Rinci::function - Metadata for your functions/methods




This document describes version 1.1.103 of Rinci::function (from Perl distribution Rinci), released on 2022-10-21.


This document describes metadata for functions/methods. Since the metadata properties describe features and the way a function works, this document also describes how a function should support those properties.

This specification is part of Rinci. Please do a read up on it first, if you have not already done so.


Enveloped result

Function should return an enveloped result to express error code/message as well as actual result. The envelope can be produced by the function itself, or added by a wrapper tool. Result envelope is modeled after HTTP or PSGI response; it is an array in the following format:


STATUS is a 3-digit integer and is the only required element, much like HTTP response status code and is explained further in "Envelope status codes". MESSAGE is a string containing error message. RESULT (or PAYLOAD) is the actual content to be returned and can be omitted or set to undef if the function does not need to return anything. META is called result metadata, a defhash containing extra data, analogous to HTTP response headers. Result metadata is specified further in Rinci::resmeta.

Some example of an enveloped results:

 [200, "OK", 42]
 [404, "Not found"]
 [200, "Account created", {id=>9323}, {}]
 [500, "Can't delete foo: permission denied", undef,
  {"cmdline.exit_code"=>300, perm_err=>1}]

As mentioned, an enveloped result can contain error code/message as well as the actual result. It can also be easily converted to HTTP response message. And it can also contain extra data, useful for things like the transaction protocol (explained in Rinci::Transaction).

Special arguments

Special arguments are some known arguments that start with dash (-) and serve special purposes. You need not specify them in the args metadata property. Examples of special arguments include -dry_run, -tx_action, -action, -confirm/*, -arg_part_start, -arg_part_len, -res_part_start, -res_part_len. Many of them will be explained in other related sections/documents, but some will be described here:

  • -content_type_ARGNAME

    This special argument can be used to describe the MIME content type of an argument, where ARGNAME is the argument name. For example, if you are creating a function that accepts data upload (in Perl):

     upload(data => '...', -content_type_data => 'image/jpeg');

Functions vs methods

Since in many programming languages (like Perl 5, Python, Ruby, PHP) static functions are not that differentiated from methods, functions and methods share the same Rinci spec. But there are certain properties that can be used to declare if a function is (also) a method or not. See is_func, is_meth, is_class_meth properties below for details.

Multiple dispatch

This specification also does not (yet) have any recommendation on how to best handle functions in languages that support multiple dispatch, like Perl 6: whether we should create multiple metadata or just one. It is more up to the tool and what you want to do with the metadata.

Envelope status codes

In general, status codes map directly to HTTP response status codes. Below are the suggestion on which codes to use (or avoid). An asterisk (*) marks which codes are not defined in HTTP specification and introduced by this specification.

  • 1xx code

    Currently not used. Some tools like Perinci::CmdLine assumes that status code is always >= 200, so you should perhaps never used status less than 200.

  • 2xx code - success

    200 should be used to mean success.

    201 (created) can be used to signal success when creating something.

    204 (no content) can also be used when there is no content to show. In HTTP, browsers won't clear current page if response is 206 (instead of showing a blank page if response is 200). Currently in Rinci there is no difference between 200 and 204.

    206 (partial content) can be used to signal partial content. See partial property in the result function property.

    207 (multistatus) can be used to signal partial success, for example a function which processed 5 items wanted to report that it successfully processed 2 items but failed to process the rest 3 items. For each item's statuses, you can use result metadata property results. Note: 207 is used by WebDAV. Note: When all functions instead of only some/partial items fail, you might want to return 4xx or 5xx instead of 207.

  • 3xx code - further actions needs to be taken by user agent (caller)

    301 (moved) can be used to redirect callers to alternate location, although this is very rare.

    304 (not modified, nothing done). Used for example by setup functions to indicate that nothing is being modified or no modifying action has been performed (see Setup::* modules in CPAN).

