In the examples, we see that graphviz adds a suffix to the filename provided based on the format:

>>> dot.render('test-output/round-table.gv', view=True)  # doctest: +SKIP
'test-output/round-table.gv.pdf'

Would it not be better to just chose the format based on the filename suffix?

Currently, I have a workaround where I write:

p = Path(filename)
if p.suffix == '.svg':
    dot.format = 'svg'

...

dot.render(p.stem)

or else the suffix is appended to the filename twice.

I'm trying to package graphviz version 0.20.1 for OpenIndiana but some tests are failing. Here is an example of failure (all tests fails with the same error):

______________________ test_render[dot-pdf-None-None-pdf] ______________________

capsys = <_pytest.capture.CaptureFixture object at 0x7fffa964c600>
tmp_path = PosixPath('/tmp/pytest-of-marcel/pytest-16/test_render_dot_pdf_None_None_0')
engine = 'dot', format_ = 'pdf', renderer = None, formatter = None
expected_suffix = 'pdf', filename = 'hello.gv'
data = b'digraph { hello -> world }'

    @pytest.mark.exe
    @pytest.mark.parametrize(
        'format_, renderer, formatter, expected_suffix',
        [('pdf', None, None, 'pdf'),
         ('plain', 'dot', 'core', 'core.dot.plain')])
    @pytest.mark.parametrize('engine', ['dot'])
    def test_render(capsys, tmp_path, engine, format_, renderer, formatter,
                    expected_suffix, filename='hello.gv',
                    data=b'digraph { hello -> world }'):
        lpath = tmp_path / filename
        assert lpath.write_bytes(data) == len(data) == lpath.stat().st_size
        rendered = lpath.with_suffix(f'{lpath.suffix}.{expected_suffix}')
    
        with pytest.deprecated_call():
            result = graphviz.render(engine, format_, str(lpath),
                                     renderer, formatter)
    
        assert result == str(rendered)
    
        assert rendered.stat().st_size
>       assert capsys.readouterr() == ('', '')
E       AssertionError: assert CaptureResult...irectories\n') == ('', '')
E         Full diff:
E         - ('', '')
E         + CaptureResult(out='', err='Fontconfig error: No writable cache directories\nFontconfig error: No writable cache directories\nFontconfig error: No writable cache directories\nFontconfig error: No writable cache directories\nFontconfig error: No writable cache directories\nFontconfig error: No writable cache directories\nFontconfig error: No writable cache directories\nFontconfig error: No writable cache directories\nFontconfig error: No writable cache directories\nFontconfig error: No writable cache directories\nFontconfig error: No writable cache directories\nFontconfig...
E         
E         ...Full output truncated (1 line hidden), use '-vv' to show

tests/backend/test_rendering.py:63: AssertionError
------------------------------ Captured log call -------------------------------
DEBUG    graphviz.backend.execute:execute.py:61 run [PosixPath('dot'), '-Kdot', '-Tpdf', '-O', 'hello.gv']

Here is the complete list of failed tests:

=========================== short test summary info ============================
FAILED tests/backend/test_rendering.py::test_render_img - AssertionError: ass...
FAILED tests/backend/test_rendering.py::test_render[dot-plain-dot-core-core.dot.plain]
FAILED tests/backend/test_rendering.py::test_render[dot-pdf-None-None-pdf] - ...
FAILED tests/backend/test_rendering.py::test_render_outfile_differnt_parent
FAILED tests/test_graphs.py::test_subgraph_render[Graph-...] - AssertionError...
FAILED tests/test_graphs.py::test_subgraph_render[Digraph-...] - AssertionErr...
FAILED tests/backend/test_piping.py::test_pipe[dot-ps-ps-core-%!PS-] - Assert...
FAILED tests/backend/test_piping.py::test_pipe[dot-svg-None-None-(?s)^<\\?xml .+</svg>\\s*$]
FAILED tests/backend/test_piping.py::test_pipe[sfdp-svg-None-None-(?s)^<\\?xml .+</svg>\\s*$]

I'm running tests as non-root user. Any idea what is wrong here?

Thank you.

only on windows i have this error: 'NoneType ' has not objects encoding line 43 in _compact.py

I fixed it commenting the lines 156 to 159 in backend.py

I was troubleshooting the Graphviz and using your python3-graphviz on Debian 10.4 Buster.

