6 Jim Tcl v0.75 - reference manual for the Jim Tcl scripting language
16 jimsh -e '<immediate-script>'
21 * <<CommandIndex,Command Reference>>
22 * <<OperatorPrecedence,Operator Precedence>>
23 * <<BuiltinVariables, Builtin Variables>>
24 * <<BackslashSequences, Backslash Sequences>>
28 Jim Tcl is a small footprint reimplementation of the Tcl scripting language.
29 The core language engine is compatible with Tcl 8.5+, while implementing
30 a significant subset of the Tcl 8.6 command set, plus additional features
31 available only in Jim Tcl.
33 Some notable differences with Tcl 8.5/8.6 are:
35 1. Object-based I/O (aio), but with a Tcl-compatibility layer
36 2. I/O: Support for sockets and pipes including udp, unix domain sockets and IPv6
38 4. Support for references (`ref`/`getref`/`setref`) and garbage collection
39 5. Builtin dictionary type (`dict`) with some limitations compared to Tcl 8.6
40 6. `env` command to access environment variables
41 7. Operating system features: `os.fork`, `os.wait`, `os.uptime`, `signal`, `alarm`, `sleep`
42 8. Much better error reporting. `info stacktrace` as a replacement for '$errorInfo', '$errorCode'
43 9. Support for "static" variables in procedures
44 10. Threads and coroutines are not supported
45 11. Command and variable traces are not supported
46 12. Built-in command line editing
47 13. Expression shorthand syntax: +$(...)+
48 14. Modular build allows many features to be omitted or built as dynamic, loadable modules
49 15. Highly suitable for use in an embedded environment
50 16. Support for UDP, IPv6, Unix-Domain sockets in addition to TCP sockets
54 Changes between 0.74 and 0.75
55 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
56 1. `binary`, `pack` and `unpack` now support floating point
57 2. `file copy` '-force' handles source and target as the same file
58 3. `format` now supports +%b+ for binary conversion
59 4. `lsort` now supports '-unique' and '-real'
60 5. Add support for half-close with `aio close` ?r|w?
61 6. Add `socket pair` for a bidirectional pipe
62 7. Add --random-hash to randomise hash tables for greater security
63 8. `dict` now supports 'for', 'values', 'incr', 'append', 'lappend', 'update', 'info' and 'replace'
64 9. `file stat` no longer requires the variable name
66 Changes between 0.73 and 0.74
67 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
68 1. Numbers with leading zeros are treated as decimal, not octal
70 3. Add LFS (64 bit) support for `aio seek`, `aio tell`, `aio copyto`, `file copy`
71 4. `string compare` and `string equal` now support '-length'
72 5. `glob` now supports '-directory'
74 Changes between 0.72 and 0.73
75 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
76 1. Built-in regexp now support non-capturing parentheses: (?:...)
77 2. Add `string replace`
78 3. Add `string totitle`
80 5. Add +build-jim-ext+ for easy separate building of loadable modules (extensions)
81 6. `local` now works with any command, not just procs
82 7. Add `info alias` to access the target of an alias
83 8. UTF-8 encoding past the basic multilingual plane (BMP) is supported
86 11. Most extensions are now enabled by default
87 12. Add support for namespaces and the `namespace` command
90 Changes between 0.71 and 0.72
91 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
92 1. procs now allow 'args' and optional parameters in any position
93 2. Add Tcl-compatible expr functions, `rand()`, `srand()` and `pow()`
94 3. Add support for the '-force' option to `file delete`
95 4. Better diagnostics when `source` fails to load a script with a missing quote or bracket
96 5. New +tcl_platform(pathSeparator)+
97 6. Add support settings the modification time with `file mtime`
98 7. `exec` is now fully supported on win32 (mingw32)
99 8. `file join`, `pwd`, `glob` etc. now work for mingw32
100 9. Line editing is now supported for the win32 console (mingw32)
101 10. Add `aio listen` command
103 Changes between 0.70 and 0.71
104 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
105 1. Allow 'args' to be renamed in procs
106 2. Add +$(...)+ shorthand syntax for expressions
107 3. Add automatic reference variables in procs with +&var+ syntax
108 4. Support +jimsh --version+
109 5. Additional variables in +tcl_platform()+
110 6. `local` procs now push existing commands and `upcall` can call them
111 7. Add `loop` command (TclX compatible)
112 8. Add `aio buffering` command
113 9. `info complete` can now return the missing character
114 10. `binary format` and `binary scan` are now (optionally) supported
115 11. Add `string byterange`
116 12. Built-in regexp now support non-greedy repetition (*?, +?, ??)
120 Tcl stands for 'tool command language' and is pronounced 'tickle.'
121 It is actually two things: a language and a library.
123 First, Tcl is a simple textual language, intended primarily for
124 issuing commands to interactive programs such as text editors,
125 debuggers, illustrators, and shells. It has a simple syntax and is also
126 programmable, so Tcl users can write command procedures to provide more
127 powerful commands than those in the built-in set.
129 Second, Tcl is a library package that can be embedded in application
130 programs. The Tcl library consists of a parser for the Tcl language,
131 routines to implement the Tcl built-in commands, and procedures that
132 allow each application to extend Tcl with additional commands specific
133 to that application. The application program generates Tcl commands and
134 passes them to the Tcl parser for execution. Commands may be generated
135 by reading characters from an input source, or by associating command
136 strings with elements of the application's user interface, such as menu
137 entries, buttons, or keystrokes.
139 When the Tcl library receives commands it parses them into component
140 fields and executes built-in commands directly. For commands implemented
141 by the application, Tcl calls back to the application to execute the
142 commands. In many cases commands will invoke recursive invocations of the
143 Tcl interpreter by passing in additional strings to execute (procedures,
144 looping commands, and conditional commands all work in this way).
146 An application program gains three advantages by using Tcl for its command
147 language. First, Tcl provides a standard syntax: once users know Tcl,
148 they will be able to issue commands easily to any Tcl-based application.
149 Second, Tcl provides programmability. All a Tcl application needs
150 to do is to implement a few application-specific low-level commands.
151 Tcl provides many utility commands plus a general programming interface
152 for building up complex command procedures. By using Tcl, applications
153 need not re-implement these features.
155 Third, Tcl can be used as a common language for communicating between
156 applications. Inter-application communication is not built into the
157 Tcl core described here, but various add-on libraries, such as the Tk
158 toolkit, allow applications to issue commands to each other. This makes
159 it possible for applications to work together in much more powerful ways
160 than was previously possible.
162 Fourth, Jim Tcl includes a command processor, +jimsh+, which can be
163 used to run standalone Tcl scripts, or to run Tcl commands interactively.
165 This manual page focuses primarily on the Tcl language. It describes
166 the language syntax and the built-in commands that will be available
167 in any application based on Tcl. The individual library procedures are
168 described in more detail in separate manual pages, one per procedure.
170 JIMSH COMMAND INTERPRETER
171 -------------------------
172 A simple, but powerful command processor, +jimsh+, is part of Jim Tcl.
173 It may be invoked in interactive mode as:
177 or to process the Tcl script in a file with:
181 It may also be invoked to execute an immediate script with:
187 Interactive mode reads Tcl commands from standard input, evaluates
188 those commands and prints the results.
191 Welcome to Jim version 0.73, Copyright (c) 2005-8 Salvatore Sanfilippo
194 . lsort [info commands p*]
195 package parray pid popen proc puts pwd
196 . foreach i {a b c} {
203 invalid command name "bad"
207 If +jimsh+ is configured with line editing (it is by default) and a VT-100-compatible
208 terminal is detected, Emacs-style line editing commands are available, including:
209 arrow keys, +\^W+ to erase a word, +\^U+ to erase the line, +^R+ for reverse incremental search
210 in history. Additionally, the +h+ command may be used to display the command history.
212 Command line history is automatically saved and loaded from +~/.jim_history+
214 In interactive mode, +jimsh+ automatically runs the script +~/.jimrc+ at startup
219 The central data structure in Tcl is an interpreter (C type 'Jim_Interp').
220 An interpreter consists of a set of command bindings, a set of variable
221 values, and a few other miscellaneous pieces of state. Each Tcl command
222 is interpreted in the context of a particular interpreter.
224 Some Tcl-based applications will maintain multiple interpreters
225 simultaneously, each associated with a different widget or portion of
226 the application. Interpreters are relatively lightweight structures.
227 They can be created and deleted quickly, so application programmers should
228 feel free to use multiple interpreters if that simplifies the application.
232 Tcl supports only one type of data: strings. All commands, all arguments
233 to commands, all command results, and all variable values are strings.
235 Where commands require numeric arguments or return numeric results,
236 the arguments and results are passed as strings. Many commands expect
237 their string arguments to have certain formats, but this interpretation
238 is up to the individual commands. For example, arguments often contain
239 Tcl command strings, which may get executed as part of the commands.
240 The easiest way to understand the Tcl interpreter is to remember that
241 everything is just an operation on a string. In many cases Tcl constructs
242 will look similar to more structured constructs from other languages.
243 However, the Tcl constructs are not structured at all; they are just
244 strings of characters, and this gives them a different behaviour than
245 the structures they may look like.
247 Although the exact interpretation of a Tcl string depends on who is doing
248 the interpretation, there are three common forms that strings take:
249 commands, expressions, and lists. The major sections below discuss
250 these three forms in more detail.
254 The Tcl language has syntactic similarities to both the Unix shells
255 and Lisp. However, the interpretation of commands is different
256 in Tcl than in either of those other two systems.
257 A Tcl command string consists of one or more commands separated
258 by newline characters or semi-colons.
259 Each command consists of a collection of fields separated by
260 white space (spaces or tabs).
261 The first field must be the name of a command, and the
262 additional fields, if any, are arguments that will be passed to
263 that command. For example, the command:
267 has three fields: the first, `set`, is the name of a Tcl command, and
268 the last two, 'a' and '22', will be passed as arguments to
269 the `set` command. The command name may refer either to a built-in
270 Tcl command, an application-specific command bound in with the library
271 procedure 'Jim_CreateCommand', or a command procedure defined with the
272 `proc` built-in command.
274 Arguments are passed literally as text strings. Individual commands may
275 interpret those strings in any fashion they wish. The `set` command,
276 for example, will treat its first argument as the name of a variable
277 and its second argument as a string value to assign to that variable.
278 For other commands arguments may be interpreted as integers, lists,
279 file names, or Tcl commands.
281 Command names should normally be typed completely (e.g. no abbreviations).
282 However, if the Tcl interpreter cannot locate a command it invokes a
283 special command named `unknown` which attempts to find or create the
286 For example, at many sites `unknown` will search through library
287 directories for the desired command and create it as a Tcl procedure if
288 it is found. The `unknown` command often provides automatic completion
289 of abbreviated commands, but usually only for commands that were typed
292 It's probably a bad idea to use abbreviations in command scripts and
293 other forms that will be re-used over time: changes to the command set
294 may cause abbreviations to become ambiguous, resulting in scripts that
299 If the first non-blank character in a command is +\#+, then everything
300 from the +#+ up through the next newline character is treated as
301 a comment and ignored. When comments are embedded inside nested
302 commands (e.g. fields enclosed in braces) they must have properly-matched
303 braces (this is necessary because when Tcl parses the top-level command
304 it doesn't yet know that the nested field will be used as a command so
305 it cannot process the nested comment character as a comment).
307 GROUPING ARGUMENTS WITH DOUBLE-QUOTES
308 -------------------------------------
309 Normally each argument field ends at the next white space, but
310 double-quotes may be used to create arguments with embedded space.
312 If an argument field begins with a double-quote, then the argument isn't
313 terminated by white space (including newlines) or a semi-colon (see below
314 for information on semi-colons); instead it ends at the next double-quote
315 character. The double-quotes are not included in the resulting argument.
316 For example, the command
318 set a "This is a single argument"
320 will pass two arguments to `set`: 'a' and 'This is a single argument'.
322 Within double-quotes, command substitutions, variable substitutions,
323 and backslash substitutions still occur, as described below. If the
324 first character of a command field is not a quote, then quotes receive
325 no special interpretation in the parsing of that field.
327 GROUPING ARGUMENTS WITH BRACES
328 ------------------------------
329 Curly braces may also be used for grouping arguments. They are similar
330 to quotes except for two differences. First, they nest; this makes them
331 easier to use for complicated arguments like nested Tcl command strings.
332 Second, the substitutions described below for commands, variables, and
333 backslashes do *not* occur in arguments enclosed in braces, so braces
334 can be used to prevent substitutions where they are undesirable.
336 If an argument field begins with a left brace, then the argument ends
337 at the matching right brace. Tcl will strip off the outermost layer
338 of braces and pass the information between the braces to the command
339 without any further modification. For example, in the command
341 set a {xyz a {b c d}}
343 the `set` command will receive two arguments: 'a'
346 When braces or quotes are in effect, the matching brace or quote need
347 not be on the same line as the starting quote or brace; in this case
348 the newline will be included in the argument field along with any other
349 characters up to the matching brace or quote. For example, the `eval`
350 command takes one argument, which is a command string; `eval` invokes
351 the Tcl interpreter to execute the command string. The command
358 will assign the value '22' to 'a' and '33' to 'b'.
360 If the first character of a command field is not a left
361 brace, then neither left nor right
362 braces in the field will be treated specially (except as part of
363 variable substitution; see below).
365 COMMAND SUBSTITUTION WITH BRACKETS
366 ----------------------------------
367 If an open bracket occurs in a field of a command, then command
368 substitution occurs (except for fields enclosed in braces). All of the
369 text up to the matching close bracket is treated as a Tcl command and
370 executed immediately. Then the result of that command is substituted
371 for the bracketed text. For example, consider the command
375 When the `set` command has only a single argument, it is the name of a
376 variable and `set` returns the contents of that variable. In this case,
377 if variable 'b' has the value 'foo', then the command above is equivalent
382 Brackets can be used in more complex ways. For example, if the variable
383 'b' has the value 'foo' and the variable 'c' has the value 'gorp',
386 set a xyz[set b].[set c]
388 is equivalent to the command
393 A bracketed command may contain multiple commands separated by newlines
394 or semi-colons in the usual fashion. In this case the value of the last
395 command is used for substitution. For example, the command
400 is equivalent to the command
405 If a field is enclosed in braces then the brackets and the characters
406 between them are not interpreted specially; they are passed through to
407 the argument verbatim.
409 VARIABLE SUBSTITUTION WITH $
410 ----------------------------
411 The dollar sign (+$+) may be used as a special shorthand form for
412 substituting variable values. If +$+ appears in an argument that isn't
413 enclosed in braces then variable substitution will occur. The characters
414 after the +$+, up to the first character that isn't a number, letter,
415 or underscore, are taken as a variable name and the string value of that
416 variable is substituted for the name.
418 For example, if variable 'foo' has the value 'test', then the command
422 is equivalent to the command
426 There are two special forms for variable substitution. If the next
427 character after the name of the variable is an open parenthesis, then
428 the variable is assumed to be an array name, and all of the characters
429 between the open parenthesis and the next close parenthesis are taken as
430 an index into the array. Command substitutions and variable substitutions
431 are performed on the information between the parentheses before it is
434 For example, if the variable 'x' is an array with one element named
435 'first' and value '87' and another element named '14' and value 'more',
438 set a xyz$x(first)zyx
440 is equivalent to the command
444 If the variable 'index' has the value '14', then the command
446 set a xyz$x($index)zyx
448 is equivalent to the command
452 For more information on arrays, see VARIABLES AND ARRAYS below.
454 The second special form for variables occurs when the dollar sign is
455 followed by an open curly brace. In this case the variable name consists
456 of all the characters up to the next curly brace.
458 Array references are not possible in this form: the name between braces
459 is assumed to refer to a scalar variable. For example, if variable
460 'foo' has the value 'test', then the command
464 is equivalent to the command
469 Variable substitution does not occur in arguments that are enclosed in
470 braces: the dollar sign and variable name are passed through to the
473 The dollar sign abbreviation is simply a shorthand form. +$a+ is
474 completely equivalent to +[set a]+; it is provided as a convenience
477 SEPARATING COMMANDS WITH SEMI-COLONS
478 ------------------------------------
479 Normally, each command occupies one line (the command is terminated by a
480 newline character). However, semi-colon (+;+) is treated as a command
481 separator character; multiple commands may be placed on one line by
482 separating them with a semi-colon. Semi-colons are not treated as
483 command separators if they appear within curly braces or double-quotes.
485 BACKSLASH SUBSTITUTION
486 ----------------------
487 Backslashes may be used to insert non-printing characters into command
488 fields and also to insert special characters like braces and brackets
489 into fields without them being interpreted specially as described above.
491 The backslash sequences understood by the Tcl interpreter are
492 listed below. In each case, the backslash
493 sequence is replaced by the given character:
494 [[BackslashSequences]]
505 Carriage-return (0xd).
528 +{backslash}<space>+::
529 Space ( ): doesn't terminate argument.
532 Semi-colon: doesn't terminate command.
537 +{backslash}<newline>+::
538 Nothing: this joins two lines together
539 into a single line. This backslash feature is unique in that
540 it will be applied even when the sequence occurs within braces.
542 +{backslash}{backslash}+::
543 Backslash ('{backslash}').
546 The digits +'ddd'+ (one, two, or three of them) give the octal value of
547 the character. Note that Jim supports null characters in strings.
550 +{backslash}u\{nnn\}+::
551 +{backslash}Unnnnnnnn+::
552 The UTF-8 encoding of the unicode codepoint represented by the hex digits, +'nnnn'+, is inserted.
553 The 'u' form allows for one to four hex digits.
554 The 'U' form allows for one to eight hex digits.
555 The 'u\{nnn\}' form allows for one to eight hex digits, but makes it easier to insert
556 characters UTF-8 characters which are followed by a hex digit.
558 For example, in the command
562 the second argument to `set` will be +{x[ yza+.
564 If a backslash is followed by something other than one of the options
565 described above, then the backslash is transmitted to the argument
566 field without any special processing, and the Tcl scanner continues
567 normal processing with the next character. For example, in the
572 The first argument to `set` will be +{backslash}*a+ and the second
573 argument will be +{backslash}{foo+.
575 If an argument is enclosed in braces, then backslash sequences inside
576 the argument are parsed but no substitution occurs (except for
577 backslash-newline): the backslash
578 sequence is passed through to the argument as is, without making
579 any special interpretation of the characters in the backslash sequence.
580 In particular, backslashed braces are not counted in locating the
581 matching right brace that terminates the argument.
587 the second argument to `set` will be +{backslash}{abc+.
589 This backslash mechanism is not sufficient to generate absolutely
590 any argument structure; it only covers the
591 most common cases. To produce particularly complicated arguments
592 it is probably easiest to use the `format` command along with
593 command substitution.
595 STRING AND LIST INDEX SPECIFICATIONS
596 ------------------------------------
598 Many string and list commands take one or more 'index' parameters which
599 specify a position in the string relative to the start or end of the string/list.
601 The index may be one of the following forms:
604 A simple integer, where '0' refers to the first element of the string
607 +integer+integer+ or::
609 The sum or difference of the two integers. e.g. +2+3+ refers to the 5th element.
610 This is useful when used with (e.g.) +$i+1+ rather than the more verbose
614 The last element of the string or list.
617 The 'nth-from-last' element of the string or list.
621 1. A command is just a string.
622 2. Within a string commands are separated by newlines or semi-colons
623 (unless the newline or semi-colon is within braces or brackets
625 3. A command consists of fields. The first field is the name of the command.
626 The other fields are strings that are passed to that command as arguments.
627 4. Fields are normally separated by white space.
628 5. Double-quotes allow white space and semi-colons to appear within
630 Command substitution, variable substitution, and backslash substitution
631 still occur inside quotes.
632 6. Braces defer interpretation of special characters.
633 If a field begins with a left brace, then it consists of everything
634 between the left brace and the matching right brace. The
635 braces themselves are not included in the argument.
636 No further processing is done on the information between the braces
637 except that backslash-newline sequences are eliminated.
638 7. If a field doesn't begin with a brace then backslash,
639 variable, and command substitution are done on the field. Only a
640 single level of processing is done: the results of one substitution
641 are not scanned again for further substitutions or any other
642 special treatment. Substitution can
643 occur on any field of a command, including the command name
644 as well as the arguments.
645 8. If the first non-blank character of a command is a +\#+, everything
646 from the +#+ up through the next newline is treated as a comment
651 The second major interpretation applied to strings in Tcl is
652 as expressions. Several commands, such as `expr`, `for`,
653 and `if`, treat one or more of their arguments as expressions
654 and call the Tcl expression processors ('Jim_ExprLong',
655 'Jim_ExprBoolean', etc.) to evaluate them.
657 The operators permitted in Tcl expressions are a subset of
658 the operators permitted in C expressions, and they have the
659 same meaning and precedence as the corresponding C operators.
660 Expressions almost always yield numeric results
661 (integer or floating-point values).
