#4

Please update pypi package.

Hello,

Is there any plans for Python3 version of react-python? Thanks.

Hey guys,
I tried to get some support on the IRC channel but couldn't find anyone using PyReact.

My main question is that in my project right now we are passing in a context variable to render our templates.

Is there a recommended way to pass context data to React.js so that it can render components with that data?

Feel free to email me if you would rather not have this one the github page.

Please upgrade to React 0.13.1.

Hi there, firstly thanks fro this useful plugin!

I'm having problems getting it working with django-pipeline however collectstatic renders the following:

Post-processed '/Users/rosscdh/Projects/my_app/toolkit/static/js/matter_search.js' as '/Users/rosscdh/Projects/my_app/toolkit/static/js/matter_search.js'

0 static files copied to '/Users/rosscdh/Projects/my_app/toolkit/static', 13115 unmodified, 1 post-processed.

The paths appear incorrect as it should be referencing matter_search.jsx (as in the settings.py files below.

Tried variations of:

PIPELINE_JS = {
    'reactjs': {
        'source_filenames': (
          '{SITE_ROOT}/toolkit/apps/matter/static/js/matter_search.jsx'.format(SITE_ROOT=SITE_ROOT),
        ),
        'output_filename': '{STATIC_ROOT}/js/matter_search.js'.format(STATIC_ROOT=STATIC_ROOT),
    }
}
PIPELINE_COMPILERS = (
  'react.utils.pipeline.JSXCompiler',
)

And

PIPELINE_JS = {
    'reactjs': {
        'source_filenames': (
          'toolkit/apps/matter/static/js/matter_search.jsx',
        ),
        'output_filename': 'toolkit/apps/matter/static/js/matter_search.js',
    }
}
PIPELINE_COMPILERS = (
  'react.utils.pipeline.JSXCompiler',
)

However, it works perfectly if I create a custom management command and issue manual JSX transform.

I seem to have missed something or am overlooking a tree in the forest.

Thanks in advance

That'll likely be the last in the 0.11.x line, so might be good to do a final update.

React v0.120 is releases. Please update to last version.

BTW, this project does not include react.js and the others. did you add that to the .gitignore file ?

It would be really nice if PyReact could offer a generic WSGI app like the react-middleware project or an API where one could give it a path to a React component and it will return a string so the controller can include it to the response string.

The documented example of:

jsx.transform('path/to/input/file.jsx', 'path/to/output/file.js')

Does not work. I think the transform functions should be changed as below.

I think transform functions should be:

def transform(jsx_path, *args, **opts):
    return JSXTransformer().transform(jsx_path, *args, **opts)

def transform_string(jsx, *args, **opts):
    return JSXTransformer().transform_string(jsx, *args, **opts)

When handling an execjs ProgramError, the code attempts to read the message attribute from the raised exception. The exception doesn't have any such attribute (causing a nested exception).

I'm not sure what the right fix is. I worked around be just doing str(e), instead of e.message which at least unblocks me.

See https://github.com/doloopwhile/PyExecJS/issues/14

Windows 7 64-bit
pyexecjs 1.1.0
pyv8 1.0
NodeJS 4.0.0

Using execjs with pyv8 (it seems that PyV8 takes precedence):

import execjs

execjs.get().name
Out[2]: u'PyV8'

from react import jsx
jsx.transform('helloworld.jsx')
Traceback (most recent call last):

  File "<ipython-input-4-ba80c49e788e>", line 1, in <module>
    jsx.transform('helloworld.jsx')

  File "C:\Python27\python-2.7.6.amd64\lib\site-packages\react\jsx.py", line 64, in transform
    return JSXTransformer().transform(jsx_path, **opts)

  File "C:\Python27\python-2.7.6.amd64\lib\site-packages\react\jsx.py", line 51, in transform
    js = self.transform_string(i.read(), **kwargs)

  File "C:\Python27\python-2.7.6.amd64\lib\site-packages\react\jsx.py", line 45, in transform_string
    raise TransformError(e.message[7:])

TransformError: ()

'JScript' was my original runtime:

import os
os.environ['EXECJS_RUNTIME'] = 'JScript'
jsx.transform('helloworld.jsx')
Traceback (most recent call last):

  File "<ipython-input-7-ba80c49e788e>", line 1, in <module>
    jsx.transform(helloworld.jsx')

  File "C:\Python27\python-2.7.6.amd64\lib\site-packages\react\jsx.py", line 64, in transform
    return JSXTransformer().transform(jsx_path, **opts)

  File "C:\Python27\python-2.7.6.amd64\lib\site-packages\react\jsx.py", line 51, in transform
    js = self.transform_string(i.read(), **kwargs)

  File "C:\Python27\python-2.7.6.amd64\lib\site-packages\react\jsx.py", line 45, in transform_string
    raise TransformError(e.message[7:])

TransformError: or: 'undefined' is null or not an object

But installing NodeJS and using it as the runtime works:

os.environ['EXECJS_RUNTIME'] = 'Node'

In [9]: jsx.transform('helloworld.jsx')
Out[9]: u'var HelloMessage = React.createClass({displayName: \'HelloMessage\',\n  render: function() {\n    return React.createElement("div", null, "Hello ", this.props.name);\n  }\n});\n\nReact.render(React.createElement(HelloMessage, {name: "John"}), mountNode);'

So I guess that should be added to the readme. I have uninstalled PyV8 and now Node is the default runtime under execjs.

This way pip install -e ... would work out of the box.

  File "C:\wc\...\lib\site-packages\react\jsx.py", line 50, in transform
    with open(jsx_path, 'rU') as i:
IOError: [Errno 22] invalid mode ('rU') or filename: u"'C:\\...\\staticfiles\\jsx\\my_jsx_file.jsx'"

Single quotes (') surround the jsx_path and js_path sent in from pipeline, as you can see from the error message thrown above. It comes from their shlex.quote call here: https://github.com/cyberdelia/django-pipeline/blob/master/pipeline/compilers/__init__.py#L43

I currently hacked around it by replacing the transform(...) in C:\wc...\lib\site-packages\react\jsx.py with this:

def transform(self, jsx_path, js_path=None, **kwargs):
    if jsx_path.startswith("'"):
        jsx_path = jsx_path[1:-1]
    with open(jsx_path, 'rU') as i:
        js = self.transform_string(i.read(), **kwargs)
        if js_path:
            if js_path.startswith("'"):
                js_path = js_path[1:-1]
            with open(js_path, 'wb') as o:
                o.write(js.encode('utf8'))
        return js

JSXTransformer() should accept an options object, where you can enable harmony transformations.

In README.md you wrote:

# For a single file, you can use a shortcut method.
jsx.transform('path/to/input/file.jsx', 'path/to/output/file.js')

But jsx.transform takes only one argument:

def transform(jsx_path, **opts):
    return JSXTransformer().transform(jsx_path, **opts)

Regards,
Henri