andprint and input to place the output at a specified positionbellbeep)notrectanglecircles, rectangles or trianglesline-commandfor-, do-, repeat- or while-loopdo-loopif-statementif-statementif-statement2.71828182864circles, rectangles or trianglesfor-loopforeign_buffer_ and foreign_function_goto, gosub or restoreinkey$gosub-targetsgoto-targets3.14159print colordata-statementsdimrepeat-loopdata-pointerclear screen clears the text windowif-statementwhile-loopThis document describes yabasic. You will find information about the yabasic interpreter (the program yabasic under Unix or yabasic.exe under Windows) as well as the language (which is, of course, a sort of basic) itself.
This document applies to version 2.91 of yabasic
However, it does not contain the latest news about yabasic or a FAQ. As such information tends to change rapidly, it is presented online only at www.yabasic.de.
Although basic has its reputation as a language for beginning programmers, this is not an introduction to programming at large. Rather this text assumes, that the reader has some (moderate) experience with writing and starting computer programs.
yabasic is a traditional basic interpreter. It understands most of the typical basic-constructs, like goto, gosub, line numbers, read, data or string-variables with a trailing '$'. But on the other hand, yabasic implements some more advanced programming-constructs like subroutines or libraries (but not objects). yabasic works much the same under Unix and Windows.
yabasic puts emphasis on giving results quickly and easily; therefore simple commands are provided to open a graphic window, print the graphics or control the console screen and get keyboard or mouse information. The example below opens a window, draws a circle and prints the graphic:
open window 100,100 open printer circle 50,50,40 text 10,50,"Press any key to get a printout" clear screen inkey$ close printer close window
This example has fewer lines, than it would have in many other programming languages. In the end however yabasic lacks behind more advanced and modern programming languages like C++ or Java. But as far as it goes it tends to give you results more quickly and easily.
Once, yabasic has been set up correctly, there are three ways to start it:
Right click on your desktop: The desktop menu appears with a submenu named new. From this submenu choose yabasic. This will create a new icon on your desktop. If you right click on this icon, its context menu will appear; choose Execute to execute the program.
As a variant of the way described above, you may simply create a file with the ending .yab (e.g. with your favorite editor). Everything else then works as described above.
From the start-menu: Choose yabasic from your start-menu. A console-window will open and you will be asked to type in your program. Once you are finished, you need to type return twice, and yabasic will parse and execute your program.
This is not the preferred way of starting yabasic ! Simply because the program, that you have typed, can not be saved and will be lost inevitably ! There is no such thing as a save-command and therefore no way to conserve the program, that you have typed. This mode is only intended for quick hacks, and short programs.
Under Windows yabasic will mostly be invoked by double-clicking on an appropriate icon; this way you do not have a chance to specify any of the command line options below. However, advanced users may change the librarypath in the registry, which has the same effect as specifying it as an option on the command line.
See the chapter on options for a complete list of all options, either on Unix or Windows.
Like every other icon under Windows, the icon of every yabasic-program has a context menu offering the most frequent operations, that may be applied to a yabasic-program.
This will invoke yabasic to execute your program. The same happens, if you double click on the icon.
notepad will be invoked, allowing you to edit your program.
This will present the embedded documentation of your program. Embedded documentation is created with the special comment doc.
If your system administrator (vulgo root) has installed yabasic correctly, there are three ways to start it:
You may use your favorite editor (emacs, vi ?) to put your program into a file (e.g. foo). Make sure that the very first line starts with the characters '#!' followed by the full pathname of yabasic (e.g. '#!/usr/local/bin/yabasic'). This she-bang-line ensures, that your Unix will invoke yabasic to execute your program (see also the entry for the hash-character). Moreover, you will need to change the permissions of your yabasic-program foo, e.g. chmod u+x foo. After that you may invoke yabasic to invoke your program by simply typing foo (without even mentioning yabasic). However, if your PATH-variable does not contain a single dot ('.') you will have to type the full pathname of your program: e.g. /home/ihm/foo (or at least ./foo).
Save your program into a file (e.g. foo) and type yabasic foo. This assumes, that the directory, where yabasic resides, is contained within your PATH-variable.
Finally your may simply type yabasic (maybe it will be necessary to include its full pathname). This will make yabasic come up and you will be asked to type in your program. Once you are finished, you need to type return twice, and yabasic will parse and execute your program.
This is not the preferred way of starting yabasic ! Simply because the program, that you have typed, can not be saved and will be lost inevitably ! There is no such thing as a save-command and therefore no way to conserve the program, that you have typed. This mode is only intended for quick hacks, and short programs, i.e. for using yabasic as some sort of fancy desktop calculator.
yabasic accepts a number of options on the command line.
See chapter on options for a complete list of all options, either on Unix or Windows.
If you want to set some options once for all, you may put them into your X-Windows resource file. This is usually the file .Xresources or some such within your home directory (type man X for details).
Here is a sample section, which may appear within this file:
yabasic*foreground: blue yabasic*background: gold yabasic*geometry: +10+10 yabasic*font: 9x15
This will set the foreground color of the graphic-window to blue and the background color to gold. The window will appear at position 10,10 and the text font will be 9x15.
Here are the options, that yabasic accepts on the command line (both under Unix and Windows).
All the options below may be abbreviated (and one hyphen may be dropped), as long as the abbreviation does not become ambiguous. For example, you may write -e instead of --execute.
--help or -?Prints a short help message, which itself describes two further help-options.
--versionPrints the version of yabasic.
--infolevel INFOLEVELChange the infolevel of yabasic, where INFOLEVEL can be one of debug, note, warning, error, fatal and bison (the default is warning). This option changes the amount of debugging-information yabasic produces. However, normally only the author of yabasic would want to change this.
--execute A-PROGRAM-AS-A-SINGLE-STRINGWith this option you may specify some yabasic-code to be executed right away. This is useful for very short programs, which you do not want to save to a file. If this option is given, yabasic will not read any code from a file. E.g.
yabasic -e 'for a=1 to 10:print a*a:next a'
prints the square numbers from 1 to 10.
--bind NAME-OF-STANDALONE-PROGRAMCreate a standalone program (whose name is specified by NAME-OF-STANDALONE-PROGRAM) from the yabasic-program, that is specified on the command line. Please note, that the file NAME-OF-STANDALONE-PROGRAM will be overwritten; therefore you should not give the name of your yabasic-program here. See the section about creating a standalone-program for a complete example.
--geometry +X-POSITION+Y-POSITIONSets the position of the graphic window, that is opened by open window (the size of this window, of course, is specified within the open window-command). An example would be -geometry +20+10, which would place the graphic window 10 pixels below the upper border and 20 pixels right of the left border of the screen. This value cannot be changed, once yabasic has been started.
-fg FOREGROUND-COLOR or --foreground FOREGROUND-COLORUnix only. Define the foreground color for the graphics-window (that will be opened with open window). The usual X11 color names, like red, green, … are accepted. This value cannot be changed, once yabasic has been started.
-bg BACKGROUND-COLOR or --background BACKGROUND-COLORUnix only. Define the background color for the graphics-window. The usual X11 color names are accepted. This value cannot be changed, once yabasic has been started.
--display X11-DISPLAY-SPECIFICATIONUnix only. Specify the display, where the graphics window of yabasic should appear. Normally this value will be already present within the environment variable DISPLAY.
--font NAME-OF-FONTUnder Unix. Name of the font, which will be used for text within the graphics window.
--font NAME-OF-FONTUnder Windows. Name of the font, which will be used for graphic-text; can be any of decorative, dontcare, modern, roman, script, swiss. You may append a fontsize (measured in pixels) to any of those fontnames; for example -font swiss30 chooses a swiss-type font with a size of 30 pixels.
--docu NAME-OF-A-PROGRAMPrint the embedded documentation of the named program. The embedded documentation of a program consists of all the comments within the program, which start with the special keyword doc. This documentation can also be seen by choosing the corresponding entry from the context-menu of any yabasic-program.
--checkCheck for possible compatibility problems within your yabasic-program. E.g. this option reports, if you are using a function, that has recently changed.
--librarypath DIRECTORY-WITH-LIBRARIESChange the directory, wherein libraries will be searched and imported (with the import-command). See also import for more information about the way, libraries are searched.
--Do not try to parse any further options; rather pass the subsequent words from the commandline to yabasic.
In addition to the usual decimal notation (e.g. 1234), yabasic also supports numeric literals with base 2 or 16; examples are 0b10011 (the number 19, written with base 2) or 0x34AF (the number 13487, written with base 16) respectively. Please note that these numbers (apart from the way you write them into your program) are no different from “ordinary” numbers and can be used in any place, where a normal number with base 10 would fit. E.g. you may compute the sine sin(0b110); so the base 2 (or 16) is just a different way of representation.
See also adding code during execution.
