Skip to main content

Document preprocessor for Apache FOP.

Project description

Document preprocessor for Apache FOP.

How does it works?

It does what the huge title is implying, preprocess a higher level template to generate dynamically an specific XSL-FO document, which then gets fed to Apache FOP and generate the expected output. So that means that this packages requires Java >_<', but fear not!, it is almost transparent to the python application.

In general the internal workflow is:

template ->  mako -> apply css ->  xsl-fo ->  fop -> *Document*

Installation

  1. Install pypfop:

    pip install pypfop
  2. Install Apache FOP:

    1. Download the binary package of fop1.1 either the zip or tar package.

    2. Decompress in wherever place you like and set environment variable FOP_CMD to the bin/fop of the decompressed folder. [1]

Usage

The Markup

The markup used to generate the documents is almost the same as the xsl-fo, the only difference is that is not necessary to set the xml namespace to all the elements, for example:

<fo:table>
 <fo:table-header>
    <fo:table-row>
       <fo:table-cell>
         <fo:block>Project</fo:block>
       </fo:table-cell>
    </fo:table-row>
 </fo:table-header>
 <fo:table-body>
    <fo:table-row>
      <fo:table-cell>
         <fo:block>pypfop</fo:block>
     </fo:table-cell>
    </fo:table-row>
 </fo:table-body>
</fo:table>

can be written like this:

<table>
 <table-header>
    <table-row>
       <table-cell>
         <block>Project</block>
       </table-cell>
    </table-row>
 </table-header>
 <table-body>
    <table-row>
      <table-cell>
         <block>pypfopp</block>
     </table-cell>
    </table-row>
 </table-body>
</table>

The higher level template language

At the time the only supported template language is mako it should not be very complicated to extend to your favorite template language based in the implementation of mako (which is pretty straight forward) and hopefully contribute back to the project :).

For example the previous table can be generated with this mako template assuming the header and rows variables are passed to the Document.generate method:

<table>
  <table-header>
    <table-row>
      % for name in header:
      <table-cell>
        <block>${name}</block>
      </table-cell>
      % endfor
    </table-row>
  </table-header>
  <table-body>
    % for row in rows:
       <table-row>
     % for cell in row:
        <table-cell>
          <block>${cell}</block>
        </table-cell>
     % endfor
       </table-row>
    % endfor
  </table-body>
</table>

Skeletons

The previous examples are really just a fragment a document to be able to generate a full document with metadata, paper size, margins, etc. To avoid the repetitive work to write this kind of base document pypfop come with skeletons documents, the purpose if this documents is to be inherited at each template, at the time the only implemented skeleton is in pypfop/skeletons/mako/simple-letter-base.fo.mako, which include place holders for metadata about the document and two regions, body and footer, with a few defaults, which of course can be overwritten with the appropriate parameters.

To be a fully functional template for pypfop the previous table need to be like this.

simple-table.fo.mako:

<%inherit file="simple-letter-base.fo.mako" />
<table id="main-table">
  <table-header>
    <table-row>
      % for name in header:
      <table-cell>
        <block>${name}</block>
      </table-cell>
      % endfor
    </table-row>
  </table-header>
  <table-body>
    % for row in rows:
       <table-row>
     % for cell in row:
        <table-cell>
          <block>${cell}</block>
        </table-cell>
     % endfor
       </table-row>
    % endfor
  </table-body>
</table>

The skeletons directory is set in the template directory path by default.

Format and style with CSS

Beside the higher level language that define the content and layout of the document, the style and formatting use CSS, to be more specific it can parse the rules that cssutils support, which are a very good subset of CSS2 and CSS3, for example it support things like :nth-child(X) and @import url(XX).

The properties that can be set are the same as in the specification of xsl-fo, check out the section of About XSL-FO syntax, with the only exception that you can use classes as selectors, xsl-fo does not support the class attribute, the pypfop parser is going to look for the class attribute then substitute with the specific style and then remove the class attribute.

For example I could define the style for the previous table in three files.

simple_table.css:

@import url("general.css");
@import url("colors.css");

#main-table > table-header > table-row{
    text-align: center;
    font-weight: bold;
}

#main-table > table-header table-cell{
    padding: 2mm 0 0mm;
}

general.css:

flow[flow-name="xsl-region-body"] {
    font-size: 10pt;
    font-family: Helvetica;
}

colors.css:

#main-table> table-body > table-row > table-cell:first-child{
    color: red;
}
#main-table> table-body > table-row > table-cell:nth-child(2){
    color: blue;
}
#main-table> table-body > table-row > table-cell:nth-child(3){
    color: cyan;
}
#main-table> table-body > table-row > table-cell:last-child{
    color: green;
}

Generate the document

There are a few different ways to implement the Document class, but for the sake of simplicity this is a way to generate the document:

from pypfop import Document
from pypfop.makotemplates import TemplateFactory

tfactory = TemplateFactory()
params = {'header': ['Project', 'Website', 'Language', 'Notes'],
          'rows': [('pypfop', 'https://bitbucket.org/cyraxjoe/pypfop',
                    'Python', 'Abstraction on top of Apache FOP'),
                    ('Apache FOP', 'https://xmlgraphics.apache.org/fop/',
                    'Java', '')] }
doc = Document(tfactory('simple-table.fo.mako'), 'simple_table.css')
print(doc.generate(params)) # returns the path of the generated file.

Supported Document formats

In the previous example we didn’t define the output of the Document in that case the default output of pdf is used, but the supported outputs are the same as in Apache FOP output formats.

From pypfop.__init__:

VALID_OFORMATS = frozenset(('awt', 'pdf', 'mif', 'rtf',
                            'tiff', 'png', 'pcl', 'ps', 'txt'))

The output format can be set in Document.__init__ or Document.generate

doc = Document(tfactory('simple-table.fo.mako'), 'simple_table.css',
               oformat='rtf')

or

doc.generate(params, oformat='rtf')

About XSL-FO syntax

As you should notice already it is required to know how to format xsl-fo documents which in most part are very similar to the HTML counterparts (except that anything needs to be in block tags), the best reference that I could find online is in the XML Bible.

Why!

This project it use to be part of a larger project of one of my customers, which I decide early on that I will use only python 3, terrible decision if you want to generate pdf files easily or at least at the moment. I was looking to have some kind of template to the very rigid format of the average invoice and billing order, so pypfop came to relieve that pain.

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

pypfop-0.2.0.tar.gz (11.9 kB view hashes)

Uploaded Source

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page