如何从 JINJA2 模板中获取变量及其过滤器名称

Question

我正在开发基于 GAE (python) 和 JINJA 的应用程序。我使用from_string 方法从文本字符串创建了一个JINJA 模板。即

template = JINJA.from_string(text)

结果模板如下所示：

Template(body=[Scope(body=[ScopedEvalContextModifier(options=[Keyword(key='autoescape', value=Name(name='on', ctx='load'))], body=[Output(nodes=[TemplateData(data=u'Dear '), Filter(node=Name(name='customer_name', ctx='load'), name='safe', args=[], kwargs=[], dyn_args=None, dyn_kwargs=None), TemplateData(data=u',\n\n'), Filter(node=Name(name='customer_name_new', ctx='load'), name='extra', args=[], kwargs=[], dyn_args=None, dyn_kwargs=None), TemplateData(data=u'\n                    \nThank you for choosing '), Name(name='company_name', ctx='load'), TemplateData(data=u'.\n\n')]), If(test=Name(name='start_datetime', ctx='load'), body=[Output(nodes=[TemplateData(data=u'Your '), Name(name='order_type', ctx='load'), TemplateData(data=u' is scheduled for:\n'), Filter(node=Name(name='start_datetime_block', ctx='load'), name='safe', args=[], kwargs=[], dyn_args=None, dyn_kwargs=None), TemplateData(data=u'\nYou can check out the estimated time of arrival for your '), Name(name='order_type', ctx='load'), TemplateData(data=u' using the button below\n'), Filter(node=Name(name='live_link_button', ctx='load'), name='safe', args=[], kwargs=[], dyn_args=None, dyn_kwargs=None), TemplateData(data=u'\n')])], else_=[Output(nodes=[TemplateData(data=u'Your '), Name(name='order_type', ctx='load'), TemplateData(data=u' is now placed.\n')])]), If(test=And(left=Name(name='start_datetime', ctx='load'), right=Name(name='confirmation_required', ctx='load')), body=[Output(nodes=[TemplateData(data=u'Please confirm your availability for this appointment:\n'), Filter(node=Name(name='confirmation_buttons', ctx='load'), name='safe', args=[], kwargs=[], dyn_args=None, dyn_kwargs=None), TemplateData(data=u'\n')])], else_=[]), If(test=Name(name='custom_text', ctx='load'), body=[Output(nodes=[Filter(node=Name(name='custom_text', ctx='load'), name='safe', args=[], kwargs=[], dyn_args=None, dyn_kwargs=None), TemplateData(data=u'\n')])], else_=[]), Output(nodes=[TemplateData(data=u'We look forward to seeing you. In case you have any questions please reach us at '), Name(name='company_email', ctx='load'), TemplateData(data=u'. '), Name(name='company_name', ctx='load'), TemplateData(data=u' '), Name(name='company_address', ctx='load'), TemplateData(data=u' '), Name(name='company_phone', ctx='load')])])])])

现在，我想做的是我想从这个模板中获取所有变量，特别是我关心这些变量，它们有一些过滤器，比如上面的模板预期过滤器是safe 和extra。请注意，extra 是我的自定义过滤器。

调用meta.find_undeclared_variables 方法只会给我关键字列表，而不是它们的过滤器。即

parsed_content = JINJA.parse(text)
keywords = meta.find_undeclared_variables(parsed_content)

有什么方法可以同时获取关键字和过滤器名称？

Answer 1

这是一个可能有帮助的简单解决方案。它给出了带有过滤器的变量名称（仅限具有过滤器的变量）：

from jinja2 import Environment, PackageLoader, meta, nodes

def find_filters(ast):
    """Find all the nodes of a given type.  If the type is a tuple,
    the check is performed for any of the tuple items.
    """
    for child in ast.iter_child_nodes():
        if isinstance(child, nodes.Filter):
            yield child
        else:
          for result in find_filters(child):
              yield result


def filtered_variables(ast):
  """Return variables that have filters, along with their filters. might
  return duplicate variable names with different filters
  """
  results = []
  for i, node in enumerate(find_filters(ast)):
      filters = []
      f = node
      filters.append(f.name)
      while isinstance(f.node, nodes.Filter):
        f = f.node
        filters.append(f.name)
      filters.reverse()
      results.append((f.node.name, filters))
  return results


env = Environment(loader=PackageLoader('templates'))

template = '{% extends "layout.html" %}'\
           '{% from "test.html" import a, b as c %}{{ some_variable | a | x}} {{ some_other }}'\
           '{% import "meh.html" as meh %}{{ some_variable | b | c | d}}'\
           '{% include "muh.html" %}'

ast = env.parse(template)
print(filtered_variables(ast))

输出将是：

[('some_variable', ['a', 'x']), ('some_variable', ['b', 'c', 'd'])]

您可以像这样包含没有过滤器的变量：

f_vars = filtered_variables(ast)
filtered = []
for var in f_vars:
  filtered.append(var[0])
f = [(x, []) for x in keywords if x not in filtered]
f_vars.extend(f)
print(f_vars)

输出：

[('some_variable', ['a', 'x']), ('some_variable', ['b', 'c', 'd']), ('some_other', [])]

请注意，结果可能有重复值。这可能更有用，因为每次出现的变量都可能存在不同的过滤器。