ftw.pdfgenerator 1.6.11

A library for generating PDF representations of Plone objects with LaTeX.

Introduction

ftw.pdfgenerator is meant to be used for generating PDFs from structured data using predefined LaTeX views. It is not useful for converting full HTML pages into LaTeX / PDFs, although it is able to convert small HTML chunks into LaTeX.

Certified: 01/2013

Requirements

ftw.pdfgenerator requires a TeX distribution with a pdflatex executable to be installed.

These TeX distributions are recommended:

• Mac OS: MacTeX

• Linux / Unix: TeX Live

• Windows: MiKTeX

The package is compatible with Plone 4.3 and 5.1.

Installing

Add ftw.pdfgenerator to your buildout configuration:

[instance]
eggs =
  ftw.pdfgenerator

Usage

The pdfgenerator uses LaTeX for generating the PDF. You need to provide a layout and a view for your context for being able to create a PDF.

Real world examples

Some packages using ftw.pdfgenerator:

• ftw.meeting has a PDF export of the meeting minutes: https://github.com/4teamwork/ftw.meeting/tree/master/ftw/meeting/latex

• ftw.book produces a PDF of the book recursively: https://github.com/4teamwork/ftw.book/tree/master/ftw/book/latex

Defining a layout

A layout is a multi adapter addapting context, request, builder. You can easily define a new layout using the mako templating engine (example: layout.py):

>>> from example.conference.session import ISession
>>> from ftw.pdfgenerator.interfaces import IBuilder
>>> from ftw.pdfgenerator.interfaces import ICustomizableLayout
>>> from ftw.pdfgenerator.layout.customizable import CustomizableLayout
>>> from zope.component import adapts
>>> from zope.interface import Interface
>>> from zope.interface import implements

>>> class SessionLayout(MakoLayoutBase):
...     adapts(ISession, Interface, IBuilder)
...     implements(ICustomizableLayout)
...
...     template_directories = ['session_templates']
...     template_name = 'layout.tex'
...
...     def before_render_hook(self):
...         self.use_babel()
...         self.use_package('inputenc', options='utf8')
...         self.use_package('fontenc', options='T1')

Register the layout with zcml (example: configure.zcml):

<configure
    xmlns="http://namespaces.zope.org/zope"
    xmlns:browser="http://namespaces.zope.org/browser">

    <adapter factory=".layout.SessionLayout"
             provides="ftw.pdfgenerator.interfaces.ILaTeXLayout" />

</configure>

Create a template as defined in SessionLayout.template_name (example: session_templates/layout.tex):

<%block name="documentclass">
\documentclass[a4paper,10pt]{article}
</%block>

<%block name="usePackages">
  ${packages}
</%block>

<%block name="beneathPackages">
</%block>


<%block name="aboveDocument">
</%block>

\begin{document}

<%block name="documentTop">
  % if logo:
    ${logo}
  % endif
</%block>

${content}

<%block name="documentBottom">
</%block>

\end{document}

There are more methods on the layout, see the definition in ftw.pdfgenerator.interfaces.ILaTeXLayout.

Defining a LaTeX view

For every context for which a PDF is generated a LaTeX view (ILaTeXView) is rendered. The view is a multi adapter adapting context, request, layout. There is a view based on the mako templating engine which can be extended (example: views.py):

>>> from example.conference.session import ISession
>>> from ftw.pdfgenerator.interfaces import ILaTeXLayout
>>> from ftw.pdfgenerator.interfaces import ILaTeXView
>>> from ftw.pdfgenerator.view import MakoLaTeXView
>>> from zope.component import adapts
>>> from zope.interface import Interface
>>> from zope.interface import implements

>>> class SessionLaTeXView(MakoLaTeXView):
...     adapts(ISession, Interface, ILaTeXLayout)
...     implements(ILaTeXView)
...
...     template_directories = ['session_templates']
...     template_name = 'view.tex'
...
...     def get_render_arguments(self):
...         return {'title': self.convert(self.context.Title()),
...                 'description': self.convert(self.context.description),
...                 'details': self.convert(self.context.details)}

Register the view with zcml (example: configure.zcml):

<configure
    xmlns="http://namespaces.zope.org/zope"
    xmlns:browser="http://namespaces.zope.org/browser">

    <adapter factory=".views.SessionLaTeXView"
             provides="ftw.pdfgenerator.interfaces.ILaTeXView" />

</configure>

Create a template with the name defined in the class (example: session_templates/view.tex):

\section*{${title}}
% if description:
  \small ${description}
% endif
\normalsize ${details}

Generating a PDF

When a layout and a view for the context are registered the PDF can be generated by simply calling the view @@export_pdf on the context.

