1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
|
<?xml version="1.0" encoding="UTF-8"?>
<!--
Licensed to the Apache Software Foundation (ASF) under one or more
contributor license agreements. See the NOTICE file distributed with
this work for additional information regarding copyright ownership.
The ASF licenses this file to You under the Apache License, Version 2.0
(the "License"); you may not use this file except in compliance with
the License. You may obtain a copy of the License at
http://www.apache.org/licenses/LICENSE-2.0
Unless required by applicable law or agreed to in writing, software
distributed under the License is distributed on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
See the License for the specific language governing permissions and
limitations under the License.
-->
<!-- $Id$ -->
<!DOCTYPE document PUBLIC "-//APACHE//DTD Documentation V2.0//EN" "http://forrest.apache.org/dtd/document-v20.dtd">
<document>
<header>
<title>Apache™ FOP: Intermediate Format</title>
<version>$Revision$</version>
</header>
<body>
<note>
Please note that the intermediate formats described here are
<strong>advanced features</strong> and can be ignored by most users of Apache FOP.
</note>
<section id="introduction">
<title>Introduction</title>
<p>
Apache™ FOP now provides two different so-called intermediate formats. The first one
(let's call it the area tree XML format) is basically a 1:1 XML representation of FOP's
area tree as generated by the layout engine. The area tree is conceptually defined in the
<a href="http://www.w3.org/TR/2001/REC-xsl-20011015/slice1.html#section-N742-Formatting">XSL-FO specification in chapter 1.1.2</a>.
Even though the area tree is mentioned in the XSL-FO specification, this part is not
standardized. Therefore, the area tree XML format is a FOP-proprietary XML file format.
The area tree XML can be generated through the area tree XML Renderer (the XMLRenderer).
</p>
<p>
The second intermediate format (which we shall name exactly like this: the intermediate
format)
is a recent addition which tries to meet a slightly different set of goals. It is highly
optimized for speed.
</p>
<p>
The intermediate format can be used to generate intermediate documents that are modified
before they are finally rendered to their ultimate output format. Modifications include
adjusting and changing trait values, adding or modifying area objects, inserting prefabricated
pages, overlays, imposition (n-up, rotation, scaling etc.). Multiple IF files can be combined
to a single output file.
</p>
</section>
<section id="which-if">
<title>Which Intermediate Format to choose?</title>
<p>
Both formats have their use cases, so the choice you will make will depend on your
particular situation. Here is a list of strengths and use cases for both formats:
</p>
<section id="strengths-at">
<title>Area Tree XML (AT XML)</title>
<ul>
<li>1:1 representation of FOP's area tree in XML.</li>
<li>Contains more structure information than the new intermediate format.</li>
<li>Used in FOP's layout engine test suite for regression testing.</li>
</ul>
</section>
<section id="strengths-if">
<title>Intermediate Format (IF)</title>
<ul>
<li>Highly optimized for speed.</li>
<li>Smaller XML files.</li>
<li>Easier to post-process.</li>
<li>XML Schema is available.</li>
<li>
Recommended for use cases where documents are formatted concurrently and later
concatenated to a single print job.
</li>
</ul>
</section>
<p>
More technical information about the two formats can be found on the
<a href="http://wiki.apache.org/xmlgraphics-fop/AreaTreeIntermediateXml/NewDesign">FOP Wiki</a>.
</p>
</section>
<section id="architecture">
<title>Architectural Overview</title>
<figure src="images/if-architecture-overview.png"
alt="Diagram with an architectural overview over the intermediate formats"/>
</section>
<section id="usage">
<title>Usage of the Area Tree XML format (AT XML)</title>
<p>
As already mentioned, the area tree XML format is generated by using the
<strong>XMLRenderer</strong> (MIME type: <strong>application/X-fop-areatree</strong>).
So, you basically set the right MIME type for the output format and process your FO files
as if you would create a PDF file.
</p>
<p>
However, there is an important detail to consider: The
various Renderers don't all use the same font sources. To be able to create the right
area tree for the ultimate output format, you need to create the area tree XML file using
the right font setup. This is achieved by telling the XMLRenderer to mimic another
renderer. This is done by calling the XMLRenderer's mimicRenderer() method with an
instance of the ultimate target renderer as the single parameter. This has a consequence:
An area tree XML file rendered with the Java2DRenderer may not look as expected when it
was actually generated for the PDF renderer. For renderers that use the same font setup,
this restriction does not apply (PDF and PS, for example). Generating the area tree XML
format file is the first step.
