“不同层，不同抽象”规则的另一个应用是，类的接口通常应与其实现不同：内部使用的表示形式应与接口中出现的抽象形式不同。如果两者具有相似的抽象，则该类可能不是很深。例如，在第 6 章讨论的文本编辑器项目中，大多数团队都以文本行的形式实现了文本模块，每行分别存储。一些团队还使用 getLine 和 putLine 之类的方法围绕行设计了文本类的 API。但是，这使文本类使用起来较浅且笨拙。在较高级别的用户界面代码中，通常在行中间插入文本（例如，当用户键入内容时）或删除跨行的文本范围。通过用于文本类的面向行的 API，调用者被迫拆分和合并行以实现用户界面操作。这段代码很简单，并且在用户界面的实现中被复制和散布。
文本类提供面向字符的接口时，使用起来要容易得多，例如，insert 方法可在文本的任意位置插入任意文本字符串（可能包括换行符），而 delete 方法则删除文本在文本中的两个任意位置之间。在内部，文本仍以行表示。面向字符的接口封装了文本类内部的行拆分和连接的复杂性，这使文本类更深，并简化了使用该类的高级代码。通过这种方法，文本 API 与面向行的存储机制大不相同。差异表示该类提供的有价值的功能。
Another application of the “different layer, different abstraction” rule is that the interface of a class should normally be different from its implementation: the representations used internally should be different from the abstractions that appear in the interface. If the two have similar abstractions, then the class probably isn’t very deep. For example, in the text editor project discussed in Chapter 6, most of the teams implemented the text module in terms of lines of text, with each line stored separately. Some of the teams also designed the APIs for the text class around lines, with methods such as getLine and putLine. However, this made the text class shallow and awkward to use. In the higher-level user interface code, it’s common to insert text in the middle of a line (e.g., when the user is typing) or to delete a range of text that spans lines. With a line-oriented API for the text class, callers were forced to split and join lines to implement the user-interface operations. This code was nontrivial and it was duplicated and scattered across the implementation of the user interface.
The text classes were much easier to use when they provided a character-oriented interface, such as an insert method that inserts an arbitrary string of text (which may include newlines) at an arbitrary position in the text and a delete method that deletes the text between two arbitrary positions in the text. Internally, the text was still represented in terms of lines. A character-oriented interface encapsulates the complexity of line splitting and joining inside the text class, which makes the text class deeper and simplifies higher level code that uses the class. With this approach, the text API is quite different from the line-oriented storage mechanism; the difference represents valuable functionality provided by the class.