Recursive views

When extending from ftw.pdfgenerator.view.RecursiveLaTeXView and inserting the variable latex_content in your template, the view automatically renders all children for which a ILaTeXView is found.

HTML to LaTeX conversion

ftw.pdfgenerator comes with a simple but powerful HTML to LaTeX converter which is optimized for the common WYSIWYG-Editors used in Plone.

The converter can be used:

• in views, using self.convert(html)

• in layouts, using self.get_converter().convert(html)

It uses regular expressions for the simple conversions and python subconverters for the more complicated conversions. The converter is heavily customizable.

Custom subconverters

Footnote

Generate a footnote by wrapping any text in a span with the class footnote. Specify the footnote text in the data-footnote attribute. Example:

<span class="footnote" data-footnote="text in footnote">text on the page</span>

Customizable layouts

When using multiple, independent addon packages using ftw.pdfgenerator, every package may implement a new, specific layout. This can be painful if there is a need to customize all layouts and add a logo image for example.

For making this easier all customizable layouts can be customized with one single adapter. This only works for layouts subclassing ftw.pdfgenerator.layout.customizable.CustomizableLayout. Those layouts need to follow certain concepts and provide inheritable blocks in the mako template. Ensure you follow the standards by subclassing and running the tests from ftw.pdfgenerator.tests.test_customizable_layout.TestCustomizableLayout.

Implementing customization adapter is very simple when customizable layouts are used. For example we change the logo image (assume the logo is at custom/mylogo.png):

>>> from ftw.pdfgenerator.customization import LayoutCustomization
>>> from ftw.pdfgenerator.interfaces import ILayoutCustomization
>>> from zope.interface import implements
>>>
>>> class MyCustomization(LayoutCustomization):
...     implements(ILayoutCustomization)
...
...     template_directories = ['custom']
...     template_name = 'layout_customization.tex'
...
...     def before_render_hook(self):
...         self.add_raw_template_file('mylogo.png')
...         self.layout.use_package('graphicx')
...
...     def get_render_arguments(self, args):
...         args['logo'] = r'\includegraphics{mylogo.png}'
...         return args

It is also possible to change the template and fill predefined slots (example: custom/layout_customization.tex):

<%inherit file="original_layout" />
<%block name="documentTop">
  my branding
</%block>

The layout customization adapter adapts context, request and the original layout.

Tables

ftw.pdfgenerator is able to convert HTML-Tables to LaTeX. Since HTML and LaTeX have completely different presentation concepts the convertion is limitted.

For getting the best results theese rules should be followed:

• Define the width of every column. The table will be streched to the text width in the defined proportions. Without defining the widths LaTeX is unable to insert newlines dynamically.

• Use relative widths (%).

• Define table headings using <thead> for long tables which may be splitted over multiple pages.

CSS classes:

page-break (<table>)

Force the longtable environment, allowing LaTeX to split up the table over multiple pages.

no-page-break (<table>)

Force the tabular environment, prohibiting LaTeX from splitting the table up over multiple pages. If the table is longer than the page it is truncated - content may be missing in this case.

border-grid / listing (<table>)

Display the table in a grid: every cell has a border on every side.

notListed (<table>)

When using a <caption>, do not list the table in the list of tables.

border-left (<td>, <th>)

Display a border on the left side of the cell.

border-right (<td>, <th>)

Display a border on the right side of the cell.

border-top (<td>, <th>)

Display a border on the top side of the cell.

border-bottom (<td>, <th>)

Display a border on the bottom side of the cell.

right (<td>, <th>)

Right align the content of the cell.

left (<td>, <th>)

Left align the content of the cell.

center (<td>, <th>)

Center the content of the cell.

indent2 (<td>, <th>)

Indent the content by 0.2 cm.

indent10 (<td>, <th>)

Indent the content by 1 cm.

bold (<td>, <th>)

Display cell contents in bold font.

grey (<td>, <th>)

Display cell content with grey text color.

footnotesize (<td>, <th>)

Display cell content with smaller font size (\footnotesize).

scriptsize (<td>, <th>)

Display cell content with smaller font size (\scriptsize).

