Update README.md.

README.md

@@ -53,14 +53,15 @@ xpath.div(_class="*mycl")
 ```
 
 Alternatively, you can use the `contains()` function to filter for subtrings.
-the following statement is equal to the above
+the following statement is equal to the above:
 
 ``` python
-xpath.div.contains(_class="mycl")    # //div[contains(@class, 'mycl')]
+# //div[contains(@class, 'mycl')]
+xpath.div.contains(_class="mycl")
 ```
 
 xpath supports "nested predicates", i.e. you can filter for specific sub-elements,
-while the xpath itself will point to the parent element (div in this example).
+while the xpath itself will point to the parent element (div in this example):
 
 ``` python
 # //div[./span[@class='mycl']]
@@ -75,7 +76,7 @@ xpath.div.has(xpath.span(_class="mycl")).has(xpath.p(text="hello"))
 ```
 
 If the value of an attribute starts with a hashtag (`#`), the xpath matches
-any element that has the following text as a *full word* in this attribute.
+any element that has the following text as a *full word* in this attribute:
 
 ``` python
 # //div[contains(concat(' ', normalize-space(@class), ' '), ' myclass ')]
@@ -89,6 +90,20 @@ Any combination of the features are allowed:
 xpath.span(_class="*mycl", _placeholder='huhu')
 ```
 
+An `_` in the role will be converted to a ".":
+
+``` python
+# //Android.Container[@id="huhu"]
+xp.Android_Container(_id="huhu")
+```
+
+Use double `__` if you need an `_`:
+
+``` python
+# //Android_Container[@id="huhu"]
+xp.Android__Container(_id="huhu")
+```
+
 ## Further reading
 
 - [The xpath cheat sheet](https://devhints.io/xpath)

setup.cfg

@@ -1,6 +1,6 @@
 [metadata]
 name = yaxp
-version = 0.2.4
+version = 0.2.5
 description = intuitive xpath generator
 keywords = xpath, selenium, xml
 long_description = file: README.md