def clean(text, config=get_config(include_project_config=False)):
"""Remove elements from text based on mode and markup.
Use config data to identify markup elements in the text and remove them.
:param config:
:param text:
"""
return remove_elements(text, config["markup"], config["settings"]["mode"])
def annotate(text, elements, own_validator=None, config=get_config(include_project_config=False)):
"""Find and annotate elements in text.
Create an invaltree with elements and units of text, validate the rules
to apply elements and augment the text with this result.
Args:
text (str): The first parameter is a text string.
elements (list): It is a list of element dicts like the following:
{'fox': {'value': '/wiki/fox', 'data-type': 'animal'}}
own_validator (list): A list of functions that will be applied in the
validation of an element, if it will be applied in the text.
config (dict): Load default config from etc/ or get_config the default
config andd update to your own rules.
Returns:
text (str): The annotated text.
Examples:
Basic example with config overwrite:
>>> text = 'The quick brown fox jumps over the lazy dog.'
>>> elements = [{'fox': {
'value': '/wiki/fox',
'data-type': 'animal'}
},
{'dog': {
'value': '/wiki/dog',
'data-type': 'animal'}
}]
>>> cfg = get_config()
>>> cfg['setting']['replaces_at_all'] = 1
>>> print annotate(text, elements, config=cfg)
'The quick brown <a href="/wiki/fox" data-type="animal">fox</a> jumps over the lazy dog .'
"""
settings = config["settings"]
intervaltree, units, existing_values, existing_a_tags = intervals(text, elements, settings)
to_be_applied = retrieve_hits(intervaltree, units, config, own_validator, existing_values, existing_a_tags)
# apply the items finally, but start from end ...its not like horse riding!
text = augment(text, to_be_applied)
return_applied_links = settings.get("return_applied_links")
if return_applied_links:
applied_links = [item[0].data[1][1] for item in to_be_applied]
return text, applied_links
return text
