Sometimes source code written in one programming language is converted into anotherone. A prominent target for such conversions is JavaScript, as JavaScript enables theexecution of programs in web browsers. Another important target language is C orC++. Creating intermediate C code, which is then compiled by a C compiler to nativeexecutables, has some advantages compared to direct compilation to native executables:C compilers exist for nearly all computer systems including microcontrollers and embeddedsystems, so the use of a language is not restricted to systems for which a nativecompiler backend is provided. And C as intermediate code simplifies the use ofsystem libraries, which typically provide a C-compatible interface. Due to decades ofdevelopment, C compilers generally can do better code optimizations than younglanguages may manage to do. Some people fear that intermediate C code carries theproblems of the C language, like verbosity, confusing and error-prone code, orundefined behavior, to the source languages. But these well-known concerns of C occuronly when humans write C code directly, in the same way as when humans write assemblycode directly. Automatic conversions are well-defined and well tested, which meansthey are free of errors to the same degree as direct machine code generation wouldbe. But indeed there are some small drawbacks when C or C++ is used as a backend of aprogramming language: C does not always allow direct access to all CPU instructions,which may make it difficult to generate optimal code for some special constructs likeexceptions. And C uses wrap-around arithmetic for unsigned integer types, which maynot be what modern languages desire. The current Nim implementation provides aJavaScript and a C and C++ backend. While the JavaScript backend is a designdecision to enable web development, the C and C++ backends are a more pragmaticdecision and may be later replaced or at least supported by direct native codegeneration or use of the popular LLVM backend. [10] When computer languagesare converted from one language to another, then sometimes the termtranspiler is used to differentiate the translation process from a directcompilation to a binary executable. When program code is converted between verysimilar languages with nearly the same level of abstractions, then the termtranspiler may be justified. But Nim is very different from C and has a higherabstraction level, and the Nim compiler performs many advanced optimizations. So itshould be not called a transpiler, even when compiling to JavaScript or to the C++backend.
Microcontrollers: From Assembly Language to C Using the PIC24 Family ebook rar
Download File: https://tinurli.com/2vFXvT
One important note at the end of this section: whenever we have a word with aspecific bit pattern stored in the memory of our computer, then we can not decidefrom the bit pattern directly what type of data it is. It can be a positive or anegative number, but maybe it is not a number at all but a letter or maybe somethingtotally different. As an example, consider this 8-bit word: 10000001. It could be 129if we have stored intentionally positive numbers in that storage location, or couldbe -127 if we intentionally stored a negative value. Or it could be not a number atall. Is that a problem? No, it is not as long as we use a programming language likeNim which uses static typing. Whenever we are using variables, we declare their typefirst, and so the compiler can do bookkeeping about the type of each variable storedin the computer memory. The benefit is, that we can use all the available bits toencode our actual data, and we do not have to reserve a few bits to encode the actualdata type of variables. For languages without static typing, that is not the case. Inlanguages like Python or Ruby, we can use variables without a static type, so we canassign whatever we want to them. That seems to be comfortable at first, but can beconfusing when we write larger programs and the Python or Ruby interpreter has to doall the bookkeeping at run-time, which is slow and wastes memory for the bookkeeping.
We use the reserved word const to tell the compiler that we want to declare aconstant which we have named Pi, and we assign it the numeric decimal value 3.1415.Nim has a small set of reserved words like var, const, proc, while, andothers, to tell the compiler that we want to declare a variable, a constant, aprocedure, or that we want to use a while loop for some repeated code execution.Reserved words in Nim are special symbols that have a special meaning for the compiler, and we shouldavoid using these symbols as names for other entities like variables, constants, or functions, as thatwould confuse the compiler.The= is the assignment operator in Nim, it assigns the value or expression on theright side of it to the symbol on the left. You have to understand that it isdifferent from the equal sign we may use in mathematics to express an equality relation. Some languages like Pascalinitially used the compound operator := for assignments, but that is noteasy to type on the keyboard and looks a bit angry for sensible people. And sourcecode usually contains a lot of assignments, so the use of = makes some sense.For the actual equality test of two entities, which is not used that often, weuse the compound == operator in Nim, as in most other programming languages including C and Python.Wecall = an operator. Operators are symbols that perform some basic operation,like + for the addition of two numbers, or = for the assignment of avalue to a symbol.Most operators are used as infix operators in between two arguments, asin the expression 2 * Pi to denote the multiplication of the constant Pi withthe literal number 2, resulting in the floating-point value 6.29.But operators can also be used as unitary operators, like in-Pi to invert the sign of a numeric value.When we declare named constants, we must always assign a value immediately.That value can never change, but of course, we can use the named constant inexpressions to create different values as in
Note that all the data types that are built into the language, like the primitive types int, float, or char, as wellas the built-in container types like tuple, object, seq, and string, are written in lower case, whiledata types that are defined by the Nim standard library or that we define our self, by convention, startswith a capital letter like the CountTables type defined in the tables module. Some people may regard this as an inconsistency,others may say that in this way we can differentiate built-in types from types defined by libraries. At least we may agreethat using capital notation for common types as in Int, Float, or String would be more difficult to type andwould look not that nice.
While using command-line arguments is convenient for data like filenames oroptions that we already know when we launch a program from the terminal window,often we have to provide textual user input while the program is already running.Functions for this task are provided by the io module, which is part of the system module,and which we have not to import explicitly. In one of the introducing sections of thebook, we used already the readLine() and the getch() procs for reading in a line of text from the terminaland for waiting on a single keypress event. 2ff7e9595c
Comments