Links

• Github: https://github.com/4teamwork/ftw.pdfgenerator

• Issues: https://github.com/4teamwork/ftw.pdfgenerator/issues

• Pypi: http://pypi.python.org/pypi/ftw.pdfgenerator

• Continuous integration: https://jenkins.4teamwork.ch/search?q=ftw.pdfgenerator

Copyright

This package is copyright by 4teamwork.

ftw.pdfgenerator is licensed under GNU General Public License, version 2.

Changelog

1.6.11 (2024-10-02)

• Remove line break from table environment end and add a smallbreak instead. [buchi]

1.6.10 (2021-05-06)

• Use a cookie instead of session storage to store the debug mode state. [mathias.leimgruber]

1.6.9 (2020-03-11)

• Add support for converting hexadecimal XML/HTML entities [Nachtalb]

1.6.8 (2020-02-11)

• Don’t use reference_catalog for uid look up, since it does not exist anymore in plone 5. [mathias.leimgruber]

1.6.7 (2020-02-03)

• Don’t use session storage anymore - does not work with plone 5 for anonymous users. [mathias.leimgruber]

1.6.6 (2020-01-17)

• Fix unicode error (when unicode and non-unicode strings are joined). [busykoala]

1.6.5 (2019-10-29)

• Fix escaping of underscore in hyperlink labels. [jone]

1.6.4 (2019-03-26)

• Robustified URL conversion. [Rotonen]

1.6.3 (2019-01-07)

• Drop support for plone 4.2. [njohner]

• Handle vertical tab special character. [njohner]

1.6.2 (2018-09-13)

• Fix escaping issues for dash characters in hyperlink labels. [jone]

1.6.1 (2018-05-18)

• Fix encoding error by passing charset hint to BeautifulSoup when converting tables. [lgraf]

• Add Plone 5 compatibility. [phgross]

• Fixed exporting the zip and latex for pdfs who are setting the pdf title. [phgross]

1.6.0 (2017-03-03)

• Add possibility to set the pdf title via build arguments. [phgross]

• Support file objects in Builder.add_file. [jone]

• Drop Plone 4.1 support.

1.5.0 (2016-06-08)

• Implement URL converter for supporting hyphenation in URLs. [jone]

1.4.1 (2016-03-30)

• Listing converter: avoid using beatiful soup. [jone]

• Fix recursive subconverter locking problem. [jone]

1.4 (2016-03-02)

• Added footnote subconverter. [lknoepfel]

1.3.8 (2015-12-11)

• Don’t add rows where not all cells are headcells to the tablehead. [tschanzt]

• Fix table converting when there are less cells than specified and no colspan attribute is given. [tschanzt]

1.3.7 (2015-04-13)

• Fix table converting error when there are more <col> tags than cells. [jone]

• HTML converter: fix leading quote when at beginning of string. [jone]

1.3.6 (2015-03-20)

• Workaround for too long makeindex argument causing buffer overflow. [jone]

1.3.5 (2015-02-19)

• PDF Builder: fix pdflatex after making index. When references are used in the index, e.g. the index title is in the TOC, more than one rerun is required after making the index. [jone]

1.3.4 (2014-09-30)

• Add a quoted_umlauts method to the converter. The method converts all umlauts to the quoted notation. [jone]

• Index: sort German umlauts correctly by including umlaut.ist. For this to work properly all index{} entries must escape umlauts with quotes, e.g. h"oflich. [jone]

1.3.3 (2014-07-24)

• Fix PDF rendering bug when having underscores in internal hyperlinks. [jone]

1.3.2 (2014-06-11)

• Improve table layouting.

  • Increase vertical padding of table cells.

  • Add line break and vertical space after all tables.

  [jone]

1.3.1 (2014-06-02)

• Fix backslash LaTeX to use textbackslash instead of \. This fixes problems with backslash in certain environments such as tables. [jone]

1.3.0 (2014-03-04)

• Run makeindex when *.idx files were created during PDF build. This allows to create indexes easily just by using the latex commands. The builder takes care that the index is built and the PDF rebuilt properly. [jone]

1.2.10 (2014-02-05)

• LaTeX: use math mode for asterisks (*), so they get no longer swallowed. [jone]

1.2.9 (2014-01-17)

• Update French translations. [jone]

• Hyperlinks in footnotes: use url instead of manual hyphenation hints. This fixes that “” is visible in footer urls with certain hyperref package versions. [jone]

• Table: fix error where rowspan cells made other cells swap rows. [jone]

• Tables: fix width of cells spanning multiple columns. This is especially visible when they are centered. [jone]