    331* (confirmation required). Function requires confirmation, for example if action to be taken is dangerous or requires user's attention. Confirmation message/prompt from function can be returned in the message, or in the confirm_prompt property (e.g. to provide translations). Confirmation from caller is in the form of special argument -confirm with boolean value of true (TODO: A more detailed confirmation for different actions can be specified later in the form of -confirm_XXX special arguments.) For an example of application of this, see Perinci::Tx::Manager.

  • 4xx code - client (caller) side error

    400 (bad request, bad arguments) should be returned when the function encounters invalid input. A function wrapper can return this code when the function arguments fail the argument schema validation (specified in the args property).

    401 (authentication required).

    403 (forbidden, access denied, authorization failed).

    404 (not found). Can be used for example by an object-retrieval functions (like get_user()) and the object is not found.

    For object-listing functions (like list_users()), when there are no users found matching the requested criteria, 200 code should still be returned with an empty result (like an empty array or hash).

    Also in general, an object-deletion function (like delete_user()) should also return 200 (or perhaps 304, but 200 is preferred) instead of 404 when the object specified to be deleted is not found, since the goal of the delete function is reached anyway.

    408 (request timeout).

    409 (conflict). Can be used for example by a create_user() function when receiving an already existing username.

    412 (precondition failed). Similar to 409, but can be used to indicate lack of resources, like disk space or bandwidth. For lacking authentication and authorization, use 401 and 403 respectively.

    413 (request range not satisfiable). When client requests partial result outside the range of available. See the partial property in result property for examples.

    429 (too many requests).

    (EXPERIMENTAL) 44x* codes are reserved for function-specific codes. Each function is free to define what each number means. However, this is not really encouraged and should only be used if necessary. Function should perhaps stick to predefined codes here. To return more detailed status, result metadata can be used.

    480* is general transaction error, e.g. transaction status is aborted so further requests for this transaction is ignored until transaction is aborted.

    484* (no such transaction).

  • 5xx code - server (callee) side error

    500 is the general code to use when a failure occurs during the execution of a function. for example when a delete_file() function fails to delete specified file (though in this case it can also choose to return 403 instead, which is more specific).

    501 (not implemented)

    503 (service unavailable). You can use this when service is temporarily unavailable, e.g. when system load is too high, a required service is down, etc. Users should try again at a later time.

    507 (insufficient storage)

    521 (maximum retries reached)

    531* (bad metadata) is used when there is something wrong with the metadata.

    532* (failure in recording transaction) when there is a failure in updating transaction status or in preparing/committing/rolling back the transaction.

    (EXPERIMENTAL) 54x* codes are reserved for function-specific codes. Each function is free to define what each number means. However, this is not really encouraged and should only be used if necessary. Function should perhaps stick to predefined codes here. To return more detailed status, result metadata can be used.

    Try not to use code greater than 555, as some tools use (CODE-300) for error codes that must fit in one unsigned byte (like Perinci::CmdLine).


Bool. Specify that the function can be called as a static function (i.e. procedural, not as a method). Default is true if unspecified, but becomes false if is_meth or is_class_meth is set to true.


 # specify that function can be called a method *as well as* a static function
 is_meth => 1
 is_func => 1 # if not specified, will default to false after is_meth set to 1


Bool. Specify that the function can be called as an instance (object) method. Default is false.


 # specify that function is a method
 is_meth => 1


Bool. Specify that the function can be called as a class method. Examples of class methods include the constructor, but there are others. Default is false.


 # specify that function is a class method
 is_class_meth => 1

args (function property)

Hash. Specify arguments. Property value is defhash of argument names and argument specification. Argument name must only contain letters, numbers, and underscores (and do not start with a number).

Argument specification is a hash containing these keys:

schema (argument property)

Data::Sah schema for argument value.


Any. Give default value for argument. This takes precedence over schema, which can also specify default value. This is useful if you want to share a common schema over several arguments but want to have different default for each argument. For example, you have a ticket_status schema. In create_ticket function you want the default status argument to be new, while in reply_ticket you want the default status to be answered.

summary (argument property)

Str. From DefHash. A one-line plaintext summary, much like the summary property in variable metadata.


