brechtm / rinohtype Goto Github PK

View Code? Open in Web Editor NEW

495.0 20.0 60.0 131.07 MB

The Python document processor

Home Page: http://www.mos6581.org/rinohtype

License: GNU Affero General Public License v3.0

Python 99.81% HTML 0.08% Shell 0.12%

python pdf pdf-generation pdf-generator typesetting typesetting-system restructuredtext sphinx-extension

rinohtype's Introduction

rinohtype

rinohtype is a batch-mode document processor. It renders structured documents to PDF based on a document template and a style sheet. An important design goal of rinohtype is make document layout and style customization user-friendly. Have a look at the showcase to get an idea of the level of customization that is possible. See the documentation to learn how to customize the style of your document.

Call for Contributions

Since rinohtype is a fairly sizable project and currently being maintained by a single person, your contribution can make a big difference. Specifically, the following things can help move rinohtype forward:

development of professional-looking stylesheets and document templates
volunteering to be a maintainer: fix issues that pop up when new versions of dependencies are released (Python, Sphinx, ...), or handling platform-specific regressions (development is mainly on macOS).
help with maintaining and improving the documentation
development of new features, e.g. widow/orphan handling, Knuth-Plass line breaking, mathematics typesetting, performance improvements, ...
companies might be interested in funding the development of particular features

So if you are interested in helping with any of these items, please don't hesitate to get in touch via Discussions, Gitter or [email protected]!

Features

rinohtype is still in beta, so you might run into some issues when using it. I'd highly appreciate it if you could create a ticket for any bugs you may encounter. That said, rinohtype is already quite capable. For example, it should be able to replace Sphinx's LaTeX builder in most cases. Here is an overview of the main features:

a powerful page layout system supporting columns, running headers/footers, floatable elements and footnotes
support for figures and (large) tables, optionally rendered sideways
automatic generation of table of contents and index
automatic numbering and cross-referencing of section headings, figures and tables
configure one of the included document templates or create your own
an intuitive style sheet system inspired by CSS allowing changing almost every aspect of how document elements are rendered
modular design allowing for multiple frontends (such as reStructuredText, Markdown, DocBook, ...)
handles OpenType, TrueType and Type1 fonts with support for advanced typographic features such as kerning, ligatures, small capitals and old style figures
built-in support for the 1000+ libre licensed fonts on Google Fonts
embeds PDF, PNG and JPEG images, preserving transparency and color profiles
easy to install and deploy; pure-Python with few dependencies
built on Unicode; ready for non-latin languages

rinohtype's primary input format is reStructuredText. The rinoh command line tool renders reStructuredText documents and the included Sphinx builder makes it possible to output large documents with your own style applied. Have a look at the rinohtype manual for an example of the output.

There is also a commercial DITA frontend, but it's development is currently on hold. Please contact me if you are interested in testing it.

Requirements

rinohtype supports all stable Python 3 versions that have not reached end-of-life status. For parsing reStructuredText and CommonMark documents, rinohtype depends on docutils and myst-parser respectively. pip takes care of installing these requirements when you install rinohtype.

Syntax highlighting of code blocks is enabled if Pygments is installed, which will be installed automatically with Sphinx. If you want to include images other than PDF, PNG or JPEG, you also need to install Pillow.

Getting Started

Installation is trivial:

pip install rinohtype

If you want to have access to bug fixes and features that are not available in a release, you can install the current development version:

pip install https://github.com/brechtm/rinohtype/archive/refs/heads/master.zip

reStructuredText Renderer

The easiest way to get started with rinohtype is to render a reStructuredText document (such as CHANGES.rst from this repository) using the rinoh command line tool:

rinoh CHANGES.rst

When rinoh finishes, you will find CHANGES.pdf alongside the input file.

By default rinoh renders the input document using the article template. Run rinoh --help to see how you can tell rinoh which document template and style sheet to use.

Sphinx Builder

rinohtype can be used as a drop-in replacement for the LaTeX builder (the latex_documents configuration variable has to be set). Simply select the rinoh builder when building the Sphinx project:

sphinx-build -b rinoh . _build/rinoh

Contributing

See CONTRIBUTING.rst and DEVELOPING.rst

License

All of rinohtype's source code is licensed under the Affero GPL 3.0, unless indicated otherwise in the source file (such as hyphenator.py and purepng.py).