1.2.8 (2013-10-21)

• Make hyphens before commas non-breakable. [jone]

• Use non-breaking spaces for some more abbreviations. [jone]

• Remove HTML comments when converting HTML to LaTeX. [jone]

• Use non-breaking spaces and en-dashes for formatting swiss currency. [jone]

1.2.7 (2013-05-24)

• Table width calculation: treat “px” the same as widths without measure units. [jone]

1.2.6 (2013-04-17)

• Update package classifiers. [jone]

1.2.5 (2013-01-24)

• Use local text formatting commands, such as textbf instead of {bf }. #13 [jone]

• onegov.ch approved: add badge to readme. [jone]

• Declare missing dependencies. [jone]

• Make layout annotatable. This allows to store certain informations while building the PDF. [jone]

• Plone 4.3 compatibility. [jone]

1.2.4 (2012-08-21)

• Tables: fix wrong indentation when using combined css classes “right” and “bold”. [jone]

• Tables: support width definition in “style” attribute of cells. #10 [jone]

• Fix ampersand escaping in urls. #11 [jone]

• Templating: fix template lookup for five based browser views which are wrapped into a meta class. [jone]

• List conversion: handle empty list items.

  • Removing them in ordered and unordered lists.

  • Keep them in definition lists, since two tags result in one item.

  [jone]

1.2.3 (2012-06-13)

• Table converter: fix centering of cell contents by using correct LaTeX (centering instead of centervspace{-1.5em}). [jone]

• Hyperlink:

  • Support resolveUid (with upper U) resolution.

  • Remove “./” from concatenated relative urls.

  • Support hyphenation in urls.

  [jone]

1.2.2 (2012-06-11)

• Improve support for mathematic characters (utf8 and html-entities). [jone]

• Table: support “summary” attribute as caption (TinyMCE support). [jone]

• Handle hyphenation characters and entities. [jone]

• Hyperlinks: handle underscores and hash keys in urls. [jone]

• Fix centering of table captions which used to affect contents beneath. [jone]

• Fix quotation marks around bold text. [jone]

1.2.1 (2012-05-10)

• Table converter: fix TypeError when using colspan without defining any widths. [jone]

1.2 (2012-05-09)

• HTML to LaTeX: escape $ charactor. Otherwise it enables math-mode. [jone]

• Tables: use seperate caption implementation for non-floating environments. [jone]

• Tables: fix border collapse bug when using rowspan. [jone]

• Tables: calculate widths properly respecting tabcolsep. [jone]

• Tables: support plone css classes (grid, listing, vertical). [jone]

1.1.2 (2012-05-02)

• Add convert_plain method to converter and view. [phgross]

• Add encode_htmlentities utils function. [phgross]

1.1.1 (2012-04-30)

• Listing converter: limit nested lists to four levels since LaTeX has this limit. [jone]

• Listing converter: fix bug when nested lists are not within a item. [jone]

• Add provide_request_layer util method to directly provide an interface on the request [eschmutz]

1.1 (2012-03-22)

• Improve testing speed by using ftw.testing. [jone]

• Add table converter css classes: indent2 -> indent the content 0.2cm indent10 -> indent the content 1cm bold -> display content with bold font grey -> Display cell content with grey text color. footnotesize -> Display cell content with smaller font size. scriptsize -> Display cell content with smaller font size. [jone]

• Listing converter: cleanup bad html in case when lists are nested without list items. FCKEditor somehow produces that sometimes. [jone]

• Implement customizable layouts. [jone]

1.0.2 (2012-03-08)

• Table converter: add css classes for aligning cells (“right”, “center”, “left”). [jone]

• Table converter: add cell-border functionality, using css classes “border-right”, “border-left”, “border-top” and “border-bottom”. [jone]

• Table converter: improve grid border, add classes “border-grid” and “listing” for enabling grid borders. [jone]

• Table converter: use “tabular” for small tables. #3 [jone]

1.0.1 (2012-03-05)

• Add use_babel function to layout, enabling the preferred language. [jone]

1.0b2 (2012-02-28)

• Added missing MANIFEST.in. [phgross]

1.0b1 (2012-02-24)

• Added a “debug-pdf” view, enabling the pdf debug mode temporarily. [jone]

• Added some French translations [ttschanz]

• Implement hyperlink converter. [jone]

• Implement convertion of definition lists. [jone]

• Rename as_pdf view to export_pdf. [jone]