662 For example, the expression
668 Tcl expressions differ from C expressions in the way that
669 operands are specified, and in that Tcl expressions support
670 non-numeric operands and string comparisons.
672 A Tcl expression consists of a combination of operands, operators,
675 White space may be used between the operands and operators and
676 parentheses; it is ignored by the expression processor.
677 Where possible, operands are interpreted as integer values.
679 Integer values may be specified in decimal (the normal case) or in
680 hexadecimal (if the first two characters of the operand are '0x').
681 Note that Jim Tcl does *not* treat numbers with leading zeros as octal.
683 If an operand does not have one of the integer formats given
684 above, then it is treated as a floating-point number if that is
685 possible. Floating-point numbers may be specified in any of the
686 ways accepted by an ANSI-compliant C compiler (except that the
687 'f', 'F', 'l', and 'L' suffixes will not be permitted in
688 most installations). For example, all of the
689 following are valid floating-point numbers: 2.1, 3., 6e4, 7.91e+16.
691 If no numeric interpretation is possible, then an operand is left
692 as a string (and only a limited set of operators may be applied to
695 1. Operands may be specified in any of the following ways:
697 2. As a numeric value, either integer or floating-point.
699 3. As a Tcl variable, using standard '$' notation.
700 The variable's value will be used as the operand.
702 4. As a string enclosed in double-quotes.
703 The expression parser will perform backslash, variable, and
704 command substitutions on the information between the quotes,
705 and use the resulting value as the operand
707 5. As a string enclosed in braces.
708 The characters between the open brace and matching close brace
709 will be used as the operand without any substitutions.
711 6. As a Tcl command enclosed in brackets.
712 The command will be executed and its result will be used as
715 Where substitutions occur above (e.g. inside quoted strings), they
716 are performed by the expression processor.
717 However, an additional layer of substitution may already have
718 been performed by the command parser before the expression
719 processor was called.
721 As discussed below, it is usually best to enclose expressions
722 in braces to prevent the command parser from performing substitutions
725 For some examples of simple expressions, suppose the variable 'a' has
726 the value 3 and the variable 'b' has the value 6. Then the expression
727 on the left side of each of the lines below will evaluate to the value
728 on the right side of the line:
733 {word one} < "word $a" 0
735 The valid operators are listed below, grouped in decreasing order
737 [[OperatorPrecedence]]
738 +int() double() round() abs(), rand(), srand()+::
739 Unary functions (except rand() which takes no arguments)
740 * +'int()'+ converts the numeric argument to an integer by truncating down.
741 * +'double()'+ converts the numeric argument to floating point.
742 * +'round()'+ converts the numeric argument to the closest integer value.
743 * +'abs()'+ takes the absolute value of the numeric argument.
744 * +'rand()'+ takes the absolute value of the numeric argument.
745 * +'rand()'+ returns a pseudo-random floating-point value in the range (0,1).
746 * +'srand()'+ takes an integer argument to (re)seed the random number generator. Returns the first random number from that seed.
748 +sin() cos() tan() asin() acos() atan() sinh() cosh() tanh() ceil() floor() exp() log() log10() sqrt()+::
749 Unary math functions.
750 If Jim is compiled with math support, these functions are available.
753 Unary minus, unary plus, bit-wise NOT, logical NOT. None of these operands
754 may be applied to string operands, and bit-wise NOT may be
755 applied only to integers.
758 Power. e.g. 'x^y^'. If Jim is compiled with math support, supports doubles and
759 integers. Otherwise supports integers only. (Note that the math-function form
760 has the same highest precedence)
763 Multiply, divide, remainder. None of these operands may be
764 applied to string operands, and remainder may be applied only
768 Add and subtract. Valid for any numeric operands.
771 Left and right shift, left and right rotate. Valid for integer operands only.
774 Boolean less, greater, less than or equal, and greater than or equal.
775 Each operator produces 1 if the condition is true, 0 otherwise.
776 These operators may be applied to strings as well as numeric operands,
777 in which case string comparison is used.
780 Boolean equal and not equal. Each operator produces a zero/one result.
781 Valid for all operand types. *Note* that values will be converted to integers
782 if possible, then floating point types, and finally strings will be compared.
783 It is recommended that 'eq' and 'ne' should be used for string comparison.
786 String equal and not equal. Uses the string value directly without
787 attempting to convert to a number first.
790 String in list and not in list. For 'in', result is 1 if the left operand (as a string)
791 is contained in the right operand (as a list), or 0 otherwise. The result for
792 +{$a ni $list}+ is equivalent to +{!($a in $list)}+.
795 Bit-wise AND. Valid for integer operands only.
798 Bit-wise OR. Valid for integer operands only.
801 Bit-wise exclusive OR. Valid for integer operands only.
804 Logical AND. Produces a 1 result if both operands are non-zero, 0 otherwise.
805 Valid for numeric operands only (integers or floating-point).
808 Logical OR. Produces a 0 result if both operands are zero, 1 otherwise.
809 Valid for numeric operands only (integers or floating-point).
812 If-then-else, as in C. If +'x'+
813 evaluates to non-zero, then the result is the value of +'y'+.
814 Otherwise the result is the value of +'z'+.
815 The +'x'+ operand must have a numeric value, while +'y'+ and +'z'+ can
818 See the C manual for more details on the results
819 produced by each operator.
820 All of the binary operators group left-to-right within the same
821 precedence level. For example, the expression
827 The +&&+, +||+, and +?:+ operators have 'lazy evaluation', just as
828 in C, which means that operands are not evaluated if they are not
829 needed to determine the outcome. For example, in
833 only one of +[a]+ or +[b]+ will actually be evaluated,
834 depending on the value of +$v+.
836 All internal computations involving integers are done with the C
837 type 'long long' if available, or 'long' otherwise, and all internal
838 computations involving floating-point are done with the C type
841 When converting a string to floating-point, exponent overflow is
842 detected and results in a Tcl error.
843 For conversion to integer from string, detection of overflow depends
844 on the behaviour of some routines in the local C library, so it should
845 be regarded as unreliable.
846 In any case, overflow and underflow are generally not detected
847 reliably for intermediate results.
849 Conversion among internal representations for integer, floating-point,
850 and string operands is done automatically as needed.
851 For arithmetic computations, integers are used until some
852 floating-point number is introduced, after which floating-point is used.
857 yields the result 1, while
860 5 / ( [string length "abcd"] + 0.0 )
862 both yield the result 1.25.
864 String values may be used as operands of the comparison operators,
865 although the expression evaluator tries to do comparisons as integer
866 or floating-point when it can.
867 If one of the operands of a comparison is a string and the other
868 has a numeric value, the numeric operand is converted back to
869 a string using the C 'sprintf' format specifier
870 '%d' for integers and '%g' for floating-point values.
871 For example, the expressions
876 both evaluate to 1. The first comparison is done using integer
877 comparison, and the second is done using string comparison after
878 the second operand is converted to the string '18'.
880 In general it is safest to enclose an expression in braces when
881 entering it in a command: otherwise, if the expression contains
882 any white space then the Tcl interpreter will split it
883 among several arguments. For example, the command
887 results in three arguments being passed to `expr`: +$a+,
888 \+, and +$b+. In addition, if the expression isn't in braces
889 then the Tcl interpreter will perform variable and command substitution
890 immediately (it will happen in the command parser rather than in
891 the expression parser). In many cases the expression is being
892 passed to a command that will evaluate the expression later (or
893 even many times if, for example, the expression is to be used to
894 decide when to exit a loop). Usually the desired goal is to re-do
895 the variable or command substitutions each time the expression is
896 evaluated, rather than once and for all at the beginning. For example,
899 for {set i 1} $i<=10 {incr i} {...} ** WRONG **
901 is probably intended to iterate over all values of +i+ from 1 to 10.
902 After each iteration of the body of the loop, `for` will pass
903 its second argument to the expression evaluator to see whether or not
904 to continue processing. Unfortunately, in this case the value of +i+
905 in the second argument will be substituted once and for all when the
906 `for` command is parsed. If +i+ was 0 before the `for`
907 command was invoked then the second argument of `for` will be +0\<=10+
908 which will always evaluate to 1, even though +i+ eventually
909 becomes greater than 10. In the above case the loop will never
910 terminate. Instead, the expression should be placed in braces:
912 for {set i 1} {$i<=10} {incr i} {...} ** RIGHT **
914 This causes the substitution of 'i'
915 to be delayed; it will be re-done each time the expression is
916 evaluated, which is the desired result.
920 The third major way that strings are interpreted in Tcl is as lists.
921 A list is just a string with a list-like structure
922 consisting of fields separated by white space. For example, the
927 is a list with four elements or fields.
928 Lists have the same basic structure as command strings, except
929 that a newline character in a list is treated as a field separator
930 just like space or tab. Conventions for braces and quotes
931 and backslashes are the same for lists as for commands. For example,
936 is a list with three elements: +a+, +b c+, and +d e {f g h}+.
938 Whenever an element is extracted from a list, the same rules about
939 braces and quotes and backslashes are applied as for commands. Thus in
940 the example above when the third element is extracted from the list,
945 (when the field was extracted, all that happened was to strip off
946 the outermost layer of braces). Command substitution and
947 variable substitution are never
948 made on a list (at least, not by the list-processing commands; the
949 list can always be passed to the Tcl interpreter for evaluation).
951 The Tcl commands `concat`, `foreach`, `lappend`, `lindex`, `linsert`,
952 `list`, `llength`, `lrange`, `lreplace`, `lsearch`, and `lsort` allow
953 you to build lists, extract elements from them, search them, and perform
954 other list-related functions.
956 Advanced list commands include `lrepeat`, `lreverse`, `lmap`, `lassign`, `lset`.
961 A new addition to Tcl 8.5 is the ability to expand a list into separate
962 arguments. Support for this feature is also available in Jim.
964 Consider the following attempt to exec a list:
969 This will attempt to exec the a command named "ls -l", which will clearly not
970 work. Typically eval and concat are required to solve this problem, however
971 it can be solved much more easily with +\{*\}+.
975 This will expand the following argument into individual elements and then evaluate
976 the resulting command.
978 Note that the official Tcl syntax is +\{*\}+, however +\{expand\}+ is retained
979 for backward compatibility with experimental versions of this feature.
983 Tcl provides two commands that support string matching using regular
984 expressions, `regexp` and `regsub`, as well as `switch -regexp` and
987 Regular expressions may be implemented one of two ways. Either using the system's C library
988 POSIX regular expression support, or using the built-in regular expression engine.
989 The differences between these are described below.
991 *NOTE* Tcl 7.x and 8.x use perl-style Advanced Regular Expressions (+ARE+).
993 POSIX Regular Expressions
994 ~~~~~~~~~~~~~~~~~~~~~~~~~
995 If the system supports POSIX regular expressions, and UTF-8 support is not enabled,
996 this support will be used by default. The type of regular expressions supported are
997 Extended Regular Expressions (+ERE+) rather than Basic Regular Expressions (+BRE+).
998 See REG_EXTENDED in the documentation.
1000 Using the system-supported POSIX regular expressions will typically
1001 make for the smallest code size, but some features such as UTF-8
1002 and +{backslash}w+, +{backslash}d+, +{backslash}s+ are not supported.
1004 See regex(3) and regex(7) for full details.
1006 Jim built-in Regular Expressions
1007 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1008 The Jim built-in regulare expression engine may be selected with +./configure --with-jim-regexp+
1009 or it will be selected automatically if UTF-8 support is enabled.
1011 This engine supports UTF-8 as well as some +ARE+ features. The differences with both Tcl 7.x/8.x
1012 and POSIX are highlighted below.
1014 1. UTF-8 strings and patterns are both supported
1015 2. Supported character classes: +[:alnum:]+, +[:digit:]+ and +[:space:]+
1016 3. Supported shorthand character classes: +{backslash}w+ = +[:alnum:]+, +{backslash}d+ = +[:digit:],+ +{backslash}s+ = +[:space:]+
1017 4. Character classes apply to ASCII characters only
1018 5. Supported constraint escapes: +{backslash}m+ = +{backslash}<+ = start of word, +{backslash}M+ = +{backslash}>+ = end of word
1019 6. Backslash escapes may be used within regular expressions, such as +{backslash}n+ = newline, +{backslash}uNNNN+ = unicode
1020 7. Support for the +?+ non-greedy quantifier. e.g. +*?+
1021 8. Support for non-capuring parentheses +(?:...)+
1025 Each command produces two results: a code and a string. The
1026 code indicates whether the command completed successfully or not,
1027 and the string gives additional information. The valid codes are
1028 defined in jim.h, and are:
1031 This is the normal return code, and indicates that the command completed
1032 successfully. The string gives the command's return value.
1035 Indicates that an error occurred; the string gives a message describing
1039 Indicates that the `return` command has been invoked, and that the
1040 current procedure (or top-level command or `source` command)
1041 should return immediately. The
1042 string gives the return value for the procedure or command.
1045 Indicates that the `break` command has been invoked, so the
1046 innermost loop should abort immediately. The string should always
1050 Indicates that the `continue` command has been invoked, so the
1051 innermost loop should go on to the next iteration. The string
1052 should always be empty.
1055 Indicates that a signal was caught while executing a commands.
1056 The string contains the name of the signal caught.
1057 See the `signal` and `catch` commands.
1060 Indicates that the command called the `exit` command.
1061 The string contains the exit code.
1063 Tcl programmers do not normally need to think about return codes,
1064 since +JIM_OK+ is almost always returned. If anything else is returned
1065 by a command, then the Tcl interpreter immediately stops processing
1066 commands and returns to its caller. If there are several nested
1067 invocations of the Tcl interpreter in progress, then each nested
1068 command will usually return the error to its caller, until eventually
1069 the error is reported to the top-level application code. The
1070 application will then display the error message for the user.
1072 In a few cases, some commands will handle certain `error` conditions
1073 themselves and not return them upwards. For example, the `for`
1074 command checks for the +JIM_BREAK+ code; if it occurs, then `for`
1075 stops executing the body of the loop and returns +JIM_OK+ to its
1076 caller. The `for` command also handles +JIM_CONTINUE+ codes and the
1077 procedure interpreter handles +JIM_RETURN+ codes. The `catch`
1078 command allows Tcl programs to catch errors and handle them without
1079 aborting command interpretation any further.
1081 The `info returncodes` command may be used to programmatically map between
1082 return codes and names.
1086 Tcl allows you to extend the command interface by defining
1087 procedures. A Tcl procedure can be invoked just like any other Tcl
1088 command (it has a name and it receives one or more arguments).
1089 The only difference is that its body isn't a piece of C code linked
1090 into the program; it is a string containing one or more other
1093 The `proc` command is used to create a new Tcl command procedure:
1095 +*proc* 'name arglist ?statics? body'+
1097 The new command is named +'name'+, and it replaces any existing command
1098 there may have been by that name. Whenever the new command is
1099 invoked, the contents of +'body'+ will be executed by the Tcl
1102 +'arglist'+ specifies the formal arguments to the procedure.
1103 It consists of a list, possibly empty, of the following
1104 argument specifiers:
1107 Required Argument - A simple argument name.
1110 Optional Argument - A two-element list consisting of the
1111 argument name, followed by the default value, which will
1112 be used if the corresponding argument is not supplied.
1115 Reference Argument - The caller is expected to pass the name of
1116 an existing variable. An implicit `upvar 1 'origname' 'name'` is done
1117 to make the variable available in the proc scope.
1120 Variable Argument - The special name +'args'+, which is
1121 assigned all remaining arguments (including none) as a list. The
1122 variable argument may only be specified once. Note that
1123 the syntax +args newname+ may be used to retain the special
1124 behaviour of +'args'+ with a different local name. In this case,
1125 the variable is named +'newname'+ rather than +'args'+.
1127 When the command is invoked, a local variable will be created for each of
1128 the formal arguments to the procedure; its value will be the value
1129 of corresponding argument in the invoking command or the argument's
1132 Arguments with default values need not be specified in a procedure
1133 invocation. However, there must be enough actual arguments for all
1134 required arguments, and there must not be any extra actual arguments
1135 (unless the Variable Argument is specified).
1137 Actual arguments are assigned to formal arguments as in left-to-right
1138 order with the following precedence.
1140 1. Required Arguments (including Reference Arguments)
1141 2. Optional Arguments
1142 3. Variable Argument
1144 The following example illustrates precedence. Assume a procedure declaration:
1146 proc p {{a A} args b {c C} d} {...}
1148 This procedure requires at least two arguments, but can accept an unlimited number.
1149 The following table shows how various numbers of arguments are assigned.
1150 Values marked as +-+ are assigned the default value.
1152 [width="40%",frame="topbot",options="header"]
1154 |Number of arguments|a|args|b|c|d
1162 When +'body'+ is being executed, variable names normally refer to local
1163 variables, which are created automatically when referenced and deleted
1164 when the procedure returns. One local variable is automatically created
1165 for each of the procedure's arguments. Global variables can be
1166 accessed by invoking the `global` command or via the +::+ prefix.
1170 In addition to procedure arguments, Jim procedures may declare static variables.
1171 These variables scoped to the procedure and initialised at procedure definition.
1172 Either from the static variable definition, or from the enclosing scope.
1174 Consider the following example:
1177 jim> proc a {} {a {b 2}} {
1189 The static variable +'a'+ has no initialiser, so it is initialised from
1190 the enclosing scope with the value 1. (Note that it is an error if there
1191 is no variable with the same name in the enclosing scope). However +'b'+
1192 has an initialiser, so it is initialised to 2.
1194 Unlike a local variable, the value of a static variable is retained across
1195 invocations of the procedure.
1197 See the `proc` command for information on how to define procedures
1198 and what happens when they are invoked. See also NAMESPACES.
1200 VARIABLES - SCALARS AND ARRAYS
1201 ------------------------------
1202 Tcl allows the definition of variables and the use of their values
1203 either through '$'-style variable substitution, the `set`
1204 command, or a few other mechanisms.
1206 Variables need not be declared: a new variable will automatically
1207 be created each time a new variable name is used.
1209 Tcl supports two types of variables: scalars and arrays.
1210 A scalar variable has a single value, whereas an array variable
1211 can have any number of elements, each with a name (called
1212 its 'index') and a value.
1214 Array indexes may be arbitrary strings; they need not be numeric.
1215 Parentheses are used refer to array elements in Tcl commands.
1216 For example, the command
1220 will modify the element of 'x' whose index is 'first'
1221 so that its new value is '44'.
1223 Two-dimensional arrays can be simulated in Tcl by using indexes
1224 that contain multiple concatenated values.
1225 For example, the commands
1230 set the elements of 'a' whose indexes are '2,3' and '3,6'.
1232 In general, array elements may be used anywhere in Tcl that scalar
1233 variables may be used.
1235 If an array is defined with a particular name, then there may
1236 not be a scalar variable with the same name.
1238 Similarly, if there is a scalar variable with a particular
1239 name then it is not possible to make array references to the
1242 To convert a scalar variable to an array or vice versa, remove
1243 the existing variable with the `unset` command.
1245 The `array` command provides several features for dealing
1246 with arrays, such as querying the names of all the elements of
1247 the array and converting between an array and a list.
1249 Variables may be either global or local. If a variable
1250 name is used when a procedure isn't being executed, then it
1251 automatically refers to a global variable. Variable names used
1252 within a procedure normally refer to local variables associated with that
1253 invocation of the procedure. Local variables are deleted whenever
1254 a procedure exits. Either `global` command may be used to request
1255 that a name refer to a global variable for the duration of the current
1256 procedure (this is somewhat analogous to 'extern' in C), or the variable
1257 may be explicitly scoped with the +::+ prefix. For example
1273 ARRAYS AS LISTS IN JIM
1274 ----------------------
1275 Unlike Tcl, Jim can automatically convert between a list (with an even
1276 number of elements) and an array value. This is similar to the way Tcl
1277 can convert between a string and a list.
1288 Thus `array set` is equivalent to `set` when the variable does not
1291 The reverse is also true where an array will be converted into
1294 set a(1) one; set a(2) two
1303 Tcl 8.5 introduced the dict command, and Jim Tcl has added a version
1304 of this command. Dictionaries provide efficient access to key-value
1305 pairs, just like arrays, but dictionaries are pure values. This
1306 means that you can pass them to a procedure just as a list or a
1307 string. Tcl dictionaries are therefore much more like Tcl lists,
1308 except that they represent a mapping from keys to values, rather
1309 than an ordered sequence.
