|
amrita can be used with XML as well as HTML.
amrita's HTML parser was developed very ad hoc way. I have found no major problem with it yet , but I think it's good idea to have an option for XML: a strictly defined standard.
amrita produces xhtml output from xhtml template.
amrita has a little code depend upon HTML's DTD(tag structure), and they can be detached easily. So any XML(not XHTML) document can be used as a template with amrita.
REXML has very natural API for ruby. And it is easy to make model datas for amrita from XML documents read by REXML.
see docs/Tour2
code:
require "amrita/template" include Amrita tmpl_text = <<-END <?xml version="1.0"?> <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd"> <html xmlns="http://www.w3.org/1999/xhtml"> <head> <title>xhtml sample</title> </head> <body> <h1 id=title>title</h1> <p id=body>body text</p> <hr /> </body> </html> END
data = { :title => "SAMPLE1", :body => "members of this HASH will be inserted here and title" } tmpl = TemplateText.new tmpl_text tmpl.prettyprint = true tmpl.xml = true # use REXML parser tmpl.asxml = true tmpl.expand(STDOUT, data)
output:
<?xml version="1.0"?> <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd"> <html xmlns="http://www.w3.org/1999/xhtml"> <head> <title>xhtml sample</title> </head> <body> <h1>SAMPLE1</h1> <p>members of this HASH will be inserted here and title</p> <hr /> </body> </html>
Basicaly amrita produce xhtml output from xhtml template, html4.0 from html4.0 template.So programers don't need tobother about comformation for some paticular standards or browser or devices. Only designers (template writers) do.
The only thing programers should do is set asxml flag to Template object.If this was set a single tag like <hr> will be printed like <hr />.
code:
data = { :table1=>[ { :name=>"Ruby In A Nutshell", :author=>"Yukihiro Matsumoto, David L. Reynolds", :isbn=>"0596002149" } { :name=>"Programming Ruby", :author=>"David Thomas, Andrew Hunt", :isbn=>"0201710897" }, { :name=>"The Ruby Way", :author=>"Hal Fulton", :isbn=>"0672320835" }, ] } xml_tmpl = TemplateText.new <<END <booklist> <book id="table1"> <title id="name" /> <author id="author" /> <isbn id="isbn" /> </book> </booklist> END
xml_tmpl.xml = true # use REXML-based parser puts "------------XML output ----------------------" xml_tmpl.expand(STDOUT, data)
output:
<booklist> <book> <title>Ruby In A Nutshell</title> <author>Yukihiro Matsumoto, David L. Reynolds</author> <isbn>0596002149</isbn> </book><book> <title>Programming Ruby</title> <author>David Thomas, Andrew Hunt</author> ..........
xml_tmpl.xml = true # use REXML-based parser puts "------------XML output ----------------------" xml_tmpl.expand(STDOUT, data)
amrita loads templates on demand. If xml flag is set when expand is called, amrita uses REXML based parser.
You can use single model data for two templates. So single code with amrita produce both XML output and HTML output.For detail see sample/tour/xml1.rb .
XML document(data) + HTML template ==> HTML document
This may be some kind of style-sheet.
code:
require "amrita/template" require "rexml/document" include Amrita doc = REXML::Document.new <<END <booklist> <book isbn="0596002149"> <title>Ruby In A Nutshell</title> <author>Yukihiro Matsumoto</author> <author>David L. Reynolds</author> </book> <book isbn="0201710897"> <title>Programming Ruby</title> <author>David Thomas</author> <author>Andrew Hunt</author> </book> <book isbn="0672320835"> <title>The Ruby Way</title> <author>Hal Fulton</author> </book> </booklist> END
table = doc.elements.to_a("booklist/book").collect do |book| { :title=>book.elements['title'], :authors=>book.elements.to_a('author').collect do |a| { :name=>a } end, #:isbn=>book.attributes['isbn'] :isbn=>e(:a, :href=>"http://www.amazon.com/exec/obidos/ASIN/#{book.attributes['isbn']}") { book.attributes['isbn'] } } end data = { :table1=>table }
html_tmpl = TemplateText.new <<END <table border="1"> <tr><th>title</th><th>author</th><th>ISBN</th></tr> <tr id=table1> <td id="title"> <td><span id="authors"><span id="name"></span><br></span> <td id="isbn"> </tr> </table> END html_tmpl.prettyprint = true html_tmpl.set_hint(HtmlCompiler::AnyData.new) html_tmpl.expand(STDOUT, data)
output:
<table border="1"> <tr> <th>title</th> <th>author</th> <th>ISBN</th> </tr> <tr> <td>Ruby In A Nutshell</td> <td>Yukihiro Matsumoto<br>David L. Reynolds <br> </td> <td><a href="http://www.amazon.com/exec/obidos/ASIN/0596002149">0596002149</a></td> </tr> <tr> <td>Programming Ruby</td> <td>David Thomas<br>Andrew Hunt <br> </td> <td><a href="http://www.amazon.com/exec/obidos/ASIN/0201710897">0201710897</a></td> </tr> <tr> <td>The Ruby Way</td> <td>Hal Fulton<br> </td> <td><a href="http://www.amazon.com/exec/obidos/ASIN/0672320835">0672320835</a></td> </tr> </table>
table = doc.elements.to_a("booklist/book").collect do |book| { :xxx=>..... } end
This code generate an Array of Hash on <book> element. <book> element is a REXML::Element data. So you can get any node or attribute you need by REXML's API.