The Affero GPL requires for software that builds on rinohtype to also be released as open source under this license. For building closed-source applications, you can obtain a commercial license. The author of rinohtype is also available for consultancy projects involving rinohtype.

rinohtype's People

Contributors

Stargazers

Watchers

Forkers

mariuscc rhythmus mathb janezstupar cedk ascribe frol calve itziakos rowhit erickpeirson seizmo pombredanne fizzadar xiaogh98 cwacekinv kumanoko24 barseghyanartur cantemointernal forkedjensh firefoxxy8 gaybro8777 hznupeter stephenfin dmytrokyrychuk sgaechter johnson04 desh-se z00sts lukzeg epoz normanlorrain electroniceagle alexfargus chrisjrwilliams parz3val hamzamohdzubair foster999 wxtim natashazu tavaresfilipe egamida rappdw ian-datature gatherer hartmans jemrobinson carl4 evildmp mawieland mforbes techauthoruk marian-vignau evdcush priyatham10 timm638 jwhitham th0mr pabloalexis611

rinohtype's Issues

Configure Travis continuous integration

Automatically:

run the unit tests on all supported Python versions
run regression tests, comparing PDF output pixel-by-pixel using diffpdf.sh

Make use of tox and coveralls.

Extend CONTRIBUTING.rst

See keep-contributing.

Provide standard document templates such as book and article

The Sphinx frontend includes a book template. This basically consists of page and document part templates. Turn this into a proper (customizable) book template object and create an article template in the same way.

Optimize rendering performance

Many features have been added since the rendering loop was last profiled and optimized. See where in the code most time is spent and optimize these parts.

Create a rinohtype style sheet to CSS mapper

This will allow for a quick pseudo-WYSIWYMG preview of the document style.

HTML5 tags

Rinohtype is a wonderful document generator, but I am wondering that whether you can use HTML5 tags and a CSS-like styling system for input, such as <table> instead of <tabular>. It may help publishing your document to both the web and the print.

Automatically handle placement of the date on the title page

If a Sphinx project includes a large logo (Lizmap for example), the date drops off the page. The vertical spacing between the title, subtitle, author and date could be "stretchy" to handle this better.

ClassNotFound exception in Literal_Block.lexer_getter()

Rendering the Lizmap documentation:

  ...
  File "/Users/brechtm/Documents/Code/rinohtype/src/rinoh/highlight.py", line 54, in highlight_block
    lexer = get_lexer(text, language)
  File "/Users/brechtm/Documents/Code/rinohtype/src/rinoh/frontend/sphinx/nodes.py", line 122, in lexer_getter
    lexer = get_lexer_by_name(language)
  File "/Users/brechtm/.pyenv/versions/rinoh-py35/lib/python3.5/site-packages/pygments/lexers/__init__.py", line 93, in get_lexer_by_name
    raise ClassNotFound('no lexer for alias %r found' % _alias)
pygments.util.ClassNotFound: no lexer for alias 'none' found

Adapt docutils frontend to table enhancements

:widths: https://sourceforge.net/p/docutils/patches/120/
:width: https://sourceforge.net/p/docutils/patches/121/

Tweak packaging and testing

Internally represent style log as a tree

This more easily allows for a better representation of the log:

plain text log: using ASCII/Unicode lines to visualize the document tree structure
HTML log: allow dynamically collapsing/expanding nodes (and hide all inline elements for example)

Show a processed example (link) in the README

Unless I've missed it, please add your "most beautiful" demo/example processed to the documentation.
Great work!

Complete support for code highlighting

Syntax highlighting of code blocks in Sphinx projects does not yet support:

background color (and border) for inline text
highlighting of lines
displaying line numbers

Consider replacing the Harmony CLA with one based on the GitHub CLA

The GitHub CLA is much shorter and thus less scary.

GitHub allows using the CLA text under CC-BY-4.0 terms (ref email from Mike Linksvayer).

Installation broken

While reviewing your proposal for EuroPython 2014 I came the following installation problem:

Suxmac-2:rinohtype ajung$ bin/python setup.py install
Attempting to get version number from git...
fatal: No names found, cannot describe anything.
Assume we are running from a source distribution.
Traceback (most recent call last):
File "setup.py", line 37, in
with open(VERSION_FILE) as version_file:
IOError: [Errno 2] No such file or directory: 'rinoh/version.py'

I don't think that it is of advantage writing a proposal for a module that can not be installed properly.