1311 You can nest dictionaries, so that the value for a particular key
1312 consists of another dictionary. That way you can elegantly build
1313 complicated data structures, such as hierarchical databases. You
1314 can also combine dictionaries with other Tcl data structures. For
1315 instance, you can build a list of dictionaries that themselves
1318 Dictionaries are values that contain an efficient, order-preserving
1319 mapping from arbitrary keys to arbitrary values. Each key in the
1320 dictionary maps to a single value. They have a textual format that
1321 is exactly that of any list with an even number of elements, with
1322 each mapping in the dictionary being represented as two items in
1323 the list. When a command takes a dictionary and produces a new
1324 dictionary based on it (either returning it or writing it back into
1325 the variable that the starting dictionary was read from) the new
1326 dictionary will have the same order of keys, modulo any deleted
1327 keys and with new keys added on to the end. When a string is
1328 interpreted as a dictionary and it would otherwise have duplicate
1329 keys, only the last value for a particular key is used; the others
1330 are ignored, meaning that, "apple banana" and "apple carrot apple
1331 banana" are equivalent dictionaries (with different string
1334 Note that in Jim, arrays are implemented as dictionaries.
1335 Thus automatic conversion between lists and dictionaries applies
1336 as it does for arrays.
1338 jim> dict set a 1 one
1340 jim> dict set a 2 two
1346 jim> dict set a 3 T three
1347 1 one 2 two 3 {T three}
1349 See the `dict` command for more details.
1353 Tcl added namespaces as a mechanism avoiding name clashes, especially in applications
1354 including a number of 3rd party components. While there is less need for namespaces
1355 in Jim Tcl (which does not strive to support large applications), it is convenient to
1356 provide a subset of the support for namespaces to easy porting code from Tcl.
1358 Jim Tcl currently supports "light-weight" namespaces which should be adequate for most
1359 purposes. This feature is currently experimental. See README.namespaces for more information
1360 and the documentation of the `namespace` command.
1362 GARBAGE COLLECTION, REFERENCES, LAMBDA
1363 --------------------------------------
1364 Unlike Tcl, Jim has some sophisticated support for functional programming.
1365 These are described briefly below.
1367 More information may be found at http://wiki.tcl.tk/13847
1371 A reference can be thought of as holding a value with one level of indirection,
1372 where the value may be garbage collected when unreferenced.
1373 Consider the following example:
1375 jim> set r [ref "One String" test]
1376 <reference.<test___>.00000000000000000000>
1380 The operation `ref` creates a references to the value specified by the
1381 first argument. (The second argument is a "type" used for documentation purposes).
1383 The operation `getref` is the dereferencing operation which retrieves the value
1384 stored in the reference.
1386 jim> setref $r "New String"
1391 The operation `setref` replaces the value stored by the reference. If the old value
1392 is no longer accessible by any reference, it will eventually be automatically be garbage
1397 Normally, all values in Tcl are passed by value. As such values are copied and released
1398 automatically as necessary.
1400 With the introduction of references, it is possible to create values whose lifetime
1401 transcend their scope. To support this, case, the Jim system will periodically identify
1402 and discard objects which are no longer accessible by any reference.
1404 The `collect` command may be used to force garbage collection. Consider a reference created
1407 jim> proc f {ref value} { puts "Finaliser called for $ref,$value" }
1408 jim> set r [ref "One String" test f]
1409 <reference.<test___>.00000000000
1414 Finaliser called for <reference.<test___>.00000000000,One String
1417 Note that once the reference, 'r', was modified so that it no longer
1418 contained a reference to the value, the garbage collector discarded
1419 the value (after calling the finalizer).
1421 The finalizer for a reference may be examined or changed with the `finalize` command
1425 jim> finalize $r newf
1430 Jim provides a garbage collected lambda function. This is a procedure
1431 which is able to create an anonymous procedure. Consider:
1433 jim> set f [lambda {a} {{x 0}} { incr x $a }]
1440 This create an anonymous procedure (with the name stored in 'f'), with a static variable
1441 which is incremented by the supplied value and the result returned.
1443 Once the procedure name is no longer accessible, it will automatically be deleted
1444 when the garbage collector runs.
1446 The procedure may also be delete immediately by renaming it "". e.g.
1452 If Jim is built with UTF-8 support enabled (configure --enable-utf),
1453 then most string-related commands become UTF-8 aware. These include,
1454 but are not limited to, `string match`, `split`, `glob`, `scan` and
1457 UTF-8 encoding has many advantages, but one of the complications is that
1458 characters can take a variable number of bytes. Thus the addition of
1459 `string bytelength` which returns the number of bytes in a string,
1460 while `string length` returns the number of characters.
1462 If UTF-8 support is not enabled, all commands treat bytes as characters
1463 and `string bytelength` returns the same value as `string length`.
1465 Note that even if UTF-8 support is not enabled, the +{backslash}uNNNN+ and related syntax
1466 is still available to embed UTF-8 sequences.
1468 Jim Tcl supports all currently defined unicode codepoints. That is 21 bits, up to +'U+1FFFFF'.
1472 Commands such as `string match`, `lsearch -glob`, `array names` and others use string
1473 pattern matching rules. These commands support UTF-8. For example:
1475 string match a\[\ua0-\ubf\]b "a\u00a3b"
1479 +format %c+ allows a unicode codepoint to be be encoded. For example, the following will return
1480 a string with two bytes and one character. The same as +{backslash}ub5+
1484 `format` respects widths as character widths, not byte widths. For example, the following will
1485 return a string with three characters, not three bytes.
1487 format %.3s \ub5\ub6\ub7\ub8
1489 Similarly, +scan ... %c+ allows a UTF-8 to be decoded to a unicode codepoint. The following will set
1490 +'a'+ to 181 (0xb5) and +'b'+ to 65 (0x41).
1492 scan \u00b5A %c%c a b
1494 `scan %s` will also accept a character class, including unicode ranges.
1498 `string is` has *not* been extended to classify UTF-8 characters. Therefore, the following
1499 will return 0, even though the string may be considered to be alphabetic.
1501 string is alpha \ub5Test
1503 This does not affect the string classes 'ascii', 'control', 'digit', 'double', 'integer' or 'xdigit'.
1505 Case Mapping and Conversion
1506 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
1507 Jim provides a simplified unicode case mapping. This means that case conversion
1508 and comparison will not increase or decrease the number of characters in a string.
1509 (Although it may change the number of bytes).
1511 `string toupper` will convert any lowercase letters to their uppercase equivalent.
1512 Any character which is not a letter or has no uppercase equivalent is left unchanged.
1513 Similarly for `string tolower` and `string totitle`.
1515 Commands which perform case insensitive matches, such as `string compare -nocase`
1516 and `lsearch -nocase` fold both strings to uppercase before comparison.
1518 Invalid UTF-8 Sequences
1519 ~~~~~~~~~~~~~~~~~~~~~~~
1520 Some UTF-8 character sequences are invalid, such as those beginning with '0xff',
1521 those which represent character sequences longer than 3 bytes (greater than U+FFFF),
1522 and those which end prematurely, such as a lone '0xc2'.
1524 In these situations, the offending bytes are treated as single characters. For example,
1525 the following returns 2.
1527 string bytelength \xff\xff
1531 If UTF-8 support is enabled, the built-in regular expression engine will be
1532 selected which supports UTF-8 strings and patterns.
1534 See REGULAR EXPRESSIONS
1538 The Tcl library provides the following built-in commands, which will
1539 be available in any application using Tcl. In addition to these
1540 built-in commands, there may be additional commands defined by each
1541 application, plus commands defined as Tcl procedures.
1543 In the command syntax descriptions below, words in +*boldface*+ are
1544 literals that you type verbatim to Tcl.
1546 Words in +'italics'+ are meta-symbols; they serve as names for any of
1547 a range of values that you can type.
1549 Optional arguments or groups of arguments are indicated by enclosing them
1550 in +?question-marks?+.
1552 Ellipses (+\...+) indicate that any number of additional
1553 arguments or groups of arguments may appear, in the same format
1554 as the preceding argument(s).
1565 Delivers the +SIGALRM+ signal to the process after the given
1566 number of seconds. If the platform supports 'ualarm(3)' then
1567 the argument may be a floating point value. Otherwise it must
1570 Note that unless a signal handler for +SIGALRM+ has been installed
1571 (see `signal`), the process will exit on this signal.
1575 +*alias* 'name args\...'+
1577 Creates a single word alias (command) for one or more words. For example,
1578 the following creates an alias for the command `info exists`.
1585 `alias` returns +'name'+, allowing it to be used with `local`.
1587 See also `proc`, `curry`, `lambda`, `local`, `info alias`, `exists -alias`
1591 +*append* 'varName value ?value value ...?'+
1593 Append all of the +'value'+ arguments to the current value
1594 of variable +'varName'+. If +'varName'+ doesn't exist,
1595 it is given a value equal to the concatenation of all the
1596 +'value'+ arguments.
1598 This command provides an efficient way to build up long
1599 variables incrementally.
1600 For example, "`append a $b`" is much more efficient than
1601 "`set a $a$b`" if +$a+ is long.
1605 +*apply* 'lambdaExpr ?arg1 arg2 \...?'+
1607 The command `apply` provides for anonymous procedure calls,
1608 similar to `lambda`, but without command name being created, even temporarily.
1610 The function +'lambdaExpr'+ is a two element list +{args body}+
1611 or a three element list +{args body namespace}+. The first element
1612 args specifies the formal arguments, in the same form as the `proc` and `lambda` commands.
1616 +*array* 'option arrayName ?arg\...?'+
1618 This command performs one of several operations on the
1619 variable given by +'arrayName'+.
1621 Note that in general, if the named array does not exist, the +'array'+ command behaves
1622 as though the array exists but is empty.
1624 The +'option'+ argument determines what action is carried out by the
1625 command. The legal +'options'+ (which may be abbreviated) are:
1627 +*array exists* 'arrayName'+::
1628 Returns 1 if arrayName is an array variable, 0 if there is
1629 no variable by that name. This command is essentially
1630 identical to `info exists`
1632 +*array get* 'arrayName ?pattern?'+::
1633 Returns a list containing pairs of elements. The first
1634 element in each pair is the name of an element in arrayName
1635 and the second element of each pair is the value of the
1636 array element. The order of the pairs is undefined. If
1637 pattern is not specified, then all of the elements of the
1638 array are included in the result. If pattern is specified,
1639 then only those elements whose names match pattern (using
1640 the matching rules of string match) are included. If arrayName
1641 isn't the name of an array variable, or if the array contains
1642 no elements, then an empty list is returned.
1644 +*array names* 'arrayName ?pattern?'+::
1645 Returns a list containing the names of all of the elements
1646 in the array that match pattern. If pattern is omitted then
1647 the command returns all of the element names in the array.
1648 If pattern is specified, then only those elements whose
1649 names match pattern (using the matching rules of string
1650 match) are included. If there are no (matching) elements
1651 in the array, or if arrayName isn't the name of an array
1652 variable, then an empty string is returned.
1654 +*array set* 'arrayName list'+::
1655 Sets the values of one or more elements in arrayName. list
1656 must have a form like that returned by array get, consisting
1657 of an even number of elements. Each odd-numbered element
1658 in list is treated as an element name within arrayName, and
1659 the following element in list is used as a new value for
1660 that array element. If the variable arrayName does not
1661 already exist and list is empty, arrayName is created with
1662 an empty array value.
1664 +*array size* 'arrayName'+::
1665 Returns the number of elements in the array. If arrayName
1666 isn't the name of an array then 0 is returned.
1668 +*array unset* 'arrayName ?pattern?'+::
1669 Unsets all of the elements in the array that match pattern
1670 (using the matching rules of string match). If arrayName
1671 isn't the name of an array variable or there are no matching
1672 elements in the array, no error will be raised. If pattern
1673 is omitted and arrayName is an array variable, then the
1674 command unsets the entire array. The command always returns
1681 This command may be invoked only inside the body of a loop command
1682 such as `for` or `foreach` or `while`. It returns a +JIM_BREAK+ code
1683 to signal the innermost containing loop command to return immediately.
1687 +*case* 'string' ?in? 'patList body ?patList body ...?'+
1689 +*case* 'string' ?in? {'patList body ?patList body ...?'}+
1691 *Note* that the `switch` command should generally be preferred unless compatibility
1692 with Tcl 6.x is desired.
1694 Match +'string'+ against each of the +'patList'+ arguments
1695 in order. If one matches, then evaluate the following +'body'+ argument
1696 by passing it recursively to the Tcl interpreter, and return the result
1697 of that evaluation. Each +'patList'+ argument consists of a single
1698 pattern or list of patterns. Each pattern may contain any of the wild-cards
1699 described under `string match`.
1701 If a +'patList'+ argument is +default+, the corresponding body will be
1702 evaluated if no +'patList'+ matches +'string'+. If no +'patList'+ argument
1703 matches +'string'+ and no default is given, then the `case` command returns
1706 Two syntaxes are provided.
1708 The first uses a separate argument for each of the patterns and commands;
1709 this form is convenient if substitutions are desired on some of the
1710 patterns or commands.
1712 The second form places all of the patterns and commands together into
1713 a single argument; the argument must have proper list structure, with
1714 the elements of the list being the patterns and commands.
1716 The second form makes it easy to construct multi-line case commands,
1717 since the braces around the whole list make it unnecessary to include a
1718 backslash at the end of each line.
1720 Since the +'patList'+ arguments are in braces in the second form,
1721 no command or variable substitutions are performed on them; this makes
1722 the behaviour of the second form different than the first form in some
1725 Below are some examples of `case` commands:
1727 case abc in {a b} {format 1} default {format 2} a* {format 3}
1737 will return '1', and
1752 +*catch* ?-?no?'code \...'? ?--? 'command ?resultVarName? ?optionsVarName?'+
1754 The `catch` command may be used to prevent errors from aborting
1755 command interpretation. `catch` evaluates +'command'+, and returns a
1756 +JIM_OK+ code, regardless of any errors that might occur while
1757 executing +'command'+ (with the possible exception of +JIM_SIGNAL+ -
1760 The return value from `catch` is a decimal string giving the code
1761 returned by the Tcl interpreter after executing +'command'+. This
1762 will be '0' (+JIM_OK+) if there were no errors in +'command'+; otherwise
1763 it will have a non-zero value corresponding to one of the exceptional
1764 return codes (see jim.h for the definitions of code values, or the
1765 `info returncodes` command).
1767 If the +'resultVarName'+ argument is given, then it gives the name
1768 of a variable; `catch` will set the value of the variable to the
1769 string returned from +'command'+ (either a result or an error message).
1771 If the +'optionsVarName'+ argument is given, then it gives the name
1772 of a variable; `catch` will set the value of the variable to a
1773 dictionary. For any return code other than +JIM_RETURN+, the value
1774 for the key +-code+ will be set to the return code. For +JIM_RETURN+
1775 it will be set to the code given in `return -code`. Additionally,
1776 for the return code +JIM_ERR+, the value of the key +-errorinfo+
1777 will contain the current stack trace (the same result as `info stacktrace`),
1778 the value of the key +-errorcode+ will contain the
1779 same value as the global variable $::errorCode, and the value of
1780 the key +-level+ will be the current return level (see `return -level`).
1781 This can be useful to rethrow an error:
1783 if {[catch {...} msg opts]} {
1784 ...maybe do something with the error...
1786 return {*}$opts $msg
1789 Normally `catch` will +'not'+ catch any of the codes +JIM_EXIT+, +JIM_EVAL+ or +JIM_SIGNAL+.
1790 The set of codes which will be caught may be modified by specifying the one more codes before
1793 e.g. To catch +JIM_EXIT+ but not +JIM_BREAK+ or +JIM_CONTINUE+
1795 catch -exit -nobreak -nocontinue -- { ... }
1797 The use of +--+ is optional. It signifies that no more return code options follow.
1799 Note that if a signal marked as `signal handle` is caught with `catch -signal`, the return value
1800 (stored in +'resultVarName'+) is name of the signal caught.
1806 Change the current working directory to +'dirName'+.
1808 Returns an empty string.
1810 This command can potentially be disruptive to an application, so it may
1811 be removed in some applications.
1816 Returns the current time as seconds since the epoch.
1818 +*clock format* 'seconds' ?*-format* 'format?'+::
1819 Format the given time (seconds since the epoch) according to the given
1820 format. See strftime(3) for supported formats.
1821 If no format is supplied, "%c" is used.
1823 +*clock scan* 'str' *-format* 'format'+::
1824 Scan the given time string using the given format string.
1825 See strptime(3) for supported formats.
1833 Closes the file given by +'fileId'+.
1834 +'fileId'+ must be the return value from a previous invocation
1835 of the `open` command; after this command, it should not be
1842 Normally reference garbage collection is automatically performed periodically.
1843 However it may be run immediately with the `collect` command.
1845 See GARBAGE COLLECTION, REFERENCES, LAMBDA for more detail.
1849 +*concat* 'arg ?arg \...?'+
1851 This command treats each argument as a list and concatenates them
1852 into a single list. It permits any number of arguments. For example,
1855 concat a b {c d e} {f {g h}}
1867 This command may be invoked only inside the body of a loop command such
1868 as `for` or `foreach` or `while`. It returns a +JIM_CONTINUE+ code to
1869 signal the innermost containing loop command to skip the remainder of
1870 the loop's body but continue with the next iteration of the loop.
1874 +*alias* 'args\...'+
1876 Similar to `alias` except it creates an anonymous procedure (lambda) instead of
1879 the following creates a local, unnamed alias for the command `info exists`.
1881 set e [local curry info exists]
1886 `curry` returns the name of the procedure.
1888 See also `proc`, `alias`, `lambda`, `local`.
1892 +*dict* 'option ?arg\...?'+
1894 Performs one of several operations on dictionary values.
1896 The +'option'+ argument determines what action is carried out by the
1897 command. The legal +'options'+ are:
1899 +*dict create* '?key value \...?'+::
1900 Create and return a new dictionary value that contains each of
1901 the key/value mappings listed as arguments (keys and values
1902 alternating, with each key being followed by its associated
1905 +*dict exists* 'dictionary key ?key \...?'+::
1906 Returns a boolean value indicating whether the given key (or path
1907 of keys through a set of nested dictionaries) exists in the given
1908 dictionary value. This returns a true value exactly when `dict get`
1909 on that path will succeed.
1911 +*dict get* 'dictionary ?key \...?'+::
1912 Given a dictionary value (first argument) and a key (second argument),
1913 this will retrieve the value for that key. Where several keys are
1914 supplied, the behaviour of the command shall be as if the result
1915 of "`dict get $dictVal $key`" was passed as the first argument to
1916 dict get with the remaining arguments as second (and possibly
1917 subsequent) arguments. This facilitates lookups in nested dictionaries.
1918 If no keys are provided, dict would return a list containing pairs
1919 of elements in a man- ner similar to array get. That is, the first
1920 element of each pair would be the key and the second element would
1921 be the value for that key. It is an error to attempt to retrieve
1922 a value for a key that is not present in the dictionary.
1924 +*dict keys* 'dictionary ?pattern?'+::
1925 Returns a list of the keys in the dictionary.
1926 If pattern is specified, then only those keys whose
1927 names match +'pattern'+ (using the matching rules of string
1928 match) are included.
1930 +*dict merge* ?'dictionary \...'?+::
1931 Return a dictionary that contains the contents of each of the
1932 +'dictionary'+ arguments. Where two (or more) dictionaries
1933 contain a mapping for the same key, the resulting dictionary
1934 maps that key to the value according to the last dictionary on
1935 the command line containing a mapping for that key.
1937 +*dict set* 'dictionaryName key ?key \...? value'+::
1938 This operation takes the +'name'+ of a variable containing a dictionary
1939 value and places an updated dictionary value in that variable
1940 containing a mapping from the given key to the given value. When
1941 multiple keys are present, this operation creates or updates a chain
1942 of nested dictionaries.
1944 +*dict size* 'dictionary'+::
1945 Return the number of key/value mappings in the given dictionary value.
1947 +*dict unset* 'dictionaryName key ?key \...? value'+::
1948 This operation (the companion to `dict set`) takes the name of a
1949 variable containing a dictionary value and places an updated
1950 dictionary value in that variable that does not contain a mapping
1951 for the given key. Where multiple keys are present, this describes
1952 a path through nested dictionaries to the mapping to remove. At
1953 least one key must be specified, but the last key on the key-path
1954 need not exist. All other components on the path must exist.
1956 +*dict with* 'dictionaryName key ?key \...? script'+::
1957 Execute the Tcl script in +'script'+ with the value for each
1958 key in +'dictionaryName'+ mapped to a variable with the same
1959 name. Where one or more keys are given, these indicate a chain
1960 of nested dictionaries, with the innermost dictionary being the
1961 one opened out for the execution of body. Making +'dictionaryName'+
1962 unreadable will make the updates to the dictionary be discarded,
1963 and this also happens if the contents of +'dictionaryName'+ are
1964 adjusted so that the chain of dictionaries no longer exists.
1965 The result of `dict with` is (unless some kind of error occurs)
1966 the result of the evaluation of body.
1968 The variables are mapped in the scope enclosing the `dict with`;
1969 it is recommended that this command only be used in a local
1970 scope (procedure). Because of this, the variables set by
1971 `dict with` will continue to exist after the command finishes (unless
1972 explicitly unset). Note that changes to the contents of +'dictionaryName'+
1973 only happen when +'script'+ terminates.