The version that I have is:

$ dot  -V
dot - graphviz version 2.44.2~dev.20200927.0217 (20200927.0217)

The behavior (actual failing result) is:

PyDot version: 1.4.1
Traceback (most recent call last):
  File "/home/steve/work/datascience/WSGraphDemo/bug-gv.py", line 6, in <module>
    print('Graphviz version:', graphviz.version())
  File "/home/steve/work/datascience/WSGraphDemo/venv/lib/python3.7/site-packages/graphviz/backend.py", line 265, in version
    raise RuntimeError('cannot parse %r output: %r' % (cmd, out))
RuntimeError: cannot parse ['dot', '-V'] output: 'dot - graphviz version 2.44.2~dev.20200927.0217 (20200927.0217)\n'

Process finished with exit code 1

The impacted line is in backend.py line 263.

The new code that made it work has been pushed in #110

In Ubuntu 20.04, the default graphviz (i.e. C code, not python) version is 2.43.0. In https://gitlab.com/graphviz/graphviz/-/issues/1605 someone reported a bug, that the scale was wrongly calculated (introduced in 2.42.0 and fixed in 2.42.4, not sure, how the versioning works, but 2.43.0 is between them).

I am not in the position to update the C graphviz, but I can easily update the Python graphviz (i.e. this package). Are you interested in a PR, that has a workaround to fix this?

I observed this bug, when I tried to visualize a Graph in a Jupyter Lab Notebook. While the small examples worked, the larger ones got broken.

Here is the image from https://gitlab.com/graphviz/graphviz/-/issues/1605 it should give you an idea, what happened in my notebook (Sorry, my examples were too complicated to post them here): When the graph gets to large, I see something like the first image, while small examples produce the correct output.

At the moment I use the following wrapper to fix the visualization, because I cannot replace the C lib.

class FixDotViewer:
    def __init__(self, dot):
        # SVG backend of dot broken since 2.42.0 and fixed in 2.42.4.
        # Not sure, how the versioning works, but it was reported for
        # 2.43.0, which is also broken. 
        # Ubuntu 20.04 has 2.43.0, hence, we need a workaround.
        # Bug reported in https://gitlab.com/graphviz/graphviz/-/issues/1605
        
        backend = 'svg fix scale'

        if backend == 'svg':
            def _repr_svg_():
                return dot._repr_svg_()
            self._repr_svg_ = _repr_svg_ 
        elif backend == 'svg fix scale':
            def _repr_svg_():
                svg = dot._repr_svg_()
                r = re.compile('scale\((\d+\.\d*) (\d+\.\d*)\)')

                m = r.search(svg)
                if m:
                    scale_x, scale_y = m.groups()
                    scale_x = float(scale_x)
                    scale_y = float(scale_y)
                    if scale_x > 1 or scale_y > 1:
                        if scale_x > 1:
                            scale_x = 1 / scale_x
                        if scale_y > 1:
                            scale_y = 1 / scale_y
                        svg = re.sub('scale\(\d+\.\d* \d+\.\d*\)', f'scale({scale_x} {scale_y})', svg)
                
                return svg
            self._repr_svg_ = _repr_svg_ 
            
        elif backend == 'pdf->svg':
            def _repr_svg_():
                import tempfile
                from pathlib import Path
                import subprocess
                pdf_data = dot.pipe(format='pdf')
                with tempfile.TemporaryDirectory() as tmpdir:
                    tmpdir = Path(tmpdir)
                    pdf = tmpdir / 'image.pdf'
                    pdf.write_bytes(pdf_data)
                    subprocess.run([
                        'inkscape',
                        '--without-gui',
                        f'--file={pdf}',
                        '--export-plain-svg=main.svg',
                    ], cwd=tmpdir)
                    return Path(tmpdir / 'main.svg').read_text()
            self._repr_svg_ = _repr_svg_ 
        elif backend == 'png':
            def _repr_png_():
                return dot.pipe(format='png')
            self._repr_png_ = _repr_png_ 
        elif backend == 'jpeg':
            def _repr_jpeg_():
                return dot.pipe(format='jpeg')
            self._repr_jpeg_ = _repr_jpeg_ 
        else:
            raise ValueError(dot_backend)

I am not sure, which workaround would be the best solution. Examples:

• Add global option for default Jupyter visualization backend, e.g. SVG, PDF, PNG, JPEG (Not sure, if more works)
  • Raise a warning for some graphviz versions, that the visualization may be broken and report how to fix it.
• Fix the SVG Code with regex for some versions or, when a global flag is set.

Write code as this

from graphviz import Digraph

g = Digraph("CFG")
g.node('0', label='\\" shape=box color=red] # "=12')
g.render()

You'll get that

digraph CFG {
	0 [label="\\" shape=box color=red] # \"=12"]
}

Hi!