Consider using pbr to replace ad-hoc code in setup.py

pbr helps managing version numbers based on git tags, but also offer other common functionality related to packaging.

Turn templates into resources

Just like style sheets and typefaces, templates could be uniquely identified by name and automatically installed from PyPI.

Not sure how to handle TemplateConfiguration. Should this also be a Resource?

Long strings in a code block disappear

Example: in the "HTTP Verbs" section of the requests documentation.

Code block::

    >>> import requests
    >>> r = requests.get('https://api.github.com/repos/kennethreitz/requests/git/commits/a050faf084662f3a352dd1a941f2c7c9f886d4ad')

Host documentation for all releases

Using sphinxcontrib-versioning should allow for hosting rinohtype-generated PDFs as well.

RecursionError: maximum recursion depth exceeded while calling a Python object

demo.txt:

.. [#label] Footnotes may be numbered, either manually (as in [1]_) or
   automatically using a "#"-prefixed label.  This footnote has a
   label so it can be referred to from multiple places, both as a
   footnote reference ([#label]_) and as a hyperlink reference
   (label_).

$ rinoh demo.txt
qq.txt:1: (ERROR/3) Unknown target name: "1".
RinohType 0.1.3-154-g264f (Feb 28 2016, 09:57:09)  Copyright (c) Brecht Machiels
This program comes with ABSOLUTELY NO WARRANTY. Its use is subject
to the terms of the GNU Affero General Public License version 3.
No style 'table of contents title' found in stylesheet
No style 'table of contents title' found in stylesheet
No style 'table of contents title' found in stylesheet
Traceback (most recent call last):
  File "env3/bin/rinoh", line 5, in <module>
    main()
  File "env3/lib/python3.5/site-packages/rinoh/tool.py", line 44, in main
    document.render(input_root)
  File "env3/lib/python3.5/site-packages/rinoh/document.py", line 360, in render
    section_num_pages = self.render_pages(_sections)
  File "env3/lib/python3.5/site-packages/rinoh/document.py", line 386, in render_pages
    section_page_count = section.render(sum(section_page_counts))
  File "env3/lib/python3.5/site-packages/rinoh/document.py", line 235, in render
    part_page_count = part.render(doc_page_count + section_page_count)
  File "env3/lib/python3.5/site-packages/rinoh/document.py", line 171, in render
    chains_requiring_new_page = set(chain for chain in page.render())
  File "env3/lib/python3.5/site-packages/rinoh/document.py", line 171, in <genexpr>
    chains_requiring_new_page = set(chain for chain in page.render())
  File "env3/lib/python3.5/site-packages/rinoh/document.py", line 117, in render
    for chain in super().render(rerender=index > 0):
  File "env3/lib/python3.5/site-packages/rinoh/layout.py", line 178, in render
    for chain in child.render(rerender):
  File "env3/lib/python3.5/site-packages/rinoh/layout.py", line 178, in render
    for chain in child.render(rerender):
  File "env3/lib/python3.5/site-packages/rinoh/layout.py", line 278, in render
    last_descender=last_descender):
  File "env3/lib/python3.5/site-packages/rinoh/layout.py", line 557, in render
    self._state.flowable_state)
  File "env3/lib/python3.5/site-packages/rinoh/flowable.py", line 125, in flow
    state=state, **kwargs)
  File "env3/lib/python3.5/site-packages/rinoh/flowable.py", line 348, in render
    render_content(maybe_container, last_descender)
  File "env3/lib/python3.5/site-packages/rinoh/flowable.py", line 334, in render_content
    descender, state=state)
  File "env3/lib/python3.5/site-packages/rinoh/flowable.py", line 125, in flow
    state=state, **kwargs)
  File "env3/lib/python3.5/site-packages/rinoh/flowable.py", line 244, in render
    state=state.first_flowable_state, **kwargs)
  File "env3/lib/python3.5/site-packages/rinoh/flowable.py", line 125, in flow
    state=state, **kwargs)
  ...
  File "env3/lib/python3.5/site-packages/rinoh/paragraph.py", line 260, in render
    line_width = float(container.width)
  File "env3/lib/python3.5/site-packages/rinoh/dimension.py", line 129, in __float__
    return float(self.minuend) - float(self.subtrahend)
  File "env3/lib/python3.5/site-packages/rinoh/dimension.py", line 129, in __float__
    return float(self.minuend) - float(self.subtrahend)
  File "env3/lib/python3.5/site-packages/rinoh/dimension.py", line 129, in __float__
    return float(self.minuend) - float(self.subtrahend)
  File "env3/lib/python3.5/site-packages/rinoh/dimension.py", line 129, in __float__
    return float(self.minuend) - float(self.subtrahend)
  File "env3/lib/python3.5/site-packages/rinoh/dimension.py", line 129, in __float__
    return float(self.minuend) - float(self.subtrahend)
  File "env3/lib/python3.5/site-packages/rinoh/dimension.py", line 129, in __float__
    return float(self.minuend) - float(self.subtrahend)
  File "env3/lib/python3.5/site-packages/rinoh/dimension.py", line 129, in __float__
    return float(self.minuend) - float(self.subtrahend)
  File "env3/lib/python3.5/site-packages/rinoh/dimension.py", line 129, in __float__
    return float(self.minuend) - float(self.subtrahend)
  File "env3/lib/python3.5/site-packages/rinoh/dimension.py", line 129, in __float__
    return float(self.minuend) - float(self.subtrahend)
  File "env3/lib/python3.5/site-packages/rinoh/dimension.py", line 129, in __float__
    return float(self.minuend) - float(self.subtrahend)
  File "env3/lib/python3.5/site-packages/rinoh/dimension.py", line 129, in __float__
    return float(self.minuend) - float(self.subtrahend)
  File "env3/lib/python3.5/site-packages/rinoh/dimension.py", line 109, in __float__
    return float(self._value)
RecursionError: maximum recursion depth exceeded while calling a Python object