Bool. Specify that argument is required (although its value can be undef/null). Default is false.

description (argument property)

Str. From DefHash. A longer description of marked up text, much like the description property. It is suggested to format the text to 74 columns.

tags (argument property)

Array of (str|hash). From DefHash. An array of tags, can be used by tools to categorize arguments. Not unlike the tags property.


Non-negative integer. Argument position when specified in an ordered fashion, e.g. in an array. Starts from zero.


Bool. Only relevant if pos is specified, specify whether argument should gobble up all remaining values in an ordered argument list into an array.

Old alias: "greedy"


Old alias for "slurpy". Will be removed in Rinci 1.2.

partial (argument property)

Bool. Whether or not argument value can be sent partially. Only argument types that have a length notion can be set as partial (e.g. str where the unit is character, buf where the unit is byte, and array where the unit is element).

There should at most be one argument with this property set to true.

To send partial argument, you can use the special arguments -arg_len (total argument length), -arg_part_start (start position of the part that is being sent), -arg_part_len (length of part that is being sent). Example (in Perl):

 # function metadata
     v => 1.1,
     summary => "Upload a file",
     args => {
         name => {schema=>"str*", req=>1},
         data => {schema=>"buf*", req=>1, partial=>1},

 # function usage example: send the first 10MiB of data
 upload_file(name=>"myvideo.mp4", data=>substr($data, 0, 10_000_000),
             -arg_len        => 24_500_000,
             -arg_part_start =>          0,
             -arg_part_len   => 10_000_000);
 # send the next 10MiB
 upload_file(name=>"myvideo.mp4", data=>substr($data, 10_000_000, 10_000_000),
             -arg_len        => 24_500_000,
             -arg_part_start => 10_000_000,
             -arg_part_len   => 10_000_000);
 # send the last 4.5 MiB
 upload_file(name=>"myvideo.mp4", data=>substr($data, 20_000_000),
             -arg_len        => 24_500_000,
             -arg_part_start => 20_000_000,
             -arg_part_len   =>  4_500_000);

stream (argument property)

Bool. By setting this property to true, function can specify that it accepts streaming data for this argument. It is useful when argument value is large or of undetermined/infinite length. To send value as a stream, caller must send a subroutine reference (callback) instead which the function will call repeatedly until it gets undef to signify exhaustion of data.


Hash. Specify aliases for use in command-line options (or other possibly suitable situation where arguments are parsed from command-line-like options). Keys are alias names, values are itself hashes (alias specification). Valid alias specification keys: summary (a string, optional), schema (optional, defaults to argument's schema), is_flag (bool, optional, if set to 1 then it is a shortcut for specifying schema to ["bool", {"is":1}]), code (a code to set argument value, optional, will be given (\%args, $val)); if not set, the default behavior is simply to set the argument value).


Code. A hook that will be called when argument is specified as a command-line option. In Perl, hook will be called with a hash argument containing this key: arg (str, argument name), value (str, option value), args (hash, the argument hash defined so far).

This can be useful if you want to process a command-line option directly on a per-option basis instead of getting the final resulting argument value. For example (in Perl):

 args => {
     library => {
         schema            => ['array*' => of => 'str*'],
         cmdline_aliases   => { I => {} },
         cmdline_on_getopt => sub {
             my %args = @_;
             require lib;
     module => {
         schema            => ['array*' => of => 'str*'],
         cmdline_aliases   => { M => {} },
         cmdline_on_getopt => sub {
             my %args = @_;
             require Module::Load;

With command-line argument like this:

 -I dir1 -M mod1 -I dir2 -M mod2

Without any cmdline_on_getopt hooks, the function will receive this argument hash:

 { library => ['dir1', 'dir2'], module => ['mod1', 'mod2'] }

but there is no way to know the order of options being specified in the command-line. With the hooks, the function can load modules correctly (e.g. loading mod1 won't search in dir2 as that directory has not been added by -I).

completion (argument property)

Code. A code to supply argument value completion. Will be explained in the examples.


Code. A code to supply argument element index completion. Applicable to the following argument types:

hash: for completing hash pair keys. (See also element_completion for completing hash pair values).

See examples for how to use this property.


Code. A code to supply argument element value completion. Applicable to the following argument types:

array: for completing array element values.

hash: for completing hash pair values. (See also index_completion for completing hash keys).

See examples for how to use this property.


Experimental. Bool. Describe that argument holds password. Programs can react to this in several ways, for example they can turn off echoing to terminal when asking value from standard input. Or they can redact values to **** when logging.


Str. Normally in a command-line program, command-line options and arguments will map to function arguments, e.g. --arg 2 will set the foo argument to 2 and positional arguments (argument which specifies the pos property and optionally also a slurpy property with true value) will get or slurp command-line arguments.

In some cases, this is not convenient. When supplying larger amount of data, a complex structures, or a stream, we might want to use other sources. The cmdline_src property can be set to one of the following value for this purpose:

  • file

    Command-line option/argument value will be treated as filename and function argument will be set to content of the file (or in the case of streaming argument, to a callback which can be used to get the file's content).

  • stdin

    This means that program should get function argument from the whole standard input until EOF.

    Note that only one argument can have its source set to stdin or stdin_or_file or stdin_or_files or stdin_or_args.

  • stdin_or_file

    This means that program should get argument value from content of file (the name of which is given from the first remaining command-line argument after all other arguments/options have been processed), or, if none is provided, from standard input.

  • stdin_or_files

    This means that program should get argument value from content of files (the names of which are taken from the remaining command-line arguments after all other arguments/options have been processed), or, or if no command-line argument remains, from standard input. This behavior is similar to the <> (diamond operator) in Perl.

  • stdin_or_args

    This means that program should get argument value from remaining command-line argument(s), or if no command-line argument remains, from standard input.

  • stdin_line

    This means that program should get argument value from a line of standard input; newline ending will be removed from the argument value.

    Arguments with cmdline_src of stdin_line will be processed before stdin/stdin_or_file/stdin_or_files/stdin_or_args.

Other sources might be defined in the future.

TODO: Define web_src property?

TODO: A way to define record separator?


Str. String to display when asking for argument value from stdin (if cmdline_src property value is stdin_line.

TODO: cmdline_prompt_template?


Experimental. Hash. This allows specifying argument submetadata, used e.g. when dealing with forms (a form field/widget can be a subform). Value is Rinci function metadata.


Experimental. Hash. This allows specifying argument element submetadata, used e.g. when dealing with forms (a form field/widget can contain an array of records/subforms). Value is Rinci function metadata.

deps (argument property)

Hash. This property specifies argument's dependencies to other arguments (but possibly to other things too, in the future). This is similar to function's "deps (function property)" property. It is a hash or dep types and values. The most important dep type is arg (dependency to another argument). Some dep types are special: all, any, none. Example:

 # function metadata property
 args => {
     delete => {
     force => {
         summary => 'Force deletion',
         schema => 'bool',
         deps => {arg=>'delete'},

The above example states that argument force "depends on" delete. What it means (usually) is that specifying force only makes sense when delete is also specified. In a CLI context:

 % prog --delete --force

Specifying --force without --delete doesn't make sense.

To depend on any one of several arguments, use the dep type any, e.g.:

 # function metadata property
 args => {
     delete => {
     replace => {
     force => {
         summary => 'Force deletion',
         schema => 'bool',
         deps => {any => [

In CLI context, this means --force can be specified when --delete or --replace is specified.

To depend on several arguments, use the dep type all, e.g.:

 # function metadata property
 args => {
     red => {
         pos => 0,
     green => {
         pos => 1,
     blue => {
         pos => 2,
     rgb16 => {
         schema => 'Use 3x16bits = 48bits to represent RGB value instead of the default 3x8 = 24bits',
         deps => {all => [

In CLI context, this means specifying --rgb16 only makes sense when all of --ed, --green, --blue are specified.


Experimental. Array of string, or code. Filters to apply before argument is converted from text and validated.

examples (argument property)

Array. Each element is sample argument values. But if it is a hash, will be assumed as a DefHash with the actual value put in the value property (so if your sample argument is a hash like {} (in JSON), you have to specify it as {"value":{}, "summary":"Optional summary..."} (in JSON).

Note that a Sah schema can also have an examples clause which you can use to put examples in. And Rinci function metadata also has examples property too.

Example function metadata and its implementation in Perl:

 $SPEC{multiply2} = {
     v => 1.1,
     summary => 'Multiple two numbers',
     args => {
         a => {
             summary => 'The first operand',
             description => '... a longer description ...',
             schema=>['float*', {
                 examples => [1, -10, 0, 3.333],
             pos => 0,
             tags => ['category:operand'],
         b => {
             summary => 'The second operand',
             description => '... a longer description ...',
             schema => 'float*',
             pos => 1,
             tags => ['category:operand'],
             examples => [
                 1, -10, 0, 3.333,
                 {value => 1e-10, summary => 'Blah blah'},
         round => {
             summary => 'Whether to round result',
             description => '... a longer description ...',
             schema => [bool => {default=>0}],
             pos => 2,
             tags => ['category:options'],
             cmdline_aliases => {
                 R=>{summary=>'Equivalent to --round=0',
                     code=>sub { my ($args, $val) = @_; $args->{round}=0 }},
 sub multiply2 {
     my %args = @_;
     my $res = $args{a} * $args{b};
     $res = int($res) if $round;
     [200, "OK", $res];

By default, without any wrapper, the function is called with a named hash style:

 multiply2(a=>4, b=>3);  # 12

But with the information from the metadata, a wrapper tool like Perinci::Sub::Wrapper is able to change the calling style to positional:

 multiply2(4, 3.1, 1);  # 12

A command-line tool will also enable the function to be called named options as well as positional arguments:

 % multiply2 --a 2 --b 3
 % multiply2 2 --b 3
 % multiply2 2 3

As mentioned earlier, cmdline_alises is parsed by command-line option parser:

 % multiply2 2 3.5 -r ; # equivalent to multiply2 2 3 --round
 % multiply2 2 3.5 -R ; # equivalent to multiply2 2 3 --noround (--round=0)

Aliases in cmdline_aliases are not recognized as real arguments:

 multiply2(a=>4, b=>3, r=>0);  # unknown argument r

Another example (demonstrates cmdline_aliases):

 $SPEC{smtpd} = {
     v => 1.1,
     summary => 'Control SMTP daemon',
     args    => {
         action => {
             schema => ['str*' => {in=>[qw/status start stop restart/]}],
             pos    => 0,
             req    => 1,
             cmdline_aliases => {
                 status => {
                     schema    => [bool=>{is=>1}],
                     summary   => 'Alias for setting action=status',
                     code      => sub { $_[0]{action} = 'status' },
                 start => {
                     schema    => [bool=>{is=>1}],
                     summary   => 'Alias for setting action=start',
                     code      => sub { $_[0]{action} = 'start' },
                 stop => {
                     schema    => [bool=>{is=>1}],
                     summary   => 'Alias for setting action=stop',
                     code      => sub { $_[0]{action} = 'stop' },
                 restart => {
                     schema    => [bool=>{is=>1}],
                     summary   => 'Alias for setting action=restart',
                     code      => sub { $_[0]{action} = 'restart' },
         force => {
             schema => 'bool',

Another example (demonstrates slurpy):

 $SPEC{multiply_many} = {
     v => 1.1,
     summary => 'Multiple numbers',
     args    => {
         nums   => {
             schema => ['array*' => {of=>'num*', min_len=>1}],
             pos    => 0,
             slurpy => 1
 sub multiply_many {
     my %args = @_;
     my $nums = $args{nums};

     my $ans = 1;
     $ans *= $_ for @$nums;
     [200, "OK", $ans];

After wrapping, in positional mode it can then be called:

 multiply_many(2, 3, 4);  # 24

which is the same as (in normal named-argument style):

 multiply_many(nums => [2, 3, 4]);  # 24

In command-line:

 % multiply-many 2 3 4

in addition to the normal:

 % multiply-many --nums '[2, 3, 4]'

completion. This argument specification key specifies how to complete argument value (e.g. in shell or Riap::HTTP) and is supplied an anonymous function as value. The function will be called with arguments: word=>... (which is the formed word so far, ci=>0|1 (whether completion should be done case-insensitively). The function should return an array containing a list of possible candidates, or a hash containing these keys: completion (array, list of possible candidates) and extra keys for formatting hints e.g. is_path (bool, whether the list of completion is path-like, meaning it can be traversed/dug to multiple levels) path_sep (string, path separator character), type (string, either filename, env, or other types). For an example of implementation for this, see Perinci::Sub::Complete in Perl which provides tab completion for argument values. Example:

 $SPEC{delete_user} = {
     v => 1.1,
     args => {
         username => {
             schema     => 'str*',
             pos        => 0,
             completion => sub {
                 my %args = @_;
                 my $word = $args{word} // "";

                 # find users beginning with $word
                 local $CWD = "/home";
                 return [grep {-d && $_ ~~ /^\Q$word/} <*>];
         force => {schema=>[bool => {default=>0}]},

When delete_user is executed over the command line and the Tab key is pressed:

 $ delete-user --force --username fo<tab>
 $ delete-user fo<tab>

then bash will try to complete with usernames starting with fo.

element_completion. This is like completion, but for array or hash elements. Argument type must be array or hash. Example for array:

 $SPEC{delete_users} = {
     v => 1.1,
     args => {
         usernames => {
             schema     => ['array*' => of => 'str*'],
             req        => 1,
             pos        => 0,
             slurpy     => 1,
             element_completion => sub {
                 my %args = @_;
                 my $word = $args{word} // "";

                 # find users beginning with $word
                 local $CWD = "/home";
                 my $res = [grep {-d && $_ ~~ /^\Q$word/} <*>];

                 # exclude users already mentioned by user
                 my $ary = $args{args}{usernames};
                 $res = [grep {!($_ ~~ @$ary)}] @$res;

                 return $res;

When delete_users is executed over the command line:

 $ delete-users c<tab> ; # will complete with all users beginning with c
 $ delete-users charlie c<tab> ; # will complete with users but exclude charlie
 $ delete-users charlie chucky <tab> ; # and so on

Example for hash (as well as index_completion property to complete hash keys):

 $SPEC{create_file} = {
     v => 1.1,
     args => {
         filename => {
             schema => 'str*',
             req => 1,
             pos => 0,
         content => {
             schema => 'buf*',
             req => 1,
             pos => 1,
         mode => {
             summary => 'Permission mode',
             schema => 'posint*',
         extra_attrs => {
             '' => 1,
             '' => 'extra_attr',
             schema     => ['hash*' => of => 'str*'],
             index_completion => sub {
                 # complete with list of known attributes
                 my %args = @_;
                 require Complete::Util;
                     word => $args{word},
                     array => [qw/mtime ctime owner group/],
             element_completion => sub {
                 my %args = @_;
                 my $word  = $args{word} // "";
                 my $index = $args{index};

                 if ($index eq 'owner') {
                     require Complete::Unix;
                     return Complete::Unix::complete_user(word=>$word);
                 } elsif ($index eq 'group') {
                     require Complete::Unix;
                     return Complete::Unix::complete_group(word=>$word);
                 } else {
                     return undef;

When create_file is executed over the command line:

 $ create-file file1 "hello filesystem" --extra-attr <tab>; # will complete with list of known attributes
 $ create-file file1 "hello filesystem" --extra-attr owner=<tab>; # will complete with list of Unix users


Hash. This property is used to expression relationships between arguments. The value is actually Sah schema hash clause set (see hash type in Sah::Type). The arguments are represented as a hash, and you can use the various Sah clauses to express relationships between the arguments (hash keys) because the Sah hash type supports such clauses, e.g. choose_one, choose_all, req_one, req_all, dep_any, dep_all, req_dep_any, req_dep_all.


 args_rels => {
     choose_one => ['delete', 'add', 'edit'],
     choose_all => ['red', 'green', 'blue'],

The above example says that only one of delete, add, edit can be specified. And if any of red, green, blue is specified then all must be specified. In CLI context this translates to:

 % prog --delete item
 % prog --delete --add item ; # error, both --delete and --add specified

 % prog --red 255 --green 255 --blue 0
 % prog --red 255 --blue 0 ; # error, --green is missing

Another example:



Str. Specify in what form the function expects the arguments. The value is actually implementation-specific since it describes the function implementation. For example in Perinci for Perl, these values are recognized: array, hash, arrayref, hashref. This property is useful for wrapper to be able to convert one form to another.

The default value is also left to the implementation.

For interimplementation communication (e.g. via Riap::HTTP or Riap::TCP), named arguments are always used so this property is irrelevant.


DefHash. Specify function return value. It is a defhash containing keys:

  • summary (result property)

    From DefHash. Like the summary property in variable metadata.

  • description (result property)

    From DefHash. Like the description property. Suggested to be formatted to 78 columns.

  • schema (result property)

    A Sah schema to validate the result (the third element in the envelope result). This schema should only be tested if status is 200. See also: statuses.

  • statuses

    Hash. Can be used to specify different result schema for different statuses. For example:

     statuses => {
         206 => {
             schema => 'str*',
  • stream (result property)

    Bool. Specify that function returns streaming output. Note that function can also signify streaming output by setting result metadata property stream to true.

    Function must then return a subroutine reference (callback) as its actual result which the caller can call repeatedly until it gets undef to signify exhaustion.

  • partial (result property)

    Bool. If set to true, specify that it is possible to request partial result. An example is in a function that reads contents from (potentially large) files:

     # function metadata
         v => 1.1,
         summary => 'Read file contents',
         args => {
             name => {
                 summary => 'File name',
                 schema  => 'str*',
                 req     => 1,
         result => {schema=>'buf*', partial=>1},
     # example function usage: request to read first 10MiB of file content,
     # -result_part_start defaults to 0.
     my $res = read_file(name=>'myvideo.mp4', -res_part_len=>10000000);
     # => [206, "Partial content", "data...", {len=>24500000, part_start=>0, part_len=>10000000}]
     # request the next 10MiB
     my $res = read_file(name=>'myvideo.mp4', -res_part_start=>10000000, -res_part_len=>10000000);
     # => [206, "Partial content", "data...", {len=>24500000, part_start=>10000000, part_len=>10000000}]
     # request the next 10MiB, since this actual file size is only 24500000,
     # function should return 416 status
     my $res = read_file(name=>'myvideo.mp4', -res_part_start=>20000000, -res_part_len=>10000000);
     # => [416, "Request range not satisfiable, file size is only 24500000"]
     # request the next 4.5MiB, this time succeeds
     my $res = read_file(name=>'myvideo.mp4', -res_part_start=>20000000, -res_part_len=>4500000);
     # => [206, "Partial content", "data...", {len=>24500000, part_start=>20000000, part_len=>4500000}]

    Partial result request to a function which does not support partial result might have the effect of the whole content being returned (status 200) or status 416.

Note that since functions normally return enveloped result, instead of returning:


your functions normally have to return an enveloped result:



 # result is an integer
 result => {schema => 'int*'}

 # result is an integer starting from zero
 result => {schema => ['int*' => {ge=>0}]}

 # result is an array of records
 result => {
     summary => 'Matching addressbook entries',
     schema => ['array*' => {
         summary => 'blah blah blah ...',
         of      => ['hash*' => {allowed_keys=>[qw/name age address/]} ]


Bool. If set to true, specify that function does not envelope its results. The default is false, to encourage functions to create envelopes. However, wrapper should be able to create or strip envelope if needed. For example, if you have "traditional" functions which does not do envelopes, you can set this property to true, and the wrapper can generate the envelope for the functions.

examples (function property)

Array. This property allows you to put examples in a detailed and structured way, as an alternative to putting everything in description.

