The ttree
script is used to process entire
directory trees containing template files. The resulting output from
processing each file is then written to a corresponding file in a
destination directory. The script compares the modification times of
source and destination files (where they already exist) and processes
only those files that have been modified. In other words, it is the
equivalent of 'make' for the Template Toolkit.
It supports a number of options which can be used to configure behaviour,
define locations and set Template Toolkit options. The script first reads
the .ttreerc
configuration file in the HOME
directory, or an alternative file specified in the TTREERC environment
variable. Then, it processes any command line arguments, including any
additional configuration files specified via the -f
(file)
option.
When you run ttree
for the first time it will
ask you if you want it to create a .ttreerc
file for you. This will be created in your home directory.
$ ttree Do you want me to create a sample '.ttreerc' file for you? (file: /home/abw/.ttreerc) [y/n]: y /home/abw/.ttreerc created. Please edit accordingly and re-run ttree
The purpose of this file is to set any global configuration
options that you want applied every time ttree
is run. For example, you can use the
ignore
and copy
option to provide regular
expressions that specify which files should be ignored and which should
be copied rather than being processed as templates. You may also want to
set flags like verbose
and recurse
according to
your preference.
A minimal .ttreerc
:
# ignore these files ignore = \b(CVS|RCS)\b ignore = ^# ignore = ~$ # copy these files copy = \.(gif|png|jpg|pdf)$ # recurse into directories recurse # provide info about what's going on verbose
In most cases, you'll want to create a different ttree
configuration file for each project you're
working on. The cfg
option allows you to specify a directory
where ttree
can find further configuration
files.
cfg = /home/abw/.ttree
The -f
command line option can be used to specify which
configuration file should be used. You can specify a filename using an
absolute or relative path:
$ ttree -f /home/abw/web/example/etc/ttree.cfg $ ttree -f ./etc/ttree.cfg $ ttree -f ../etc/ttree.cfg
If the configuration file does not begin with /
or
.
or something that looks like a MS-DOS absolute path (e.g.
C:\\etc\\ttree.cfg
) then ttree
will look for it in the directory specified by the cfg
option.
$ ttree -f test1 # /home/abw/.ttree/test1
The cfg
option can only be used in the .ttreerc
file. All the other options can be used in
the .ttreerc
or any other ttree
configuration file. They can all also be
specified as command line options.
Remember that .ttreerc
is always processed
before any configuration file specified with the -f
option. Certain options like lib
can be used any number of
times and accumulate their values.
For example, consider the following configuration files:
/home/abw/.ttreerc
:
cfg = /home/abw/.ttree lib = /usr/local/tt2/templates
/home/abw/.ttree/myconfig
:
lib = /home/abw/web/example/templates/lib
When ttree
is invoked as follows:
$ ttree -f myconfig
the lib
option will be set to the following directories:
/usr/local/tt2/templates /home/abw/web/example/templates/lib
Any templates located under /usr/local/tt2/templates
will be used in preference
to those located under /home/abw/web/example/templates/lib
. This may be what
you want, but then again, it might not. For this reason, it is good
practice to keep the .ttreerc
as simple as
possible and use different configuration files for each ttree
project.
The src
option is used to define the directory containing
the source templates to be processed. It can be provided as a command
line option or in a configuration file as shown here:
src = /home/abw/web/example/templates/src
Each template in this directory typically corresponds to a single web page or other document.
The dest
option is used to specify the destination directory
for the generated output.
dest = /home/abw/web/example/html
The lib
option is used to define one or more directories
containing additional library templates. These templates are not
documents in their own right and typically comprise of smaller, modular
components like headers, footers and menus that are incorporated into
pages templates.
lib = /home/abw/web/example/templates/lib lib = /usr/local/tt2/templates
The lib
option can be used repeatedly to add further
directories to the search path.
A list of templates can be passed to ttree
as
command line arguments.