1975 +*dict for, values, incr, append, lappend, update, info, replace*+ to be documented...
1979 +*env* '?name? ?default?'+
1981 If +'name'+ is supplied, returns the value of +'name'+ from the initial
1982 environment (see getenv(3)). An error is returned if +'name'+ does not
1983 exist in the environment, unless +'default'+ is supplied - in which case
1984 that value is returned instead.
1986 If no arguments are supplied, returns a list of all environment variables
1987 and their values as +{name value \...}+
1989 See also the global variable +::env+
1997 Returns 1 if an end-of-file condition has occurred on +'fileId'+,
2000 +'fileId'+ must have been the return value from a previous call to `open`,
2001 or it may be +stdin+, +stdout+, or +stderr+ to refer to one of the
2002 standard I/O channels.
2006 +*error* 'message ?stacktrace?'+
2008 Returns a +JIM_ERR+ code, which causes command interpretation to be
2009 unwound. +'message'+ is a string that is returned to the application
2010 to indicate what went wrong.
2012 If the +'stacktrace'+ argument is provided and is non-empty,
2013 it is used to initialize the stacktrace.
2015 This feature is most useful in conjunction with the `catch` command:
2016 if a caught error cannot be handled successfully, +'stacktrace'+ can be used
2017 to return a stack trace reflecting the original point of occurrence
2022 error $errMsg [info stacktrace]
2024 See also `errorInfo`, `info stacktrace`, `catch` and `return`
2028 +*errorInfo* 'error ?stacktrace?'+
2030 Returns a human-readable representation of the given error message and stack trace.
2033 if {[catch {...} error]} {
2034 puts stderr [errorInfo $error [info stacktrace]]
2042 +*eval* 'arg ?arg\...?'+
2044 `eval` takes one or more arguments, which together comprise a Tcl
2045 command (or collection of Tcl commands separated by newlines in the
2046 usual way). `eval` concatenates all its arguments in the same
2047 fashion as the `concat` command, passes the concatenated string to the
2048 Tcl interpreter recursively, and returns the result of that
2049 evaluation (or any error generated by it).
2053 +*exec* 'arg ?arg\...?'+
2055 This command treats its arguments as the specification
2056 of one or more UNIX commands to execute as subprocesses.
2057 The commands take the form of a standard shell pipeline;
2058 +|+ arguments separate commands in the
2059 pipeline and cause standard output of the preceding command
2060 to be piped into standard input of the next command (or +|&+ for
2061 both standard output and standard error).
2063 Under normal conditions the result of the `exec` command
2064 consists of the standard output produced by the last command
2067 If any of the commands in the pipeline exit abnormally or
2068 are killed or suspended, then `exec` will return an error
2069 and the error message will include the pipeline's output followed by
2070 error messages describing the abnormal terminations.
2072 If any of the commands writes to its standard error file,
2073 then `exec` will return an error, and the error message
2074 will include the pipeline's output, followed by messages
2075 about abnormal terminations (if any), followed by the standard error
2078 If the last character of the result or error message
2079 is a newline then that character is deleted from the result
2080 or error message for consistency with normal
2083 An +'arg'+ may have one of the following special forms:
2086 The standard output of the last command in the pipeline
2087 is redirected to the file. In this situation `exec`
2088 will normally return an empty string.
2091 As above, but append to the file.
2094 The standard output of the last command in the pipeline is
2095 redirected to the given (writable) file descriptor (e.g. stdout,
2096 stderr, or the result of `open`). In this situation `exec`
2097 will normally return an empty string.
2100 The standard error of the last command in the pipeline
2101 is redirected to the file.
2104 As above, but append to the file.
2107 The standard error of the last command in the pipeline is
2108 redirected to the given (writable) file descriptor.
2111 The standard error of the last command in the pipeline is
2112 redirected to the same file descriptor as the standard output.
2115 Both the standard output and standard error of the last command
2116 in the pipeline is redirected to the file.
2119 As above, but append to the file.
2122 The standard input of the first command in the pipeline
2123 is taken from the file.
2126 The standard input of the first command is taken as the
2127 given immediate value.
2130 The standard input of the first command in the pipeline
2131 is taken from the given (readable) file descriptor.
2133 If there is no redirection of standard input, standard error
2134 or standard output, these are connected to the corresponding
2135 input or output of the application.
2137 If the last +'arg'+ is +&+ then the command will be
2138 executed in background.
2139 In this case the standard output from the last command
2140 in the pipeline will
2141 go to the application's standard output unless
2142 redirected in the command, and error output from all
2143 the commands in the pipeline will go to the application's
2144 standard error file. The return value of exec in this case
2145 is a list of process ids (pids) in the pipeline.
2147 Each +'arg'+ becomes one word for a command, except for
2148 +|+, +<+, +<<+, +>+, and +&+ arguments, and the
2149 arguments that follow +<+, +<<+, and +>+.
2151 The first word in each command is taken as the command name;
2152 the directories in the PATH environment variable are searched for
2153 an executable by the given name.
2155 No `glob` expansion or other shell-like substitutions
2156 are performed on the arguments to commands.
2158 If the command fails, the global $::errorCode (and the -errorcode
2159 option in `catch`) will be set to a list, as follows:
2161 +*CHILDKILLED* 'pid sigName msg'+::
2162 This format is used when a child process has been killed
2163 because of a signal. The pid element will be the process's
2164 identifier (in decimal). The sigName element will be the
2165 symbolic name of the signal that caused the process to
2166 terminate; it will be one of the names from the include
2167 file signal.h, such as SIGPIPE. The msg element will be a
2168 short human-readable message describing the signal, such
2169 as "write on pipe with no readers" for SIGPIPE.
2171 +*CHILDSUSP* 'pid sigName msg'+::
2172 This format is used when a child process has been suspended
2173 because of a signal. The pid element will be the process's
2174 identifier, in decimal. The sigName element will be the
2175 symbolic name of the signal that caused the process to
2176 suspend; this will be one of the names from the include
2177 file signal.h, such as SIGTTIN. The msg element will be a
2178 short human-readable message describing the signal, such
2179 as "background tty read" for SIGTTIN.
2181 +*CHILDSTATUS* 'pid code'+::
2182 This format is used when a child process has exited with a
2183 non-zero exit status. The pid element will be the process's
2184 identifier (in decimal) and the code element will be the
2185 exit code returned by the process (also in decimal).
2187 The environment for the executed command is set from $::env (unless
2188 this variable is unset, in which case the original environment is used).
2192 +*exists ?-var|-proc|-command|-alias?* 'name'+
2194 Checks the existence of the given variable, procedure, command
2195 or alias respectively and returns 1 if it exists or 0 if not. This command
2196 provides a more simplified/convenient version of `info exists`,
2197 `info procs` and `info commands`.
2199 If the type is omitted, a type of '-var' is used. The type may be abbreviated.
2203 +*exit* '?returnCode?'+
2205 Terminate the process, returning +'returnCode'+ to the
2206 parent as the exit status.
2208 If +'returnCode'+ isn't specified then it defaults
2211 Note that exit can be caught with `catch`.
2217 Calls the expression processor to evaluate +'arg'+, and returns
2218 the result as a string. See the section EXPRESSIONS above.
2220 Note that Jim supports a shorthand syntax for `expr` as +$(\...)+
2221 The following two are identical.
2223 set x [expr {3 * 2 + 1}]
2228 +*file* 'option name ?arg\...?'+
2230 Operate on a file or a file name. +'name'+ is the name of a file.
2232 +'option'+ indicates what to do with the file name. Any unique
2233 abbreviation for +'option'+ is acceptable. The valid options are:
2235 +*file atime* 'name'+::
2236 Return a decimal string giving the time at which file +'name'+
2237 was last accessed. The time is measured in the standard UNIX
2238 fashion as seconds from a fixed starting time (often January 1, 1970).
2239 If the file doesn't exist or its access time cannot be queried then an
2242 +*file copy ?-force?* 'source target'+::
2243 Copies file +'source'+ to file +'target'+. The source file must exist.
2244 The target file must not exist, unless +-force+ is specified.
2246 +*file delete ?-force?* 'name\...'+::
2247 Deletes file or directory +'name'+. If the file or directory doesn't exist, nothing happens.
2248 If it can't be deleted, an error is generated. Non-empty directories will not be deleted
2249 unless the +-force+ options is given. In this case no errors will be generated, even
2250 if the file/directory can't be deleted.
2252 +*file dirname* 'name'+::
2253 Return all of the characters in +'name'+ up to but not including
2254 the last slash character. If there are no slashes in +'name'+
2255 then return +.+ (a single dot). If the last slash in +'name'+ is its first
2256 character, then return +/+.
2258 +*file executable* 'name'+::
2259 Return '1' if file +'name'+ is executable by
2260 the current user, '0' otherwise.
2262 +*file exists* 'name'+::
2263 Return '1' if file +'name'+ exists and the current user has
2264 search privileges for the directories leading to it, '0' otherwise.
2266 +*file extension* 'name'+::
2267 Return all of the characters in +'name'+ after and including the
2268 last dot in +'name'+. If there is no dot in +'name'+ then return
2271 +*file isdirectory* 'name'+::
2272 Return '1' if file +'name'+ is a directory,
2275 +*file isfile* 'name'+::
2276 Return '1' if file +'name'+ is a regular file,
2279 +*file join* 'arg\...'+::
2280 Joins multiple path components. Note that if any components is
2281 an absolute path, the preceding components are ignored.
2282 Thus +"`file` join /tmp /root"+ returns +"/root"+.
2284 +*file lstat* 'name varName'+::
2285 Same as 'stat' option (see below) except uses the +'lstat'+
2286 kernel call instead of +'stat'+. This means that if +'name'+
2287 refers to a symbolic link the information returned in +'varName'+
2288 is for the link rather than the file it refers to. On systems that
2289 don't support symbolic links this option behaves exactly the same
2290 as the 'stat' option.
2292 +*file mkdir* 'dir1 ?dir2\...?'+::
2293 Creates each directory specified. For each pathname +'dir'+ specified,
2294 this command will create all non-existing parent directories
2295 as well as +'dir'+ itself. If an existing directory is specified,
2296 then no action is taken and no error is returned. Trying to
2297 overwrite an existing file with a directory will result in an
2298 error. Arguments are processed in the order specified, halting
2299 at the first error, if any.
2301 +*file mtime* 'name ?time?'+::
2302 Return a decimal string giving the time at which file +'name'+
2303 was last modified. The time is measured in the standard UNIX
2304 fashion as seconds from a fixed starting time (often January 1, 1970).
2305 If the file doesn't exist or its modified time cannot be queried then an
2306 error is generated. If +'time'+ is given, sets the modification time
2307 of the file to the given value.
2309 +*file normalize* 'name'+::
2310 Return the normalized path of +'name'+. See 'realpath(3)'.
2312 +*file owned* 'name'+::
2313 Return '1' if file +'name'+ is owned by the current user,
2316 +*file readable* 'name'+::
2317 Return '1' if file +'name'+ is readable by
2318 the current user, '0' otherwise.
2320 +*file readlink* 'name'+::
2321 Returns the value of the symbolic link given by +'name'+ (i.e. the
2322 name of the file it points to). If
2323 +'name'+ isn't a symbolic link or its value cannot be read, then
2324 an error is returned. On systems that don't support symbolic links
2325 this option is undefined.
2327 +*file rename* 'oldname' 'newname'+::
2328 Renames the file from the old name to the new name.
2330 +*file rootname* 'name'+::
2331 Return all of the characters in +'name'+ up to but not including
2332 the last '.' character in the name. If +'name'+ doesn't contain
2333 a dot, then return +'name'+.
2335 +*file size* 'name'+::
2336 Return a decimal string giving the size of file +'name'+ in bytes.
2337 If the file doesn't exist or its size cannot be queried then an
2340 +*file stat* 'name ?varName?'+::
2341 Invoke the 'stat' kernel call on +'name'+, and return the result
2342 as a dictionary with the following keys: 'atime',
2343 'ctime', 'dev', 'gid', 'ino', 'mode', 'mtime',
2344 'nlink', 'size', 'type', 'uid'.
2345 Each element except 'type' is a decimal string with the value of
2346 the corresponding field from the 'stat' return structure; see the
2347 manual entry for 'stat' for details on the meanings of the values.
2348 The 'type' element gives the type of the file in the same form
2349 returned by the command `file type`.
2350 If +'varName'+ is specified, it is taken to be the name of an array
2351 variable and the values are also stored into the array.
2353 +*file tail* 'name'+::
2354 Return all of the characters in +'name'+ after the last slash.
2355 If +'name'+ contains no slashes then return +'name'+.
2357 +*file tempfile* '?template?'+::
2358 Creates and returns the name of a unique temporary file. If +'template'+ is omitted, a
2359 default template will be used to place the file in /tmp. See 'mkstemp(3)' for
2360 the format of the template and security concerns.
2362 +*file type* 'name'+::
2363 Returns a string giving the type of file +'name'+, which will be
2364 one of +file+, +directory+, +characterSpecial+,
2365 +blockSpecial+, +fifo+, +link+, or +socket+.
2367 +*file writable* 'name'+::
2368 Return '1' if file +'name'+ is writable by
2369 the current user, '0' otherwise.
2371 The `file` commands that return 0/1 results are often used in
2372 conditional or looping commands, for example:
2374 if {![file exists foo]} {
2375 error {bad file name}
2382 +*finalize* 'reference ?command?'+
2384 If +'command'+ is omitted, returns the finalizer command for the given reference.
2386 Otherwise, sets a new finalizer command for the given reference. +'command'+ may be
2387 the empty string to remove the current finalizer.
2389 The reference must be a valid reference create with the `ref`
2392 See GARBAGE COLLECTION, REFERENCES, LAMBDA for more detail.
2400 Flushes any output that has been buffered for +'fileId'+. +'fileId'+ must
2401 have been the return value from a previous call to `open`, or it may be
2402 +stdout+ or +stderr+ to access one of the standard I/O streams; it must
2403 refer to a file that was opened for writing. This command returns an
2408 +*for* 'start test next body'+
2410 `for` is a looping command, similar in structure to the C `for` statement.
2411 The +'start'+, +'next'+, and +'body'+ arguments must be Tcl command strings,
2412 and +'test'+ is an expression string.
2414 The `for` command first invokes the Tcl interpreter to execute +'start'+.
2415 Then it repeatedly evaluates +'test'+ as an expression; if the result is
2416 non-zero it invokes the Tcl interpreter on +'body'+, then invokes the Tcl
2417 interpreter on +'next'+, then repeats the loop. The command terminates
2418 when +'test'+ evaluates to 0.
2420 If a `continue` command is invoked within +'body'+ then any remaining
2421 commands in the current execution of +'body'+ are skipped; processing
2422 continues by invoking the Tcl interpreter on +'next'+, then evaluating
2423 +'test'+, and so on.
2425 If a `break` command is invoked within +'body'+ or +'next'+, then the `for`
2426 command will return immediately.
2428 The operation of `break` and `continue` are similar to the corresponding
2431 `for` returns an empty string.
2435 +*foreach* 'varName list body'+
2437 +*foreach* 'varList list ?varList2 list2 \...? body'+
2439 In this command, +'varName'+ is the name of a variable, +'list'+
2440 is a list of values to assign to +'varName'+, and +'body'+ is a
2441 collection of Tcl commands.
2443 For each field in +'list'+ (in order from left to right), `foreach` assigns
2444 the contents of the field to +'varName'+ (as if the `lindex` command
2445 had been used to extract the field), then calls the Tcl interpreter to
2448 If instead of being a simple name, +'varList'+ is used, multiple assignments
2449 are made each time through the loop, one for each element of +'varList'+.
2451 For example, if there are two elements in +'varList'+ and six elements in
2452 the list, the loop will be executed three times.
2454 If the length of the list doesn't evenly divide by the number of elements
2455 in +'varList'+, the value of the remaining variables in the last iteration
2456 of the loop are undefined.
2458 The `break` and `continue` statements may be invoked inside +'body'+,
2459 with the same effect as in the `for` command.
2461 `foreach` returns an empty string.
2465 +*format* 'formatString ?arg \...?'+
2467 This command generates a formatted string in the same way as the
2468 C 'sprintf' procedure (it uses 'sprintf' in its
2469 implementation). +'formatString'+ indicates how to format
2470 the result, using +%+ fields as in 'sprintf', and the additional
2471 arguments, if any, provide values to be substituted into the result.
2473 All of the 'sprintf' options are valid; see the 'sprintf'
2474 man page for details. Each +'arg'+ must match the expected type
2475 from the +%+ field in +'formatString'+; the `format` command
2476 converts each argument to the correct type (floating, integer, etc.)
2477 before passing it to 'sprintf' for formatting.
2479 The only unusual conversion is for +%c+; in this case the argument
2480 must be a decimal string, which will then be converted to the corresponding
2481 ASCII (or UTF-8) character value.
2483 In addition, Jim Tcl provides basic support for conversion to binary with +%b+.
2485 `format` does backslash substitution on its +'formatString'+
2486 argument, so backslash sequences in +'formatString'+ will be handled
2487 correctly even if the argument is in braces.
2489 The return value from `format` is the formatted string.
2493 +*getref* 'reference'+
2495 Returns the string associated with +'reference'+. The reference must
2496 be a valid reference create with the `ref` command.
2498 See GARBAGE COLLECTION, REFERENCES, LAMBDA for more detail.
2502 +*gets* 'fileId ?varName?'+
2504 +'fileId' *gets* '?varName?'+
2506 Reads the next line from the file given by +'fileId'+ and discards
2507 the terminating newline character.
2509 If +'varName'+ is specified, then the line is placed in the variable
2510 by that name and the return value is a count of the number of characters
2511 read (not including the newline).
2513 If the end of the file is reached before reading
2514 any characters then -1 is returned and +'varName'+ is set to an
2517 If +'varName'+ is not specified then the return value will be
2518 the line (minus the newline character) or an empty string if
2519 the end of the file is reached before reading any characters.
2521 An empty string will also be returned if a line contains no characters
2522 except the newline, so `eof` may have to be used to determine
2523 what really happened.
2525 If the last character in the file is not a newline character, then
2526 `gets` behaves as if there were an additional newline character
2527 at the end of the file.
2529 +'fileId'+ must be +stdin+ or the return value from a previous
2530 call to `open`; it must refer to a file that was opened
2535 +*glob* ?*-nocomplain*? ?*-directory* 'dir'? ?*--*? 'pattern ?pattern \...?'+
2537 This command performs filename globbing, using csh rules. The returned
2538 value from `glob` is the list of expanded filenames.
2540 If +-nocomplain+ is specified as the first argument then an empty
2541 list may be returned; otherwise an error is returned if the expanded
2542 list is empty. The +-nocomplain+ argument must be provided
2543 exactly: an abbreviation will not be accepted.
2545 If +-directory+ is given, the +'dir'+ is understood to contain a
2546 directory name to search in. This allows globbing inside directories
2547 whose names may contain glob-sensitive characters. The returned names
2548 are specified relative to this directory.
2553 +*global* 'varName ?varName \...?'+
2555 This command is ignored unless a Tcl procedure is being interpreted.
2556 If so, then it declares each given +'varName'+ to be a global variable
2557 rather than a local one. For the duration of the current procedure
2558 (and only while executing in the current procedure), any reference to
2559 +'varName'+ will be bound to a global variable instead
2562 An alternative to using `global` is to use the +::+ prefix
2563 to explicitly name a variable in the global scope.
2567 +*if* 'expr1' ?*then*? 'body1' *elseif* 'expr2' ?*then*? 'body2' *elseif* \... ?*else*? ?'bodyN'?+
2569 The `if` command evaluates +'expr1'+ as an expression (in the same way
2570 that `expr` evaluates its argument). The value of the expression must
2571 be numeric; if it is non-zero then +'body1'+ is executed by passing it to
2572 the Tcl interpreter.
2574 Otherwise +'expr2'+ is evaluated as an expression and if it is non-zero
2575 then +'body2'+ is executed, and so on.
2577 If none of the expressions evaluates to non-zero then +'bodyN'+ is executed.
2579 The +then+ and +else+ arguments are optional "noise words" to make the
2580 command easier to read.
2582 There may be any number of +elseif+ clauses, including zero. +'bodyN'+
2583 may also be omitted as long as +else+ is omitted too.
2585 The return value from the command is the result of the body script that
2586 was executed, or an empty string if none of the expressions was non-zero
2587 and there was no +'bodyN'+.
2591 +*incr* 'varName ?increment?'+
2593 Increment the value stored in the variable whose name is +'varName'+.
2594 The value of the variable must be integral.
2596 If +'increment'+ is supplied then its value (which must be an
2597 integer) is added to the value of variable +'varName'+; otherwise
2598 1 is added to +'varName'+.
2600 The new value is stored as a decimal string in variable +'varName'+
2601 and also returned as result.
2603 If the variable does not exist, the variable is implicitly created
2604 and set to +0+ first.
