Load yaml data as utf-8 strings into a dictionary

[dukebody@…]

Hello, I'm looking for support, but I don't know if this is the right place to ask.

I'm using PyYAML to load some nested data from an utf-8 encoded file in Python with:

import yaml
stream=file('data.yaml','r')
data=yaml.load(stream)

The data variable becomes a dictionary with unicode values where is needed. What I want is to put utf-8 strings instead of unicode values to use this data with another library: Cheetah. If I try to use the unicode-type dictionary generated by default with PyYAML I get an UnicodeDecodeError, because de Cheetah strings are in iso-8859-15 and Python tries to decode them to Unicode using ASCII charset tables, so it obiously fails.

Is there any way to get an utf-8 coded dictionary?

[xi]

You need to specify an alternative constructor for the tag !!str. You can do it as follows:

import yaml
def custom_str_constructor(loader, node):
    return loader.construct_scalar(node).encode('utf-8')
yaml.add_constructor(u'tag:yaml.org,2002:str', custom_str_constructor)

Example:

>>> import yaml
>>> yaml.load("Кирилл")
u'\u041a\u0438\u0440\u0438\u043b\u043b'
>>> def custom_str_constructor(loader, node):
...     return loader.construct_scalar(node).encode('utf-8')
... 
>>> yaml.add_constructor(u'tag:yaml.org,2002:str', custom_str_constructor)
>>> yaml.load("Кирилл")
'\xd0\x9a\xd0\xb8\xd1\x80\xd0\xb8\xd0\xbb\xd0\xbb'