This chapter presents some general concepts and terms, which deserve a description on their own, but are not associated with a single command or function in yabasic. Most of these topics do not lend themselves to be read alone, rather they might be read (or skimmed) as background material if an entry from the alphabetical list of commands refers to them.
Logical shortcuts are no special language construct and there is no keyword for them; they are just a way to evaluate logical expressions. Logical expressions (i.e. a series of conditions or comparisons joined by and or or) are only evaluated until the final result of the expression can be determined. An example:
if (a<>0 and b/a>2) print "b is at least twice as big as a"
The logical expression a<>0 and b/a>2 consists of two comparisons, both of which must be true, if the print statement should be executed. Now, if the first comparison (a<>0) is false, the whole logical expression can never be true and the second comparison (b/a>2) need not be evaluated.
This is exactly, how yabasic behaves: The evaluation of a composed logical expressions is terminated immediately, as soon as the final result can be deduced from the already evaluated parts.
In practice, this has the following consequences:
If two or more comparisons are joined with and and one comparison results in false, the logical expression is evaluated no further and the overall result is false.
If two or more comparisons are joined with or and one comparison results in true, the logical expression is evaluated no further and the result is true.
“Nice, but whats this good for ?”, I hear you say. Well, just have another look at the example, especially the second comparison (b/a>2); dividing b by a is potentially hazardous: If a equals zero, the expression will cause an error and your program will terminate. To avoid this, the first part of the comparison (a<>0) checks, if the second one can be evaluated without risk. This pre-checking is the most common usage and primary motivation for logical shortcuts (and the reason why most programming languages implement them).
Well, bottomline there is no difference or distinction between conditions and expressions, at least as yabasic is concerned. So you may assign the result of comparisons to variables or use an arithmetic expression or a simple variable within a condition (e.g. within an if-statement). So the constructs shown in the example below are all totally valid:
input "Please enter a number between 1 and 10: " a rem Assigning the result of a comparison to a variable okay=a>=1 and a<=10 rem Use a variable within an if-statement if (not okay) error "Wrong, wrong !"
So conditions and expressions are really the same thing (at least as long as yabasic is concerned). Therefore the terms conditions and expression can really be used interchangeably, at least in theory. In reality the term condition is used in connection with if or while whereas the term expression tends to be used more often within arithmetic context.
Yabasic, of course, allows to compare strings with strings and numbers with numbers; <, <=, > and >= compare their left-hand side to their right-hand side as usual; nothing new here and examples can be found throughout this manual.
More interesting, the equality-operator (for numbers as well as for strings) can be written in two different ways: either as = (traditional) or as == (more modern). The second form has the advantage of beeing visually distinct from the assignment-operator, which is the single =. One may argue therefore, that using == results in code, that is easier to understand and read; This manual however sticks to tradition and mostly uses the single = for equality-check.
Finally, inequality can be checked with <> or !=; both operators behave identically and so it is only a matter of taste, which one to use.
References on arrays are the only way to refer to an array as a whole and to pass it to subroutines or functions like arraydim or arraysize.
While (for example) a(2) designates the second element of the array a, a() (with empty braces) refers to the array a itself. a() is called an array reference. A nice example is the bultin function split, that accepts an array-reference and modifies the content of this array.
You may also pass to and use array reference within your own subroutines; these subroutines will then be able to modify the array you have passed in often this is intended.
Passing an array reference does not create a copy of the array; this has some interesting consequences:
Speed and space: Creating a copy of an array would be a time and memory consuming operation; passing just a reference is cheap and fast.
Returning many values: A subroutine, that wants to give back more than one value, may require an array reference among its arguments and then store its many return values within this array. This is the only way to return more than one value from a subroutine.
The following program creates two subroutines (print_words and upcase_words), that operate on an array of words (words$() below):
dim words$(4)
for i=1 to 4
read words$(i)
next i
print_words(words$())
upcase_words(words$())
print_words(words$())
sub print_words(w$())
local i
for i=1 to arraysize(w$(),1)
print w$(i)," ";
next i
print
end sub
sub upcase_words(w$())
local i
for i=1 to arraysize(w$(),1)
w$(i) = upper$(w$(i))
next i
end sub
data "case","does","not","matter"
If you run this program, you will get this output:
case does not matter
CASE DOES NOT MATTER
As you probably know, windows uses the character '\' to separate the directories within a pathname; an example would be C:\programs\yabasic\yabasic.exe (a possible location of the yabasic executable). However, the very same character '\' is used to construct escape sequences, not only in yabasic but in most other programming languages.
Therefore e.g. the string "C:\t.dat" does not specify the file t.dat within the directory C:; this is because the sequence '\t' is translated into the tab-character (and thus giving a rather unusal filename containing a tab). To specify this filename, you need to use the string "C:\\t.dat" (note the double slash '\\').
Escape-sequences are the preferred way of specifying 'special' characters. They are introduced by the '\'-character and followed by one of a few regular letters, e.g. '\n' or '\r' (see the table below).
Escape-sequences may occur within any string at any position; they are replaced at parsetime (opposed to runtime), i.e. as soon as yabasic discovers the string, with their corresponding special character. As a consequence of this len("\a") returns 1, because yabasic replaces "\a" with the matching special character just before the program executes.
| Escape Sequence | Matching special character |
|---|---|
\n | newline |
\t | tabulator |
\v | vertical tabulator |
\b | backspace |
\r | carriage return |
\f | formfeed |
\a | alert (i.e. a beeping sound) |
\\ | backslash |
\' | single quote |
\" | double quote |
\xHEX | chr$(HEX) (see below) |
Note, that an escape sequences of the form \xHEX allows one to encode arbitrary
characters as long as you know their position (as a hex-number) within the
ascii-charset:
For example \x012 is transformed into the character chr$(18) (or chr$(dec("12",16)). Note that \x requires a
hexa-decimal number (and the hexadecimal string "12" corresponds to the decimal number 18).
Nobody wants to repeat oneself and therefore yabasic allows to collect arbitrary code into subroutines, so that you may call it from multiple locations within you program. To this end, two conditions must be fulfilled:
The subroutine neeeds to know details about what to do; that's why subroutines have parameters. E.g. in the overly simple subroutine sub add(a,b) (see the example below) the parameters would be a and b, specifying, which numbers to add.
Remark: In certain cases a subroutine may want to find out, how many parameters it has been called with, by querying the special variable numparams.
The subroutine needs to run without messing up the state of the program, at the point where it has been called. That's why many subroutines use local variables, which are different and isolated from all other variables in your program, even if they happen to have the same name. parameters (as described above) are, in addition to their primary function, also local variables.
Remark: If a subroutine wants to remember some information between invocations, it may declare some of its variables as static instead of local function
To see these concepts explained in more detail (complete with examples), follow the links at the end of this section.
Remark: You may notice, that other programming language may use other terms than subroutine for the same concept: function or procedure have been popular for pieces of code, that either return a value or not, and in other languages def is used to name both. And the term method is used in object-oriented languages. However yabasic is not object oriented, and regardless, if a piece of code produces a value or not, it can be encapsulated in a subroutine, so yabasic uses the function sub throughout.
The short program below does nothing more than to add two numbers; for this purpose, it even defines a subroutine. This admittedly is more overhead, than you would normally take.
print "About to add two numbers." input "Please enter first number: " x input "Please enter second number: " y print "Their sum is: ", add(x,y) sub add(a,b) return a+b end sub
If you run it, you would see:
About to add two numbers.
Please enter first number: 2
Please enter second number: 3
Their sum is: 5
Again, see the link at the end of this section for more explanations and examples (e.g. on local or static, which have only been mentioned but not shown at work so far).
All commands for subroutines, where you will find links to the individual keywords related.
Libraries build upon subroutines and take the concept of code-reuse one step further: They allow code to be shared between different programs (as compared to subroutines, which on their own allow code-reuse within a single program only). Moreover, it is possible and in fact common, that the author of a library and the author of a program using that library, are different persons, each writing their respective code on their own.
Here is a program, that asks the user for two numbers and then uses a library adder to add those:
import adder print "About to add two numbers." input "Please enter first number: " x input "Please enter second number: " y print "Their sum is: ", adder.add(x,y)
The statement import adder pulls in code from a very simple library adder.yab:
sub add(a,b) return a+b end sub
If you run it, you might see:
About to add two numbers.
Please enter first number: 3
Please enter second number: 4
Their sum is: 7
Compared with the very similar example for subroutines, there are two differences:
The code of the subroutine add has been moved to its own file adder.yab.
The subroutine add needs to be called as adder.add, which consists of filename (adder.yab but without the ending .yab) and the name of the subroutine (add) within that file.
This is an example of namespaces.
When the executable code is devided between a main program file and (multiple) libraries it is important to keep their subroutines and variables seperate. To this end yabasic internally prefixes the subroutines and variables defined in a library with the shortened name of the library. E.g. in the example above, the subroutine add from the library adder.yab is prefixed by this library-name and ends up as beeing defined as adder.add.
Subroutines and variables defined within the main program are prefixed with main; this prefix is fixed and not related to the actual filename of the main-program. Normally, however, there is no need to use this prefix explicitly; it only helps yabasic to keep everything apart.
Normally, you write programs in yabasic and specify all the necessary logic and calculations within your program. Once you are done, you invoke it, probably multiple times; and while it is running, it does not change.
However, there are some commands within yabasic, that allow to blur the line between writing and execution. Namely eval, eval$, compile, execute and execute$ allow to create and execute new yabasic-code while your program is running. This comes in handy, if the code to be used comes from the user of your program and will only be known after your program has started. A simple example is the yabasic-code to calculate the maximum of a user-supplied expression within a given range; find it as an example for eval. In the same way, one may write a program to plot an arithmetic function, whose definition is entered by the user.
Note: Even if the commands listed above allow to change the yabasic-program, that is currently running, the file where the program is stored, does not change. Therefore, the changes to the running program are not permanent.
The most simple functions are eval and eval$; they compile an expression (with a numeric or string result), e.g. and execute it right away. The compiled code is remembered, so that it need not be compiled again, when the sames expression is executed again; this caters efficiency. However these functions only accept a single expression and nothing else.
If you need more complex computation and logic, the process needs to be split: First create a new subroutine with the compile-command, then execute this subroutine (maybe multiple times) via execute or execute$. This allows to use the broad logic available in subroutines (e.g. conditions, loops, local variables or even other subroutines) and therefore much more complex calculations than with eval. If you want to use compile multiple times within your program (e.g. in a loop), you may want to enumerate the functions you create to avoid name-clashes (as shown in the examples of compile).
To invoke the subroutines created, you need to execute them with execute or execute$, which require the name of the function (a string) as their first argument.
Sometimes you may want to give one of your yabasic-programs to other people. However, what if those other people do not have yabasic installed ? In that case you may create a standalone-program from your yabasic-program, i.e. an executable, that may be executed on its own, standalone, even (and especially !) on computers, that do not have yabasic installed. Having created a standalone program, you may pass it around like any other program (e.g. one written in C) and you can be sure that your program will execute right away.
Such a standalone-program is simply created by copying the full yabasic-interpreter and your yabasic-program (plus all the libraries that it may import) together into a single, new program, whose name might be chosen at will (under windows of course it should have the ending .exe). If you decide to create a standalone-program, there are three facilities in yabasic, that you may use:
The bind-command, which does the actual job of creating the standalone program from the yabasic-interpreter and your program.
The command-line Option --bind (see options), which does the same from the command-line.
The special peek("isbound"), which may be used to check, if the yabasic-program containing this peek is bound to the interpreter as part of a standalone program.
With these bits you know enough to create a standalone-program. Actually there are two ways to do this: on the command line and from within your program.
Let's say you have the following very simple program within the file foo.yab:
print "Hello World !"
Normally you would start this yabasic-program by typing yabasic foo.yab and as a result the string Hello World ! would appear on your screen. However, to create a standalone-program from foo.yab you would type:
yabasic -bind foo.exe foo.yab
This command does not execute your program foo.yab but rather create a standalone-program foo.exe. Note: under Unix you would probably name the standalone program foo or such, omitting the windows-specific ending .exe.
Yabasic will confirm by printing something like: ---Info: Successfully bound 'yabasic' and 'foo.yab' into 'foo.exe'.
After that you will find a program foo.exe (which must be made executable with the chmod-command under Unix first). Now, executing this program foo.exe (or foo under Unix) will produce the output Hello World !.
This newly created program foo.exe might be passed around to anyone, even if he does not have yabasic installed.
It is possible to write a yabasic-program, that binds itself to the yabasic-interpreter. Here is an example:
if (!peek("isbound")) then
bind "foo"
print "Successfully created the standalone executable 'foo' !"
exit
endif
print "Hello World !"
If you run this program (which may be saved in the file foo.yab) via yabasic foo.yab, the peek("isbound") in the first line will check, if the program is already part of a standalone-program. If not (i.e. if the yabasic-interpreter and the yabasic-program are separate files) the bind-command will create a standalone program foo containing both. As a result you would see the output Successfully created the standalone executable 'foo' !. Note: Under Windows you would probably choose the filename foo.exe.
Now, if you run this standalone executable foo (or foo.exe), the very same yabasic-program that is shown above will be executed again. However, this time the peek("isbound") will return TRUE and therefore the condition of the if-statement is false and the three lines after then are not executed. Rather the last print-statement will run, and you will see the output Hello World !.
That way a yabasic-program may turn itself into a standalone-program.
The new standalone program will be at least as big as the interpreter itself, which is typically a few hundred kilobytes.
There is no easy way to extract your yabasic-program from within the standalone program: If you ever want to change it, you should keep it around as a separate file.
If a new version of yabasic becomes available, you might want to recreate your standalone program to take advantage of bugfixes and improvements.
The bind-command, the peek-function and the command line options.
Under Unix, depending on the way yabasic has been built, this feature might have been disabled; the error message in this case will read like this build of yabasic does not support calling foreign libraries. To resolve this issue, you are invited to contact the maintainer.
This is interesting, but somewhat advanced stuff. You will need a good understanding of various concepts of the C-language, especially pointers and structures as well as allocating and freeing blocks of memory. Please be aware, that mistakes or errors during calls to foreign functions or buffers may easily crash yabasic.
Yabasic allows to employ functionality from
an external library; i.e. from a library, which is not written in yabasic, but rather in C; such a library is called a foreign library, as opposed to a library written in yabasic itself. Calling out to a foreign library can be useful, if such a library provides functionality, that can not be replicated in yabasic itself and for which a commandline-interface (which could be used via system) does not exist or is too cumbersome or slow. Examples would be libraries libVLC or libcurl which offer the functionality of vlc or curl to other programs, especially programs written in yabasic.
The foreign function interface of yabasic relies on the great libffi-library, a library making it easy to call other libraries dynamicaaly and the established standard for this task.
Libraries (e.g. libcurl) are meant to provide functionality to other programs (here: yabasic and your yabasic-program). Libraries and programs must be linked together; this can happen either statically at compile-time or dynamically during the excution of the program. For yabasic as the program, static linking happens at the time, yabasic itself is build, whereas dynamic linking happen during the execution and under control of your yabasic-program. So the foreign function interface deals with dynamic (or runtime) linking to external libraries. This linking is done by yabasic behind the scene, when you invoke foreign_function_call; this function, after loading the library, directly calls the specified function therein.
Which functions are available differs from library to library and you should already have this information before you try the library with yabasic.
If you want to use a function from a foreign library, you will need to deal with the fact, that each function requires several parameters and returns exactly one. These parameters have a wide variety of types which need to be mapped to the two types (numbers and strings) known by yabasic. Here are the types available for foreign functions, grouped by the way, they are handled in yabasic:
uint8,int8,uint16,int16,uint32,int32,uint64,int64,float,double,char,short,int,longIn C all these types are used to represent numbers (integer or floating point) with various degrees of precision. When invoking foreign_function_call you need to pass strings, which specify the right type as well as the actual yabasic-value, which will then be converted accordingly. Which types a foreign function expects can be looked up e.g. from its manpage.
stringStrings for foreign functions directly map to strings of yabasic. So if you specify this type for a parameter of a foreign function, the matching value is simply a yabasic string. If you specify string as the return value of a foreign function you should use the variant of calling it, which returns a string, i.e. foreign_function_call$.
bufferYou should specify a buffer as the type of a parameter or the return type, if the foreign function expects a structure or a pointer to a memory area; see structures and buffers for details.
This first example prints the cosine of 2, not by using yabasics own cos-function but by calling out to the standard C-library:
if peek$("os")="windows" then
lib$ = "msvcrt.dll"
else
lib$ = "libm.so.6"
endif
print "cos(2): ",foreign_function_call(lib$,"double","cos","double",2)
The first lines determine the name of the library, which is different under Unix and Windows. The call to foreign_function_call than just states the name of the library, the return type ("double") of the function and then its name ("cos"), as well as type and value (2) of its argument. The final result -0.416147 then is the same as from the the internal cos-function, which is no surprise, because yabasic is already statically linked to the standard C-library and uses its function to compute the cosine.
A second example:
if peek$("os")="windows" then
lib$ = "msvcrt.dll"
else
lib$ = "libm.so.6"
endif
print foreign_function_call$(lib$,"string","strstr","string","foobar","string","ob","options","copy_string_result")
This example calls the strstr-function from the standard C-library; this function accepts two string arguments and returns a string, which is the first (if any) appearance of the second string within the first one (remark: this function makes more sense in C than in yabasic). Please note the option "copy_string_result", which advices yabasic to return a copy of the result of strstr; otherwise your program might crash, because strstr simply returns a pointer to a part of its first argument, a string that will later be freed by yabasic.
This example is windows only; it shows a standard Windows message box with the given title and message:
message_box("Hello World !","Message from yabasic")
sub message_box(message$, title$)
msgptr$ = foreign_buffer_alloc$(len(message$)+1)
foreign_buffer_set msgptr$, 0, message$
titleptr$ = foreign_buffer_alloc$(len(title$)+1)
foreign_buffer_set titleptr$, 0, title$
hwnd$ = foreign_function_call$("user32.dll", "buffer", "GetActiveWindow")
ret = foreign_function_call("user32.dll", "int32", "MessageBoxA", "buffer", hwnd$, "buffer", msgptr$, "buffer", titleptr$, "uint32", 0)
foreign_buffer_free msgptr$
foreign_buffer_free titleptr$
return ret
end sub
The relevant Windows-function MessageBoxA is found within the library user32.dll; most of the example deals with properly allocating, handling and freeing buffers to hold the supplied text-snippets. Thanx to Jean-Marc Duro for this example.
A remark on libffi: this is the library which allows yabasic to call functions from other libraries libffi is used by many other programming-languages for the same purpose; in yabasic it is linked statically (rather than dynamically) so that its functionality is available right from the start. Summing up: libffi itself need not be loaded but helps to call functions from other loaded libraries.
Here is the sequence of events during a foreign function call (e.g. foreign_function_call):
Yabasic parses the type specifications and argument values provided and collects the necessary information for libffi.
The named library is loaded with the appropriate call (which is different under Windows and Unix). This step might easily fail, e.g. if you misspelled the name of the library or your system cannot find the library.
With the help of libffi the named function is invoked.
If you specified the option unload_library, the library that has been loaded is unloaded again.
The return value of the function is converted to a form suitable for yabasic and your program continues.
Errors are reported during every step.
Yabasics functions for dealing with foreign libraries start with foreign_function or foreign_buffer (e.g. foreign_buffer_alloc). To help in typing, these names can all be abbreviated by contracting foreign_function into frnfn and foreign_buffer into frnbf. In the examples below, both forms appear.
The C-language provides a wide variety of simple datatypes (like numbers an strings) and allows to aggregate simple datatypes to structures such a structure contains a set of simple types arranged without overlap (but sometimes with gaps). Yabasic on itself does not know the internals of a structure but rather treats it as a uniform buffer. Structure and buffer are just flipsides of the same memory area viewed either from C or yabasic. For your yabasic-program a buffer is represented by a handle, which is just a simple printable string (containing the size and the memory adress).
The detailed knowledge about the simple types within a structure must be coded into your program, which uses the command (or function) foreign_buffer_set and foreign_buffer_get. Both functions require type and offset (which needs to be looked up in documentation of the foreign library) of the simple type within the structure and a handle to the buffer, which contains the structure.
Beeing essentially a memory area, a buffer is created with foreign_buffer_alloc and destroyed with foreign_buffer_free if needed no more.
Besides representing a structure, a buffer can also provide room to store raw areas of memory for use by the foreign library; example might be image- or sound-content.
The example below deals with the time functions from the standard C-library; some of them deal with the tm structure for keeping the segmented time; to understand the example it is good to have the tm-structure at hand; see below. In addition it might be helpful to consult the manpages of the various C-functions (e.g. localtime)involved.
struct tm {
int tm_sec; /* Seconds (0-60) */
int tm_min; /* Minutes (0-59) */
int tm_hour; /* Hours (0-23) */
int tm_mday; /* Day of the month (1-31) */
int tm_mon; /* Month (0-11) */
int tm_year; /* Year - 1900 */
int tm_wday; /* Day of the week (0-6, Sunday = 0) */
int tm_yday; /* Day in the year (0-365, 1 Jan = 0) */
int tm_isdst; /* Daylight saving time */
};
The example plays with the two forms of keeping the time, either as unix-time (number of seconds since epoch) or as a segmented time (sec, min, etc.). The six steps are each introduced by comments, please see below.
# First: Determine the correct library depending on OS
#
if peek$("os")="windows" then
lib$ = "msvcrt.dll"
else
lib$ = "libm.so.6"
endif
# Second: Get the unix-time
#
# time() has a pointer argument to store the result (in addition to returning it)
# we pass NULL, so only the return value is relevant
#
null$ = foreign_buffer_alloc$(-1)
now = foreign_function_call(lib$,"int","time","buffer",null$)
print "Seconds since the epoch: ",now
# Third: Convert the unix-time to a segmented time
#
# localtime() does not accept the time-value as an argument, but rather requires a pointer
# to the time-value, so we construct a buffer for one int and put in our value
now$ = foreign_buffer_alloc$(foreign_function_size("int"))
foreign_buffer_set now$,0,"int",now
# Dump the buffer for educational purpose
print "Dump of buffer: ",foreign_buffer_dump$(now$)
# localtime() returns a structure with the componentes (year, day, sec, etc.) as elements
local$ = foreign_function_call$(lib$,"buffer","localtime","buffer",now$)
# Fourth: Get the current year from the resulting buffer
#
# assuming, that year is the sixth element of the structure
# so offset is 5
offset = 5 * foreign_function_size("int")
year = foreign_buffer_get(local$,offset,"int")
print "Current year: ", year + 1900
# Fifth: manipulate the segmented time
#
# set year to something else
foreign_buffer_set local$,offset,"int",year-50
# Sixth: convert time-structure from localtime into ascii
#
print "50 years back: ", foreign_function_call$(lib$,"string","asctime","buffer",local$)
On my computer this program produces the following output:
Seconds since the epoch: 1559014899 Dump of buffer: F3ADEC5C Current year: 2019 50 years back: Tue May 28 05:41:39 1969
This final example just invokes libcurl to report its version. This is somewhat involved, because the matching function curl_version_info (see its man-page) returns a structure, which contains a pointer to a string, as can be seen from the structures definition:
typedef struct {
CURLversion age; /* see description below */
const char *version; /* human readable string */
unsigned int version_num; /* numeric representation */
const char *host; /* human readable string */
int features; /* bitmask, see below */
char *ssl_version; /* human readable string */
long ssl_version_num; /* not used, always zero */
const char *libz_version; /* human readable string */
const char * const *protocols; /* protocols */
... /* more lines omitted */
} curl_version_info_data;
Please note, that the yabasic-code below uses abbreviations (e.g. frnfn_call instead of foreign_function_call.
# Get structure with version info
info$ = frnfn_call$("libcurl.so.4","buffer","curl_version_info","int",1)
# dump it for reference
print frnbf_dump$(info$,32)
# assume, that the pointer to version string is at offset 8
sinfo$ = frnbf_get_buffer$(info$,8)
# print readable version
print frnbf_get$(sinfo$,0,10)
The printing of frnbf_dump gives a hint on the internal offsets within the structure and helps to determine that offset of 8 for the next call.
On my system this program produces 7.61.1 for the version of curl.
foreign_function_call, foreign_function_call2, foreign_function_size, foreign_buffer_alloc, foreign_buffer_free, foreign_buffer_size, foreign_buffer_dump, foreign_buffer_set, foreign_buffer_set_buffer, foreign_buffer_get, foreign_buffer_get2, foreign_buffer_get_buffer, system
andprint and input to place the output at a specified positionbellbeep)notrectanglecircles, rectangles or trianglesline-commandfor-, do-, repeat- or while-loopdo-loopif-statementif-statementif-statement2.71828182864circles, rectangles or trianglesfor-loopforeign_buffer_ and foreign_function_goto, gosub or restoreinkey$gosub-targetsgoto-targets3.14159print colordata-statementsdimrepeat-loopdata-pointerclear screen clears the text windowif-statementwhile-loopabs() — returns the absolute value of its numeric argument
y=abs(x)
If the argument of the abs-function is positive (e.g. 2) it is returned unchanged, if the argument is negative (e.g. -1) it is returned as a positive value (e.g. 1).
print abs(-2),abs(2)
This example will print 2 2
acos() — returns the arcus cosine of its numeric argument
x=acos(angle)
acos is the arcus cosine-function, i.e. the inverse of the cos-function. Or, more elaborate: It Returns the angle (in radians, not degrees !), which, fed to the cosine-function will produce the argument passed to the acos-function.
print acos(0.5),acos(cos(pi))
This example will print 1.0472 3.14159 which are π/3 and π respectively.
and — logical and, used in conditions
if a and b … while a and b …
Used in conditions (e.g within if, while or until) to join two expressions. Returns true, if and only if its left and right argument are both true and false otherwise.
Note, that logical shortcuts may take place.
input "Please enter a number" a
if (a>=1 and a<=9) print "your input is between 1 and 9"
and()
— the bitwise arithmetic and
x=and(a,b)
Used to compute the bitwise and of both its argument. Both arguments are treated as binary numbers (i.e. a sequence of digits 0 and 1); a bit of the resulting value will then be 1, if both arguments have a 1 at this position in their binary representation.
Note, that both arguments are silently converted to integer values and that negative numbers have their own binary representation and may lead to unexpected results when passed to and.
print and(6,3)
This will print 2. This result is clear, if you note, that the binary representation of 6 and 3 are 110 and 011 respectively; this will yield 010 in binary representation or 2 as decimal.
arraydim() — returns the dimension of the array, which is passed as an array reference
a=arraydim(b())
If you apply the arraydim()-function on a one-dimensional array (i.e. a vector) it will return 1, on a two-dimensional array (i.e. a matrix) it will return 2, and so on.
This is mostly used within subroutines, which expect an array among their parameters. Such subroutines tend to use the arraydim-function to check, if the array which has been passed, has the right dimension. E.g. a subroutine to multiply two matrices may want to check, if it really is invoked with two 2-dimensional arrays.
dim a(10,10),b(10)
print arraydim(a()),arraydim(b())
This will print 2 1, which are the dimension of the arrays a() and b(). You may check out the function arraysize for a full example.
arraysize() — returns the size of a dimension of an array
x=arraysize(a(),b)
The arraysize-function computes the size of the specified dimension of a given array. Here, size stands for the maximum number, that may be used as an index for this array. The first argument to this function must be an reference to an array, the second one specifies, which of the multiple dimensions of the array should be taken to calculate the size. Please note, that arraysize returns the value that has been used in the actual dim-statement, the real (internal) size of the array is allocated one larger in each dimension to have a first element at index 0; however this is not reflected by the output of arraysize.
An Example involving subroutines: Let's say, an array has been declared as dim a(10,20) (that is a two-dimensional array or a matrix). If this array is passed as an array reference to a subroutine, this sub will not know, what sort of array has been passed. With the arraydim-function the sub will be able to find the dimension of the array, with the arraysize-function it will be able to find out the size of this array in its two dimensions, which will be 10 and 20 respectively.
Our sample array is two dimensional; if you envision it as a matrix this matrix has 10 lines and 20 columns (see the dim-statement above. To state it more formally: The first dimension (lines) has a size of 10, the second dimension (columns) has a size of 20; these numbers are those returned by arraysize(a(),1) and arraysize(a(),2) respectively. Refer to the example below for a typical usage.
rem
rem This program adds two matrices elementwise.
rem
dim a(10,20),b(10,20),c(10,20)
rem initialization of the arrays a() and b()
for y=1 to 10:for x=1 to 20
a(y,x)=int(ran(4)):b(y,x)=int(ran(4))
next x:next y
matadd(a(),b(),c())
print "Result:"
for x=1 to 20
for y=10 to 1 step -1
print c(y,x)," ";
next y
print
next x
sub matadd(m1(),m2(),r())
rem This sub will add the matrices m1() and m2()
rem elementwise and store the result within r()
rem This is not very useful but easy to implement.
rem However, this sub excels in checking its arguments
rem with arraydim() and arraysize()
local x:local y
if (arraydim(m1())<>2 or arraydim(m2())<>2 or arraydim(r())<>2) then
error "Need two dimensional arrays as input"
endif
y=arraysize(m1(),1):x=arraysize(m1(),2)
if (arraysize(m2(),1)<>y or arraysize(m2(),2)<>x) then
error "The two matrices cannot be added elementwise"
endif
if (arraysize(r(),1)<>y or arraysize(r(),2)<>x) then
error "The result cannot be stored in the third argument"
endif
local xx:local yy
for xx=1 to x
for yy=1 to y
r(yy,xx)=m1(yy,xx)+m2(yy,xx)
next yy
next xx
end sub
asc() — accepts a string and returns the position of its first character within the ascii charset
a=asc(char$)
The asc-function accepts a string, takes its first character and looks it up within the ascii-charset; this position will be returned. The asc-function is the opposite of the chr$-function. There are valid uses for asc, however, comparing strings (i.e. to bring them into alphabetical sequence) is not among them; in such many cases you might consider to compare strings directly with <, = and > (rather than converting a string to a number and comparing this number).
input "Please enter a letter between 'a' and 'y': " a$
if (a$<"a" or a$>"y") print a$," is not in the proper range":end
print "The letter after ",a$," is ",chr$(asc(a$)+1)
asin() — returns the arcus sine of its numeric argument
angle=asin(x)
acos is the arcus sine-function, i.e. the inverse of the sin-function. Or, more elaborate: It Returns the angle (in radians, not degrees !), which, fed to the sine-function will produce the argument passed to the asin-function.
print asin(0.5),asin(sin(pi))
This will print 0.523599 -2.06823e-13 which is π/6 and almost 0 respectively.
clear screen … print at(a,b) input at(a,b) "Your input: " a$ print @(a,b)
The at-clause takes two numeric arguments (e.g. at(2,3)) and can be inserted after the print-keyword. at() can be used only if clear screen has been executed at least once within the program (otherwise you will get an error).
The two numeric arguments of the at-function may range from 0 to the width of your terminal minus 1, and from 0 to the height of your terminal minus 1; if any argument exceeds these values, it will be truncated accordingly. However, yabasic has no influence on the size of your terminal (80x25 is a common, but not mandatory), the size of your terminal and the maximum values acceptable within the at-clause may vary. To get the size of your terminal you may use the peek-function: peek("screenwidth") returns the width of your terminal and peek("screenheight") its height.
clear screen
maxx=peek("screenwidth")-1:maxy=peek("screenheight")-1
for x=0 to maxx
print at(x,maxy*(0.5+sin(2*pi*x/maxx)/2)) "*"
next x
This example plots a full period of the sine-function across the screen.
atan() — returns the arctangent of its numeric argument
angle=atan(a,b) angle=atan(a)
atan is the arctangent-function, i.e. the inverse of the tan-function. Or, more elaborate: It Returns the angle (in radians, not degrees !), which, fed to the tan-function will produce the argument passed to the atan-function.
The atan-function has a second form, which accepts two arguments: atan(a,b) which is (mostly) equivalent to atan(a/b) except for the fact, that the two-argument-form returns an angle in the range -π to π, whereas the one-argument-form returns an angle in the range -π/2 to π/2. To understand this you have to be good at math.
print atan(1),atan(tan(pi)),atan(-0,-1),atan(-0,1)
This will print 0.463648 2.06823e-13 -3.14159 3.14159 which is π/4, almost 0, -π and π respectively.
color — change color for background of graphic window
backcolour red,green,blue backcolour "red,green,blue"
Change the color, that becomes visible, if any portion of the window is erased, e.g. after clear window or clear line. Note however, that parts of the window, that show the old background color will not change.
As with the color-command, the new background color can either be specified as a triple of three numbers or as a single string, that contains those three numbers separated by commas.
Note, that the command backcolor can be written as backcolour too and vice versa.
open window 255,255
for x=10 to 235 step 10:for y=10 to 235 step 10
backcolour x,y,0
clear window
sleep 1
next y:next x
This changes the background colour of the graphic window repeatedly and clears it every time, so that it is filled with the new background colour.
beep
— ring the bell within your computer; a synonym for bell
beep
The bell-command rings the bell within your computer once. This command is not a sound-interface, so you can neither vary the length or the height of the sound (technically, it just prints \a). bell is exactly the same as beep.
beep:print "This is a problem ..."
bell
— ring the bell within your computer (just as beep)
bell
The beep-command rings the bell within your computer once. beep is a synonym for bell.
print "This is a problem ...":beep
bin$() — converts a number into a sequence of binary digits
hexadecimal$=bin$(decimal)
The bin$-function takes a single numeric argument an converts it into a string of binary digits (i.e. zeroes and ones). If you pass a negative number to bin$, the resulting string will be preceded by a '-'.
If you want to convert the other way around (i.e. from binary to decimal) you may use the dec-function.
for a=1 to 100
print bin$(a)
next a
This example prints the binary representation of all digits between 1 and 100.
bind() — binds a yabasic-program and the yabasic-interpreter together into a standalone program
bind("foo.exe")
The bind-command combines your own yabasic-program (plus all the libraries it does import) and the interpreter by copying them into a new file, whose name is passed as an argument. This new program may then be executed on any computer, even if it does not have yabasic installed.
Please see the section about creating a standalone-program for details.
if (!peek("isbound")) then
bind "foo"
print "Successfully created the standalone executable 'foo' !"
exit
endif
print "Hello World !"
This example creates a standalone program foo from itself.
The section about creating a standalone-program, the peek-function and the command line options.
bitnot()
— the bitwise arithmetic not
x=bitnot(a)
This function is used to compute the bitwise not of its single argument. The argument is treated as binary number (i.e. a sequence of digits 0 and 1); a bit of the resulting value will be 1, if the argument has a 0 at this position in its binary representation; if the bit in the argument is 1, the bit in the result will be 0.
Note, that its argument is silently converted to a positive integer value and that negative numbers have their own binary representation and may lead to unexpected results when passed to bitnot.
A note on naming: This one-argument-function is named bitnot to distinguish it from the one-argument-function not, which operates on logical expressions. For the similar functions and and or this distinction between logical and bitwise function is done implicitly through the number of arguments (1 and 2, respectively).
print bin$(not(17))
This will print 11111111111111111111111111101110. This result is clear, if you note, that the binary representation of 17 is 10001, which inverted will give the long binary number given before.
break — breaks out of one or more loops or switch statements
break
break 2
break transfers control immediately outside the enclosing loop or switch statement. This is the preferred way of leaving a such a statement (rather than goto, which is still possible in most cases). An optional digit allows one to break out of multiple levels, e.g. to leave a loop from within a switch statement. Please note, that only a literal (e.g. 2) is allowed at this location.
for a=1 to 10
break
print "Hi"
next a
while 1
break
print "Hi"
wend
repeat
break
print "Hi"
until 0
switch 1
case 1:break
case 2:case 3:print "Hi"
end switch
This example prints nothing at all, because each of the loops (and the switch-statement) does an immediate break (before it could print any "Hi").
case
— mark the different cases within a switch-statement
switch a case 1 case 2 … end switch … switch a$ case "a" case "b" … end switch
Please see the switch-statement.
input a
switch(a)
case 1:print "one":break
case 2:print "two":break
default:print "more"
end switch
Depending on your input (a number is expected) this code will print one or two or otherwise more.
ceil() — compute the ceiling for its (float) argument
print ceil(x)
The ceil-function returns the smallest integer number, that is larger or equal than its argument.
print ceil(1.5),floor(1.5)
print ceil(2),floor(2)
Comparing functions ceil and floor, gives a first line of output (1 2), showing that ceil is less or equal than floor; but as the second line of output (2 2) shows, the two functions give equal results for integer arguments.
chomp$() — remove a single trailing newline from its string-argument; if the string does not end in a newline, the string is returned unchanged
print chomp$("Hallo !\n")
The chomp$-function checks, if its string-argument ends in a newline and removes it eventually; for this purpose chomp$ can replace an if-statement. This can be especially useful, when you deal with input from external sources like system$.
You may apply chomp$ freely, as it only acts, if there is a newline to remove; note however, that user-input, that comes from the normal input-statement, does not need such a treatment, because it already comes without a newline.
The following yabasic-program uses the unix-command whoami to get the username of the current user in order to greet him personally. This is done twice: First with the chomp$-function and then again with with an equivalent if-statement:
print "Hello " + chomp$(system$("whoami")) + " !"
user$ = system$("whoami")
if (right$(user$,1)="\n") user$=left$(user$,len(user$)-1)
print "Hello again " + user$ + " !"
chr$() — accepts a number and returns the character at this position within the ascii charset
character$=chr$(ascii)
The chr$-function is the opposite of the asc-function. It looks up and returns the character at the given position within the ascii-charset. It's typical use is to construct nonprintable characters which do not occur on your keyboard.
Nevertheless you won't use chr$ as often as you might think, because the most important nonprintable characters can be constructed using escape-sequences using the \-character (e.g. you might use \n instead of chr$(10) wherever you want to use the newline-character).
print "a",chr$(10),"b"
This will print the letters 'a' and 'b' in different lines because of the intervening newline-character, which is returned by chr$(10).
circle — draws a circle in the graphic-window
circle x,y,r clear circle x,y,r fill circle x,y,r clear fill circle x,y,r
The circle-command accepts three parameters: The x- and y-coordinates of the center and the radius of the circle.
Some more observations related with the circle-command:
The graphic-window must have been opened already.
The circle may well extend over the boundaries of the window.
If you have issued open printer before, the circle
will finally appear in the printed hard copy of the window.
fill circle will draw a filled (with black ink) circle.
clear circle will erase (or clear) the outline of the circle.
clear fill circle or fill clear circle will erase the full area of the circle.
open window 200,200
for n=1 to 2000
x=ran(200)
y=ran(200)
fill circle x,y,10
clear fill circle x,y,8
next n
This code will open a window and draw 2000 overlapping circles within. Each circle is drawn in two steps: First it is filled with black ink (fill circle x,y,10), then most of this circle is erased again (clear fill circle x,y,8). As a result each circle is drawn with an opaque white interior and a 2-pixel outline (2-pixel, because the radii differ by two).
clear
— erase circles, rectangles or triangles
clear rectangle 10,10,90,90 clear fill circle 50,50,20 clear triangle 10,10,20,20,50,30
May be used within the circle, rectangle or triangle command and causes these shapes to be erased (i.e. be drawn in the colour of the background).
fill can be used in conjunction with and wherever the fill-clause may appear. Used alone, clear will erase the outline (not the interior) of the shape (circle, rectangle or triangle); together with fill the whole shape (including its interior) is erased.
open window 200,200
fill circle 100,100,50
clear fill rectangle 10,10,90,90
This opens a window and draws a pacman-like figure.
clear screen — erases the text window
clear screen
clear screen erases the text window (the window where the output of print appears).
It must be issued at least once, before some advanced screen-commands (e.g. print at or inkey$) may be called; this requirement is due to some limitations of the curses-library, which is used by yabasic under Unix for some commands.
clear screen
print "Please press a key : ";
a$=inkey$
print a$
The clear screen command is essential here; if it would be omitted, yabasic would issue an error ("need to call 'clear screen' first") while trying to execute the inkey$-function.
clear window — clear the graphic window and begin a new page, if printing is under way
clear window
clear window clears the graphic window. If you have started printing the graphic via open printer, the clear window-command starts a new page as well.
open window 200,200
open printer "t.ps"
for a=1 to 10
if (a>1) clear window
text 100,100,"Hallo "+str$(a)
next a
close printer
close window
This example prints 10 pages, with the text "Hello 1", "Hello 2", … and so on. The clear screen-command clears the graphics window and starts a new page.
close — close a file, which has been opened before
close filenum close # filenum
The close-command closes an open file. You should issue this command as soon as you are done with reading from or writing to a file.
open "my.data" for reading as 1
input #1 a
print a
close 1
This program opens the file "my.data", reads a number from it, prints this number and closes the file again.
close curve
— close a curve, that has been drawn by the line-command
new curve line to x1,y1 … close curve
The close curve-command closes a sequence of lines, that has been drawn by repeated line to-commands.
open window 200,200
new curve
line to 100,50
line to 150,150
line to 50,150
close curve
This example draws a triangle: The three line to-commands draw two lines; the final line is however not drawn explicitly, but drawn by the close curve-command.
close printer — stops printing of graphics
close printer
The close printer-command ends the printing graphics. Between open printer and close printer everything you draw (e.g. circles, lines …) is sent to your printer. close printer puts an end to printing and will make your printer eject the page.
open window 200,200
open printer
circle 100,100,50
close printer
close window
As soon as close printer is executed, your printer will eject a page with a circle on it.
close window — close the graphics-window
close window
The close window-command closes the graphics-window, i.e. it makes it disappear from your screen. It includes an implicit close printer, if a printer has been opened previously.
open window 200,200
circle 100,100,50
close window
This example will open a window, draw a circle and close the window again; all this without any pause or delay, so the window will be closed before you may regard the circle..
color — change color for any subsequent drawing-command
colour red,green,blue colour "red,green,blue"
Change the color, in which lines, dots, circles, rectangles or triangles are drawn. The color-command accepts three numbers in the range 0 … 255 (as in the first line of the synopsis above). Those numbers specify the intensity for the primary colors red, green and blue respectively. As an example 255,0,0 is red and 255,255,0 is yellow.
Alternatively you may specify the color with a single string (as in the second line of the synopsis above); this string should contain three numbers, separated by commas. As an example "255,0,255" would be violet. Using this variant of the colour-command, you may use symbolic names for colours:
open window 100,100 yellow$="255,255,0" color yellow$ text 50,50,"Hallo"
, which reads much clearer.
open window 255,255
for x=10 to 235 step 10:for y=10 to 235 step 10
colour x,y,0
fill rectangle x,y,x+10,y+10
next y:next x
This fills the window with colored rectangles. However, none of the used colours contains any shade of blue, because the color-command has always 0 as a third argument.
Note, that the command color can be written as colour too and vice versa.
compile — compile a string with yabasic-code on the fly
compile(code$)
This is an advanced command (closely related with the execute-command). It allows you to compile a string of yabasic-code (which is the only argument). Afterwards the compiled code is a normal part of your program.
Note, that there is no way to remove the compiled code.
compile("sub mysub(a):print a:end sub")
mysub(2)
This example creates a function named mysub, which simply prints its single argument.
This next example combines the functions compile and execute:
count = 1
subname$ = "foo" + str$(count)
compile("sub "+ subname$ + "(a):print a:end sub")
execute(subname$,2)
This example creates and executes a function, whose name (foo1) is stored within the variable subname$; the newly created function simply prints its single argument. This example could be executed multiple times within a single yabasic-program, simply by incrementing the variable count; by doing that, multiple subroutines (foo1, foo2, …) could be created and executed in succession.
adding code during execution, execute, execute$, eval, eval$
continue
— start the next iteration of a for-, do-, repeat- or while-loop
continue
You may use continue within any loop to start the next iteration immediately. Depending on the type of the loop, the loop-condition will or will not be checked. Especially: for- and while-loops will evaluate their respective conditions, do- and repeat-loops will not.
Remark: Another way to change the flow of execution within a loop, is the break-command.
for a=1 to 100
if mod(a,2)=0 continue
print a
next a
This example will print all odd numbers between 1 and 100.
data — introduces a list of data-items
data 9,"world" … read b,a$
The data-keyword introduces a list of comma-separated list of strings or numbers, which may be retrieved with the read-command.
The data-command itself does nothing; it just stores data. A single data-command may precede an arbitrarily long list of values, in which strings or numbers may be mixed at will.
yabasic internally uses a data-pointer to keep track of the current location within the data-list; this pointer may be reset with the restore-command.
do
restore
for a=1 to 4
read num$,num
print num$,"=",num
next a
loop
data "eleven",11,"twelve",12,"thirteen",13,"fourteen",14
This example just prints a series of lines eleven=11 up to fourteen=14 and so on without end.
The restore-command ensures that the list of data-items is read from the start with every iteration.
date$ — returns a string with various components of the current date
a$=date$
The date$-function (which must be called without parentheses; i.e. date$() would be an error) returns a string containing various components of a date; an example would be 4-05-27-2004-Thu-May. This string consists of various fields separated by hyphens ("-"):
The day within the week as a number in the range 0 (=Sunday) to 6 (=Saturday) (in the example above: 4, i.e. Thursday).
The month as a number in the range 1 (=January) to 12 (=December) (in the example: 5 which stands for May).
The day within the month as a number in the range 1 to 31 (in the example: 27).
The full, 4-digit year (in the example: 2004, which reminds me that I should adjust the clock within my computer …).
The abbreviated name of the day within the week (Mon to Sun).
The abbreviated name of the month (Jan to Dec).
Therefore the whole example above (4-05-27-2004-Thu-May) would read: day 4 in the week (counting from 0), May 27 in the year 2004, which is a Thursday in May.
Note, that all fields within the string returned by date$ have a fixed with (numbers are padded with zeroes); therefore it is easy to extract the various fields of a date format with mid$.
rem Two ways to print the same ...
print mid$(date$,3,10)
dim fields$(6)
a=split(date$,fields$(),"-")
print fields$(2),"-",fields$(3),"-",fields$(4)
This example shows two different techniques to extract components from the value returned by date$. The mid$-function is the preferred way, but you could just as well split the return-value of date$ at every "-" and store the result within an array of strings.
dec() — convert a base 2 or base 16 number into decimal form
a=dec(number$) a=dec(number$,base)
The dec-function takes the string-representation of a base-2 or base-16 (which is the default) number and converts it into a decimal number. The optional second argument (base) might be used to specify a base other than 16. However, currently only base 2 or base 16 are supported. Please note, that for base 16 and 2 you may write literals in the usual way, by preceding them with 0x or 0b respectively, e.g. like
print 0xff + 0b11
; this may save you from applying the dec altogether.
input "Please enter a binary number: " a$
print a$," is ",dec(a$)
default
— mark the default-branch within a switch-statement
switch a+3 case 1 … case 2 … default … end switch
The default-clause is an optional part of the switch-statement (see there for more information). It introduces a series of statements, that should be executed, if none of the cases matches, that have been specified before (each with its own case-clause).
So default specifies a default to be executed, if none of the explicitly named cases matches; hence its name.
print "Please enter a number between 0 and 6,"
print "specifying a day in the week."
input d
switch d
case 0:print "Monday":break
case 1:print "Tuesday":break
case 2:print "Wednesday":break
case 3:print "Thursday":break
case 4:print "Friday":break
case 5:print "Saturday":break
case 6:print "Sunday":break
default:print "Hey you entered something invalid !"
end switch
This program translates a number between 0 and 6 into the name of a weekday; the default-case is used to detect (and complain about) invalid input.
dim — create an array prior to its first use
dim array(x,y) dim array$(x,y)
The dim-command prepares one or more arrays (of either strings or numbers) for later use. This command can also be used to enlarges an existing array.
When an array is created with the dim-statement, memory is allocated and all elements are initialized with either 0 (for numerical arrays) or "" (for string arrays). Please be aware, that the dim reserves room for one element more than actually specified, e.g. dim(10) reserves memory for 11 elements. This makes it possible to access element 0 as well as element 10, which serves the conventions of C as well as basic.
If the array already existed, and the dim-statement specifies a larger size than the current size, the array is enlarged and any old content is preserved. But note, that dim cannot be used to shrink an array: If you specify a size, that is smaller than the current size, the dim-command does nothing.
Finally: To create an array, that is only known within a single subroutine, you should use the command local, which creates local variables as well as local arrays.
dim a(5,5)
for x=1 to 5:for y=1 to 5
a(x,y)=int(ran(100))
next y:next x
printmatrix(a())
dim a(7,7)
printmatrix(a())
sub printmatrix(ar())
local x,y,p,q
x=arraysize(ar(),1)
y=arraysize(ar(),2)
for q=1 to y
for p=1 to y
print ar(p,q),"\t";
next p
print
next q
end sub
This example creates a 2-dimensional array (i.e. a matrix) with the dim-statement and fills it with random numbers. The second dim-statement enlarges the array, all new elements are filled with 0.
The subroutine printmatrix just does, what its name says.
do
— start a (conditionless) do-loop
do … loop
Starts a loop, which is terminated by loop; everything between do and loop will be repeated forever. This loop has no condition, so it is an infinite loop; note however, that a break- or goto-statement might be used to leave this loop anytime.
do
a=a+1
print a
if (a>100) break
loop
This example prints the numbers between 1 and 101. The break-statement is used to leave the loop.
doc — special comment, which might be retrieved by the program itself
doc This is a comment docu This is another comment
Introduces a comment, which spans up to the end of the line. But other than the rem-comment, any docu-comment is collected within the special docu$-array and might be retrieved later on. Moreover you might invoke yabasic -docu foo.yab on the command line to retrieve the embedded documentation within the program foo.yab.
Instead of doc you may just as well write docu or even documentation.
rem Hi, this has been written by me
rem
doc This program asks for a number and
doc prints this number multiplied with 2
rem
rem Print out the above message
rem
for a=1 to arraysize(docu$(),1):print docu$(a):next a
rem Read and print the number
input "Please input a number: " x
print x*2
This program uses the comments within its code to print out a help message for the user; if you run this program, you get this output:
This program asks for a number and prints this number multiplied with 2 Please input a number: 2 4
The contents of the doc-lines are retrieved from the docu$-array; if you do not want a comment to be collected within this array, use the rem-statement instead.
docu$ — special array, containing the contents of all docu-statement within the program
a$=docu$(1)
Before your program is executed, yabasic collects the content of all the doc-statements within your program within this 1-dimensional array (well only those within the main-program, libraries are skipped).
You may use the arraysize function to find out, how many lines it contains.
docu
docu This program reads two numbers
docu and adds them.
docu
rem retrieve and print the embedded documentation
for a=1 to arraysize(docu$(),1)
print docu$(a)
next a
input "First number: " b
input "Second number: " c
print "The sum of ",b," and ",c," is ",b+c
This program uses the embedded documentation to issue a usage-message.
dot — draw a dot in the graphic-window
dot x,y clear dot x,y
Draws a dot at the specified coordinates within your graphic-window. If printing is in effect, the dot appears on your printout too.
Use the functions peek("winheight") or peek("winwidth") to get the size of your window and hence the boundaries of the coordinates specified for the dot-command.
open window 200,200
circle 100,100,100
do
x=ran(200):y=ran(200)
dot x,y
total=total+1
if (sqrt((x-100)^2+(y-100)^2)<100) in=in+1
print 4*in/total
loop
This program uses a well known algorithm to compute π.
else
— mark an alternative within an if-statement
if (…) then … else … endif
The else-statement introduces the alternate branch of an if-statement. I.e. it starts the sequence of statements, which is executed, if the condition of the if-statement is not true.
input "Please enter a number: " a
if (mod(a,2)=1) then
print a," is odd."
else
print a," is even."
endif
This program detects, if the number you have entered is even or odd.
elsif
— starts an alternate condition within an if-statement
if (…) then … elseif (…) … elsif (…) then … else … endif
The elsif-statement is used to select a single alternative among a series of choices.
With each elsif-statement you may specify a condition, which is tested, if the main condition (specified with the if-statement) has failed. Note that elsif might be just as well written as elseif.
Within the example below, two variables a and b are tested against a range of values. The variable a is tested with the elsif-statement. The very same tests are performed for the variable b too; but here an involved series of if-else-statements is employed, making the tests much more obscure.
input "Please enter a number: " a
if (a<0) then
print "less than 0"
elseif (a<=10) then
print "between 0 and 10"
elsif (a<=20)
print "between 11 and 20"
else
print "over 20"
endif
input "Please enter another number: " b
if (b<0) then
print "less than 0"
else
if (b<=10) then
print "between 0 and 10"
else
if (b<=20) then
print "between 11 and 20"
else
print "over 20"
endif
endif
endif
Note, that the very same tests are performed for the variables a and b, but can be stated much more clearly with the elsif-statement.
Note, that elsif might be written as elseif too, and that the keyword then is optional.
end — terminate your program
end
Terminate your program. Much (but not exactly) like the exit command.
Note, that end may not end your program immediately; if you have opened a window or called clear screen, yabasic assumes, that your user wants to study the output of your program after it has ended; therefore it issues the line ---Program done, press RETURN--- and waits for a key to be pressed. If you do not like this behaviour, consider using exit.
print "Do you want to continue ?"
input "Please answer y(es) or n(o): " a$
if (lower$(left$(a$,1))="n") then
print "bye"
end
fi
endif
— ends an if-statement
if (…) then … endif
The endif-statement closes (or ends) an if-statement.
Note, that endif may be written in a variety of other ways: end if, end-if or even fi.
The endif-statement must be omitted, if the if-statement does not contain the keyword then (see the example below). Such an if-statement without endif extends only over a single line.
input "A number please: " a
if (a<10) then
print "Your number is less than 10."
endif
REM and now without endif
input "A number please: " a
if (a<10) print "Your number is less than 10."
end sub — ends a subroutine definition
sub foo(…) … end sub
Marks the end of a subroutine-definition (which starts with the sub-keyword).
The whole concept of subroutines is explained within the entry for sub.
print foo(3)
sub foo(a)
return a*2
end sub
This program prints out 6. The subroutine foo simply returns twice its argument.
eof — check, if an open file contains data
open 1,"foo.bar" if (eof(1)) then … end if
The eof-function checks, if there is still data left within an open file. As an argument it expects the file-number as returned by (or used within) the open-function (or statement).
As a special case, if the argument is zero: test if input from stdin is available.
a=open("foo.bar")
while not eof(a)
input #a,a$
print a$
end while
This example will print the contents of the file "foo.bar". The eof-function will terminate the loop, if there is no more data left within the file.
eor() — compute the bitwise exclusive or of its two arguments
print eor(a,b)
The eor-function takes two arguments and computes their bitwise exclusive or. I.e. treat each arguments as a sequence of bits and compare theses two sequences bit by bit to produce the result. If the bits from the arguments are equal, the resulting bit will be 0, otherwise 1.
The xor-function is the same as the eor function; both are synonymous; however they have each their own description, so you may check out the entry of xor for a slightly different view.
for a=0 to 3
for b=0 to 3
print fill$(bin$(a))," eor ",fill$(bin$(b))," = ",fill$(bin$(eor(a,b)))
next b
next a
sub fill$(a$)
return right$("0"+a$,2)
end sub
This example prints a table, from which you may figure, how the eor-function is computed.
error — raise an error and terminate your program
error "Wrong, wrong, wrong !!"
Produces the same kind or error messages, that yabasic itself produces (e.g. in case of a syntax-error). The single argument is issued along with the current line-number.
input "Please enter a number between 1 and 10: " a
if (a<1 or a>10) error "Oh no ..."
This program is very harsh in checking the users input; instead of just asking again, the program terminates with an error, if the user enters something wrong.
The error message would look like this:
---Error in t.yab, line 2: Oh no ... ---Error: Program stopped due to an error
Well, there should be a corresponding called warning; unfortunately ther is none yet.
euler
— another name for the constant 2.71828182864
foo=euler
euler is the well known constant named after Leonard Euler; its value is 2.71828182864. euler is not a function, so parens are not allowed (i.e. euler() will produce an error). Finally, you may not assign to euler; it wouldn't sense anyway, because it is a constant.
print euler
eval() — compile and execute a single numeric expression
print eval("1+2")
eval accepts a string, which should be the text of a single numeric expression; it processes the expression and returns the result. All numeric functions and arithmetic operators of yabasic can be used as well as any variables known.
The string passed to eval is first compiled and then executed right away. The compilation happens just before the execution and may cause compilation errors, if you pass an invalid expression. eval might come handy, if you want to calculate an expression, that is not known at the start of your program, e.g. because it is read from the user; see the example below.
input "Please enter an aritmetic expression involving the variable x: " expr$
first = true
for x=0 to 100 step 0.01
result = eval(expr$)
if (first or result > maximum) maximum = result: xmaximum = x
first = false
next x
print "In the range 0 to 100, expression " + expr$ + " has its maximum of " + str$(maximum) + " at " + str$(xmaximum) + " (approximately)"
The example above reads an arithmetic expression from the user and steps through the range 0 … 100 to find its maximum. If the user types e.g. -(x-50)**2, the program would find a maximum of around zero (e.g. -1.90013e-24) at 50.
adding code during execution, eval$, compile, execute, execute$
eval$() — compile and execute a single string-expression
print eval$("a$ + b$")
eval$ accepts a string, which should be the text of a single string-expression; it processes the expression and returns the result. All string-functions and string-operators of yabasic can be used as well as any variables known.
The string passed to eval$ is first compiled and then executed right away. The compilation happens right before the execution and may cause compilation errors, if you pass an invalid expression. See the example below for two interesting use-cases.
The example below allows to apply the quoting rules of yabasic to user-input:
input "Please enter a string with some excape-sequences (e.g. \\r,\\n,\\t): " a$
print eval$("\"" + a$ + "\"")
If the user types abc\ndef at the prompt, the text is echoed like this:
abc
def
The next example shows the subroutine evemex$ (for eval embedded expression) that allows to embed expressions into a string, simply by enclosing them with {{ and }}:
input "Please enter your name: " name$
print evemex$("Hello {{name$}}, your name has {{len(name$)}} characters.")
sub evemex$(evemex_str$)
local evemex_pos1, evemex_pos2, evemex_res$
evemex_pos1 = 1
evemex_pos2 = 1
evemex_res$ = ""
while (evemex_pos1 < len(evemex_str$))
if (mid$(evemex_str$, evemex_pos1, 2) = "{{") then
evemex_res$ = evemex_res$ + mid$(evemex_str$, evemex_pos2, evemex_pos1 - evemex_pos2)
evemex_pos1 = evemex_pos1 + 2
evemex_pos2 = evemex_pos1
while (evemex_pos2 < len(evemex_str$))
if (mid$(evemex_str$, evemex_pos2, 2) = "}}") then
rem
rem See the use of eval in the next line
rem
evemex_res$ = evemex_res$ + eval$("str$(" + mid$(evemex_str$, evemex_pos1, evemex_pos2 - evemex_pos1) + ")")
evemex_pos2 = evemex_pos2 + 2
evemex_pos1 = evemex_pos2
break
else
evemex_pos2 = evemex_pos2 + 1
endif
wend
else
evemex_pos1 = evemex_pos1 + 1
endif
wend
evemex_res$ = evemex_res$ + mid$(evemex_str$, evemex_pos2, evemex_pos1 - evemex_pos2 + 1)
return evemex_res$
end sub
If the user when prompted types Marc, he is greeted with Hello Marc, your name has 4 characters. The program uses eval$ only once, and it adds str$ around the embedded expression to ensure, that the result is always a string and can be concatenated with the other strings.
Please note, that the subroutine prefixes its local variables with evemex (for eval embedded expression) to avoid name clashes with any variable that might be used in expressions within the string passed.
adding code during execution, eval, compile, execute, execute$
execute() — execute a user defined subroutine, which must return a number
print execute("bar","arg1","arg2")
The execute-function is the counterpart of the execute$-function (please see there for some caveats). execute may be used to execute subroutines, which return a number.
print execute("bar",2,3)
sub bar(a,b)
return a+b
end sub
This example would print out 5.
adding code during execution, compile, execute$, eval, eval$
execute$() — execute a user defined subroutine, which must return a string
print execute$("foo$","arg1","arg2")
execute$ can be used to execute a user defined subroutine, whose name is specified as a string expression.
This function allows to execute a subroutine, whose name is not known by the time you write your program. This might happen, if you want to execute a subroutine, which is compiled (using the compile command) as late as of execution of your program.
Note however, that the execute$-function is not the preferred method to execute a user defined subroutine; in almost all cases you should just execute a subroutine by writing down its name within your yabasic program (see the example below).
print ex