2609 +*info* 'option ?arg\...?'+::
2611 Provide information about various internals to the Tcl interpreter.
2612 The legal +'option'+'s (which may be abbreviated) are:
2614 +*info args* 'procname'+::
2615 Returns a list containing the names of the arguments to procedure
2616 +'procname'+, in order. +'procname'+ must be the name of a
2617 Tcl command procedure.
2619 +*info alias* 'command'+::
2620 +'command'+ must be an alias created with `alias`. In which case the target
2621 command and arguments, as passed to `alias` are returned. See `exists -alias`
2623 +*info body* 'procname'+::
2624 Returns the body of procedure +'procname'+. +'procname'+ must be
2625 the name of a Tcl command procedure.
2628 Returns a list of all open file handles from `open` or `socket`
2630 +*info commands* ?'pattern'?+::
2631 If +'pattern'+ isn't specified, returns a list of names of all the
2632 Tcl commands, including both the built-in commands written in C and
2633 the command procedures defined using the `proc` command.
2634 If +'pattern'+ is specified, only those names matching +'pattern'+
2635 are returned. Matching is determined using the same rules as for
2638 +*info complete* 'command' ?'missing'?+::
2639 Returns 1 if +'command'+ is a complete Tcl command in the sense of
2640 having no unclosed quotes, braces, brackets or array element names,
2641 If the command doesn't appear to be complete then 0 is returned.
2642 This command is typically used in line-oriented input environments
2643 to allow users to type in commands that span multiple lines; if the
2644 command isn't complete, the script can delay evaluating it until additional
2645 lines have been typed to complete the command. If +'varName'+ is specified, the
2646 missing character is stored in the variable with that name.
2648 +*info exists* 'varName'+::
2649 Returns '1' if the variable named +'varName'+ exists in the
2650 current context (either as a global or local variable), returns '0'
2653 +*info frame* ?'number'?+::
2654 If +'number'+ is not specified, this command returns a number
2655 which is the same result as `info level` - the current stack frame level.
2656 If +'number'+ is specified, then the result is a list consisting of the procedure,
2657 filename and line number for the procedure call at level +'number'+ on the stack.
2658 If +'number'+ is positive then it selects a particular stack level (1 refers
2659 to the top-most active procedure, 2 to the procedure it called, and
2660 so on); otherwise it gives a level relative to the current level
2661 (0 refers to the current procedure, -1 to its caller, and so on).
2662 The level has an identical meaning to `info level`.
2664 +*info globals* ?'pattern'?+::
2665 If +'pattern'+ isn't specified, returns a list of all the names
2666 of currently-defined global variables.
2667 If +'pattern'+ is specified, only those names matching +'pattern'+
2668 are returned. Matching is determined using the same rules as for
2672 An alias for `os.gethostname` for compatibility with Tcl 6.x
2674 +*info level* ?'number'?+::
2675 If +'number'+ is not specified, this command returns a number
2676 giving the stack level of the invoking procedure, or 0 if the
2677 command is invoked at top-level. If +'number'+ is specified,
2678 then the result is a list consisting of the name and arguments for the
2679 procedure call at level +'number'+ on the stack. If +'number'+
2680 is positive then it selects a particular stack level (1 refers
2681 to the top-most active procedure, 2 to the procedure it called, and
2682 so on); otherwise it gives a level relative to the current level
2683 (0 refers to the current procedure, -1 to its caller, and so on).
2684 See the `uplevel` command for more information on what stack
2687 +*info locals* ?'pattern'?+::
2688 If +'pattern'+ isn't specified, returns a list of all the names
2689 of currently-defined local variables, including arguments to the
2690 current procedure, if any. Variables defined with the `global`
2691 and `upvar` commands will not be returned. If +'pattern'+ is
2692 specified, only those names matching +'pattern'+ are returned.
2693 Matching is determined using the same rules as for `string match`.
2695 +*info nameofexecutable*+::
2696 Returns the name of the binary file from which the application
2697 was invoked. A full path will be returned, unless the path
2698 can't be determined, in which case the empty string will be returned.
2700 +*info procs* ?'pattern'?+::
2701 If +'pattern'+ isn't specified, returns a list of all the
2702 names of Tcl command procedures.
2703 If +'pattern'+ is specified, only those names matching +'pattern'+
2704 are returned. Matching is determined using the same rules as for
2707 +*info references*+::
2708 Returns a list of all references which have not yet been garbage
2711 +*info returncodes* ?'code'?+::
2712 Returns a list representing the mapping of standard return codes
2713 to names. e.g. +{0 ok 1 error 2 return \...}+. If a code is given,
2714 instead returns the name for the given code.
2717 If a Tcl script file is currently being evaluated (i.e. there is a
2718 call to 'Jim_EvalFile' active or there is an active invocation
2719 of the `source` command), then this command returns the name
2720 of the innermost file being processed. Otherwise the command returns an
2723 +*info source* 'script'+::
2724 Returns the original source location of the given script as a list of
2725 +{filename linenumber}+. If the source location can't be determined, the
2726 list +{{} 0}+ is returned.
2728 +*info stacktrace*+::
2729 After an error is caught with `catch`, returns the stack trace as a list
2730 of +{procedure filename line \...}+.
2732 +*info statics* 'procname'+::
2733 Returns a dictionary of the static variables of procedure
2734 +'procname'+. +'procname'+ must be the name of a Tcl command
2735 procedure. An empty dictionary is returned if the procedure has
2736 no static variables.
2739 Returns the version number for this version of Jim in the form +*x.yy*+.
2741 +*info vars* ?'pattern'?+::
2742 If +'pattern'+ isn't specified,
2743 returns a list of all the names of currently-visible variables, including
2744 both locals and currently-visible globals.
2745 If +'pattern'+ is specified, only those names matching +'pattern'+
2746 are returned. Matching is determined using the same rules as for
2751 +*join* 'list ?joinString?'+
2753 The +'list'+ argument must be a valid Tcl list. This command returns the
2754 string formed by joining all of the elements of +'list'+ together with
2755 +'joinString'+ separating each adjacent pair of elements.
2757 The +'joinString'+ argument defaults to a space character.
2761 +*kill* ?'SIG'|*-0*? 'pid'+
2763 Sends the given signal to the process identified by +'pid'+.
2765 The signal may be specified by name or number in one of the following forms:
2773 The signal name may be in either upper or lower case.
2775 The special signal name +-0+ simply checks that a signal +'could'+ be sent.
2777 If no signal is specified, SIGTERM is used.
2779 An error is raised if the signal could not be delivered.
2783 +*lambda* 'args ?statics? body'+
2785 The `lambda` command is identical to `proc`, except rather than
2786 creating a named procedure, it creates an anonymous procedure and returns
2787 the name of the procedure.
2789 See `proc` and GARBAGE COLLECTION, REFERENCES, LAMBDA for more detail.
2793 +*lappend* 'varName value ?value value \...?'+
2795 Treat the variable given by +'varName'+ as a list and append each of
2796 the +'value'+ arguments to that list as a separate element, with spaces
2799 If +'varName'+ doesn't exist, it is created as a list with elements given
2800 by the +'value'+ arguments. `lappend` is similar to `append` except that
2801 each +'value'+ is appended as a list element rather than raw text.
2803 This command provides a relatively efficient way to build up large lists.
2808 is much more efficient than
2810 set a [concat $a [list $b]]
2816 +*lassign* 'list varName ?varName \...?'+
2818 This command treats the value +'list'+ as a list and assigns successive elements from that list to
2819 the variables given by the +'varName'+ arguments in order. If there are more variable names than
2820 list elements, the remaining variables are set to the empty string. If there are more list ele-
2821 ments than variables, a list of unassigned elements is returned.
2823 jim> lassign {1 2 3} a b; puts a=$a,b=$b
2829 +*local* 'cmd ?arg\...?'+
2831 First, `local` evaluates +'cmd'+ with the given arguments. The return value must
2832 be the name of an existing command, which is marked as having local scope.
2833 This means that when the current procedure exits, the specified
2834 command is deleted. This can be useful with `lambda`, local procedures or
2835 to automatically close a filehandle.
2837 In addition, if a command already exists with the same name,
2838 the existing command will be kept rather than deleted, and may be called
2839 via `upcall`. The previous command will be restored when the current
2840 procedure exits. See `upcall` for more details.
2842 In this example, a local procedure is created. Note that the procedure
2843 continues to have global scope while it is active.
2846 # proc ... returns "inner" which is marked local
2847 local proc inner {} {
2848 # will be deleted when 'outer' exits
2855 In this example, the lambda is deleted at the end of the procedure rather
2856 than waiting until garbage collection.
2859 set x [lambda inner {args} {
2860 # will be deleted when 'outer' exits
2862 # Use 'function' here which simply returns $x
2871 +*loop* 'var first limit ?incr? body'+
2873 Similar to `for` except simpler and possibly more efficient.
2874 With a positive increment, equivalent to:
2876 for {set var $first} {$var < $limit} {incr var $incr} $body
2878 If +'incr'+ is not specified, 1 is used.
2879 Note that setting the loop variable inside the loop does not
2880 affect the loop count.
2884 +*lindex* 'list index'+
2886 Treats +'list'+ as a Tcl list and returns element +'index'+ from it
2887 (0 refers to the first element of the list).
2888 See STRING AND LIST INDEX SPECIFICATIONS for all allowed forms for +'index'+.
2890 In extracting the element, +'lindex'+ observes the same rules concerning
2891 braces and quotes and backslashes as the Tcl command interpreter; however,
2892 variable substitution and command substitution do not occur.
2894 If +'index'+ is negative or greater than or equal to the number of elements
2895 in +'value'+, then an empty string is returned.
2899 +*linsert* 'list index element ?element element \...?'+
2901 This command produces a new list from +'list'+ by inserting all
2902 of the +'element'+ arguments just before the element +'index'+
2903 of +'list'+. Each +'element'+ argument will become
2904 a separate element of the new list. If +'index'+ is less than
2905 or equal to zero, then the new elements are inserted at the
2906 beginning of the list. If +'index'+ is greater than or equal
2907 to the number of elements in the list, then the new elements are
2908 appended to the list.
2910 See STRING AND LIST INDEX SPECIFICATIONS for all allowed forms for +'index'+.
2915 +*list* 'arg ?arg \...?'+
2917 This command returns a list comprised of all the arguments, +'arg'+. Braces
2918 and backslashes get added as necessary, so that the `lindex` command
2919 may be used on the result to re-extract the original arguments, and also
2920 so that `eval` may be used to execute the resulting list, with
2921 +'arg1'+ comprising the command's name and the other args comprising
2922 its arguments. `list` produces slightly different results than
2923 `concat`: `concat` removes one level of grouping before forming
2924 the list, while `list` works directly from the original arguments.
2925 For example, the command
2927 list a b {c d e} {f {g h}}
2931 a b {c d e} {f {g h}}
2933 while `concat` with the same arguments will return
2941 Treats +'list'+ as a list and returns a decimal string giving
2942 the number of elements in it.
2946 +*lset* 'varName ?index ..? newValue'+
2948 Sets an element in a list.
2950 The `lset` command accepts a parameter, +'varName'+, which it interprets
2951 as the name of a variable containing a Tcl list. It also accepts
2952 zero or more indices into the list. Finally, it accepts a new value
2953 for an element of varName. If no indices are presented, the command
2956 lset varName newValue
2958 In this case, newValue replaces the old value of the variable
2961 When presented with a single index, the `lset` command
2962 treats the content of the varName variable as a Tcl list. It addresses
2963 the index'th element in it (0 refers to the first element of the
2964 list). When interpreting the list, `lset` observes the same rules
2965 concerning braces and quotes and backslashes as the Tcl command
2966 interpreter; however, variable substitution and command substitution
2967 do not occur. The command constructs a new list in which the
2968 designated element is replaced with newValue. This new list is
2969 stored in the variable varName, and is also the return value from
2972 If index is negative or greater than or equal to the number of
2973 elements in $varName, then an error occurs.
2975 See STRING AND LIST INDEX SPECIFICATIONS for all allowed forms for +'index'+.
2977 If additional index arguments are supplied, then each argument is
2978 used in turn to address an element within a sublist designated by
2979 the previous indexing operation, allowing the script to alter
2980 elements in sublists. The command,
2984 replaces element 2 of sublist 1 with +'newValue'+.
2986 The integer appearing in each index argument must be greater than
2987 or equal to zero. The integer appearing in each index argument must
2988 be strictly less than the length of the corresponding list. In other
2989 words, the `lset` command cannot change the size of a list. If an
2990 index is outside the permitted range, an error is reported.
2995 +*lmap* 'varName list body'+
2997 +*lmap* 'varList list ?varList2 list2 \...? body'+
2999 `lmap` is a "collecting" `foreach` which returns a list of its results.
3003 jim> lmap i {1 2 3 4 5} {expr $i*$i}
3005 jim> lmap a {1 2 3} b {A B C} {list $a $b}
3008 If the body invokes `continue`, no value is added for this iteration.
3009 If the body invokes `break`, the loop ends and no more values are added.
3015 Loads the dynamic extension, +'filename'+. Generally the filename should have
3016 the extension +.so+. The initialisation function for the module must be based
3017 on the name of the file. For example loading +hwaccess.so+ will invoke
3018 the initialisation function, +Jim_hwaccessInit+. Normally the `load` command
3019 should not be used directly. Instead it is invoked automatically by `package require`.
3023 +*lrange* 'list first last'+
3025 +'list'+ must be a valid Tcl list. This command will return a new
3026 list consisting of elements +'first'+ through +'last'+, inclusive.
3028 See STRING AND LIST INDEX SPECIFICATIONS for all allowed forms for +'first'+ and +'last'+.
3030 If +'last'+ is greater than or equal to the number of elements
3031 in the list, then it is treated as if it were +end+.
3033 If +'first'+ is greater than +'last'+ then an empty string
3036 Note: +"`lrange` 'list first first'"+ does not always produce the
3037 same result as +"`lindex` 'list first'"+ (although it often does
3038 for simple fields that aren't enclosed in braces); it does, however,
3039 produce exactly the same results as +"`list` [`lindex` 'list first']"+
3044 +*lreplace* 'list first last ?element element \...?'+
3046 Returns a new list formed by replacing one or more elements of
3047 +'list'+ with the +'element'+ arguments.
3049 +'first'+ gives the index in +'list'+ of the first element
3052 If +'first'+ is less than zero then it refers to the first
3053 element of +'list'+; the element indicated by +'first'+
3054 must exist in the list.
3056 +'last'+ gives the index in +'list'+ of the last element
3057 to be replaced; it must be greater than or equal to +'first'+.
3059 See STRING AND LIST INDEX SPECIFICATIONS for all allowed forms for +'first'+ and +'last'+.
3061 The +'element'+ arguments specify zero or more new arguments to
3062 be added to the list in place of those that were deleted.
3064 Each +'element'+ argument will become a separate element of
3067 If no +'element'+ arguments are specified, then the elements
3068 between +'first'+ and +'last'+ are simply deleted.
3072 +*lrepeat* 'number element1 ?element2 \...?'+
3074 Build a list by repeating elements +'number'+ times (which must be
3075 a positive integer).
3084 Returns the list in reverse order.
3086 jim> lreverse {1 2 3}
3091 +*lsearch* '?options? list pattern'+
3093 This command searches the elements +'list'+ to see if one of them matches +'pattern'+. If so, the
3094 command returns the index of the first matching element (unless the options +-all+, +-inline+ or +-bool+ are
3095 specified.) If not, the command returns -1. The option arguments indicates how the elements of
3096 the list are to be matched against pattern and must have one of the values below:
3098 *Note* that this command is different from Tcl in that default match type is +-exact+ rather than +-glob+.
3101 +'pattern'+ is a literal string that is compared for exact equality against each list element.
3102 This is the default.
3105 +'pattern'+ is a glob-style pattern which is matched against each list element using the same
3106 rules as the string match command.
3109 +'pattern'+ is treated as a regular expression and matched against each list element using
3110 the rules described by `regexp`.
3112 +*-command* 'cmdname'+::
3113 +'cmdname'+ is a command which is used to match the pattern against each element of the
3114 list. It is invoked as +'cmdname' ?*-nocase*? 'pattern listvalue'+ and should return 1
3115 for a match, or 0 for no match.
3118 Changes the result to be the list of all matching indices (or all matching values if
3119 +-inline+ is specified as well). If indices are returned, the indices will be in numeric
3120 order. If values are returned, the order of the values will be the order of those values
3121 within the input list.
3124 The matching value is returned instead of its index (or an empty string if no value
3125 matches). If +-all+ is also specified, then the result of the command is the list of all
3126 values that matched. The +-inline+ and +-bool+ options are mutually exclusive.
3129 Changes the result to '1' if a match was found, or '0' otherwise. If +-all+ is also specified,
3130 the result will be a list of '0' and '1' for each element of the list depending upon whether
3131 the corresponding element matches. The +-inline+ and +-bool+ options are mutually exclusive.
3134 This negates the sense of the match, returning the index (or value
3135 if +-inline+ is specified) of the first non-matching value in the
3136 list. If +-bool+ is also specified, the '0' will be returned if a
3137 match is found, or '1' otherwise. If +-all+ is also specified,
3138 non-matches will be returned rather than matches.
3141 Causes comparisons to be handled in a case-insensitive manner.
3145 +*lsort* ?*-index* 'listindex'? ?*-nocase!-integer|-real|-command* 'cmdname'? ?*-unique*? ?*-decreasing*|*-increasing*? 'list'+
3147 Sort the elements of +'list'+, returning a new list in sorted order.
3148 By default, ASCII (or UTF-8) sorting is used, with the result in increasing order.
3150 If +-nocase+ is specified, comparisons are case-insenstive.
3152 If +-integer+ is specified, numeric sorting is used.
3154 If +-real+ is specified, floating point number sorting is used.
3156 If +-command 'cmdname'+ is specified, +'cmdname'+ is treated as a command
3157 name. For each comparison, +'cmdname $value1 $value2+' is called which
3158 should compare the values and return an integer less than, equal
3159 to, or greater than zero if the +'$value1'+ is to be considered less
3160 than, equal to, or greater than +'$value2'+, respectively.
3162 If +-decreasing+ is specified, the resulting list is in the opposite
3163 order to what it would be otherwise. +-increasing+ is the default.
3165 If +-unique+ is specified, then only the last set of duplicate elements found in the list will be retained.
3166 Note that duplicates are determined relative to the comparison used in the sort. Thus if +-index 0+ is used,
3167 +{1 a}+ and +{1 b}+ would be considered duplicates and only the second element, +{1 b}+, would be retained.
3169 If +-index 'listindex'+ is specified, each element of the list is treated as a list and
3170 the given index is extracted from the list for comparison. The list index may
3171 be any valid list index, such as +1+, +end+ or +end-2+.
3175 +*open* 'fileName ?access?'+
3177 +*open* '|command-pipeline ?access?'+
3179 Opens a file and returns an identifier
3180 that may be used in future invocations
3181 of commands like `read`, `puts`, and `close`.
3182 +'fileName'+ gives the name of the file to open.
3184 The +'access'+ argument indicates the way in which the file is to be accessed.
3185 It may have any of the following values:
3188 Open the file for reading only; the file must already exist.
3191 Open the file for both reading and writing; the file must
3195 Open the file for writing only. Truncate it if it exists. If it doesn't
3196 exist, create a new file.
3199 Open the file for reading and writing. Truncate it if it exists.
3200 If it doesn't exist, create a new file.
3203 Open the file for writing only. The file must already exist, and the file
3204 is positioned so that new data is appended to the file.
3207 Open the file for reading and writing. If the file doesn't
3208 exist, create a new empty file. Set the initial access position
3209 to the end of the file.
3211 +'access'+ defaults to 'r'.
3213 If a file is opened for both reading and writing, then `seek`
3214 must be invoked between a read and a write, or vice versa.
3216 If the first character of +'fileName'+ is "|" then the remaining
3217 characters of +'fileName'+ are treated as a list of arguments that
3218 describe a command pipeline to invoke, in the same style as the
3219 arguments for exec. In this case, the channel identifier returned
3220 by open may be used to write to the command's input pipe or read
3221 from its output pipe, depending on the value of +'access'+. If write-only
3222 access is used (e.g. +'access'+ is 'w'), then standard output for the
3223 pipeline is directed to the current standard output unless overridden
3224 by the command. If read-only access is used (e.g. +'access'+ is r),
3225 standard input for the pipeline is taken from the current standard
3226 input unless overridden by the command.
3228 The `pid` command may be used to return the process ids of the commands
3229 forming the command pipeline.
3231 See also `socket`, `pid`, `exec`
3235 +*package provide* 'name ?version?'+
3237 Indicates that the current script provides the package named +'name'+.
3238 If no version is specified, '1.0' is used.
3240 Any script which provides a package may include this statement
3241 as the first statement, although it is not required.