Hide the index chapter when there are no index entries

Even when there are no index entries in the document, the Index chapter is still being included in the document.

Avoid a second rendering pass if it is not needed

If the document contains no page references, a second rendering pass is not required. We can still save the .rtc though.

If a table of contents or index is present, there will always be page references though.

Article template: optionally place the title on the same page as the body text

Now, the article template includes a separate page for the title. It should be possible to place the title on the same page as the body text (spanning columns).

Undeclared purepng > 0.1 dependency

setup.py should declare the dependency on purepng, in install_requires. Note that currently this dependency is on an unreleased version of purepng.

Running rinoh in a fresh virtualenv. :

(env) alex@martha:~/src/rinohtype$ pip install -e .
...
Successfully installed RinohType
Cleaning up...
(env) alex@martha:~/src/rinohtype$ rinoh
Traceback (most recent call last):
  File "/home/alex/src/rinohtype/env/bin/rinoh", line 6, in <module>
    exec(compile(open(__file__).read(), __file__, 'exec'))
  File "/home/alex/src/rinohtype/bin/rinoh", line 4, in <module>
    from rinoh.tool import main
  File "/home/alex/src/rinohtype/rinoh/__init__.py", line 37, in <module>
    module = import_module('.' + name, __name__)
  File "/usr/lib/python3.4/importlib/__init__.py", line 109, in import_module
    return _bootstrap._gcd_import(name[level:], package, level)
  File "/home/alex/src/rinohtype/rinoh/document.py", line 31, in <module>
    from .backend import pdf
  File "/home/alex/src/rinohtype/rinoh/backend/pdf/__init__.py", line 19, in <module>
    from .png import PNGReader
  File "/home/alex/src/rinohtype/rinoh/backend/pdf/png.py", line 11, in <module>
    import png as purepng
ImportError: No module named 'png'

Install the latest released version of purepng (0.1.0), running rinoh raises an AttributeError:

