3 SPDX-License-Identifier: BSD-2-Clause
5 Copyright (c) 2018-2020 Gavin D. Howard and contributors.
7 Redistribution and use in source and binary forms, with or without
8 modification, are permitted provided that the following conditions are met:
10 * Redistributions of source code must retain the above copyright notice, this
11 list of conditions and the following disclaimer.
13 * Redistributions in binary form must reproduce the above copyright notice,
14 this list of conditions and the following disclaimer in the documentation
15 and/or other materials provided with the distribution.
17 THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
18 AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
19 IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
20 ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS BE
21 LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
22 CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
23 SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
24 INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
25 CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
26 ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
27 POSSIBILITY OF SUCH DAMAGE.
33 bc - arbitrary-precision arithmetic language and calculator
37 **bc** [**-ghilPqsvVw**] [**--global-stacks**] [**--help**] [**--interactive**] [**--mathlib**] [**--no-prompt**] [**--quiet**] [**--standard**] [**--warn**] [**--version**] [**-e** *expr*] [**--expression**=*expr*...] [**-f** *file*...] [**-file**=*file*...]
42 bc(1) is an interactive processor for a language first standardized in 1991 by
43 POSIX. (The current standard is [here][1].) The language provides unlimited
44 precision decimal arithmetic and is somewhat C-like, but there are differences.
45 Such differences will be noted in this document.
47 After parsing and handling options, this bc(1) reads any files given on the
48 command line and executes them before reading from **stdin**.
50 This bc(1) is a drop-in replacement for *any* bc(1), including (and
51 especially) the GNU bc(1).
55 The following are the options that bc(1) accepts.
57 **-g**, **--global-stacks**
59 Turns the globals **ibase**, **obase**, and **scale** into stacks.
61 This has the effect that a copy of the current value of all three are pushed
62 onto a stack for every function call, as well as popped when every function
63 returns. This means that functions can assign to any and all of those
64 globals without worrying that the change will affect other functions.
65 Thus, a hypothetical function named **output(x,b)** that simply printed
66 **x** in base **b** could be written like this:
68 define void output(x, b) {
75 define void output(x, b) {
83 This makes writing functions much easier.
85 However, since using this flag means that functions cannot set **ibase**,
86 **obase**, or **scale** globally, functions that are made to do so cannot
87 work anymore. There are two possible use cases for that, and each has a
90 First, if a function is called on startup to turn bc(1) into a number
91 converter, it is possible to replace that capability with various shell
94 alias d2o="bc -e ibase=A -e obase=8"
95 alias h2b="bc -e ibase=G -e obase=2"
97 Second, if the purpose of a function is to set **ibase**, **obase**, or
98 **scale** globally for any other purpose, it could be split into one to
99 three functions (based on how many globals it sets) and each of those
100 functions could return the desired value for a global.
102 If the behavior of this option is desired for every run of bc(1), then users
103 could make sure to define **BC_ENV_ARGS** and include this option (see the
104 **ENVIRONMENT VARIABLES** section for more details).
106 If **-s**, **-w**, or any equivalents are used, this option is ignored.
108 This is a **non-portable extension**.
112 : Prints a usage message and quits.
114 **-i**, **--interactive**
116 : Forces interactive mode. (See the **INTERACTIVE MODE** section.)
118 This is a **non-portable extension**.
120 **-l**, **--mathlib**
122 : Sets **scale** (see the **SYNTAX** section) to **20** and loads the included
123 math library before running any code, including any expressions or files
124 specified on the command line.
126 To learn what is in the library, see the **LIBRARY** section.
128 **-P**, **--no-prompt**
130 : Disables the prompt in TTY mode. (The prompt is only enabled in TTY mode.
131 See the **TTY MODE** section) This is mostly for those users that do not
132 want a prompt or are not used to having them in bc(1). Most of those users
133 would want to put this option in **BC_ENV_ARGS** (see the
134 **ENVIRONMENT VARIABLES** section).
136 This is a **non-portable extension**.
140 : Do not print copyright header. bc(1) will also suppress the header in
141 non-interactive mode.
143 This is mostly for compatibility with the [GNU bc(1)][2].
145 This is a **non-portable extension**.
147 **-s**, **--standard**
149 : Process exactly the language defined by the [standard][1] and error if any
152 This is a **non-portable extension**.
154 **-v**, **-V**, **--version**
156 : Print the version information (copyright header) and exit.
158 This is a **non-portable extension**.
162 : Like **-s** and **--standard**, except that warnings (and not errors) are
163 printed for non-standard extensions and execution continues normally.
165 This is a **non-portable extension**.
167 **-e** *expr*, **--expression**=*expr*
169 : Evaluates *expr*. If multiple expressions are given, they are evaluated in
170 order. If files are given as well (see below), the expressions and files are
171 evaluated in the order given. This means that if a file is given before an
172 expression, the file is read in and evaluated first.
174 In other bc(1) implementations, this option causes the program to execute
175 the expressions and then exit. This bc(1) does not, unless the
176 **BC_EXPR_EXIT** is defined (see the **ENVIRONMENT VARIABLES** section).
178 This is a **non-portable extension**.
180 **-f** *file*, **--file**=*file*
182 : Reads in *file* and evaluates it, line by line, as though it were read
183 through **stdin**. If expressions are also given (see above), the
184 expressions are evaluated in the order given.
186 In other bc(1) implementations, this option causes the program to execute
187 the files and then exit. This bc(1) does not, unless the
188 **BC_EXPR_EXIT** is defined (see the **ENVIRONMENT VARIABLES** section).
190 This is a **non-portable extension**.
192 All long options are **non-portable extensions**.
196 Any non-error output is written to **stdout**.
198 **Note**: Unlike other bc(1) implementations, this bc(1) will issue a fatal
199 error (see the **EXIT STATUS** section) if it cannot write to **stdout**, so if
200 **stdout** is closed, as in **bc <file> >&-**, it will quit with an error. This
201 is done so that bc(1) can report problems when **stdout** is redirected to a
204 If there are scripts that depend on the behavior of other bc(1) implementations,
205 it is recommended that those scripts be changed to redirect **stdout** to
210 Any error output is written to **stderr**.
212 **Note**: Unlike other bc(1) implementations, this bc(1) will issue a fatal
213 error (see the **EXIT STATUS** section) if it cannot write to **stderr**, so if
214 **stderr** is closed, as in **bc <file> 2>&-**, it will quit with an error. This
215 is done so that bc(1) can exit with an error code when **stderr** is redirected
218 If there are scripts that depend on the behavior of other bc(1) implementations,
219 it is recommended that those scripts be changed to redirect **stderr** to
224 The syntax for bc(1) programs is mostly C-like, with some differences. This
225 bc(1) follows the [POSIX standard][1], which is a much more thorough resource
226 for the language this bc(1) accepts. This section is meant to be a summary and a
227 listing of all the extensions to the standard.
229 In the sections below, **E** means expression, **S** means statement, and **I**
232 Identifiers (**I**) start with a lowercase letter and can be followed by any
233 number (up to **BC_NAME_MAX-1**) of lowercase letters (**a-z**), digits
234 (**0-9**), and underscores (**\_**). The regex is **\[a-z\]\[a-z0-9\_\]\***.
235 Identifiers with more than one character (letter) are a
236 **non-portable extension**.
238 **ibase** is a global variable determining how to interpret constant numbers. It
239 is the "input" base, or the number base used for interpreting input numbers.
240 **ibase** is initially **10**. If the **-s** (**--standard**) and **-w**
241 (**--warn**) flags were not given on the command line, the max allowable value
242 for **ibase** is **36**. Otherwise, it is **16**. The min allowable value for
243 **ibase** is **2**. The max allowable value for **ibase** can be queried in
244 bc(1) programs with the **maxibase()** built-in function.
246 **obase** is a global variable determining how to output results. It is the
247 "output" base, or the number base used for outputting numbers. **obase** is
248 initially **10**. The max allowable value for **obase** is **BC_BASE_MAX** and
249 can be queried in bc(1) programs with the **maxobase()** built-in function. The
250 min allowable value for **obase** is **2**. Values are output in the specified
253 The *scale* of an expression is the number of digits in the result of the
254 expression right of the decimal point, and **scale** is a global variable that
255 sets the precision of any operations, with exceptions. **scale** is initially
256 **0**. **scale** cannot be negative. The max allowable value for **scale** is
257 **BC_SCALE_MAX** and can be queried in bc(1) programs with the **maxscale()**
260 bc(1) has both *global* variables and *local* variables. All *local*
261 variables are local to the function; they are parameters or are introduced in
262 the **auto** list of a function (see the **FUNCTIONS** section). If a variable
263 is accessed which is not a parameter or in the **auto** list, it is assumed to
264 be *global*. If a parent function has a *local* variable version of a variable
265 that a child function considers *global*, the value of that *global* variable in
266 the child function is the value of the variable in the parent function, not the
267 value of the actual *global* variable.
269 All of the above applies to arrays as well.
271 The value of a statement that is an expression (i.e., any of the named
272 expressions or operands) is printed unless the lowest precedence operator is an
273 assignment operator *and* the expression is notsurrounded by parentheses.
275 The value that is printed is also assigned to the special variable **last**. A
276 single dot (**.**) may also be used as a synonym for **last**. These are
277 **non-portable extensions**.
279 Either semicolons or newlines may separate statements.
283 There are two kinds of comments:
285 1. Block comments are enclosed in **/\*** and **\*/**.
286 2. Line comments go from **#** until, and not including, the next newline. This
287 is a **non-portable extension**.
291 The following are named expressions in bc(1):
294 2. Array Elements: **I[E]**
298 6. **last** or a single dot (**.**)
300 Number 6 is a **non-portable extension**.
302 Variables and arrays do not interfere; users can have arrays named the same as
303 variables. This also applies to functions (see the **FUNCTIONS** section), so a
304 user can have a variable, array, and function that all have the same name, and
305 they will not shadow each other, whether inside of functions or not.
307 Named expressions are required as the operand of **increment**/**decrement**
308 operators and as the left side of **assignment** operators (see the *Operators*
313 The following are valid operands in bc(1):
315 1. Numbers (see the *Numbers* subsection below).
316 2. Array indices (**I[E]**).
317 3. **(E)**: The value of **E** (used to change precedence).
318 4. **sqrt(E)**: The square root of **E**. **E** must be non-negative.
319 5. **length(E)**: The number of significant decimal digits in **E**.
320 6. **length(I[])**: The number of elements in the array **I**. This is a
321 **non-portable extension**.
322 7. **scale(E)**: The *scale* of **E**.
323 8. **abs(E)**: The absolute value of **E**. This is a **non-portable
325 9. **I()**, **I(E)**, **I(E, E)**, and so on, where **I** is an identifier for
326 a non-**void** function (see the *Void Functions* subsection of the
327 **FUNCTIONS** section). The **E** argument(s) may also be arrays of the form
328 **I[]**, which will automatically be turned into array references (see the
329 *Array References* subsection of the **FUNCTIONS** section) if the
330 corresponding parameter in the function definition is an array reference.
331 10. **read()**: Reads a line from **stdin** and uses that as an expression. The
332 result of that expression is the result of the **read()** operand. This is a
333 **non-portable extension**.
334 11. **maxibase()**: The max allowable **ibase**. This is a **non-portable
336 12. **maxobase()**: The max allowable **obase**. This is a **non-portable
338 13. **maxscale()**: The max allowable **scale**. This is a **non-portable
343 Numbers are strings made up of digits, uppercase letters, and at most **1**
344 period for a radix. Numbers can have up to **BC_NUM_MAX** digits. Uppercase
345 letters are equal to **9** + their position in the alphabet (i.e., **A** equals
346 **10**, or **9+1**). If a digit or letter makes no sense with the current value
347 of **ibase**, they are set to the value of the highest valid digit in **ibase**.
349 Single-character numbers (i.e., **A** alone) take the value that they would have
350 if they were valid digits, regardless of the value of **ibase**. This means that
351 **A** alone always equals decimal **10** and **Z** alone always equals decimal
356 The following arithmetic and logical operators can be used. They are listed in
357 order of decreasing precedence. Operators in the same group have the same
362 : Type: Prefix and Postfix
366 Description: **increment**, **decrement**
374 Description: **negation**, **boolean not**
382 Description: **power**
390 Description: **multiply**, **divide**, **modulus**
398 Description: **add**, **subtract**
400 **=** **+=** **-=** **\*=** **/=** **%=** **\^=**
406 Description: **assignment**
408 **==** **\<=** **\>=** **!=** **\<** **\>**
414 Description: **relational**
422 Description: **boolean and**
430 Description: **boolean or**
432 The operators will be described in more detail below.
436 : The prefix and postfix **increment** and **decrement** operators behave
437 exactly like they would in C. They require a named expression (see the
438 *Named Expressions* subsection) as an operand.
440 The prefix versions of these operators are more efficient; use them where
445 : The **negation** operator returns **0** if a user attempts to negate any
446 expression with the value **0**. Otherwise, a copy of the expression with
447 its sign flipped is returned.
451 : The **boolean not** operator returns **1** if the expression is **0**, or
454 This is a **non-portable extension**.
458 : The **power** operator (not the **exclusive or** operator, as it would be in
459 C) takes two expressions and raises the first to the power of the value of
462 The second expression must be an integer (no *scale*), and if it is
463 negative, the first value must be non-zero.
467 : The **multiply** operator takes two expressions, multiplies them, and
468 returns the product. If **a** is the *scale* of the first expression and
469 **b** is the *scale* of the second expression, the *scale* of the result is
470 equal to **min(a+b,max(scale,a,b))** where **min()** and **max()** return
475 : The **divide** operator takes two expressions, divides them, and returns the
476 quotient. The *scale* of the result shall be the value of **scale**.
478 The second expression must be non-zero.
482 : The **modulus** operator takes two expressions, **a** and **b**, and
483 evaluates them by 1) Computing **a/b** to current **scale** and 2) Using the
484 result of step 1 to calculate **a-(a/b)\*b** to *scale*
485 **max(scale+scale(b),scale(a))**.
487 The second expression must be non-zero.
491 : The **add** operator takes two expressions, **a** and **b**, and returns the
492 sum, with a *scale* equal to the max of the *scale*s of **a** and **b**.
496 : The **subtract** operator takes two expressions, **a** and **b**, and
497 returns the difference, with a *scale* equal to the max of the *scale*s of
500 **=** **+=** **-=** **\*=** **/=** **%=** **\^=**
502 : The **assignment** operators take two expressions, **a** and **b** where
503 **a** is a named expression (see the *Named Expressions* subsection).
505 For **=**, **b** is copied and the result is assigned to **a**. For all
506 others, **a** and **b** are applied as operands to the corresponding
507 arithmetic operator and the result is assigned to **a**.
509 **==** **\<=** **\>=** **!=** **\<** **\>**
511 : The **relational** operators compare two expressions, **a** and **b**, and
512 if the relation holds, according to C language semantics, the result is
513 **1**. Otherwise, it is **0**.
515 Note that unlike in C, these operators have a lower precedence than the
516 **assignment** operators, which means that **a=b\>c** is interpreted as
519 Also, unlike the [standard][1] requires, these operators can appear anywhere
520 any other expressions can be used. This allowance is a
521 **non-portable extension**.
525 : The **boolean and** operator takes two expressions and returns **1** if both
526 expressions are non-zero, **0** otherwise.
528 This is *not* a short-circuit operator.
530 This is a **non-portable extension**.
534 : The **boolean or** operator takes two expressions and returns **1** if one
535 of the expressions is non-zero, **0** otherwise.
537 This is *not* a short-circuit operator.
539 This is a **non-portable extension**.
543 The following items are statements:
546 2. **{** **S** **;** ... **;** **S** **}**
547 3. **if** **(** **E** **)** **S**
548 4. **if** **(** **E** **)** **S** **else** **S**
549 5. **while** **(** **E** **)** **S**
550 6. **for** **(** **E** **;** **E** **;** **E** **)** **S**
551 7. An empty statement
557 13. A string of characters, enclosed in double quotes
558 14. **print** **E** **,** ... **,** **E**
559 15. **I()**, **I(E)**, **I(E, E)**, and so on, where **I** is an identifier for
560 a **void** function (see the *Void Functions* subsection of the
561 **FUNCTIONS** section). The **E** argument(s) may also be arrays of the form
562 **I[]**, which will automatically be turned into array references (see the
563 *Array References* subsection of the **FUNCTIONS** section) if the
564 corresponding parameter in the function definition is an array reference.
566 Numbers 4, 9, 11, 12, 14, and 15 are **non-portable extensions**.
568 Also, as a **non-portable extension**, any or all of the expressions in the
569 header of a for loop may be omitted. If the condition (second expression) is
570 omitted, it is assumed to be a constant **1**.
572 The **break** statement causes a loop to stop iterating and resume execution
573 immediately following a loop. This is only allowed in loops.
575 The **continue** statement causes a loop iteration to stop early and returns to
576 the start of the loop, including testing the loop condition. This is only
579 The **if** **else** statement does the same thing as in C.
581 The **quit** statement causes bc(1) to quit, even if it is on a branch that will
582 not be executed (it is a compile-time command).
584 The **halt** statement causes bc(1) to quit, if it is executed. (Unlike **quit**
585 if it is on a branch of an **if** statement that is not executed, bc(1) does not
588 The **limits** statement prints the limits that this bc(1) is subject to. This
589 is like the **quit** statement in that it is a compile-time command.
591 An expression by itself is evaluated and printed, followed by a newline.
595 The "expressions" in a **print** statement may also be strings. If they are, there
596 are backslash escape sequences that are interpreted specially. What those
597 sequences are, and what they cause to be printed, are shown below:
611 Any other character following a backslash causes the backslash and character to
614 Any non-string expression in a print statement shall be assigned to **last**,
615 like any other expression that is printed.
617 ## Order of Evaluation
619 All expressions in a statment are evaluated left to right, except as necessary
620 to maintain order of operations. This means, for example, assuming that **i** is
621 equal to **0**, in the expression
625 the first (or 0th) element of **a** is set to **1**, and **i** is equal to **2**
626 at the end of the expression.
628 This includes function arguments. Thus, assuming **i** is equal to **0**, this
629 means that in the expression
633 the first argument passed to **x()** is **0**, and the second argument is **1**,
634 while **i** is equal to **2** before the function starts executing.
638 Function definitions are as follows:
648 Any **I** in the parameter list or **auto** list may be replaced with **I[]** to
649 make a parameter or **auto** var an array, and any **I** in the parameter list
650 may be replaced with **\*I[]** to make a parameter an array reference. Callers
651 of functions that take array references should not put an asterisk in the call;
652 they must be called with just **I[]** like normal array parameters and will be
653 automatically converted into references.
655 As a **non-portable extension**, the opening brace of a **define** statement may
656 appear on the next line.
658 As a **non-portable extension**, the return statement may also be in one of the
662 2. **return** **(** **)**
665 The first two, or not specifying a **return** statement, is equivalent to
666 **return (0)**, unless the function is a **void** function (see the *Void
667 Functions* subsection below).
671 Functions can also be **void** functions, defined as follows:
674 define void I(I,...,I){
681 They can only be used as standalone expressions, where such an expression would
682 be printed alone, except in a print statement.
684 Void functions can only use the first two **return** statements listed above.
685 They can also omit the return statement entirely.
687 The word "void" is not treated as a keyword; it is still possible to have
688 variables, arrays, and functions named **void**. The word "void" is only
689 treated specially right after the **define** keyword.
691 This is a **non-portable extension**.
695 For any array in the parameter list, if the array is declared in the form
701 it is a **reference**. Any changes to the array in the function are reflected,
702 when the function returns, to the array that was passed in.
704 Other than this, all function arguments are passed by value.
706 This is a **non-portable extension**.
710 All of the functions below are available when the **-l** or **--mathlib**
711 command-line flags are given.
715 The [standard][1] defines the following functions for the math library:
719 : Returns the sine of **x**, which is assumed to be in radians.
721 This is a transcendental function (see the *Transcendental Functions*
726 : Returns the cosine of **x**, which is assumed to be in radians.
728 This is a transcendental function (see the *Transcendental Functions*
733 : Returns the arctangent of **x**, in radians.
735 This is a transcendental function (see the *Transcendental Functions*
740 : Returns the natural logarithm of **x**.
742 This is a transcendental function (see the *Transcendental Functions*
747 : Returns the mathematical constant **e** raised to the power of **x**.
749 This is a transcendental function (see the *Transcendental Functions*
754 : Returns the bessel integer order **n** (truncated) of **x**.
756 This is a transcendental function (see the *Transcendental Functions*
759 ## Transcendental Functions
761 All transcendental functions can return slightly inaccurate results (up to 1
762 [ULP][4]). This is unavoidable, and [this article][5] explains why it is
763 impossible and unnecessary to calculate exact results for the transcendental
766 Because of the possible inaccuracy, I recommend that users call those functions
767 with the precision (**scale**) set to at least 1 higher than is necessary. If
768 exact results are *absolutely* required, users can double the precision
769 (**scale**) and then truncate.
771 The transcendental functions in the standard math library are:
782 When bc(1) encounters an error or a signal that it has a non-default handler
783 for, it resets. This means that several things happen.
785 First, any functions that are executing are stopped and popped off the stack.
786 The behavior is not unlike that of exceptions in programming languages. Then
787 the execution point is set so that any code waiting to execute (after all
788 functions returned) is skipped.
790 Thus, when bc(1) resets, it skips any remaining code waiting to be executed.
791 Then, if it is interactive mode, and the error was not a fatal error (see the
792 **EXIT STATUS** section), it asks for more input; otherwise, it exits with the
793 appropriate return code.
795 Note that this reset behavior is different from the GNU bc(1), which attempts to
796 start executing the statement right after the one that caused an error.
800 Most bc(1) implementations use **char** types to calculate the value of **1**
801 decimal digit at a time, but that can be slow. This bc(1) does something
804 It uses large integers to calculate more than **1** decimal digit at a time. If
805 built in a environment where **BC_LONG_BIT** (see the **LIMITS** section) is
806 **64**, then each integer has **9** decimal digits. If built in an environment
807 where **BC_LONG_BIT** is **32** then each integer has **4** decimal digits. This
808 value (the number of decimal digits per large integer) is called
811 The actual values of **BC_LONG_BIT** and **BC_BASE_DIGS** can be queried with
812 the **limits** statement.
814 In addition, this bc(1) uses an even larger integer for overflow checking. This
815 integer type depends on the value of **BC_LONG_BIT**, but is always at least
816 twice as large as the integer type used to store digits.
820 The following are the limits on bc(1):
824 : The number of bits in the **long** type in the environment where bc(1) was
825 built. This determines how many decimal digits can be stored in a single
826 large integer (see the **PERFORMANCE** section).
830 : The number of decimal digits per large integer (see the **PERFORMANCE**
831 section). Depends on **BC_LONG_BIT**.
835 : The max decimal number that each large integer can store (see
836 **BC_BASE_DIGS**) plus **1**. Depends on **BC_BASE_DIGS**.
840 : The max number that the overflow type (see the **PERFORMANCE** section) can
841 hold. Depends on **BC_LONG_BIT**.
845 : The maximum output base. Set at **BC_BASE_POW**.
849 : The maximum size of arrays. Set at **SIZE_MAX-1**.
853 : The maximum **scale**. Set at **BC_OVERFLOW_MAX-1**.
857 : The maximum length of strings. Set at **BC_OVERFLOW_MAX-1**.
861 : The maximum length of identifiers. Set at **BC_OVERFLOW_MAX-1**.
865 : The maximum length of a number (in decimal digits), which includes digits
866 after the decimal point. Set at **BC_OVERFLOW_MAX-1**.
870 : The maximum allowable exponent (positive or negative). Set at
875 : The maximum number of vars/arrays. Set at **SIZE_MAX-1**.
877 The actual values can be queried with the **limits** statement.
879 These limits are meant to be effectively non-existent; the limits are so large
880 (at least on 64-bit machines) that there should not be any point at which they
881 become a problem. In fact, memory should be exhausted before these limits should
884 # ENVIRONMENT VARIABLES
886 bc(1) recognizes the following environment variables:
890 : If this variable exists (no matter the contents), bc(1) behaves as if
891 the **-s** option was given.
895 : This is another way to give command-line arguments to bc(1). They should be
896 in the same format as all other command-line arguments. These are always
897 processed first, so any files given in **BC_ENV_ARGS** will be processed
898 before arguments and files given on the command-line. This gives the user
899 the ability to set up "standard" options and files to be used at every
900 invocation. The most useful thing for such files to contain would be useful
901 functions that the user might want every time bc(1) runs.
903 The code that parses **BC_ENV_ARGS** will correctly handle quoted arguments,
904 but it does not understand escape sequences. For example, the string
905 **"/home/gavin/some bc file.bc"** will be correctly parsed, but the string
906 **"/home/gavin/some \"bc\" file.bc"** will include the backslashes.
908 The quote parsing will handle either kind of quotes, **'** or **"**. Thus,
909 if you have a file with any number of single quotes in the name, you can use
910 double quotes as the outside quotes, as in **"some 'bc' file.bc"**, and vice
911 versa if you have a file with double quotes. However, handling a file with
912 both kinds of quotes in **BC_ENV_ARGS** is not supported due to the
913 complexity of the parsing, though such files are still supported on the
914 command-line where the parsing is done by the shell.
918 : If this environment variable exists and contains an integer that is greater
919 than **1** and is less than **UINT16_MAX** (**2\^16-1**), bc(1) will output
920 lines to that length, including the backslash (**\\**). The default line
925 : If this variable exists (no matter the contents), bc(1) will exit
926 immediately after executing expressions and files given by the **-e** and/or
927 **-f** command-line options (and any equivalents).
931 bc(1) returns the following exit statuses:
939 : A math error occurred. This follows standard practice of using **1** for
940 expected errors, since math errors will happen in the process of normal
943 Math errors include divide by **0**, taking the square root of a negative
944 number, attempting to convert a negative number to a hardware integer,
945 overflow when converting a number to a hardware integer, and attempting to
946 use a non-integer where an integer is required.
948 Converting to a hardware integer happens for the second operand of the power
949 (**\^**) operator and the corresponding assignment operator.
953 : A parse error occurred.
955 Parse errors include unexpected **EOF**, using an invalid character, failing
956 to find the end of a string or comment, using a token where it is invalid,
957 giving an invalid expression, giving an invalid print statement, giving an
958 invalid function definition, attempting to assign to an expression that is
959 not a named expression (see the *Named Expressions* subsection of the
960 **SYNTAX** section), giving an invalid **auto** list, having a duplicate
961 **auto**/function parameter, failing to find the end of a code block,
962 attempting to return a value from a **void** function, attempting to use a
963 variable as a reference, and using any extensions when the option **-s** or
964 any equivalents were given.
968 : A runtime error occurred.
970 Runtime errors include assigning an invalid number to **ibase**, **obase**,
971 or **scale**; give a bad expression to a **read()** call, calling **read()**
972 inside of a **read()** call, type errors, passing the wrong number of
973 arguments to functions, attempting to call an undefined function, and
974 attempting to use a **void** function call as a value in an expression.
978 : A fatal error occurred.
980 Fatal errors include memory allocation errors, I/O errors, failing to open
981 files, attempting to use files that do not have only ASCII characters (bc(1)
982 only accepts ASCII characters), attempting to open a directory as a file,
983 and giving invalid command-line options.
985 The exit status **4** is special; when a fatal error occurs, bc(1) always exits
986 and returns **4**, no matter what mode bc(1) is in.
988 The other statuses will only be returned when bc(1) is not in interactive mode
989 (see the **INTERACTIVE MODE** section), since bc(1) resets its state (see the
990 **RESET** section) and accepts more input when one of those errors occurs in
991 interactive mode. This is also the case when interactive mode is forced by the
992 **-i** flag or **--interactive** option.
994 These exit statuses allow bc(1) to be used in shell scripting with error
995 checking, and its normal behavior can be forced by using the **-i** flag or
996 **--interactive** option.
1000 Per the [standard][1], bc(1) has an interactive mode and a non-interactive mode.
1001 Interactive mode is turned on automatically when both **stdin** and **stdout**
1002 are hooked to a terminal, but the **-i** flag and **--interactive** option can
1003 turn it on in other cases.
1005 In interactive mode, bc(1) attempts to recover from errors (see the **RESET**
1006 section), and in normal execution, flushes **stdout** as soon as execution is
1007 done for the current input.
1011 If **stdin**, **stdout**, and **stderr** are all connected to a TTY, bc(1) turns
1014 TTY mode is required for history to be enabled (see the **COMMAND LINE HISTORY**
1015 section). It is also required to enable special handling for **SIGINT** signals.
1017 The prompt is enabled in TTY mode.
1019 TTY mode is different from interactive mode because interactive mode is required
1020 in the [bc(1) specification][1], and interactive mode requires only **stdin**
1021 and **stdout** to be connected to a terminal.
1025 Sending a **SIGINT** will cause bc(1) to stop execution of the current input. If
1026 bc(1) is in TTY mode (see the **TTY MODE** section), it will reset (see the
1027 **RESET** section). Otherwise, it will clean up and exit.
1029 Note that "current input" can mean one of two things. If bc(1) is processing
1030 input from **stdin** in TTY mode, it will ask for more input. If bc(1) is
1031 processing input from a file in TTY mode, it will stop processing the file and
1032 start processing the next file, if one exists, or ask for input from **stdin**
1033 if no other file exists.
1035 This means that if a **SIGINT** is sent to bc(1) as it is executing a file, it
1036 can seem as though bc(1) did not respond to the signal since it will immediately
1037 start executing the next file. This is by design; most files that users execute
1038 when interacting with bc(1) have function definitions, which are quick to parse.
1039 If a file takes a long time to execute, there may be a bug in that file. The
1040 rest of the files could still be executed without problem, allowing the user to
1043 **SIGTERM** and **SIGQUIT** cause bc(1) to clean up and exit, and it uses the
1044 default handler for all other signals. The one exception is **SIGHUP**; in that
1045 case, when bc(1) is in TTY mode, a **SIGHUP** will cause bc(1) to clean up and
1048 # COMMAND LINE HISTORY
1050 bc(1) supports interactive command-line editing. If bc(1) is in TTY mode (see
1051 the **TTY MODE** section), history is enabled. Previous lines can be recalled
1052 and edited with the arrow keys.
1054 **Note**: tabs are converted to 8 spaces.
1062 bc(1) is compliant with the [IEEE Std 1003.1-2017 (“POSIX.1-2017”)][1]
1063 specification. The flags **-efghiqsvVw**, all long options, and the extensions
1064 noted above are extensions to that specification.
1066 Note that the specification explicitly says that bc(1) only accepts numbers that
1067 use a period (**.**) as a radix point, regardless of the value of
1072 None are known. Report bugs at https://git.yzena.com/gavin/bc.
1076 Gavin D. Howard <yzena.tech@gmail.com> and contributors.
1078 [1]: https://pubs.opengroup.org/onlinepubs/9699919799/utilities/bc.html
1079 [2]: https://www.gnu.org/software/bc/
1080 [3]: https://en.wikipedia.org/wiki/Rounding#Round_half_away_from_zero
1081 [4]: https://en.wikipedia.org/wiki/Unit_in_the_last_place
1082 [5]: https://people.eecs.berkeley.edu/~wkahan/LOG10HAF.TXT
1083 [6]: https://en.wikipedia.org/wiki/Rounding#Rounding_away_from_zero