3243 +*package require* 'name ?version?'*+
3245 Searches for the package with the given +'name'+ by examining each path
3246 in '$::auto_path' and trying to load '$path/$name.so' as a dynamic extension,
3247 or '$path/$name.tcl' as a script package.
3249 The first such file which is found is considered to provide the the package.
3250 (The version number is ignored).
3252 If '$name.so' exists, it is loaded with the `load` command,
3253 otherwise if '$name.tcl' exists it is loaded with the `source` command.
3255 If `load` or `source` fails, `package require` will fail immediately.
3256 No further attempt will be made to locate the file.
3264 The first form returns the process identifier of the current process.
3266 The second form accepts a handle returned by `open` and returns a list
3267 of the process ids forming the pipeline in the same form as `exec ... &`.
3268 If 'fileId' represents a regular file handle rather than a command pipeline,
3269 the empty string is returned instead.
3271 See also `open`, `exec`
3275 +*proc* 'name args ?statics? body'+
3277 The `proc` command creates a new Tcl command procedure, +'name'+.
3278 When the new command is invoked, the contents of +'body'+ will be executed.
3279 Tcl interpreter. +'args'+ specifies the formal arguments to the procedure.
3280 If specified, +'static'+, declares static variables which are bound to the
3283 See PROCEDURES for detailed information about Tcl procedures.
3285 The `proc` command returns +'name'+ (which is useful with `local`).
3287 When a procedure is invoked, the procedure's return value is the
3288 value specified in a `return` command. If the procedure doesn't
3289 execute an explicit `return`, then its return value is the value
3290 of the last command executed in the procedure's body.
3292 If an error occurs while executing the procedure body, then the
3293 procedure-as-a-whole will return that same error.
3297 +*puts* ?*-nonewline*? '?fileId? string'+
3299 +'fileId' *puts* ?*-nonewline*? 'string'+
3301 Writes the characters given by +'string'+ to the file given
3302 by +'fileId'+. +'fileId'+ must have been the return
3303 value from a previous call to `open`, or it may be
3304 +stdout+ or +stderr+ to refer to one of the standard I/O
3305 channels; it must refer to a file that was opened for
3308 In the first form, if no +'fileId'+ is specified then it defaults to +stdout+.
3309 `puts` normally outputs a newline character after +'string'+,
3310 but this feature may be suppressed by specifying the +-nonewline+
3313 Output to files is buffered internally by Tcl; the `flush`
3314 command may be used to force buffered characters to be output.
3320 Returns the path name of the current working directory.
3324 +*rand* '?min? ?max?'+
3326 Returns a random integer between +'min'+ (defaults to 0) and +'max'+
3327 (defaults to the maximum integer).
3329 If only one argument is given, it is interpreted as +'max'+.
3333 +*range* '?start? end ?step?'+
3335 Returns a list of integers starting at +'start'+ (defaults to 0)
3336 and ranging up to but not including +'end'+ in steps of +'step'+ defaults to 1).
3349 +*read* ?*-nonewline*? 'fileId'+
3351 +'fileId' *read* ?*-nonewline*?+
3353 +*read* 'fileId numBytes'+
3355 +'fileId' *read* 'numBytes'+
3358 In the first form, all of the remaining bytes are read from the file
3359 given by +'fileId'+; they are returned as the result of the command.
3360 If the +-nonewline+ switch is specified then the last
3361 character of the file is discarded if it is a newline.
3363 In the second form, the extra argument specifies how many bytes to read;
3364 exactly this many bytes will be read and returned, unless there are fewer than
3365 +'numBytes'+ bytes left in the file; in this case, all the remaining
3368 +'fileId'+ must be +stdin+ or the return value from a previous call
3369 to `open`; it must refer to a file that was opened for reading.
3373 +*regexp ?-nocase? ?-line? ?-indices? ?-start* 'offset'? *?-all? ?-inline? ?--?* 'exp string ?matchVar? ?subMatchVar subMatchVar \...?'+
3375 Determines whether the regular expression +'exp'+ matches part or
3376 all of +'string'+ and returns 1 if it does, 0 if it doesn't.
3378 See REGULAR EXPRESSIONS above for complete information on the
3379 syntax of +'exp'+ and how it is matched against +'string'+.
3381 If additional arguments are specified after +'string'+ then they
3382 are treated as the names of variables to use to return
3383 information about which part(s) of +'string'+ matched +'exp'+.
3384 +'matchVar'+ will be set to the range of +'string'+ that
3385 matched all of +'exp'+. The first +'subMatchVar'+ will contain
3386 the characters in +'string'+ that matched the leftmost parenthesized
3387 subexpression within +'exp'+, the next +'subMatchVar'+ will
3388 contain the characters that matched the next parenthesized
3389 subexpression to the right in +'exp'+, and so on.
3391 Normally, +'matchVar'+ and the each +'subMatchVar'+ are set to hold the
3392 matching characters from `string`, however see +-indices+ and
3395 If there are more values for +'subMatchVar'+ than parenthesized subexpressions
3396 within +'exp'+, or if a particular subexpression in +'exp'+ doesn't
3397 match the string (e.g. because it was in a portion of the expression
3398 that wasn't matched), then the corresponding +'subMatchVar'+ will be
3399 set to +"-1 -1"+ if +-indices+ has been specified or to an empty
3402 The following switches modify the behaviour of +'regexp'+
3405 Causes upper-case and lower-case characters to be treated as
3406 identical during the matching process.
3409 Use newline-sensitive matching. By default, newline
3410 is a completely ordinary character with no special meaning in
3411 either REs or strings. With this flag, +[^+ bracket expressions
3412 and +.+ never match newline, a +^+ anchor matches the null
3413 string after any newline in the string in addition to its normal
3414 function, and the +$+ anchor matches the null string before any
3415 newline in the string in addition to its normal function.
3418 Changes what is stored in the subMatchVars. Instead of
3419 storing the matching characters from string, each variable
3420 will contain a list of two decimal strings giving the indices
3421 in string of the first and last characters in the matching
3422 range of characters.
3424 +*-start* 'offset'+::
3425 Specifies a character index offset into the string at which to start
3426 matching the regular expression. If +-indices+ is
3427 specified, the indices will be indexed starting from the
3428 absolute beginning of the input string. +'offset'+ will be
3429 constrained to the bounds of the input string.
3432 Causes the regular expression to be matched as many times as possible
3433 in the string, returning the total number of matches found. If this
3434 is specified with match variables, they will contain information
3435 for the last match only.
3438 Causes the command to return, as a list, the data that would otherwise
3439 be placed in match variables. When using +-inline+, match variables
3440 may not be specified. If used with +-all+, the list will be concatenated
3441 at each iteration, such that a flat list is always returned. For
3442 each match iteration, the command will append the overall match
3443 data, plus one element for each subexpression in the regular
3447 Marks the end of switches. The argument following this one will be
3448 treated as +'exp'+ even if it starts with a +-+.
3452 +*regsub ?-nocase? ?-all? ?-line? ?-start* 'offset'? ?*--*? 'exp string subSpec ?varName?'+
3454 This command matches the regular expression +'exp'+ against
3455 +'string'+ using the rules described in REGULAR EXPRESSIONS
3458 If +'varName'+ is specified, the commands stores +'string'+ to +'varName'+
3459 with the substitutions detailed below, and returns the number of
3460 substitutions made (normally 1 unless +-all+ is specified).
3461 This is 0 if there were no matches.
3463 If +'varName'+ is not specified, the substituted string will be returned
3466 When copying +'string'+, the portion of +'string'+ that
3467 matched +'exp'+ is replaced with +'subSpec'+.
3468 If +'subSpec'+ contains a +&+ or +{backslash}0+, then it is replaced
3469 in the substitution with the portion of +'string'+ that
3472 If +'subSpec'+ contains a +{backslash}n+, where +'n'+ is a digit
3473 between 1 and 9, then it is replaced in the substitution with
3474 the portion of +'string'+ that matched the +'n'+\'-th
3475 parenthesized subexpression of +'exp'+.
3476 Additional backslashes may be used in +'subSpec'+ to prevent special
3477 interpretation of +&+ or +{backslash}0+ or +{backslash}n+ or
3480 The use of backslashes in +'subSpec'+ tends to interact badly
3481 with the Tcl parser's use of backslashes, so it's generally
3482 safest to enclose +'subSpec'+ in braces if it includes
3485 The following switches modify the behaviour of +'regsub'+
3488 Upper-case characters in +'string'+ are converted to lower-case
3489 before matching against +'exp'+; however, substitutions
3490 specified by +'subSpec'+ use the original unconverted form
3494 All ranges in +'string'+ that match +'exp'+ are found and substitution
3495 is performed for each of these ranges, rather than only the
3496 first. The +&+ and +{backslash}n+ sequences are handled for
3497 each substitution using the information from the corresponding
3501 Use newline-sensitive matching. By default, newline
3502 is a completely ordinary character with no special meaning in
3503 either REs or strings. With this flag, +[^+ bracket expressions
3504 and +.+ never match newline, a +^+ anchor matches the null
3505 string after any newline in the string in addition to its normal
3506 function, and the +$+ anchor matches the null string before any
3507 newline in the string in addition to its normal function.
3509 +*-start* 'offset'+::
3510 Specifies a character index offset into the string at which to
3511 start matching the regular expression. +'offset'+ will be
3512 constrained to the bounds of the input string.
3515 Marks the end of switches. The argument following this one will be
3516 treated as +'exp'+ even if it starts with a +-+.
3520 +*ref* 'string tag ?finalizer?'+
3522 Create a new reference containing +'string'+ of type +'tag'+.
3523 If +'finalizer'+ is specified, it is a command which will be invoked
3524 when the a garbage collection cycle runs and this reference is
3525 no longer accessible.
3527 The finalizer is invoked as:
3529 finalizer reference string
3531 See GARBAGE COLLECTION, REFERENCES, LAMBDA for more detail.
3535 +*rename* 'oldName newName'+
3537 Rename the command that used to be called +'oldName'+ so that it
3538 is now called +'newName'+. If +'newName'+ is an empty string
3539 (e.g. {}) then +'oldName'+ is deleted. The `rename` command
3540 returns an empty string as result.
3544 +*return* ?*-code* 'code'? ?*-errorinfo* 'stacktrace'? ?*-errorcode* 'errorcode'? ?*-level* 'n'? ?'value'?+
3546 Return immediately from the current procedure (or top-level command
3547 or `source` command), with +'value'+ as the return value. If +'value'+
3548 is not specified, an empty string will be returned as result.
3550 If +-code+ is specified (as either a number or ok, error, break,
3551 continue, signal, return or exit), this code will be used instead
3552 of +JIM_OK+. This is generally useful when implementing flow of control
3555 If +-level+ is specified and greater than 1, it has the effect of delaying
3556 the new return code from +-code+. This is useful when rethrowing an error
3557 from `catch`. See the implementation of try/catch in tclcompat.tcl for
3558 an example of how this is done.
3560 Note: The following options are only used when +-code+ is JIM_ERR.
3562 If +-errorinfo+ is specified (as returned from `info stacktrace`)
3563 it is used to initialize the stacktrace.
3565 If +-errorcode+ is specified, it is used to set the global variable $::errorCode.
3569 +*scan* 'string format varName1 ?varName2 \...?'+
3571 This command parses fields from an input string in the same fashion
3572 as the C 'sscanf' procedure. +'string'+ gives the input to be parsed
3573 and +'format'+ indicates how to parse it, using '%' fields as in
3574 'sscanf'. All of the 'sscanf' options are valid; see the 'sscanf'
3575 man page for details. Each +'varName'+ gives the name of a variable;
3576 when a field is scanned from +'string'+, the result is converted back
3577 into a string and assigned to the corresponding +'varName'+. The
3578 only unusual conversion is for '%c'. For '%c' conversions a single
3579 character value is converted to a decimal string, which is then
3580 assigned to the corresponding +'varName'+; no field width may be
3581 specified for this conversion.
3585 +*seek* 'fileId offset ?origin?'+
3587 +'fileId' *seek* 'offset ?origin?'+
3589 Change the current access position for +'fileId'+.
3590 The +'offset'+ and +'origin'+ arguments specify the position at
3591 which the next read or write will occur for +'fileId'+.
3592 +'offset'+ must be a number (which may be negative) and +'origin'+
3593 must be one of the following:
3596 The new access position will be +'offset'+ bytes from the start
3600 The new access position will be +'offset'+ bytes from the current
3601 access position; a negative +'offset'+ moves the access position
3602 backwards in the file.
3605 The new access position will be +'offset'+ bytes from the end of
3606 the file. A negative +'offset'+ places the access position before
3607 the end-of-file, and a positive +'offset'+ places the access position
3608 after the end-of-file.
3610 The +'origin'+ argument defaults to +start+.
3612 +'fileId'+ must have been the return value from a previous call to
3613 `open`, or it may be +stdin+, +stdout+, or +stderr+ to refer to one
3614 of the standard I/O channels.
3616 This command returns an empty string.
3620 +*set* 'varName ?value?'+
3622 Returns the value of variable +'varName'+.
3624 If +'value'+ is specified, then set the value of +'varName'+ to +'value'+,
3625 creating a new variable if one doesn't already exist, and return
3628 If +'varName'+ contains an open parenthesis and ends with a
3629 close parenthesis, then it refers to an array element: the characters
3630 before the open parenthesis are the name of the array, and the characters
3631 between the parentheses are the index within the array.
3632 Otherwise +'varName'+ refers to a scalar variable.
3634 If no procedure is active, then +'varName'+ refers to a global
3637 If a procedure is active, then +'varName'+ refers to a parameter
3638 or local variable of the procedure, unless the +'global'+ command
3639 has been invoked to declare +'varName'+ to be global.
3641 The +::+ prefix may also be used to explicitly reference a variable
3642 in the global scope.
3646 +*setref* 'reference string'+
3648 Store a new string in +'reference'+, replacing the existing string.
3649 The reference must be a valid reference create with the `ref`
3652 See GARBAGE COLLECTION, REFERENCES, LAMBDA for more detail.
3656 Command for signal handling.
3658 See `kill` for the different forms which may be used to specify signals.
3660 Commands which return a list of signal names do so using the canonical form:
3663 +*signal handle* ?'signals \...'?+::
3664 If no signals are given, returns a list of all signals which are currently
3666 If signals are specified, these are added to the list of signals currently
3669 +*signal ignore* ?'signals \...'?+::
3670 If no signals are given, returns a lists all signals which are currently
3672 If signals are specified, these are added to the list of signals
3673 currently being ignored. These signals are still delivered, but
3674 are not considered by `catch -signal` or `try -signal`. Use
3675 `signal check` to determine which signals have occurred but
3678 +*signal default* ?'signals \...'?+::
3679 If no signals are given, returns a lists all signals which are currently have
3680 the default behaviour.
3681 If signals are specified, these are added to the list of signals which have
3682 the default behaviour.
3684 +*signal check ?-clear?* ?'signals \...'?+::
3685 Returns a list of signals which have been delivered to the process
3686 but are 'ignored'. If signals are specified, only that set of signals will
3687 be checked, otherwise all signals will be checked.
3688 If +-clear+ is specified, any signals returned are removed and will not be
3689 returned by subsequent calls to `signal check` unless delivered again.
3691 +*signal throw* ?'signal'?+::
3692 Raises the given signal, which defaults to +SIGINT+ if not specified.
3693 The behaviour is identical to:
3697 Note that `signal handle` and `signal ignore` represent two forms of signal
3698 handling. `signal handle` is used in conjunction with `catch -signal` or `try -signal`
3699 to immediately abort execution when the signal is delivered. Alternatively, `signal ignore`
3700 is used in conjunction with `signal check` to handle signal synchronously. Consider the
3703 Prevent a processing from taking too long
3705 signal handle SIGALRM
3708 .. possibly long running process ..
3711 puts stderr "Process took too long"
3714 Handle SIGHUP to reconfigure:
3716 signal ignore SIGHUP
3718 ... handle configuration/reconfiguration ...
3719 while {[signal check -clear SIGHUP] eq ""} {
3720 ... do processing ..
3722 # Received SIGHUP, so reconfigure
3729 Pauses for the given number of seconds, which may be a floating
3730 point value less than one to sleep for less than a second, or an
3731 integer to sleep for one or more seconds.
3735 +*source* 'fileName'+
3737 Read file +'fileName'+ and pass the contents to the Tcl interpreter
3738 as a sequence of commands to execute in the normal fashion. The return
3739 value of `source` is the return value of the last command executed
3740 from the file. If an error occurs in executing the contents of the
3741 file, then the `source` command will return that error.
3743 If a `return` command is invoked from within the file, the remainder of
3744 the file will be skipped and the `source` command will return
3745 normally with the result from the `return` command.
3749 +*split* 'string ?splitChars?'+
3751 Returns a list created by splitting +'string'+ at each character
3752 that is in the +'splitChars'+ argument.
3754 Each element of the result list will consist of the
3755 characters from +'string'+ between instances of the
3756 characters in +'splitChars'+.
3758 Empty list elements will be generated if +'string'+ contains
3759 adjacent characters in +'splitChars'+, or if the first or last
3760 character of +'string'+ is in +'splitChars'+.
3762 If +'splitChars'+ is an empty string then each character of
3763 +'string'+ becomes a separate element of the result list.
3765 +'splitChars'+ defaults to the standard white-space characters.
3768 split "comp.unix.misc" .
3770 returns +'"comp unix misc"'+ and
3772 split "Hello world" {}
3774 returns +'"H e l l o { } w o r l d"'+.
3779 +*stackdump* 'stacktrace'+
3781 Creates a human readable representation of a stack trace.
3788 Returns a live stack trace as a list of +proc file line proc file line \...+.
3789 Iteratively uses `info frame` to create the stack trace. This stack trace is in the
3790 same form as produced by `catch` and `info stacktrace`
3792 See also `stackdump`.
3797 +*string* 'option arg ?arg \...?'+
3799 Perform one of several string operations, depending on +'option'+.
3800 The legal options (which may be abbreviated) are:
3802 +*string bytelength* 'string'+::
3803 Returns the length of the string in bytes. This will return
3804 the same value as `string length` if UTF-8 support is not enabled,
3805 or if the string is composed entirely of ASCII characters.
3806 See UTF-8 AND UNICODE.
3808 +*string byterange* 'string first last'+::
3809 Like `string range` except works on bytes rather than characters.
3810 These commands are identical if UTF-8 support is not enabled.
3812 +*string compare ?-nocase?* ?*-length* 'len? string1 string2'+::
3813 Perform a character-by-character comparison of strings +'string1'+ and
3814 +'string2'+ in the same way as the C 'strcmp' procedure. Return
3815 -1, 0, or 1, depending on whether +'string1'+ is lexicographically
3816 less than, equal to, or greater than +'string2'+. If +-length+
3817 is specified, then only the first +'len'+ characters are used
3818 in the comparison. If +'len'+ is negative, it is ignored.
3819 Performs a case-insensitive comparison if +-nocase+ is specified.
3821 +*string equal ?-nocase?* '?*-length* len?' 'string1 string2'+::
3822 Returns 1 if the strings are equal, or 0 otherwise. If +-length+
3823 is specified, then only the first +'len'+ characters are used
3824 in the comparison. If +'len'+ is negative, it is ignored.
3825 Performs a case-insensitive comparison if +-nocase+ is specified.
3827 +*string first* 'string1 string2 ?firstIndex?'+::
3828 Search +'string2'+ for a sequence of characters that exactly match
3829 the characters in +'string1'+. If found, return the index of the
3830 first character in the first such match within +'string2'+. If not
3831 found, return -1. If +'firstIndex'+ is specified, matching will start
3832 from +'firstIndex'+ of +'string1'+.
3834 See STRING AND LIST INDEX SPECIFICATIONS for all allowed forms for +'firstIndex'+.
3836 +*string index* 'string charIndex'+::
3837 Returns the +'charIndex'+'th character of the +'string'+
3838 argument. A +'charIndex'+ of 0 corresponds to the first
3839 character of the string.
3840 If +'charIndex'+ is less than 0 or greater than
3841 or equal to the length of the string then an empty string is
3844 See STRING AND LIST INDEX SPECIFICATIONS for all allowed forms for +'charIndex'+.
3846 +*string is* 'class' ?*-strict*? 'string'+::
3847 Returns 1 if +'string'+ is a valid member of the specified character
3848 class, otherwise returns 0. If +-strict+ is specified, then an
3849 empty string returns 0, otherwise an empty string will return 1
3850 on any class. The following character classes are recognized
3851 (the class name can be abbreviated):
3853 +alnum+;; Any alphabet or digit character.
3854 +alpha+;; Any alphabet character.
3855 +ascii+;; Any character with a value less than 128 (those that are in the 7-bit ascii range).
3856 +control+;; Any control character.
3857 +digit+;; Any digit character.
3858 +double+;; Any of the valid forms for a double in Tcl, with optional surrounding whitespace.
3859 In case of under/overflow in the value, 0 is returned.
3860 +graph+;; Any printing character, except space.