:title=>book.elements['title'],
book.elements['title'] is the first <title> element of <book> element.
:authors=>book.elements.to_a('author').collect do |a| { :name=>a } end,
In this sample, a book has one title but can have many authors. So authors shuold be treated as an Array with to_a, generate Array of Hash by Ruby's standard method collect.
:isbn=>e(:a, :href=>"http://www.amazon.com/exec/obidos/ASIN/#{book.attributes['isbn']}") { book.attributes['isbn'] }
I want to insert a direct link to amazon. e(...) { ... } generate a <tt><a><tt>tag like
<a href="http://www.amazon.com/exec/obidos/ASIN/0596002149">0596002149</a>
and insert it into template.
This idea was extended for document processing in sample/tour/xml3.rb and reached to amx. See docs/Tour2 for detaile.
When you map a ruby object to XML entry, some members are mapped to attribute and others are to sub elements.
This sample shows how to do it by generate ls -l information as an XML.
code:
amrita can use an existing class for model data. To show this ability, this sample uses Ruby's system class File::Stat. require "amrita/template" include Amrita
class File class Stat include Amrita::ExpandByMember def entry(name) a(:name=>name, :type=>ftype) { self } end
def size_or_nil size if ftype == "file" end def mode_str ret = "-rwxrwxrwx" /(.*)(.)(.)(.)(.)(.)(.)(.)(.)(.)$/ =~ sprintf("%b",mode) $~[2..10].each_with_index do |b,n| ret[n+1] = "-" if b != "1" end ret[0] = "d" if $~[1] == "100000" ret end
def unix_inf a(:dev=>dev, :uid=>uid, :gid=>gid) { self } end end end tmpl = TemplateText.new <<END <file id="filelist"> <size id="size_or_nil" /> <mode id="mode_str" /> <times> <ctime id="mtime" /> <mtime id="mtime" /> <atime id="atime" /> </times> <unix_inf id="unix_inf"> <inode id="ino" /> </unix_inf> </file> END
dir = ARGV.shift || '*' filelist = Dir[dir].collect do |f| File::stat(f).entry(f) end
data = { :filelist=>filelist } tmpl.xml = true tmpl.expand(STDOUT, data)
output:
<file name="CVS" type="directory"> <mode>drwxr-xr-x</mode> <times> <ctime>Tue Sep 03 11:07:10 JST 2002</ctime> <mtime>Tue Sep 03 11:07:10 JST 2002</mtime> <atime>Thu Sep 05 07:30:39 JST 2002</atime> </times> <unix_inf uid="1000" gid="1000" dev="770"> <inode>652250</inode> </unix_inf> </file> <file name="precompile.rb" type="file"> <size>2596</size> <mode>-rw-r--r--</mode> <times> <ctime>Mon Aug 26 09:12:11 JST 2002</ctime> <mtime>Mon Aug 26 09:12:11 JST 2002</mtime> <atime>Thu Sep 05 09:26:48 JST 2002</atime> </times> <unix_inf uid="1000" gid="1000" dev="770"> <inode>310411</inode> </unix_inf> </file> <file name="amstest.ams" type="file"> .....
def entry(name) a(:name=>name, :type=>ftype) { self } end
This method generate an AttrArray that put some value of self to XML attribute and make sub-elements with itself.
def size_or_nil size if ftype == "file" end
If the file is not a normal file, this method returns nil and The <size> element will be deleted
code:
output: