Command: num


COMMAND

num – number selected Humdrum records according to user-defined criteria


SYNOPSIS

`` num [-efT] [-a **interp] [-i n] [-n regexp] [-o n] [-O n] [-p regexp] [-P regexp] [-r regexp] [-R regexp] [-s regexp] [-S regexp] [-x regexp] [-z regexp] [-Z regexp] [inputfile ...]``


DESCRIPTION

The num command produces sequential numerical output according to user-defined numbering criteria. In its default operation, num simply inserts numbers at the beginning of each data record – beginning with the value 1, and increasing by 1 for successive data records. However, num provides various options that allow the user to specify more precisely the kinds of data and conditions under which numbering occurs.

Numbers may be inserted prior to data tokens, appended following data tokens, or inserted in the middle of data tokens. Numbers may be output only for the first input spine, or for all input spines. Alternatively, numbers may be output in a separate output spine specified by the user. Numbers may be assigned only to those data records matching a given regular expression, or may be assigned to all records other than those matching a regular expression. Numerical counts may begin at any real or integer value and may be incremented or decremented by any real or integer value. Numbering may restart at some defined value whenever a certain regular expression is matched in the input. Numbers may be output only when certain conditions are met; for example, although counting may continue, outputting of numbers may be suspended or resumed when the input data match user-specified regular expressions.

By way of example, the num command might be used to number measures, phrases, chords, notes, rests, or other musically-pertinent features. (See EXAMPLES below.)


OPTIONS

The num command provides the following options:

-a **interp append a new spine (**interp) containing the output numbers
-e place numbers at end of data tokens (rather than at the beginning)
-f number all spines (all fields) in the input
-h displays a help screen summarizing the command syntax
-i n set increment value to n (defaults to 1)
-n regexp number only those records matching regexp
-o n set initial offset value to n (defaults to 1)
-O n set offset value to n after a reset
-p regexp horizontally position the output number immediately following the
  first occurrence of regexp on the line
-P regexp horizontally position the output number immediately prior to the
  first occurrence of regexp on the line
-r regexp resume numbering records when regexp is matched
-R regexp resume numbering records after regexp is matched
-s regexp suspend numbering records when regexp is matched
-S regexp suspend numbering records after regexp is matched
-T reset counter when all spines have exclusive interpretations
-x regexp exclude numbering those records matching regexp
-z regexp reset counter when record matches regexp
-Z regexp reset counter after record matches regexp

Options are specified in the command line.

Normally, the effect of num is to add numbers to data tokens already in the input. With the -a option, num creates a new spine which is appended to the right of the input stream. The numerical outputs of num become data records in this new output spine. The user can specify the name of the output interpretation via the command line parameter **interp. The -a option cannot be used with the -p or -P options.

The -e option causes numbers generated by num to be append to the end of each appropriate data token rather than the (default) beginning of each data token. The -e and -a options are mutually exclusive. In addition, the -e option cannot be used with the -p or -P options.

The -f option causes all spines (all fields) in the input to be numbered rather than the (default) first (left-most) spine. The -f option is mutually exclusive with the -a option. In addition, the -f option cannot be used with the -p or -P options.

The -i option allows the user to set the increment value for successive numbers. The default value is 1 – meaning that successive numerical outputs are 1 greater than the previous value. Negative increment values are also permissible. For example, the user might define an initial value beginning at 100, and decrement by 5 with each successive value.

The -n option causes num to output numerical values, only when the current data record matches a specified regular expression.

The -o option is used to define an initial (offset) value from which subsequent numbers are calculated. If no offset is defined, the default value is 1.

The -O option defines an offset value to which the counter will be returned each time a reset action occurs. The -O option should be used in conjunction with one of either the -T, -z or -Z options.

The -p and -P options allow the user to place any output numerical value in a particular (horizontal) place in the output line. In the case of -p the output number is positioned immediately following the first (left-most) string matching the specified regular expression. With the -P option, the output number is positioned immediately prior to the first string matching the specified regular expression. The -p and -P options cannot be used with either the -a, -e or -f options.

The -r option defines a condition under which the outputting of numbers will resume. Specifically, the user defines a regular expression with the -r option that, when matched, causes the immediate resumption of printing.

The -R option is similar to the -r option, with the exception that outputting of numbers is resumed after any record matching the specified regular expression.

The -s option causes the outputting of numbers to be suspended when an input record matches a specified regular expression. Although the numerical values are not outputted, the numerical values continue to be incremented in accordance with the defined counting conditions.

The -S option is similar to the -s option, with the exception that the outputting of numbers is suspended after any record matching the specified regular expression.

The -T option causes the counter to be reset (to the value specified by -O) whenever exclusive interpretations are encountered in all of the input spines. If no initial offset has been specified via the -O option, then the counter is reset to the value 1.

The -x option causes records matching a given regular expression to be excluded from the counting; no output is generated for such records. Note that when used in conjunction with the -n option, both the match and don’t match criteria must be fulfilled in order for the current record to participate in the counting.

The -z option causes the counter to be reset (to the value specified by -O) whenever a data record matches a specified regular expression. If no initial offset has been specified via the -O option, then the counter is reset to the value 1.

The -Z option is similar to the -z option, with the exception that the counter is reset after any record matching the specified regular expression.


EXAMPLES

The following examples illustrate how num may be used. Consider the following input (left spine) and corresponding num output (right spine). ````

**kern **plength
=23 .
{8a .
. .
8cc .
}8ee 3
{8g# .
=24 .
8dd .
8ee .
}8ff 4
8r .
. .
=25 .
{8gn .
8cc .
}8ee 3
{8f# .
=26 .
8cc .
8dd .
}8ee- 4
*- *-

The **plength output indicates the number of notes in each phrase for the corresponding **kern spine. The output was generated using the following command:

`` num -a ‘**plength’ -z ‘{‘ -x ‘[.r=]’ -s ‘{‘ -r ‘}’ -S ‘}’``

The -x option excludes **kern rests, barlines, and null tokens from the counting. The -z option causes the counter to be reset to 1 whenever a begin-phrase signifier (`{‘) is encountered. The -s option causes suspension of output numbers to occur at the beginning of each phrase, and the -r option causes output numbers to be resumed at the end of each phrase (hence, only the phrase-end signifiers are given numbered output). The -S option ensures that numbers are not printed for notes outside of phrases; that is, it suspends outputting numbers following the end of a phrase.§ .FS § Note that this command will still fail to suppress the numbering of notes occuring prior to the first phrase. .FE The -a option causes the numbers to be output as a separate spine labelled **plength.

The command

`` num -a ‘**ordo’ koto``

outputs a new spine labelled **ordo containing successive integers beginning at 1 for each data record in the input.

`` num -n ‘^=’ -x ‘==’ -p ‘=’ -o 108 sarod``

numbers all “common system” barlines in the file sarod, beginning with measure 108. Double barlines are not numbered (due to the -x option) and numbers are positioned directly following the equals sign (due to the -p option). The -p option ensures that the number precedes pause markings and other possible barline signifiers. Note that if measure numbers already exist for a file, the measures can be renumbered by first removing the existing measure numbers using humsed.

The command

`` num -a ‘**phrase#’ -n ‘{‘ -T rebec``

outputs a spine containing numbers that number the beginning of each **kern phrase for the file rebec; if any exclusive interpretation is encountered, the phrase numbering is restarted at 1.

The command

`` num -x ‘^=’``

numbers all data records other than common system barlines.

`` num -x ‘^=’ -Z ‘=’``

numbers all data records within each common system measure – starting at the value 1 with each new measure.


PORTABILITY

DOS 2.0 and up, with the MKS Toolkit. OS/2 with the MKS Toolkit. UNIX systems supporting the Korn shell or Bourne shell command interpreters, and revised awk (1985).


SEE ALSO

nl (UNIX), **ordo (2), **regexp** (4), regexp (6), **rend** (4)


NOTES

The -O option should be used in conjunction with one of either the -T, -z or -Z options.