3861 +integer+;; Any of the valid string formats for an integer value in Tcl, with optional surrounding whitespace.
3862 +lower+;; Any lower case alphabet character.
3863 +print+;; Any printing character, including space.
3864 +punct+;; Any punctuation character.
3865 +space+;; Any space character.
3866 +upper+;; Any upper case alphabet character.
3867 +xdigit+;; Any hexadecimal digit character ([0-9A-Fa-f]).
3869 Note that string classification does +'not'+ respect UTF-8. See UTF-8 AND UNICODE
3871 +*string last* 'string1 string2 ?lastIndex?'+::
3872 Search +'string2'+ for a sequence of characters that exactly match
3873 the characters in +'string1'+. If found, return the index of the
3874 first character in the last such match within +'string2'+. If there
3875 is no match, then return -1. If +'lastIndex'+ is specified, only characters
3876 up to +'lastIndex'+ of +'string2'+ will be considered in the match.
3878 See STRING AND LIST INDEX SPECIFICATIONS for all allowed forms for +'lastIndex'+.
3880 +*string length* 'string'+::
3881 Returns a decimal string giving the number of characters in +'string'+.
3882 If UTF-8 support is enabled, this may be different than the number of bytes.
3883 See UTF-8 AND UNICODE
3885 +*string map ?-nocase?* 'mapping string'+::
3886 Replaces substrings in +'string'+ based on the key-value pairs in
3887 +'mapping'+, which is a list of +key value key value \...+ as in the form
3888 returned by `array get`. Each instance of a key in the string will be
3889 replaced with its corresponding value. If +-nocase+ is specified, then
3890 matching is done without regard to case differences. Both key and value may
3891 be multiple characters. Replacement is done in an ordered manner, so the
3892 key appearing first in the list will be checked first, and so on. +'string'+ is
3893 only iterated over once, so earlier key replacements will have no affect for
3894 later key matches. For example,
3896 string map {abc 1 ab 2 a 3 1 0} 1abcaababcabababc
3899 will return the string +01321221+.
3901 Note that if an earlier key is a prefix of a later one, it will completely mask the later
3902 one. So if the previous example is reordered like this,
3904 string map {1 0 ab 2 a 3 abc 1} 1abcaababcabababc
3907 it will return the string +02c322c222c+.
3909 +*string match ?-nocase?* 'pattern string'+::
3910 See if +'pattern'+ matches +'string'+; return 1 if it does, 0
3911 if it doesn't. Matching is done in a fashion similar to that
3912 used by the C-shell. For the two strings to match, their contents
3913 must be identical except that the following special sequences
3914 may appear in +'pattern'+:
3917 Matches any sequence of characters in +'string'+,
3918 including a null string.
3921 Matches any single character in +'string'+.
3924 Matches any character in the set given by +'chars'+.
3925 If a sequence of the form +'x-y'+ appears in +'chars'+,
3926 then any character between +'x'+ and +'y'+, inclusive,
3930 Matches the single character +'x'+. This provides a way of
3931 avoiding the special interpretation of the characters +{backslash}*?[]+
3934 Performs a case-insensitive comparison if +-nocase+ is specified.
3936 +*string range* 'string first last'+::
3937 Returns a range of consecutive characters from +'string'+, starting
3938 with the character whose index is +'first'+ and ending with the
3939 character whose index is +'last'+. An index of 0 refers to the
3940 first character of the string.
3942 See STRING AND LIST INDEX SPECIFICATIONS for all allowed forms for +'first'+ and +'last'+.
3944 If +'first'+ is less than zero then it is treated as if it were zero, and
3945 if +'last'+ is greater than or equal to the length of the string then
3946 it is treated as if it were +end+. If +'first'+ is greater than
3947 +'last'+ then an empty string is returned.
3949 +*string repeat* 'string count'+::
3950 Returns a new string consisting of +'string'+ repeated +'count'+ times.
3952 +*string replace* 'string first last ?newstring?'+::
3953 Removes a range of consecutive characters from +'string'+, starting
3954 with the character whose index is +'first'+ and ending with the
3955 character whose index is +'last'+. If +'newstring'+ is specified,
3956 then it is placed in the removed character range. If +'first'+ is
3957 less than zero then it is treated as if it were zero, and if +'last'+
3958 is greater than or equal to the length of the string then it is
3959 treated as if it were +end+. If +'first'+ is greater than +'last'+
3960 or the length of the initial string, or +'last'+ is less than 0,
3961 then the initial string is returned untouched.
3963 +*string reverse* 'string'+::
3964 Returns a string that is the same length as +'string'+ but
3965 with its characters in the reverse order.
3967 +*string tolower* 'string'+::
3968 Returns a value equal to +'string'+ except that all upper case
3969 letters have been converted to lower case.
3971 +*string totitle* 'string'+::
3972 Returns a value equal to +'string'+ except that the first character
3973 is converted to title case (or upper case if there is no UTF-8 titlecase variant)
3974 and all remaining characters have been converted to lower case.
3976 +*string toupper* 'string'+::
3977 Returns a value equal to +'string'+ except that all lower case
3978 letters have been converted to upper case.
3980 +*string trim* 'string ?chars?'+::
3981 Returns a value equal to +'string'+ except that any leading
3982 or trailing characters from the set given by +'chars'+ are
3984 If +'chars'+ is not specified then white space is removed
3985 (spaces, tabs, newlines, and carriage returns).
3987 +*string trimleft* 'string ?chars?'+::
3988 Returns a value equal to +'string'+ except that any
3989 leading characters from the set given by +'chars'+ are
3991 If +'chars'+ is not specified then white space is removed
3992 (spaces, tabs, newlines, and carriage returns).
3994 +*string trimright* 'string ?chars?'+::
3995 Returns a value equal to +'string'+ except that any
3996 trailing characters from the set given by +'chars'+ are
3998 If +'chars'+ is not specified then white space is removed
3999 (spaces, tabs, newlines, and carriage returns).
4000 Null characters are always removed.
4004 +*subst ?-nobackslashes? ?-nocommands? ?-novariables?* 'string'+
4006 This command performs variable substitutions, command substitutions,
4007 and backslash substitutions on its string argument and returns the
4008 fully-substituted result. The substitutions are performed in exactly
4009 the same way as for Tcl commands. As a result, the string argument
4010 is actually substituted twice, once by the Tcl parser in the usual
4011 fashion for Tcl commands, and again by the subst command.
4013 If any of the +-nobackslashes+, +-nocommands+, or +-novariables+ are
4014 specified, then the corresponding substitutions are not performed.
4015 For example, if +-nocommands+ is specified, no command substitution
4016 is performed: open and close brackets are treated as ordinary
4017 characters with no special interpretation.
4019 *Note*: when it performs its substitutions, subst does not give any
4020 special treatment to double quotes or curly braces. For example,
4021 the following script returns +xyz \{44\}+, not +xyz \{$a\}+.
4029 +*switch* '?options? string pattern body ?pattern body \...?'+
4031 +*switch* '?options? string {pattern body ?pattern body \...?}'+
4033 The `switch` command matches its string argument against each of
4034 the pattern arguments in order. As soon as it finds a pattern that
4035 matches string it evaluates the following body and returns the
4036 result of that evaluation. If the last pattern argument is default
4037 then it matches anything. If no pattern argument matches string and
4038 no default is given, then the `switch` command returns an empty string.
4039 If the initial arguments to switch start with - then they are treated
4040 as options. The following options are currently supported:
4043 Use exact matching when comparing string to a
4044 pattern. This is the default.
4047 When matching string to the patterns, use glob-style
4048 matching (i.e. the same as implemented by the string
4052 When matching string to the patterns, use regular
4053 expression matching (i.e. the same as implemented
4054 by the regexp command).
4056 +-command 'commandname'+::
4057 When matching string to the patterns, use the given command, which
4058 must be a single word. The command is invoked as
4059 'commandname pattern string', or 'commandname -nocase pattern string'
4060 and must return 1 if matched, or 0 if not.
4063 Marks the end of options. The argument following
4064 this one will be treated as string even if it starts
4067 Two syntaxes are provided for the pattern and body arguments. The
4068 first uses a separate argument for each of the patterns and commands;
4069 this form is convenient if substitutions are desired on some of the
4070 patterns or commands. The second form places all of the patterns
4071 and commands together into a single argument; the argument must
4072 have proper list structure, with the elements of the list being the
4073 patterns and commands. The second form makes it easy to construct
4074 multi-line `switch` commands, since the braces around the whole list
4075 make it unnecessary to include a backslash at the end of each line.
4076 Since the pattern arguments are in braces in the second form, no
4077 command or variable substitutions are performed on them; this makes
4078 the behaviour of the second form different than the first form in
4081 If a body is specified as +-+ it means that the body for the next
4082 pattern should also be used as the body for this pattern (if the
4083 next pattern also has a body of +-+ then the body after that is
4084 used, and so on). This feature makes it possible to share a single
4085 body among several patterns.
4087 Below are some examples of `switch` commands:
4089 switch abc a - b {format 1} abc {format 2} default {format 3}
4093 switch -regexp aaab {
4113 +*tailcall* 'cmd ?arg\...?'+
4115 The `tailcall` command provides an optimised way of invoking a command whilst replacing
4116 the current call frame. This is similar to 'exec' in Bourne Shell.
4118 The following are identical except the first immediately replaces the current call frame.
4122 return [uplevel 1 [list a b c]]
4124 `tailcall` is useful as a dispatch mechanism:
4127 tailcall sub_$cmd {*}$args
4138 Returns a decimal string giving the current access position in
4141 +'fileId'+ must have been the return value from a previous call to
4142 `open`, or it may be +stdin+, +stdout+, or +stderr+ to refer to one
4143 of the standard I/O channels.
4147 +*throw* 'code ?msg?'+
4149 This command throws an exception (return) code along with an optional message.
4150 This command is mostly for convenient usage with `try`.
4152 The command +throw break+ is equivalent to +break+.
4153 The command +throw 20 message+ can be caught with an +on 20 \...+ clause to `try`.
4157 +*time* 'command ?count?'+
4159 This command will call the Tcl interpreter +'count'+
4160 times to execute +'command'+ (or once if +'count'+ isn't
4161 specified). It will then return a string of the form
4163 503 microseconds per iteration
4165 which indicates the average amount of time required per iteration,
4168 Time is measured in elapsed time, not CPU time.
4172 +*try* '?catchopts? tryscript' ?*on* 'returncodes {?resultvar? ?optsvar?} handlerscript \...'? ?*finally* 'finalscript'?+
4174 The `try` command is provided as a convenience for exception handling.
4176 This interpeter first evaluates +'tryscript'+ under the effect of the catch
4177 options +'catchopts'+ (e.g. +-signal -noexit --+, see `catch`).
4179 It then evaluates the script for the first matching 'on' handler
4180 (there many be zero or more) based on the return code from the `try`
4181 section. For example a normal +JIM_ERR+ error will be matched by
4182 an 'on error' handler.
4184 Finally, any +'finalscript'+ is evaluated.
4186 The result of this command is the result of +'tryscript'+, except in the
4187 case where an exception occurs in a matching 'on' handler script or the 'finally' script,
4188 in which case the result is this new exception.
4190 The specified +'returncodes'+ is a list of return codes either as names ('ok', 'error', 'break', etc.)
4193 If +'resultvar'+ and +'optsvar'+ are specified, they are set as for `catch` before evaluating
4194 the matching handler.
4201 } on {continue break} {} {
4202 error "Unexpected break/continue"
4203 } on error {msg opts} {
4204 puts "Dealing with error"
4205 return {*}$opts $msg
4207 puts "Got signal: $sig"
4212 If break, continue or error are raised, they are dealt with by the matching
4215 In any case, the file will be closed via the 'finally' clause.
4217 See also `throw`, `catch`, `return`, `error`.
4221 +*unknown* 'cmdName ?arg arg ...?'+
4223 This command doesn't actually exist as part of Tcl, but Tcl will
4224 invoke it if it does exist.
4226 If the Tcl interpreter encounters a command name for which there
4227 is not a defined command, then Tcl checks for the existence of
4228 a command named `unknown`.
4230 If there is no such command, then the interpreter returns an
4233 If the `unknown` command exists, then it is invoked with
4234 arguments consisting of the fully-substituted name and arguments
4235 for the original non-existent command.
4237 The `unknown` command typically does things like searching
4238 through library directories for a command procedure with the name
4239 +'cmdName'+, or expanding abbreviated command names to full-length,
4240 or automatically executing unknown commands as UNIX sub-processes.
4242 In some cases (such as expanding abbreviations) `unknown` will
4243 change the original command slightly and then (re-)execute it.
4244 The result of the `unknown` command is used as the result for
4245 the original non-existent command.
4249 +*unset ?-nocomplain? ?--?* '?name name ...?'+
4252 Each +'name'+ is a variable name, specified in any of the
4253 ways acceptable to the `set` command.
4255 If a +'name'+ refers to an element of an array, then that
4256 element is removed without affecting the rest of the array.
4258 If a +'name'+ consists of an array name with no parenthesized
4259 index, then the entire array is deleted.
4261 The `unset` command returns an empty string as result.
4263 An error occurs if any of the variables doesn't exist, unless '-nocomplain'
4264 is specified. The '--' argument may be specified to stop option processing
4265 in case the variable name may be '-nocomplain'.
4269 +*upcall* 'command ?args ...?'+
4271 May be used from within a proc defined as `local` `proc` in order to call
4272 the previous, hidden version of the same command.
4274 If there is no previous definition of the command, an error is returned.
4278 +*uplevel* '?level? command ?command ...?'+
4280 All of the +'command'+ arguments are concatenated as if they had
4281 been passed to `concat`; the result is then evaluated in the
4282 variable context indicated by +'level'+. `uplevel` returns
4283 the result of that evaluation. If +'level'+ is an integer, then
4284 it gives a distance (up the procedure calling stack) to move before
4285 executing the command. If +'level'+ consists of +\#+ followed by
4286 a number then the number gives an absolute level number. If +'level'+
4287 is omitted then it defaults to +1+. +'level'+ cannot be
4288 defaulted if the first +'command'+ argument starts with a digit or +#+.
4290 For example, suppose that procedure 'a' was invoked
4291 from top-level, and that it called 'b', and that 'b' called 'c'.
4292 Suppose that 'c' invokes the `uplevel` command. If +'level'+
4293 is +1+ or +#2+ or omitted, then the command will be executed
4294 in the variable context of 'b'. If +'level'+ is +2+ or +#1+
4295 then the command will be executed in the variable context of 'a'.
4297 If +'level'+ is '3' or +#0+ then the command will be executed
4298 at top-level (only global variables will be visible).
4299 The `uplevel` command causes the invoking procedure to disappear
4300 from the procedure calling stack while the command is being executed.
4301 In the above example, suppose 'c' invokes the command
4303 uplevel 1 {set x 43; d}
4305 where 'd' is another Tcl procedure. The `set` command will
4306 modify the variable 'x' in 'b's context, and 'd' will execute
4307 at level 3, as if called from 'b'. If it in turn executes
4312 then the `set` command will modify the same variable 'x' in 'b's
4313 context: the procedure 'c' does not appear to be on the call stack
4314 when 'd' is executing. The command `info level` may
4315 be used to obtain the level of the current procedure.
4317 `uplevel` makes it possible to implement new control
4318 constructs as Tcl procedures (for example, `uplevel` could
4319 be used to implement the `while` construct as a Tcl procedure).
4323 +*upvar* '?level? otherVar myVar ?otherVar myVar ...?'+
4325 This command arranges for one or more local variables in the current
4326 procedure to refer to variables in an enclosing procedure call or
4327 to global variables.
4329 +'level'+ may have any of the forms permitted for the `uplevel`
4330 command, and may be omitted if the first letter of the first +'otherVar'+
4331 isn't +#+ or a digit (it defaults to '1').
4333 For each +'otherVar'+ argument, `upvar` makes the variable
4334 by that name in the procedure frame given by +'level'+ (or at
4335 global level, if +'level'+ is +#0+) accessible
4336 in the current procedure by the name given in the corresponding
4339 The variable named by +'otherVar'+ need not exist at the time of the
4340 call; it will be created the first time +'myVar'+ is referenced, just like
4341 an ordinary variable.
4343 `upvar` may only be invoked from within procedures.
4345 `upvar` returns an empty string.
4347 The `upvar` command simplifies the implementation of call-by-name
4348 procedure calling and also makes it easier to build new control constructs
4350 For example, consider the following procedure:
4357 'add2' is invoked with an argument giving the name of a variable,
4358 and it adds two to the value of that variable.
4359 Although 'add2' could have been implemented using `uplevel`
4360 instead of `upvar`, `upvar` makes it simpler for 'add2'
4361 to access the variable in the caller's procedure frame.
4365 +*while* 'test body'+
4367 The +'while'+ command evaluates +'test'+ as an expression
4368 (in the same way that `expr` evaluates its argument).
4369 The value of the expression must be numeric; if it is non-zero
4370 then +'body'+ is executed by passing it to the Tcl interpreter.
4372 Once +'body'+ has been executed then +'test'+ is evaluated
4373 again, and the process repeats until eventually +'test'+
4374 evaluates to a zero numeric value. `continue`
4375 commands may be executed inside +'body'+ to terminate the current
4376 iteration of the loop, and `break`
4377 commands may be executed inside +'body'+ to cause immediate
4378 termination of the `while` command.
4380 The `while` command always returns an empty string.
4385 The following extensions may or may not be available depending upon
4386 what options were selected when Jim Tcl was built.
4389 posix: os.fork, os.wait, os.gethostname, os.getids, os.uptime
4390 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
4392 Invokes 'fork(2)' and returns the result.
4394 +*os.wait -nohang* 'pid'+::
4395 Invokes waitpid(2), with WNOHANG if +-nohang+ is specified.
4396 Returns a list of 3 elements.
4398 {0 none 0} if -nohang is specified, and the process is still alive.
4400 {-1 error <error-description>} if the process does not exist or has already been waited for.
4402 {<pid> exit <exit-status>} if the process exited normally.
4404 {<pid> signal <signal-number>} if the process terminated on a signal.
4406 {<pid> other 0} otherwise (core dump, stopped, continued, etc.)
4408 +*os.gethostname*+::
4409 Invokes 'gethostname(3)' and returns the result.
4412 Returns the various user/group ids for the current process.
4415 uid 1000 euid 1000 gid 100 egid 100
4418 Returns the number of seconds since system boot. See description of 'uptime' in 'sysinfo(2)'.
4420 ANSI I/O (aio) and EVENTLOOP API
4421 --------------------------------
4422 Jim provides an alternative object-based API for I/O.
4424 See `open` and `socket` for commands which return an I/O handle.
4428 +$handle *accept*+::
4429 Server socket only: Accept a connection and return stream
4431 +$handle *buffering none|line|full*+::
4432 Sets the buffering mode of the stream.
4434 +$handle *close* ?r(ead)|w(rite)?+::
4436 The two-argument form is a "half-close" on a socket. See the +shutdown(2)+ man page.
4438 +$handle *copyto* 'tofd ?size?'+::
4439 Copy bytes to the file descriptor +'tofd'+. If +'size'+ is specified, at most
4440 that many bytes will be copied. Otherwise copying continues until the end
4441 of the input file. Returns the number of bytes actually copied.
4444 Returns 1 if stream is at eof
4446 +$handle *filename*+::
4447 Returns the original filename associated with the handle.
4448 Handles returned by `socket` give the socket type instead of a filename.
4453 +$handle *gets* '?var?'+::
4454 Read one line and return it or store it in the var
4456 +$handle *isatty*+::
4457 Returns 1 if the stream is a tty device.
4459 +$handle *ndelay ?0|1?*+::
4460 Set O_NDELAY (if arg). Returns current/new setting.
4461 Note that in general ANSI I/O interacts badly with non-blocking I/O.
4464 +$handle *puts ?-nonewline?* 'str'+::
4465 Write the string, with newline unless -nonewline
4467 +$handle *read ?-nonewline?* '?len?'+::
4468 Read and return bytes from the stream. To eof if no len.
4470 +$handle *recvfrom* 'maxlen ?addrvar?'+::
4471 Receives a message from the handle via recvfrom(2) and returns it.
4472 At most +'maxlen'+ bytes are read.
4473 If +'addrvar'+ is specified, the sending address of the message is stored in
4474 the named variable in the form 'addr:port'. See `socket` for details.
4476 +$handle *seek* 'offset' *?start|current|end?*+::
4477 Seeks in the stream (default 'current')
4479 +$handle *sendto* 'str ?addr:?port'+::
4480 Sends the string, +'str'+, to the given address via the socket using sendto(2).
4481 This is intended for udp/dgram sockets and may give an error or behave in unintended
4482 ways for other handle types.
4483 Returns the number of bytes written.
4486 Returns the current seek position
4490 +*fconfigure* 'handle' *?-blocking 0|1? ?-buffering noneline|full? ?-translation* 'mode'?+::
4491 For compatibility with Tcl, a limited form of the `fconfigure`
4492 command is supported.