(env) alex@martha:~/src/rinohtype$ pip install purepng
Downloading/unpacking purepng
  Downloading purepng-0.1.0.zip (284kB): 284kB downloaded
...
Successfully installed purepng
Cleaning up...
(env) alex@martha:~/src/rinohtype$ rinoh
Traceback (most recent call last):
  File "/home/alex/src/rinohtype/env/bin/rinoh", line 6, in <module>
    exec(compile(open(__file__).read(), __file__, 'exec'))
  File "/home/alex/src/rinohtype/bin/rinoh", line 4, in <module>
    from rinoh.tool import main
  File "/home/alex/src/rinohtype/rinoh/__init__.py", line 37, in <module>
    module = import_module('.' + name, __name__)
  File "/usr/lib/python3.4/importlib/__init__.py", line 109, in import_module
    return _bootstrap._gcd_import(name[level:], package, level)
  File "/home/alex/src/rinohtype/rinoh/document.py", line 31, in <module>
    from .backend import pdf
  File "/home/alex/src/rinohtype/rinoh/backend/pdf/__init__.py", line 19, in <module>
    from .png import PNGReader
  File "/home/alex/src/rinohtype/rinoh/backend/pdf/png.py", line 166, in <module>
    RENDERING_INTENT = {purepng.ABSOLUTE_COLORIMETRIC: ABSOLUTE_COLORIMETRIC,
AttributeError: 'module' object has no attribute 'ABSOLUTE_COLORIMETRIC'

Install an unreleased build of purepng, straight from github. rinoh runs

(env) alex@martha:~/src/rinohtype$ pip uninstall purepng
Uninstalling purepng:
...
  Successfully uninstalled purepng
(env) alex@martha:~/src/rinohtype$ pip install https://github.com/Scondo/purepng/archive/58f553e62a4b608df04b19d2ff83f12607b49f63.zip
Downloading/unpacking https://github.com/Scondo/purepng/archive/58f553e62a4b608df04b19d2ff83f12607b49f63.zip
  Downloading 58f553e62a4b608df04b19d2ff83f12607b49f63.zip (unknown size): 267kB downloaded
  Running setup.py (path:/tmp/pip-tdvayijc-build/setup.py) egg_info for package from https://github.com/Scondo/purepng/archive/58f553e62a4b608df04b19d2ff83f12607b49f63.zip

    warning: no directories found matching 'html'
    warning: no previously-included files matching '*.pyc' found anywhere in distribution
    warning: no previously-included files matching '*.DS_Store' found anywhere in distribution
Installing collected packages: purepng
  Running setup.py install for purepng

    warning: no directories found matching 'html'
    warning: no previously-included files matching '*.pyc' found anywhere in distribution
    warning: no previously-included files matching '*.DS_Store' found anywhere in distribution
Successfully installed purepng
Cleaning up...
(env) alex@martha:~/src/rinohtype$ rinoh
usage: rinoh [-h] [--paper [PAPER]] [input]

Render a reStructuredText document to PDF.

positional arguments:
  input            the reStructuredText document to render

optional arguments:
  -h, --help       show this help message and exit
  --paper [PAPER]  the paper size to render to (default: A4)
(env) alex@martha:~/src/rinohtype$

Add option to keep minimally x number of lines together

Add a ParagraphStyle attribute split_minimum_lines to configure how paragraphs are split across pages. For example, if split_minimum_lines is set to 3, a paragraph will start on the next page if less than 2 lines fit on the current page.

Travis CI: install poppler-utils 0.41.0

The poppler-utils version (0.18.4) provided by Ubuntu 12.04 cannot properly render PDFs generated by rinohtype (see #2). Poppler 0.41.0 can be built on Ubuntu, so we could use that one:

apt-get install libopenjpeg
wget https://poppler.freedesktop.org/poppler-0.41.0.tar.xz
tar xf poppler-0.41.0.tar.xz
cd poppler-0.41.0
./configure --prefix=/usr
make
make install

We should package this in a .deb so it can be easily installed before running the tests on Travis.

Non-latin characters produce "?" output instead of a proper character

demo.txt

Non-latin characters
====================

Ф - Cyrillyc

검 - Korean

ت - Arabic

$ rinoh  demo.rst
RinohType 0.1.3-154-g264f (Feb 28 2016, 09:57:09)  Copyright (c) Brecht Machiels
This program comes with ABSOLUTELY NO WARRANTY. Its use is subject
to the terms of the GNU Affero General Public License version 3.
No style 'table of contents title' found in stylesheet
No style 'table of contents title' found in stylesheet
No style 'table of contents title' found in stylesheet
TeXGyrePagella-Regular does not contain glyph for unicode index 0x0424 (Ф)
TeXGyrePagella-Regular does not contain glyph for unicode index 0xac80 (검)
TeXGyrePagella-Regular does not contain glyph for unicode index 0x062a (ت)
TeXGyreHeros-Bold does not contain glyph for unicode index 0x0009 (     )
Writing output: demo.pdf

PDF contains:

