_dir/etc/system.log. You can change this with system_log_file.
The default line separator is ctrl-A. The advantage of this separator is
that it is very unlikely to appear in any of the fields, making it easy to
split()
the line. The disadvantage is that it will not always
display, e.g. from a Unix shell, making the log harder to read casually.
You can change the separator to any sequence of characters with system_log_separator.
The time on each log line will be of the form ``seconds.microseconds'' if
you are using Time::HiRes, and simply ``seconds'' otherwise. See
Config.pm
section.
Sample Log Parser
Here is a code skeleton for parsing the various events in a log. You can also find this in eg/parselog.pl in the Mason distribution.
open(LOG,"mason.log"); while (<LOG>) { chomp; my (@fields) = split("\cA"); my ($time,$event,$pid) = splice(@fields,0,3); if ($event eq 'REQ_START') { my ($reqnum,$url) = @fields; ... } elsif ($event eq 'REQ_END') { my ($reqnum,$errflag) = @fields; ... } elsif ($event eq 'CACHE_READ') { my ($comp,$key,$hitflag) = @fields; ... } elsif ($event eq 'CACHE_STORE') { my ($comp,$key) = @fields; ... } elsif ($event eq 'COMP_LOAD') { my ($comp) = @fields; ... } else { warn "unrecognized event type: $event\n"; } }
Suggested Uses
Performance: REQUEST events are useful for analyzing the performance of all Mason requests occurring on your site, and identifying the slowest requests. eg/perflog.pl in the Mason distribution is a log parser that outputs the average compute time of each unique URL, in order from slowest to quickest.
Server activity: REQUEST events are useful for determining what your web server children are working on, especially when you have a runaway. For a given process, simply tail the log and find the last REQ_START event with that process id. (You can also use the Apache status page for this.)
Cache efficiency: CACHE events are useful for monitoring cache ``hit rates'' (number of successful reads over total number of reads) over all components that use a data cache. Because stores to a cache are more expensive than reads, a high hit rate is essential for the cache to have a beneficial effect. If a particular cache hit rate is too low, you may want to consider changing how frequently it is expired or whether to use it at all.
Load frequency: COMP_LOAD events are useful for monitoring your code cache. Too many loads may indicate that your code cache is too small. Also, if you can turn off the code cache for a short time, COMP_LOAD events will tell you which components are loaded most often and thus good candidates for preloading.
The maximum size of the cache is specified with the code_cache_max_size Interp parameter; default is 10MB. When the cache fills up, Mason frees up space by discarding a number of components. The discard algorithm is least frequently used (LFU), with a periodic decay to gradually eliminate old frequency information. In a nutshell, the components called most often in recent history should remain in the cache. Very large components (over 20% of the maximum cache size) never get cached, on the theory that they would force out too many other components.
Note that the ``size'' of a component in memory cannot literally be measured. It is estimated by the length of the source text plus some overhead. Your process growth will not match the code cache size exactly.
You can monitor the performance of the memory cache by turning on system logs and counting the COMP_LOAD events. If these are occurring frequently even for a long-running process, you may want to increase the size of your code cache.
You can prepopulate the cache with components that you know will be accessed often; see Preloading. Note that preloaded components possess no special status in the cache and can be discarded like any others.
Naturally, a cache entry is invalidated if the corresponding component source file changes.
To turn off code caching completely, set code_cache_max_size to 0.
As a secondary, longer-term cache mechanism, Mason stores a compiled form
of each component in an object file under
data_dir/obj/component-path
. Any server process can eval the object file and save time on parsing the
component source file. The object file is recreated whenever the source
file changes.
Besides improving performance, object files are essential for debugging and interpretation of errors. Line numbers in error messages are given in terms of the object file. The curious-minded can peek inside an object file to see exactly how Mason converted a given component to a Perl object.
If for some reason you don't want Mason to create object files, set the use_object_files Interp parameter to 0.
To remedy this, Mason has an accelerated mode that changes its behavior in two ways:
1. Does not check component source files at all, relying solely on object files. This means the developer or an automated system is responsible for recompiling any components that change and recreating object files, using the make_component Parser method.