4493 * `fconfigure ... -blocking` maps to `aio ndelay`
4494 * `fconfigure ... -buffering` maps to `aio buffering`
4495 * `fconfigure ... -translation` is accepted but ignored
4498 eventloop: after, vwait, update
4499 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
4501 The following commands allow a script to be invoked when the given condition occurs.
4502 If no script is given, returns the current script. If the given script is the empty, the
4505 +$handle *readable* '?readable-script?'+::
4506 Sets or returns the script for when the socket is readable.
4508 +$handle *writable* '?writable-script?'+::
4509 Sets or returns the script for when the socket is writable.
4511 +$handle *onexception* '?exception-script?'+::
4512 Sets or returns the script for when when oob data received.
4514 For compatibility with 'Tcl', these may be prefixed with `fileevent`. e.g.
4517 +fileevent $handle *readable* '\...'+
4519 Time-based execution is also available via the eventloop API.
4522 Sleeps for the given number of milliseconds. No events are
4523 processed during this time.
4525 +*after* 'ms'|*idle* script ?script \...?'+::
4526 The scripts are concatenated and executed after the given
4527 number of milliseconds have elapsed. If 'idle' is specified,
4528 the script will run the next time the event loop is processed
4529 with `vwait` or `update`. The script is only run once and
4530 then removed. Returns an event id.
4532 +*after cancel* 'id|command'+::
4533 Cancels an `after` event with the given event id or matching
4534 command (script). Returns the number of milliseconds
4535 remaining until the event would have fired. Returns the
4536 empty string if no matching event is found.
4538 +*after info* '?id?'+::
4539 If +'id'+ is not given, returns a list of current `after`
4540 events. If +'id'+ is given, returns a list containing the
4541 associated script and either 'timer' or 'idle' to indicated
4542 the type of the event. An error occurs if +'id'+ does not
4545 +*vwait* 'variable'+::
4546 A call to `vwait` is enters the eventloop. `vwait` processes
4547 events until the named (global) variable changes or all
4548 event handlers are removed. The variable need not exist
4549 beforehand. If there are no event handlers defined, `vwait`
4550 returns immediately.
4552 +*update ?idletasks?*+::
4553 A call to `update` enters the eventloop to process expired events, but
4554 no new events. If 'idletasks' is specified, only expired time events are handled,
4556 Returns once handlers have been run for all expired events.
4558 Scripts are executed at the global scope. If an error occurs during a handler script,
4559 an attempt is made to call (the user-defined command) `bgerror` with the details of the error.
4560 If the `bgerror` command does not exist, the error message details are printed to stderr instead.
4562 If a file event handler script generates an error, the handler is automatically removed
4563 to prevent infinite errors. (A time event handler is always removed after execution).
4566 Called when an event handler script generates an error. Note that the normal command resolution
4567 rules are used for bgerror. First the name is resolved in the current namespace, then in the
4572 Various socket types may be created.
4574 +*socket unix* 'path'+::
4575 A unix domain socket client.
4577 +*socket unix.server* 'path'+::
4578 A unix domain socket server.
4580 +*socket ?-ipv6? stream* 'addr:port'+::
4581 A TCP socket client. (See the forms for +'addr'+ below)
4583 +*socket ?-ipv6? stream.server* '?addr:?port'+::
4584 A TCP socket server (+'addr'+ defaults to +0.0.0.0+ for IPv4 or +[::]+ for IPv6).
4586 +*socket ?-ipv6? dgram* ?'addr:port'?+::
4587 A UDP socket client. If the address is not specified,
4588 the client socket will be unbound and 'sendto' must be used
4589 to indicated the destination.
4591 +*socket ?-ipv6? dgram.server* 'addr:port'+::
4592 A UDP socket server.
4595 A pipe. Note that unlike all other socket types, this command returns
4596 a list of two channels: {read write}
4599 A socketpair (see socketpair(2)). Like `socket pipe`, this command returns
4600 a list of two channels: {s1 s2}. These channels are both readable and writable.
4602 This command creates a socket connected (client) or bound (server) to the given
4605 The returned value is channel and may generally be used with the various file I/O
4606 commands (gets, puts, read, etc.), either as object-based syntax or Tcl-compatible syntax.
4608 set f [socket stream www.google.com:80]
4610 $f puts -nonewline "GET / HTTP/1.0\r\n\r\n"
4615 Server sockets, however support only 'accept', which is most useful in conjunction with
4618 set f [socket stream.server 80]
4620 set client [$f accept]
4623 $client puts -nonewline "HTTP/1.1 404 Not found\r\n"
4628 The address, +'addr'+, can be given in one of the following forms:
4630 1. For IPv4 socket types, an IPv4 address such as 192.168.1.1
4631 2. For IPv6 socket types, an IPv6 address such as [fe80::1234] or [::]
4634 Note that on many systems, listening on an IPv6 address such as [::] will
4635 also accept requests via IPv4.
4637 Where a hostname is specified, the +'first'+ returned address is used
4638 which matches the socket type is used.
4640 The special type 'pipe' isn't really a socket.
4642 lassign [socket pipe] r w
4644 # Must close $w after exec
4652 +*syslog* '?options? ?priority? message'+
4654 This command sends message to system syslog facility with given
4655 priority. Valid priorities are:
4657 emerg, alert, crit, err, error, warning, notice, info, debug
4659 If a message is specified, but no priority is specified, then a
4660 priority of info is used.
4662 By default, facility user is used and the value of global tcl variable
4663 argv0 is used as ident string. However, any of the following options
4664 may be specified before priority to control these parameters:
4666 +*-facility* 'value'+::
4667 Use specified facility instead of user. The following
4668 values for facility are recognized:
4670 authpriv, cron, daemon, kernel, lpr, mail, news, syslog, user,
4673 +*-ident* 'string'+::
4674 Use given string instead of argv0 variable for ident string.
4676 +*-options* 'integer'+::
4677 Set syslog options such as +LOG_CONS+, +LOG_NDELAY+. You should
4678 use numeric values of those from your system syslog.h file,
4679 because I haven't got time to implement yet another hash
4684 The optional 'pack' extension provides commands to encode and decode binary strings.
4686 +*pack* 'varName value' *-intle|-intbe|-floatle|-floatbe|-str* 'bitwidth ?bitoffset?'+::
4687 Packs the binary representation of +'value'+ into the variable
4688 +'varName'+. The value is packed according to the given type
4689 (integer/floating point/string, big-endian/little-endian), width and bit offset.
4690 The variable is created if necessary (like `append`).
4691 Ihe variable is expanded if necessary.
4693 +*unpack* 'binvalue' *-intbe|-intle|-uintbe|-uintle|-floatbe|-floatle|-str* 'bitpos bitwidth'+::
4694 Unpacks bits from +'binvalue'+ at bit position +'bitpos'+ and with +'bitwidth'+.
4695 Interprets the value according to the type (integer/floating point/string, big-endian/little-endian
4696 and signed/unsigned) and returns it. For integer types, +'bitwidth'+
4697 may be up to the size of a Jim Tcl integer (typically 64 bits). For floating point types,
4698 +'bitwidth'+ may be 32 bits (for single precision numbers) or 64 bits (for double precision).
4699 For the string type, both the width and the offset must be on a byte boundary (multiple of 8). Attempting to
4700 access outside the length of the value will return 0 for integer types, 0.0 for floating point types
4701 or the empty string for the string type.
4705 The optional, pure-Tcl 'binary' extension provides the Tcl-compatible `binary scan` and `binary format`
4706 commands based on the low-level `pack` and `unpack` commands.
4708 See the Tcl documentation at: http://www.tcl.tk/man/tcl8.5/TclCmd/binary.htm
4710 Note that 'binary format' with f/r/R specifiers (single-precision float) uses the value of Infinity
4711 in case of overflow.
4715 The optional, pure-Tcl 'oo' extension provides object-oriented (OO) support for Jim Tcl.
4717 See the online documentation (http://jim.berlios.de/documentation/oo/) for more details.
4719 +*class* 'classname ?baseclasses? classvars'+::
4720 Create a new class, +'classname'+, with the given dictionary
4721 (+'classvars'+) as class variables. These are the initial variables
4722 which all newly created objects of this class are initialised with.
4723 If a list of baseclasses is given, methods and instance variables
4726 +*super* 'method ?args \...?'+::
4727 From within a method, invokes the given method on the base class.
4728 Note that this will only call the last baseclass given.
4732 The optional, pure-Tcl 'tree' extension implements an OO, general purpose tree structure
4733 similar to that provided by tcllib ::struct::tree (http://tcllib.sourceforge.net/doc/struct_tree.html)
4735 A tree is a collection of nodes, where each node (except the root node) has a single parent
4736 and zero or more child nodes (ordered), as well as zero or more attribute/value pairs.
4739 Creates and returns a new tree object with a single node named "root".
4740 All operations on the tree are invoked through this object.
4743 Destroy the tree and all it's nodes. (Note that the the tree will also
4744 be automatically garbage collected once it goes out of scope).
4746 +$tree *set* 'nodename key value'+::
4747 Set the value for the given attribute key.
4749 +$tree *lappend* 'nodename key value \...'+::
4750 Append to the (list) value(s) for the given attribute key, or set if not yet set.
4752 +$tree *keyexists* 'nodename key'+::
4753 Returns 1 if the given attribute key exists.
4755 +$tree *get* 'nodename key'+::
4756 Returns the value associated with the given attribute key.
4758 +$tree *getall* 'nodename'+::
4759 Returns the entire attribute dictionary associated with the given key.
4761 +$tree *depth* 'nodename'+::
4762 Returns the depth of the given node. The depth of "root" is 0.
4764 +$tree *parent* 'nodename'+::
4765 Returns the node name of the parent node, or "" for the root node.
4767 +$tree *numchildren* 'nodename'+::
4768 Returns the number of child nodes.
4770 +$tree *children* 'nodename'+::
4771 Returns a list of the child nodes.
4773 +$tree *next* 'nodename'+::
4774 Returns the next sibling node, or "" if none.
4776 +$tree *insert* 'nodename ?index?'+::
4777 Add a new child node to the given node. The index is a list index
4778 such as +3+ or +end-2+. The default index is +end+.
4779 Returns the name of the newly added node.
4781 +$tree *walk* 'nodename' *dfs|bfs* {'actionvar nodevar'} 'script'+::
4782 Walks the tree starting from the given node, either breadth first (+bfs+)
4783 depth first (+dfs+).
4784 The value +"enter"+ or +"exit"+ is stored in variable +'actionvar'+.
4785 The name of each node is stored in +'nodevar'+.
4786 The script is evaluated twice for each node, on entry and exit.
4789 Dumps the tree contents to stdout
4793 The optional tclprefix extension provides the Tcl8.6-compatible 'tcl::prefix' command
4794 (http://www.tcl.tk/man/tcl8.6/TclCmd/prefix.htm) for matching strings against a table
4795 of possible values (typically commands or options).
4797 +*tcl::prefix all* 'table string'+::
4798 Returns a list of all elements in +'table'+ that begin with the prefix +'string'+.
4800 +*tcl::prefix longest* 'table string'+::
4801 Returns the longest common prefix of all elements in +'table'+ that begin with the prefix +'string'+.
4803 +*tcl::prefix match* '?options? table string'+::
4804 If +'string'+ equals one element in +'table'+ or is a prefix to
4805 exactly one element, the matched element is returned. If not, the
4806 result depends on the +-error+ option.
4808 * +*-exact*+ Accept only exact matches.
4809 * +*-message* 'string'+ Use +'string'+ in the error message at a mismatch. Default is "option".
4810 * +*-error* 'options'+ The options are used when no match is found. If +'options'+ is
4811 empty, no error is generated and an empty string is returned.
4812 Otherwise the options are used as return options when
4813 generating the error message. The default corresponds to
4818 The optional history extension provides script access to the command line editing
4819 and history support available in 'jimsh'. See 'examples/jtclsh.tcl' for an example.
4820 Note: if line editing support is not available, `history getline` acts like `gets` and
4821 the remaining subcommands do nothing.
4823 +*history load* 'filename'+::
4824 Load history from a (text) file. If the file does not exist or is not readable,
4827 +*history getline* 'prompt ?varname?'+::
4828 Displays the given prompt and allows a line to be entered. Similarly to `gets`,
4829 if +'varname'+ is given, it receives the line and the length of the line is returned,
4830 or -1 on EOF. If +'varname'+ is not given, the line is returned directly.
4832 +*history add* 'line'+::
4833 Adds the given line to the history buffer.
4835 +*history save* 'filename'+::
4836 Saves the current history buffer to the given file.
4839 Displays the current history buffer to standard output.
4843 Provides namespace-related functions. See also: http://www.tcl.tk/man/tcl8.6/TclCmd/namespace.htm
4845 +*namespace code* 'script'+::
4846 Captures the current namespace context for later execution of
4847 the script +'script'+. It returns a new script in which script has
4848 been wrapped in a +*namespace inscope*+ command.
4850 +*namespace current*+::
4851 Returns the fully-qualified name for the current namespace.
4853 +*namespace delete* '?namespace ...?'+::
4854 Deletes all commands and variables with the given namespace prefixes.
4856 +*namespace eval* 'namespace arg ?arg...?'+::
4857 Activates a namespace called +'namespace'+ and evaluates some code in that context.
4859 +*namespace origin* 'command'+::
4860 Returns the fully-qualified name of the original command to which the imported command +'command'+ refers.
4862 +*namespace parent* ?namespace?+::
4863 Returns the fully-qualified name of the parent namespace for namespace +'namespace'+, if given, otherwise
4864 for the current namespace.
4866 +*namespace qualifiers* 'string'+::
4867 Returns any leading namespace qualifiers for +'string'+
4869 +*namespace tail* 'string'+::
4870 Returns the simple name at the end of a qualified string.
4872 +*namespace upvar* 'namespace ?arg...?'+::
4873 This command arranges for zero or more local variables in the current procedure to refer to variables in +'namespace'+
4875 +*namespace which* '?-command|-variable? name'+::
4876 Looks up +'name'+ as either a command (the default) or variable and returns its fully-qualified name.
4878 [[BuiltinVariables]]
4882 The following global variables are created automatically
4886 This variable is set by Jim as an array
4887 whose elements are the environment variables for the process.
4888 Reading an element will return the value of the corresponding
4889 environment variable.
4890 This array is initialised at startup from the `env` command.
4891 It may be modified and will affect the environment passed to
4892 commands invoked with `exec`.
4895 This variable is set by Jim as an array containing information
4896 about the platform on which Jim was built. Currently this includes
4897 'os' and 'platform'.
4900 This variable contains a list of paths to search for packages.
4901 It defaults to a location based on where jim is installed
4902 (e.g. +/usr/local/lib/jim+), but may be changed by +jimsh+
4903 or the embedding application. Note that +jimsh+ will consider
4904 the environment variable +$JIMLIB+ to be a list of colon-separated
4905 list of paths to add to +*auto_path*+.
4908 This variable holds the value of the -errorcode return
4909 option set by the most recent error that occurred in this
4910 interpreter. This list value represents additional information
4911 about the error in a form that is easy to process with
4912 programs. The first element of the list identifies a general
4913 class of errors, and determines the format of the rest of
4914 the list. The following formats for -errorcode return options
4915 are used by the Tcl core; individual applications may define
4916 additional formats. Currently only `exec` sets this variable.
4917 Otherwise it will be +NONE+.
4919 The following global variables are set by jimsh.
4921 +*tcl_interactive*+::
4922 This variable is set to 1 if jimsh is started in interactive mode
4926 This variable is set by Jim as an array containing information
4927 about the platform upon which Jim was built. The following is an
4928 example of the contents of this array.
4930 tcl_platform(byteOrder) = littleEndian
4931 tcl_platform(os) = Darwin
4932 tcl_platform(platform) = unix
4933 tcl_platform(pointerSize) = 8
4934 tcl_platform(threaded) = 0
4935 tcl_platform(wordSize) = 8
4936 tcl_platform(pathSeparator) = :
4939 If jimsh is invoked to run a script, this variable contains the name
4943 If jimsh is invoked to run a script, this variable contains a list
4944 of any arguments supplied to the script.
4947 If jimsh is invoked to run a script, this variable contains the number
4948 of arguments supplied to the script.
4951 The value of argv[0] when jimsh was invoked.
4953 CHANGES IN PREVIOUS RELEASES
4954 ----------------------------
4958 1. +platform_tcl()+ settings are now automatically determined
4959 2. Add aio `$handle filename`
4960 3. Add `info channels`
4961 4. The 'bio' extension is gone. Now `aio` supports 'copyto'.
4962 5. Add `exists` command
4963 6. Add the pure-Tcl 'oo' extension
4964 7. The `exec` command now only uses vfork(), not fork()
4965 8. Unit test framework is less verbose and more Tcl-compatible
4966 9. Optional UTF-8 support
4967 10. Optional built-in regexp engine for better Tcl compatibility and UTF-8 support
4968 11. Command line editing in interactive mode, e.g. 'jimsh'
4972 1. `source` now checks that a script is complete (.i.e. not missing a brace)
4973 2. 'info complete' now uses the real parser and so is 100% accurate
4974 3. Better access to live stack frames with 'info frame', `stacktrace` and `stackdump`
4975 4. `tailcall` no longer loses stack trace information
4976 5. Add `alias` and `curry`
4977 6. `lambda`, `alias` and `curry` are implemented via `tailcall` for efficiency
4978 7. `local` allows procedures to be deleted automatically at the end of the current procedure
4979 8. udp sockets are now supported for both clients and servers.
4980 9. vfork-based exec is now working correctly
4981 10. Add 'file tempfile'
4982 11. Add 'socket pipe'
4983 12. Enhance 'try ... on ... finally' to be more Tcl 8.6 compatible
4984 13. It is now possible to `return` from within `try`
4985 14. IPv6 support is now included
4987 16. Event handlers works better if an error occurs. eof handler has been removed.
4988 17. `exec` now sets $::errorCode, and catch sets opts(-errorcode) for exit status
4989 18. Command pipelines via open "|..." are now supported
4990 19. `pid` can now return pids of a command pipeline
4991 20. Add 'info references'
4992 21. Add support for 'after +'ms'+', 'after idle', 'after info', `update`
4993 22. `exec` now sets environment based on $::env
4995 24. Add support for 'lsort -index'
4999 1. Add support to `exec` for '>&', '>>&', '|&', '2>@1'
5000 2. Fix `exec` error messages when special token (e.g. '>') is the last token
5001 3. Fix `subst` handling of backslash escapes.
5002 4. Allow abbreviated options for `subst`
5003 5. Add support for `return`, `break`, `continue` in subst
5004 6. Many `expr` bug fixes
5005 7. Add support for functions in `expr` (e.g. int(), abs()), and also 'in', 'ni' list operations
5006 8. The variable name argument to `regsub` is now optional
5007 9. Add support for 'unset -nocomplain'
5008 10. Add support for list commands: `lassign`, `lrepeat`
5009 11. Fully-functional `lsearch` is now implemented
5010 12. Add 'info nameofexecutable' and 'info returncodes'
5011 13. Allow `catch` to determine what return codes are caught
5012 14. Allow `incr` to increment an unset variable by first setting to 0
5013 15. Allow 'args' and optional arguments to the left or required arguments in `proc`
5015 17. Add 'try ... finally' command
5021 Copyright 2005 Salvatore Sanfilippo <antirez@invece.org>
5022 Copyright 2005 Clemens Hintze <c.hintze@gmx.net>
5023 Copyright 2005 patthoyts - Pat Thoyts <patthoyts@users.sf.net>
5024 Copyright 2008 oharboe - Oyvind Harboe - oyvind.harboe@zylin.com
5025 Copyright 2008 Andrew Lunn <andrew@lunn.ch>
5026 Copyright 2008 Duane Ellis <openocd@duaneellis.com>
5027 Copyright 2008 Uwe Klein <uklein@klein-messgeraete.de>
5028 Copyright 2009 Steve Bennett <steveb@workware.net.au>
5031 Redistribution and use in source and binary forms, with or without
5032 modification, are permitted provided that the following conditions
5034 1. Redistributions of source code must retain the above copyright
5035 notice, this list of conditions and the following disclaimer.
5036 2. Redistributions in binary form must reproduce the above
5037 copyright notice, this list of conditions and the following
5038 disclaimer in the documentation and/or other materials
5039 provided with the distribution.
5041 THIS SOFTWARE IS PROVIDED BY THE JIM TCL PROJECT ``AS IS'' AND ANY
5042 EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO,
5043 THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A
5044 PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE
5045 JIM TCL PROJECT OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT,
5046 INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
5047 (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
5048 OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
5049 HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT,
5050 STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
5051 ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF
5052 ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
5054 The views and conclusions contained in the software and documentation
5055 are those of the authors and should not be interpreted as representing
5056 official policies, either expressed or implied, of the Jim Tcl Project.