? - Cyrillyc
? - Korean
? - Arabic

UnboundLocalError: local variable 'width' referenced before assignment

demo.txt:

+------------------------+------------+----------+----------+
| Header row, column 1   | Header 2   | Header 3 | Header 4 |
| (header rows optional) |            |          |          |
+========================+============+==========+==========+
| body row 1, column 1   | column 2   | column 3 | column 4 |
+------------------------+------------+----------+----------+

$ rinoh demo.txt
RinohType 0.1.3-154-g264f (Feb 28 2016, 09:57:09)  Copyright (c) Brecht Machiels
This program comes with ABSOLUTELY NO WARRANTY. Its use is subject
to the terms of the GNU Affero General Public License version 3.
No style 'table of contents title' found in stylesheet
No style 'table of contents title' found in stylesheet
No style 'table of contents title' found in stylesheet
No style 'table head cell right border' found in stylesheet
No style 'table cell right border' found in stylesheet
No style 'table head cell right border' found in stylesheet
No style 'table cell right border' found in stylesheet
No style 'table head left border' found in stylesheet
No style 'table left border' found in stylesheet
No style 'table head cell left border' found in stylesheet
No style 'table cell left border' found in stylesheet
No style 'table head left border' found in stylesheet
No style 'table left border' found in stylesheet
No style 'table head cell left border' found in stylesheet
No style 'table cell left border' found in stylesheet
Traceback (most recent call last):
  File "env3/lib/python3.5/site-packages/rinoh/util.py", line 101, in function_wrapper
    return cache[args_kwargs]
KeyError: ('vertical_align', <rinohlib.templates.article.Article object at 0x7f0cad6bc4e0>)

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
  File "env3/lib/python3.5/site-packages/rinoh/util.py", line 101, in function_wrapper
    return cache[args_kwargs]
KeyError: (<rinohlib.templates.article.Article object at 0x7f0cad6bc4e0>,)

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
  File "env3/lib/python3.5/site-packages/rinoh/flowable.py", line 412, in flow
    state)
  File "env3/lib/python3.5/site-packages/rinoh/flowable.py", line 125, in flow
    state=state, **kwargs)
  File "env3/lib/python3.5/site-packages/rinoh/table.py", line 101, in render
    if render_rows(self.head) != len(self.head):
  File "env3/lib/python3.5/site-packages/rinoh/table.py", line 95, in render_rows
    rendered_rows)
  File "env3/lib/python3.5/site-packages/rinoh/table.py", line 242, in _place_rows_and_render_borders
    document)
  File "env3/lib/python3.5/site-packages/rinoh/util.py", line 103, in function_wrapper
    cache_value = function(obj, *args, **kwargs)
  File "env3/lib/python3.5/site-packages/rinoh/style.py", line 338, in get_style
    return self.get_style_recursive(attribute, flowable_target)
  File "env3/lib/python3.5/site-packages/rinoh/style.py", line 358, in get_style_recursive
    style = self._style(flowable_target)
  File "env3/lib/python3.5/site-packages/rinoh/util.py", line 103, in function_wrapper
    cache_value = function(obj, *args, **kwargs)
  File "env3/lib/python3.5/site-packages/rinoh/style.py", line 374, in _style
    return container.document.stylesheet.find_style(self, container)
AttributeError: 'Article' object has no attribute 'document'

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
  File "env3/bin/rinoh", line 5, in <module>
    main()
  File "env3/lib/python3.5/site-packages/rinoh/tool.py", line 44, in main
    document.render(input_root)
  File "env3/lib/python3.5/site-packages/rinoh/document.py", line 360, in render
    section_num_pages = self.render_pages(_sections)
  File "env3/lib/python3.5/site-packages/rinoh/document.py", line 386, in render_pages
    section_page_count = section.render(sum(section_page_counts))
  File "env3/lib/python3.5/site-packages/rinoh/document.py", line 235, in render
    part_page_count = part.render(doc_page_count + section_page_count)
  File "env3/lib/python3.5/site-packages/rinoh/document.py", line 171, in render
    chains_requiring_new_page = set(chain for chain in page.render())
  File "env3/lib/python3.5/site-packages/rinoh/document.py", line 171, in <genexpr>
    chains_requiring_new_page = set(chain for chain in page.render())
  File "env3/lib/python3.5/site-packages/rinoh/document.py", line 117, in render
    for chain in super().render(rerender=index > 0):
  File "env3/lib/python3.5/site-packages/rinoh/layout.py", line 178, in render
    for chain in child.render(rerender):
  File "env3/lib/python3.5/site-packages/rinoh/layout.py", line 178, in render
    for chain in child.render(rerender):
  File "env3/lib/python3.5/site-packages/rinoh/layout.py", line 278, in render
    last_descender=last_descender):
  File "env3/lib/python3.5/site-packages/rinoh/layout.py", line 557, in render
    self._state.flowable_state)
  File "env3/lib/python3.5/site-packages/rinoh/flowable.py", line 417, in flow
    self._align(align_container, width)
UnboundLocalError: local variable 'width' referenced before assignment

