|
|
|
|
|
Given the fundamental language of communication with a computer—namely, machine language—there exists an almost intolerable barrier between the person who desires to solve some problem using the computer and the description of the solution in terms of the machine language.
|
|
|
|
The development of symbolic languages and assemblers was followed closely by the development of autocoders, in which the programmer's language was more closely related to mathematical notation than to the machine operations. However, the autocoder required the development of more sophisticated conversion processors, thus leading to a study of the general translatory process.
|
|
|
|
The methods of program conversion and subsequent execution of a user's program can be classified into two basic techniques: compilation and execution, or interpretation.
|
|
|
|
translate To transform statements from one language to another without significantly changing the meaning. In this sense, a programmer must insist that a compiler or the conversion process within an interpreter should be a translator, since the meaning associated with a source program must be carried over to the target language program.
|
|
|
|
compile To prepare a machine language program from a computer program written in another programming language by making use of the overall logic structure of the program, or generating more than one machine ［language］ instruction for each symbolic statement, or both, as well as performing the function of an assembler.
|
|
|
|
interpreter A computer program that translates and executes each source language statement before translating and executing the next one.
|
|
|
|
In practice, there rarely exists a compiler or an interpreter which adheres precisely to these definitions； most languages possess certain features that cannot be compiled, and most interpreters initially preprocess the source program into some intermediate form from which the original program can be reconstructed, and operate upon that code rather than the original form.
|
|
|
|
|
|
如果只基于计算机通信的原始语言——机器语言，则在用计算机解决某一问题的人与用机器语言来对解决该问题的方法进行的描述之间存在一个几乎令人不能容忍的障碍。
|
|
|
|
随着符号语言和汇编语言的研制与开发，紧接着出现了自动编码器，这样，程序员用的语言与数学表达的联系比与机器操作的联系更加紧密。然而，自动编码器需要开发更复杂的转换处理程序，于是导致了对于一般翻译过程的研究。
|
|
|
|
程序翻译及随后的用户程序的执行方法，可归类为两大基本技术，即编译和执行或解释。
|
|
|
|
翻译：对语句进行从一种语言到另一种语言的变换，而不改变（句子的）意义。在这一意义上，程序员必须坚持做到使一个编译程序或一个解释程序中的翻译过程是一个翻译器，因为与源程序有关的意义必须转移到目的语言程序中。
|
|
|
|
编译：利用程序的总体逻辑结构，将用其他语言书写的计算机程序翻译成机器语言程序；或为每一符号化语句产生多于一个的机器语言指令，或同时具有以上两种形式，并执行汇编程序的功能。
|
|
|
|
解释程序：在翻译并执行下一语句前，翻译并执行每一当前源语言程序语句的计算机程序。
|
|
|
|
实际上，很少有完全遵循这些定义的编译程序或解释程序；大多数语言具有某些不能被编译的特征，而大多数解释程序首先对源程序进行预处理，处理成某一中间形式，再由该中间形式重新构造原始程序，并在该重新构造的代码上而不是在原始形式上运行。
|
|
|