</p>
<p>
The second step is to reparse the file using the <strong>AreaTreeParser</strong> which is
found in the org.apache.fop.area package. The pages retrieved from the area tree XML file
are added to an AreaTreeModel instance from where they are normally rendered using one of
the available Renderer implementations. You can find examples for the area tree XML
processing in the
<a href="http://svn.apache.org/viewvc/xmlgraphics/fop/trunk/examples/embedding/java/embedding/intermediate/"><code>examples/embedding</code></a>
directory in the FOP distribution.
</p>
<p>
The basic pattern to parse the area tree XML format looks like this:
</p>
<source><![CDATA[
FopFactory fopFactory = FopFactory.newInstance();
// Setup output
OutputStream out = new java.io.FileOutputStream(pdffile);
out = new java.io.BufferedOutputStream(out);
try {
//Setup fonts and user agent
FontInfo fontInfo = new FontInfo();
FOUserAgent userAgent = fopFactory.newFOUserAgent();
//Construct the AreaTreeModel that will received the individual pages
AreaTreeModel treeModel = new RenderPagesModel(userAgent,
MimeConstants.MIME_PDF, fontInfo, out);
//Parse the area tree file into the area tree
AreaTreeParser parser = new AreaTreeParser();
Source src = new StreamSource(myIFFile);
parser.parse(src, treeModel, userAgent);
//Signal the end of the processing. The renderer can finalize the target document.
treeModel.endDocument();
} finally {
out.close();
}]]></source>
<p>
This example simply reads an area tree file and renders it to a PDF file. Please note, that in normal
FOP operation you're shielded from having to instantiate the FontInfo object yourself. This
is normally a task of the AreaTreeHandler which is not present in this scenario. The same
applies to the AreaTreeModel instance, in this case an instance of a subclass called
RenderPagesModel. RenderPagesModel is ideal in this case as it has very little overhead
processing the individual pages. An important line in the example is the call to
<code>endDocument()</code> on the AreaTreeModel. This lets the Renderer know that the processing
is now finished.
</p>
<p>
The area tree XML format can also be used from the <a href="running.html#standalone-start">command-line</a>
by using the "-atin" parameter for specifying the area tree XML as input file. You can also
specify a "mimic renderer" by inserting a MIME type between "-at" and the output file.
</p>
<section id="concat">
<title>Concatenating Documents</title>
<p>
This initial example is obviously not very useful. It would be faster to create the PDF file
directly. As the <a href="http://svn.apache.org/repos/asf/xmlgraphics/fop/trunk/examples/embedding/java/embedding/atxml/ExampleConcat.java">ExampleConcat.java</a>
example shows you can easily parse multiple area tree files in a row and add the parsed pages to the
same AreaTreeModel instance which essentially concatenates all the input document to one single
output document.
</p>
</section>
<section id="modifying">
<title>Modifying Documents</title>
<p>
One of the most important use cases for this format is obviously modifying the area
tree XML before finally rendering it to the target format. You can easily use XSLT to process
the AT XML file according to your needs. Please note, that we will currently not formally describe
the area tree XML format. You need to have a good understanding its structure so you don't
create any non-parseable files. We may add an XML Schema and more detailed documentation at a
later time. You're invited to help us with that.
</p>
<note>
The area tree XML format is sensitive to changes in whitespace. If you're not careful,
the modified file may not render correctly.
</note>
</section>
<section id="advanced">
<title>Advanced Use</title>
<p>
The generation of the area tree format as well as it parsing process has been designed to allow
for maximum flexibility and optimization. Please note that you can call <code>setTransformerHandler()</code> on
XMLRenderer to give the XMLRenderer your own TransformerHandler instance in case you would like to
do custom serialization (to a W3C DOM, for example) and/or to directly modify the area tree using
XSLT. The AreaTreeParser on the other side allows you to retrieve a ContentHandler instance where
you can manually send SAX events to to start the parsing process (see <code>getContentHandler()</code>).
</p>
</section>
</section>
<section id="usage-if">
<title>Usage of the Intermediate Format (IF)</title>
<p>
The Intermediate Format (IF) is generated by the <strong>IFSerializer</strong>
(MIME type: <strong>application/X-fop-intermediate-format</strong>).
So, you basically set the right MIME type for the output format and process your FO files
as if you would create a PDF file.
</p>
<p>
The IFSerializer is an implementation of the <strong>IFDocumentHandler</strong> and
<strong>IFPainter</strong> interfaces. The <strong>IFRenderer</strong> class is responsible
for converting FOP's area tree into calls against these two interfaces.
</p>
<ul>
<li>
IFDocumentHandler: This interface is used on the document-level and defines the
overall structure of the Intermediate Format.