$ ttree foo.html bar.html
It looks for these templates in the src
directory and
processes them through the Template Toolkit, using any additional
template components from the lib
directories. The generated
output is then written to the corresponding file in the dest
directory.
If ttree
is invoked without explicitly
specifying any templates to be processed then it will process every file
in the src
directory. If the -r
(recurse)
option is set then it will additionally iterate down through
sub-directories and process and other template files it finds therein.
$ ttree -r
If a template has been processed previously, ttree
will compare the modification times of the
source and destination files. If the source template (or one it is
dependant on) has not been modified more recently than the generated
output file then ttree
will not process it. The
-a
(all) option can be used to force ttree
to process all files regardless of modification
time.
$ ttree -a
Any templates explicitly named as command line argument are always processed and the modification time checking is bypassed.
The ignore
, copy
and accept
options are used to specify Perl regexen to filter file names. Files that
match any of the ignore
options will not be processed.
Remaining files that match any of the copy
regexen will be
copied to the destination directory. Remaining files that then match any
of the accept
criteria are then processed via the Template
Toolkit. If no accept
parameter is specified then all files
will be accepted for processing if not already copied or ignored.
# ignore these files ignore = \b(CVS|RCS)\b ignore = ^# ignore = ~$ # copy these files copy = \.(gif|png|jpg|pdf)$ # accept only .tt2 templates accept = \.tt2$
The suffix
option is used to define mappings between the
file extensions for source templates and the generated output files. The
following example specifies that source templates with a
.tt2
suffix should be output as .html
files:
suffix tt2=html
Or on the command line,
--suffix tt2=html
You can provide any number of different suffix mappings by repeating this option.
The depend
and depend_file
options allow you to
specify how any given template file depends on another file or group of
files. The depend
option is used to express a single
dependency.
$ ttree --depend foo=bar,baz
This command line example shows the --depend
option being
used to specify that the foo
file is dependant
on the bar
and baz
templates. This option can be used many time on the command line:
$ ttree --depend foo=bar,baz --depend crash=bang,wallop
or in a configuration file:
depend foo=bar,baz depend crash=bang,wallop
The file appearing on the left of the =
is specified
relative to the src
or lib
directories. The
file(s) appearing on the right can be specified relative to any of these
directories or as absolute file paths.
For example:
$ ttree --depend foo=bar,/tmp/baz
To define a dependency that applies to all files, use *
on
the left of the =
.
$ ttree --depend *=header,footer
or in a configuration file:
depend *=header,footer
Any templates that are defined in the pre_process
,
post_process
, process
or wrapper
options will automatically be added to the list of global dependencies
that apply to all templates.
The depend_file
option can be used to specify a file that
contains dependency information.
$ ttree --depend_file=/home/abw/web/example/etc/ttree.dep
Here is an example of a dependency file:
# This is a comment. It is ignored. index.html: header footer menubar header: titlebar hotlinks menubar: menuitem # spanning multiple lines with the backslash another.html: header footer menubar \ sidebar searchform
Lines beginning with the #
character are comments and are
ignored. Blank lines are also ignored. All other lines should provide a
filename followed by a colon and then a list of dependant files separated
by whitespace, commas or both. Whitespace around the colon is also
optional. Lines ending in the \
character are continued onto
the following line.
Files that contain spaces can be quoted. That is only necessary for files after the colon (':'). The file before the colon may be quoted if it contains a colon.
As with the command line options, the *
character can be
used as a wildcard to specify a dependency for all templates.
* : config,header
Andy Wardley <abw@wardley.org>
With contributions from Dylan William Hardison (support for
dependencies), Bryce Harrington (absolute
and
relative
options), Mark Anderson (suffix
and
debug
options), Harald Joerg and Leon Brocard who gets
everywhere, it seems.
Copyright (C) 1996-2007 Andy Wardley. All Rights Reserved.
This module is free software; you can redistribute it and/or modify it under the same terms as Perl itself.