Looks like a degradation in https://github.com/xflr6/graphviz/commit/a29eaa0bb52fb7a1ffeb8ad6cb05b07b059a064d My script uses '\l' in lables [it's DOT syntax], and now it is broken, as backslash is escaped itself. [script is here: https://git.qemu.org/?p=qemu.git;a=blob_plain;f=scripts/render_block_graph.py;hb=HEAD;h=656f0388ad6b05840f1a77b1e150284c5ae89736;hb=HEAD ]

Hi！When I run “make_dot ()”,it appears“failed to execute ['dot', '-Tsvg'], make sure the Graphviz executables are on your systems' PATH ” error. I install graphviz offline by the command of pip insatll graphviz-0.10.1-py2.py3-none-any.whl. Thanks!

To make it easier for potential users to demo this package, the repo can be configured so that that a pre-configured Jupyter notebook server can be launched from it and the example notebook(s) run directly using MyBinder.

Changes required:

• add a binder directory;
• to binder/apt.txt add the graphviz requirement;
• to binder/requirements.txt add the . requirement to install the package from the repo root.

I am not sure if GraphViz is only needed for rendering, but pure Python packages are those that do not require any binary dependencies except Python itself.

somehow on my system (openSUSE Tumbleweed, i.e. rolling distro), the test suite of this package (version 0.19) fails with:

> python3 run-tests.py 
pytest.main([])
============================================= test session starts ==============================================
platform linux -- Python 3.8.12, pytest-6.2.5, py-1.10.0, pluggy-1.0.0
rootdir: /home/abuild/rpmbuild/BUILD/graphviz-0.19, configfile: setup.cfg, testpaths: README.rst, docs, graphviz, tests
plugins: mock-3.6.1, cov-3.0.0
collected 321 items                                                                                            
INTERNALERROR> Traceback (most recent call last):
INTERNALERROR>   File "/usr/lib/python3.8/site-packages/_pytest/config/__init__.py", line 1463, in getoption
INTERNALERROR>     val = getattr(self.option, name)
INTERNALERROR> AttributeError: 'Namespace' object has no attribute '--only-exe'
INTERNALERROR> 
INTERNALERROR> The above exception was the direct cause of the following exception:
INTERNALERROR> 
INTERNALERROR> Traceback (most recent call last):
INTERNALERROR>   File "/usr/lib/python3.8/site-packages/_pytest/main.py", line 269, in wrap_session
INTERNALERROR>     session.exitstatus = doit(config, session) or 0
INTERNALERROR>   File "/usr/lib/python3.8/site-packages/_pytest/main.py", line 322, in _main
INTERNALERROR>     config.hook.pytest_collection(session=session)
INTERNALERROR>   File "/usr/lib/python3.8/site-packages/pluggy/_hooks.py", line 265, in __call__
INTERNALERROR>     return self._hookexec(self.name, self.get_hookimpls(), kwargs, firstresult)
INTERNALERROR>   File "/usr/lib/python3.8/site-packages/pluggy/_manager.py", line 80, in _hookexec
INTERNALERROR>     return self._inner_hookexec(hook_name, methods, kwargs, firstresult)
INTERNALERROR>   File "/usr/lib/python3.8/site-packages/pluggy/_callers.py", line 60, in _multicall
INTERNALERROR>     return outcome.get_result()
INTERNALERROR>   File "/usr/lib/python3.8/site-packages/pluggy/_result.py", line 60, in get_result
INTERNALERROR>     raise ex[1].with_traceback(ex[2])
INTERNALERROR>   File "/usr/lib/python3.8/site-packages/pluggy/_callers.py", line 39, in _multicall
INTERNALERROR>     res = hook_impl.function(*args)
INTERNALERROR>   File "/usr/lib/python3.8/site-packages/_pytest/main.py", line 333, in pytest_collection
INTERNALERROR>     session.perform_collect()
INTERNALERROR>   File "/usr/lib/python3.8/site-packages/_pytest/main.py", line 637, in perform_collect
INTERNALERROR>     hook.pytest_collection_modifyitems(
INTERNALERROR>   File "/usr/lib/python3.8/site-packages/pluggy/_hooks.py", line 265, in __call__
INTERNALERROR>     return self._hookexec(self.name, self.get_hookimpls(), kwargs, firstresult)
INTERNALERROR>   File "/usr/lib/python3.8/site-packages/pluggy/_manager.py", line 80, in _hookexec
INTERNALERROR>     return self._inner_hookexec(hook_name, methods, kwargs, firstresult)
INTERNALERROR>   File "/usr/lib/python3.8/site-packages/pluggy/_callers.py", line 60, in _multicall
INTERNALERROR>     return outcome.get_result()
INTERNALERROR>   File "/usr/lib/python3.8/site-packages/pluggy/_result.py", line 60, in get_result
INTERNALERROR>     raise ex[1].with_traceback(ex[2])
INTERNALERROR>   File "/usr/lib/python3.8/site-packages/pluggy/_callers.py", line 39, in _multicall
INTERNALERROR>     res = hook_impl.function(*args)
INTERNALERROR>   File "/home/abuild/rpmbuild/BUILD/graphviz-0.19/tests/conftest.py", line 16, in pytest_collection_modifyitems
INTERNALERROR>     if config.getoption(ONLY_EXE) or config.getoption(SKIP_EXE):
INTERNALERROR>   File "/usr/lib/python3.8/site-packages/_pytest/config/__init__.py", line 1474, in getoption
INTERNALERROR>     raise ValueError(f"no option named {name!r}") from e
INTERNALERROR> ValueError: no option named '--only-exe'