A document without a table of contents or without a chapter crashes to build

The minimal working document is:

.. contents:: Table of Contents

Structural Elements
===================

If I remove any line from here I will get the following traceback:

Traceback (most recent call last):
  File "/tmp/env/bin/rinoh", line 5, in <module>
    main()
  File "/tmp/env/lib/python3.5/site-packages/rinoh/tool.py", line 43, in main
    document.render(input_base)
  File "/tmp/env/lib/python3.5/site-packages/rinoh/document.py", line 313, in render
    section_num_pages = self.render_pages()
  File "/tmp/env/lib/python3.5/site-packages/rinoh/document.py", line 333, in render_pages
    section_page_count = section.render(sum(section_page_counts))
  File "/tmp/env/lib/python3.5/site-packages/rinoh/document.py", line 206, in render
    part_page_count = part.render(doc_page_count + section_page_count)
  File "/tmp/env/lib/python3.5/site-packages/rinoh/document.py", line 152, in render
    chains_requiring_new_page = set(chain for chain in page.render())
  File "/tmp/env/lib/python3.5/site-packages/rinoh/document.py", line 152, in <genexpr>
    chains_requiring_new_page = set(chain for chain in page.render())
  File "/tmp/env/lib/python3.5/site-packages/rinoh/document.py", line 109, in render
    for chain in super().render(rerender=index > 0):
  File "/tmp/env/lib/python3.5/site-packages/rinoh/layout.py", line 221, in render
    for chain in child.render(rerender):
  File "/tmp/env/lib/python3.5/site-packages/rinoh/layout.py", line 225, in render
    height, last_descender = flowable.flow(self, last_descender)
  File "/tmp/env/lib/python3.5/site-packages/rinoh/flowable.py", line 123, in flow
    state=state, **kwargs)
  File "/tmp/env/lib/python3.5/site-packages/rinoh/paragraph.py", line 293, in render
    word = state.next_word()
  File "/tmp/env/lib/python3.5/site-packages/rinoh/paragraph.py", line 238, in next_word
    word = next(self._words)
  File "/tmp/env/lib/python3.5/site-packages/rinoh/paragraph.py", line 200, in spans_to_words
    for chars in span.split(container):
  File "/tmp/env/lib/python3.5/site-packages/rinoh/reference.py", line 241, in split
    if section else None) or ''
  File "/tmp/env/lib/python3.5/site-packages/rinoh/reference.py", line 42, in get_id
    return self.id or document.ids_by_element[self]
AttributeError: 'TableOfContentsSection' object has no attribute 'ids_by_element'

Discover fonts installed on the system

Scan the OS-specific font paths for fonts and make them available for use in rinohtype. This might require caching as each font needs to be parsed to find out its name.

Use more descriptive style labels in the matcher and style sheets

body ➡️ body paragraph
emphasis ➡️ emphasized text
...

currently Python 3.4 is required

According to the README 3.2 and up should be supported, but at least base64.a85{en,de}code are new in 3.4.

Extend the rinoh command line tool's options

Allow for:

selecting a style sheet (by name or by path)
choosing a document template
setting template options
automatic detection of the input format (each frontend should provide a function to test a file)
...

It is impossible to verbally communicate the name without commentary regarding the spelling.

Where does "rinoh" come from? In English at least, it is impossible to tell someone the name of this program without adding explanations regarding the spelling, and expect the listener to know how to spell it.

Render equations

The old ideas.txt might contain some relevant information/URLs.

Footnotes do not work properly

Rinohtype has trouble rendering the reStructuredText Demonstration document demo.txt.

Test cases should be created to detect regressions.