2. Rather than continuously checking whether object files have changed,
Mason monitors a ``reload file'' containing an ever-growing list of
components that have changed. Whenever a component changes, the developer
or an automated system is responsible for appending the component path to
the reload file. The reload file is kept in
data_dir/etc/reload.lst
.
You can activate this mode with the use_reload_file Interp method.
The advantage of using this mode is that Mason stats one file per request instead of ten or twenty. The disadvantage is a increase in maintenance costs as the object and reload files have to be kept up-to-date. Automated editorial tools, and cron jobs that periodically scan the component hierarchy for changes, are two possible solutions. The Mason content management system automatically handles this task.
The priorities for the staging site are rapid development and easy debugging, while the main priority for the production site is performance. This section describes various ways to adapt Mason for each case.
Batch mode has the advantage of better error handling. Suppose an error occurs in the middle of a page. In stream mode, the error message interrupts existing output, often appearing in an awkward HTML context such as the middle of a table which never gets closed. In batch mode, the error message is output neatly and alone.
Batch mode also offers more flexibility in controlling HTTP headers. See sending_http_headers.
Stream mode may help get data to the browser more quickly, allowing server and browser to work in parallel. It also prevents memory buildup for very large responses.
Since Apache does its own buffering, stream mode does not entail immediate
delivery of output to the client. You must set $|=1 to turn off Apache
buffering completely (generally not a good idea) or call
$m->flush_buffer
to flush the buffer selectively.
In terms of making your server seem responsive, the initial bytes are most
important. You can send these early by calling $m->flush_buffer
in key locations such as the common page header. However, this dilutes the
advantages of batch mode mentioned above. Tradeoffs...
You control output mode by setting interp->out_mode
to ``batch'' or ``stream''.
# httpd.conf PerlRequire /usr/local/mason/handler.pl
# Web site #1 <VirtualHost www.site1.com> DocumentRoot /usr/local/www/htdocs/site1 <Location /> SetHandler perl-script PerlHandler HTML::Mason </Location> </VirtualHost>
# Web site #2 <VirtualHost www.site2.com> DocumentRoot /usr/local/www/htdocs/site2 <Location /> SetHandler perl-script PerlHandler HTML::Mason </Location> </VirtualHost>
In handler.pl:
my $interp = new HTML::Mason::Interp (parser=>$parser, comp_root=>'/usr/local/www/htdocs' data_dir=>'/usr/local/mason/');
The directory structure for this scenario might look like:
/usr/local/www/htdocs/ # component root +- shared/ # shared components +- site1/ # DocumentRoot for first site +- site2/ # DocumentRoot for second site
Incoming URLs for each site can only request components in their respective DocumentRoots, while components internally can call other components anywhere in the component space. The shared/ directory is a private directory for use by components, inaccessible from the Web.
my (%interp,%ah); foreach my $site qw(...) { $interp{$site} = new HTML::Mason::Interp (comp_root=>"/usr/local/www/$site",...); $ah{$site} = new HTML::Mason::ApacheHandler (interp=>$interp{$site},...); }
...
sub handler { my $site = $r->dir_config('site'); $ah{$site}->handle_request($r); }
We assume each virtual server configuration section has a
PerlSetVar site <site_name>
Above we pre-create all Mason objects in the parent. Another scheme is to create objects on demand in the child:
my (%interp,%ah);
...
sub handler { my $site = $r->dir_config('site'); unless exists($interp{$site}) { # get comp_root from PerlSetVar as well my $comp_root = $r->dir_config('comp_root'); $interp{$site} = new HTML::Mason::Interp(comp_root=>$comp_root,...); $ah{$site} = new HTML::Mason::ApacheHandler(interp=>$interp{$site},...); } }
The advantage of the second scheme is that you don't have to hardcode as much information in the handler.pl. The disadvantage is a slight memory and performance impact. On development servers this shouldn't matter; on production servers you may wish to profile the two schemes.