Used pytest 6.2.5

The graphviz library uses type annotations widely, which is very useful. However, currently when using master, attempting to use mypy on a file that imports graphviz gives the following error:

$ echo "import graphviz" > foo.py
$ mypy foo.py
foo.py:1: error: Skipping analyzing "graphviz": module is installed, but missing library stubs or py.typed marker
foo.py:1: note: See https://mypy.readthedocs.io/en/stable/running_mypy.html#missing-imports
Found 1 error in 1 file (checked 1 source file)

To make the graphviz type annotations accessible to dependent projects, I added py.typed files to graphviz, graphviz.backend, and graphviz.parameters as described in the MyPy documentation.

Hi, I need to add a functionality to use the "-y" flag which inverts the axis as documented here: https://graphviz.org/doc/info/command.html#-y

I added the command to the "cmd" clause while debugging and looks like it works.

If I want to create a PR for it, do we want to go along the design of: https://github.com/xflr6/graphviz/issues/131 which means adding another optional parameter?

Thanks in advance.

I'm starting this based on this Prefect issue: https://github.com/PrefectHQ/prefect/issues/5656

Long story short, in graphviz library's quoting.py, there is this regex:

HTML_STRING = re.compile(r'<.*>$', re.DOTALL)

...which incorrectly matches strings like '<lambda> <map>'.

To reproduce:

mkdir graphviz_test
cd graphviz_test
python3.9 -m venv env
. env/bin/activate
pip install -U pip
pip install graphviz
cat >test.py <<EOF
from graphviz import Digraph

if __name__ == '__main__':

    graph = Digraph()
    graph.node('a', '<lambda> <map>')
    graph.view()
EOF
python test.py 

...this results in:

graphviz.backend.execute.CalledProcessError: Command '[PosixPath('dot'), '-Kdot', '-Tpdf', '-O', 'Digraph.gv']' returned non-zero exit status 1. [stderr: b"Error: Digraph.gv: syntax error in line 2 near ']'\n"]

...and the generated Digraph.gv looks like:

digraph {
    a [label=<lambda> <map>]
}

Hi,

I've been using your nice graphviz library as part of a largely asyncio-based side project. Overall it works great - so thanks! For this side project, in the interest of getting things done (i.e., an excuse for poorly-made choices) I had just patched over some methods of graphviz.Digraph to allow for asynchronous subprocess calls. While others may find deferring to a thread sufficient here and using the existing methods, I like the ability to keep things in asyncio-land as much as possible.

So, to get to my question - would the developers here be interested in a contribution to support asyncio in the library itself? If so, would you have a suggested naming scheme or any API specifics? If not, feel free to close this.

PyGraphviz allows for this, but I think leans pretty heavily on the underlying Graphviz native library, which is sometimes a pain to install as a dependency. It would be super neato keen if there was a pure Python version of this functionality.

Architecturally, this would likely be a substantial departure from the current approach. I'm assuming one would want to define and maintain an internal representation of the graph(s), and then translate them to the dot format at write-time (rather than at edit time, which is what appears to be done now). One would have to be careful to ensure that the same set of inputs still resulted in the same dot source each time.

Once such functionality existed, one could write a reader as well, which would support a fairly powerful workflow:

1. Read a dot file from an external source.
2. Edit it (add/remove nodes or edges, tweak styles/attrs, etc.).
3. Create a new dot structure reflecting the edited structure, and optionally render it.