Images that don't fit are still placed on the page

The Lizmap documentation contains many images and illustrates this problem many times.

Python 3.5 is not supported

When I run RinohType on Python 3.5 (I haven't tried 3.4 or earlier versions), I get the traceback:

  File "/tmp/env/lib/python3.5/site-packages/rinoh/backend/pdf/cos.py", line 75, in <module>
    class Null(Object, NoneType):
TypeError: type 'NoneType' is not an acceptable base type

Add support for Markdown files to rinoh

rinoh should be able to render CommonMark files besides reStructuredText files.

autodetection based on file extension
each frontend should provide a list of extensions to help with this
option to manually specify the input file format

We could even detect the input file format by simply parsing the input by each parser and see if it errors...

Support hyperlinked images

reStructuredText allows for images to link to a target:

.. image:: images/biohazard.png
   :target: http://www.python.org

Rinohtype does not yet create the hyperlink in the PDF.

Handle images more like the LaTeX Sphinx builder?

Some of the images in the Lizmap documentation are displayed much larger than in the LaTeX-generated PDF.

Notes:

~~this is perhaps only the case for images that do not have a DPI set~~
LaTeX's \includegraphics or Sphinx's \sphinxincludegraphics may always scale down the image if it doesn't fit the available width
setting an .. image:'s :scale: 80% seems to scale the image to 80% of the text width (not 80% of the image's original size)

While it is convenient to have the images fit the available width if they are too wide, it is probably not a good idea to scale relative to the text width. If supported by Sphinx/docutils, the user should set :width: 80%.

The Into Pdf seems to be broken

At last it does not look like i think it should on my computer in neither evince nor firefox. I'm talking about this file

Investigate whether memory usage can be reduced

Rendering the Sphinx documentation takes about 2GB of memory.

Top 10 lines
#1: pdf/__init__.py:143: 400.0 MiB
    parent_canvas.write(self.getvalue())
#2: rinoh/util.py:115: 189.0 MiB
    cache[args_kwargs] = cache_value
#3: rinoh/dimension.py:156: 88.0 MiB
    self.addends = list(addends)
#4: rinoh/layout.py:218: 58.6 MiB
    self._self_cursor = Dimension(0)  # initialized at container's top edge
#5: pdf/__init__.py:136: 58.2 MiB
    self.fonts = {}
#6: rinoh/style.py:290: 54.2 MiB
    if not hasattr(styled, attr) or getattr(styled, attr) != value:
#7: rinoh/style.py:317: 48.1 MiB
    return type(self)(*(child_selector for selector in self.selectors
#8: pdf/__init__.py:137: 47.5 MiB
    self.images = {}
#9: packaging/version.py:353: 41.2 MiB
    reversed(release),
#10: pdf/__init__.py:344: 35.9 MiB
    self.annotation = annotation
11976 other: 819.5 MiB
Total allocated size: 1840.2 MiB

See:

Make Travis deploy releases to PyPI

The tox-travis project does this, for example.

References:

https://docs.travis-ci.com/user/deployment/pypi
tox-travis's .travis.yml

Traceback (most recent call last):
  File "/Users/taku/dev/src/github.com/brechtm/rinohtype/venv/bin/rinoh", line 6, in <module>
    exec(compile(open(__file__).read(), __file__, 'exec'))
  File "/Users/taku/dev/src/github.com/brechtm/rinohtype/bin/rinoh", line 5, in <module>
    main()
  File "/Users/taku/dev/src/github.com/brechtm/rinohtype/rinoh/tool.py", line 89, in main
    options = ArticleOptions(page_size=page_size)
  File "/Users/taku/dev/src/github.com/brechtm/rinohtype/rinoh/template.py", line 248, in __init__
    raise AttributeError('No such document option: {}'.format(name))
AttributeError: No such document option: page_size

Run tests on AppVeyor

The appveyor branch already includes a basic configuration. What needs to be done:

diffpdf.shneeds to be converted to a Python script to allow running on Windows (and this would make it easier to maintain)
ImageMagick's convert and Poppler's pdftoppm and pdfinfo need to be installed
have each build matrix entry build only the relevant Tox environment

The AppVeyor web hook needs to be enabled in GitHub's Settings > Webhooks & services.

support markdown, e.g. the one used by authorea

Copying text from the PDF output is broken

Small capitals and ligatures are not mapped back to unicode characters, for example.