Behave step matcher

behave 提供3中step匹配模式
'parse'
'cfparse' 基于parse的扩展, 支持cardinality field syntax?
're' 支持在step中定义正则表达式

'parse' 是默认的step mathcer, 他被使用最多, 有以下特点

上手容易, 易读性好, 好理解
支持预定义的数据类型和用户自定义类型
可以在自定义数据类型中使用re, 在step_impl中隐藏了re, 可读性好

'cfparse' 是parse的扩展, 设计初衷是替代parse, 它有以下特点

继承parse, 支持 the cardinality field syntax
自动创建缺少的类型转换函数for fields with cardinality field part
基于parse_type

're'有以下特点

addresses some cases that cannot be solved otherwise (currently)
is backward compatible to cucumber (uses regular expressions)
is less ambiguous compared to the “parse” matcher (currently)

定义step matcher的两种方法

在environment.py中定义默认的matcher

# -- FILE: features/environment.py
from behave import use_step_matcher

# -- SELECT DEFAULT STEP MATCHER: Use "re" matcher as default.
# use_step_matcher("parse")
# use_step_matcher("cfparse")
use_step_matcher("re")

在step definition文件中切换step matcher, 同样使用use_step_matcher("re")
从切换step matcher行后的所有step_impl都使用你切换的step matcher, 除非你再次切换

正则匹配

# 简单的group捕获, 并赋值给P<test>
# -- SIMPLE GROUP: foo
@when(u'I try to match "(?P<foo>foo)"')
def step_when_I_try_to_match_foo(context, foo):
    context.foo = foo

# -- SIMPLE GROUP: anything else
@when(u'I try to match "(?P<anything>.*)"')
def step_when_I_try_to_match_anything_else(context, anything):
    context.anything = anything

# 可选的的group:  (?P<an_>an )?
@when(u'I try to match (?P<an_>an )?optional "(?P<foo>foo)"')
def step_when_I_try_to_match_an_optional_foo(context, an_, foo):
    context.foo = foo
    context.an_ = an_

# 套嵌的正则

@when(u'I try to match nested "(?P<foo>foo(?P<bar>bar)?)"')
def step_when_I_try_to_match_nested_foobar(context, foo, bar):
    context.foo = foo
    context.bar = bar