</li>
<li>
IFPainter: This interface is used to generate graphical page content like text, images
and borders.
</li>
</ul>
<p>
As with the AT XML, there is an important detail to consider: The various output
implementations don't all use the same font sources. To be able
to create the right IF for the ultimate output file, you need to create the IF file using
the right font setup. This is achieved by telling the IFRenderer (responsible for
converting the area tree into calls to the IFDocumentHandler and IFPainter interfaces)
to mimic another renderer. This is done by calling the IFSerializer's
mimicDocumentHandler() method with an instance of the ultimate target document handler
as the single parameter. This has a consequence: An IF file rendered with the
Java2DDocumentHandler may not look as expected when it was actually generated for the PDF
implementation. For implementations that use the same font setup,
this restriction does not apply (PDF and PS, for example). Generating the Intermediate
Format file is the first step.
</p>
<p>
The second step is to reparse the file using the <strong>IFParser</strong> which is
found in the org.apache.fop.render.intermediate package. The IFParser simply takes an
IFDocumentHandler instance against which it generates the appropriate calls. The IFParser
is implemented as a SAX ContentHandler so you're free to choose the method for
post-processing the IF file(s). You can use XSLT or write SAX- or DOM-based code to
manipulate the contents. You can find examples for the Intermediate Format
processing in the
<a href="http://svn.apache.org/viewvc/xmlgraphics/fop/trunk/examples/embedding/java/embedding/intermediate/"><code>examples/embedding</code></a>
directory in the FOP distribution.
</p>
<p>
The basic pattern to parse the intermediate format looks like this:
</p>
<source><![CDATA[
FopFactory fopFactory = FopFactory.newInstance();
// Setup output
OutputStream out = new java.io.FileOutputStream(pdffile);
out = new java.io.BufferedOutputStream(out);
try {
//Setup user agent
FOUserAgent userAgent = fopFactory.newFOUserAgent();
//Create IFDocumentHandler instance
IFDocumentHandler targetHandler;
String mime = MimeConstants.MIME_PDF;
targetHandler = fopFactory.getRendererFactory().createDocumentHandler(
userAgent, mime);
//Setup fonts
IFUtil.setupFonts(targetHandler);
//Tell the target handler where to write the PDF to
targetHandler.setResult(new StreamResult(pdffile));
//Parse the IF file
IFParser parser = new IFParser();
Source src = new StreamSource(myIFFile);
parser.parse(src, targetHandler, userAgent);
} finally {
out.close();
}]]></source>
<p>
This example simply reads an intermediate file and renders it to a PDF file. Here
IFParser.parse() is used, but you can also just get a SAX ContentHandler by using the
IFParser.getContentHandler() method.
</p>
<section id="concat-if">
<title>Concatenating Documents</title>
<p>
This initial example is obviously not very useful. It would be faster to create the PDF file
directly (without the intermediate step). As the
<a href="http://svn.apache.org/repos/asf/xmlgraphics/fop/trunk/examples/embedding/java/embedding/intermediate/ExampleConcat.java">ExampleConcat.java</a>
example shows you can easily parse multiple intermediate files in a row and use the
IFConcatenator class to concatenate page sequences from multiple source files to a single
output file. This particular example does the concatenation on the level of the
IFDocumentHandler interface. You could also do this in XSLT or using SAX on the XML level.
Whatever suits your process best.
</p>
</section>
<section id="modifying-if">
<title>Modifying Documents</title>
<p>
One of the most important use cases for this format is obviously modifying the
intermediate format before finally rendering it to the target format. You can easily use
XSLT to process the IF file according to your needs.
</p>
<p>
There is an XML Schema (located under
<a href="http://svn.apache.org/viewvc/xmlgraphics/fop/trunk/src/documentation/intermediate-format-ng/">src/documentation/intermediate-format-ng</a>)
that helps you verify that your modified content is correct.
</p>
<p>
For certain output formats there's a caveat: Formats like AFP and PCL do not support
arbitrary transformations on the IF's "viewport" and "g" elements. Possible are
only rotations in 90 degree steps and translations.
</p>
</section>
<section id="advanced-if">
<title>Advanced Use</title>
<p>
The generation of the intermediate format as well as it parsing process has been
designed to allow for maximum flexibility and optimization. So rather than just passing
in a StreamResult to IFSerializer's setResult() method, you can also use a SAXResult
or a DOMResult. And as you've already seen , the IFParser on the other side allows you
to retrieve a ContentHandler instance where you can manually send SAX events to
start the parsing process (see <code>getContentHandler()</code>).
</p>
</section>
</section>
</body>
</document>
|
