Translator Types (Cambridge (CIE) A Level Computer Science): Revision Note
Exam code: 9618
What is a translator?
A translator is a program that translates program source code into machine code so that it can executed directly by a processor
Low-level languages such as assembly code are translated using an assembler
High-level languages such as Python are translated using a compiler or interpreter
Assembler, compiler, interpreter
What is an assembler?
An assembler translates mnemonics written in assembly language (low-level) in to machine code
Each lime of assembly language is assembled into a single machine code instruction
Assemblers have been used less and less since high-level languages were introduced
Advantages | Disadvantages |
|---|---|
Speed of execution | Difficult to write due to limited and hard to understand commands |
Optimises the code | Changes mean it must be reassembled |
Original source code will not be seen | Designed solely for one specific processor |
What is a compiler?
A compiler translates high-level languages into machine code all in one go
Compilers translate inputs into machine code directly
Compilers are generally used when a program is finished and has been checked for syntax errors
Compiled code can be distributed (creates an executable) and run without the need for translation software
If compiled code contains any errors, after fixing, it will need re-compiling
Advantages | Disadvantages |
|---|---|
Speed of execution | Can be memory intensive |
Optimises the code | Difficult to debug |
Original source code will not be seen | Changes mean it must be recompiled |
| It is designed solely for one specific processor |
What is an interpreter?
An interpreter translates high-level languages into machine code one line at a time
Each line is executed after translation and if any errors are found, the process stops
Interpreters do not generate machine code directly, appropriate machine code subroutines are called
Interpreters are generally used when a program is being written in the development stage
Interpreted code is more difficult to distribute as translation software is needed for it to run
Advantages | Disadvantages |
|---|---|
Stops when it finds a specific syntax error in the code | Slower execution |
Easier to debug | Every time the program is run it has to be translated |
Require less RAM to process the code | Executed as is, no optimisation |
Compiler vs interpreter
Aspect | Compiler | Interpreter |
|---|---|---|
Software needed by user | End users only need the executable file, not the compiler – so there’s no extra cost or setup | End users need an interpreter or compiler to translate and run the source code – may involve setup or cost |
Access to source code | Users do not receive the source code, so they can’t modify or extend the program | Users receive the full source code, making it easier to modify or extend the program themselves |
Control and monetisation | Developers keep control of the code and can charge for upgrades or customisations | Developers lose control over their code, making it harder to monetise upgrades or prevent copying |
Speed of execution | Compiled programs run faster, as translation is done in advance and optimised machine code is used | Interpreted programs run slower, since each line is translated every time it runs |
Error handling | Compiled programs don’t run until all syntax and semantic errors are fixed | Errors are caught one line at a time, making it easier for the developer to test and fix during development |
Portability | Compiled code can be run on different machines, regardless of where it was compiled | Interpreted programs must be interpreted on the same type of machine – less portable |
Debugging experience | More difficult to test sections – developers need to write special routines to check partial results | Developers can view partial results as they go, making it easier to test, debug, and refine ideas |
Crash risk | Untested compiled programs may crash the system if run with errors | Interpreted programs are safer during testing – errors are caught before the system can crash |
User independence | Users are reliant on the developer for updates or modifications – they can’t edit the code themselves | Users have full access to code and libraries, giving them freedom to customise or fix the program |
Mixed mode translation
What is mixed mode translation?
Mixed mode translation combines features of both a compiler and an interpreter
Instead of fully compiling or interpreting the code, the program is partially compiled into an intermediate form, which is then interpreted at runtime
It works by:
The source code is compiled into intermediate code, not full machine code
This intermediate code (e.g. bytecode) is interpreted by a virtual machine (e.g. the Java Virtual Machine – JVM)
Optionally, some parts may later be just-in-time (JIT) compiled into machine code for better performance
Java uses mixed mode translation:
Java source code is compiled to bytecode (.class file)
Bytecode is interpreted (or JIT compiled) by the JVM on any platform
Feature | Compiler | Interpreter | Mixed mode translation |
|---|---|---|---|
Translation | Entire program at once | One line at a time | Compiles to intermediate code |
Speed of execution | Fastest (fully compiled) | Slowest | Medium (JIT makes it faster over time) |
Portability | Not portable | Not portable | Highly portable (across platforms) |
Errors shown | After full compilation | One at a time | During compile or runtime |
Requires runtime system | No | Yes | Yes (e.g. JVM) |
Example languages | C, C++ | Python, JavaScript | Java, C# (with .NET CLR) |
Worked Example
Jennifer uses an Integrated Development Environment (IDE) to write her computer program.
The IDE allows Jennifer to use both an interpreter and a compiler while creating her computer program.
Describe the ways in which Jennifer can use both a compiler and an interpreter while developing the program.[4]
Answer
Interpreter:
Use an interpreter while writing the program [1 mark]
… to test/debug the partially completed program [1 mark]
… because errors can be corrected and processing continue from where the execution stopped // errors can be corrected in real time // errors are identified one at a time [1 mark]
Compiler:
Use the compiler after the program is complete [1 mark]
… to create an executable file [1 mark]
Use the compiler to repeatedly test the same (completed) section [1 mark]
… without having to re-interpret every time // compiler not needed at runtime [1 mark]
You've read 0 of your 5 free revision notes this week
Unlock more, it's free!
Did this page help you?