Each example is a defhash, it specifies what arguments are used, what the results are, and some description. It can be used when generating API/usage documentation, as well as for testing data. It can also be used for testing (function will be run with specified arguments and the result will be matched against expected result). Known properties:

  • args (function example property)

    Hash. Arguments used to produce result. Can be converted to argv by tool, e.g. when displaying command-line examples.

  • argv

    Array. An alternative to args, for example when function is run from the command-line. Can be converted to args most of the time when wanting to display examples in Perl instead of command-line.

  • src

    Str. An alternative to args or argv, to provide raw source code. See also: src_plang. This can be used to show more general examples. For example, you can show how a function is used in an expression or code block, or how a command-line program is used in a shell script.

    Exactly one of args, argv, or src must be specified.

  • src_plang

    Str. The programming language the examples source code src is written in. Valid values include: perl, bash.

    Command-line interface tools will typically only show examples written in bash or other shells, while Perl module tools will typically only show perl examples.

    Required if src is specified.

  • status

    Int. Status from envelope. If unspecified, assumed to be 200.

  • result

    Any. Expected result.

  • env_result

    Array. Expected enveloped result.

  • naked_result

    Any. Expected naked result (payload).

  • summary (function example property)

    From DefHash. A one-line summary of the example You should describe, in one phrase or sentence, what the example tries to demonstrate. You can skip the summary if the example is pretty basic or things are already clear from the args alone.

  • description (function example property)

    From DefHash. Longer marked up text about the example (e.g. discussion or things to note), suggested to be formatted to 72 columns.

  • tags (function example property)

    From DefHash.

  • test

    Bool. Defaults to true. Whether to actually test example or not. Examples are by default run as tests by a test module (e.g. Perl module Test::Rinci. Setting this to 0 disables this example from being included in a test.

    TODO: more detailed testing instruction (e.g. only test in release candidate, or under certain environment flag, etc).


 # part of metadata for Math::is_prime function
 examples => [
         args => {num=>10},
         result => 0,
         # summary no needed here, already clear.
         args => {},
         result => 400,
         summary => 'Num argument is required',

         argv => [-5],
         result => 1,
         summary => 'Also works for negative integers',

Another example demonstrating src for a function called list_countries:

 examples => [
         src => 'for c in `list-countries`; do wget$c; done',
         src_plang => 'bash',
         src => <<'EOT',
 my $res = list_countries(detail => 1, sort=>['-popsize']);
 die "Can't list countries: $res->[0] - $res->[1]" unless $res->[0] == 200;
 my $i = 0;
 for my $c (@{ $res->[2] }) { $i++; say "$i. $_->{name}'s population: $_->{popsize}";
         src_plang => 'perl',


DefHash. Allows functions to express their features. Each hash key contains feature name, which must only contain letters/numbers/underscores.

Below is the list of defined features. New feature names may be defined by extension.

  • reverse

    Bool. Default: false. If set to true, specifies that function supports reverse operation. To reverse, caller can add special argument -reverse. For example:

     $SPEC{triple} = {
         v => 1.1,
         args     => {num=>{schema=>'num*'}},
         features => {reverse=>1}
     sub triple {
         my %args = @_;
         my $num  = $args{num};
         [200, "OK", $args{-reverse} ? $num/3 : $num*3];
     triple(num=>12);              # => 36
     triple(num=>12, -reverse=>1); # =>  4
  • tx

    Hash. Default is none. Specify transactional support, as specified in Rinci::Transaction. Value is a hash containing these keys: v (int, protocol version, default if not specified is 1).

    Please see Rinci::Transaction for more details on transaction.

  • dry_run

    Bool, or hash. If set to a true value, specifies that function supports dry-run (simulation) mode. Can also be set to a hash like this: <{default=1}>> to mean that function supports dry-run and the default is dry-run mode. Example:

     use Log::ger;
     $SPEC{rmre} = {
         summary  => 'Delete files in curdir matching a regex',
         args     => {re=>{schema=>'str*'}},
         features => {dry_run=>1}
     sub rmre {
         my %args    = @_;
         my $re      = qr/$args{re}/;
         my $dry_run = $args{-dry_run};
         opendir my($dir), ".";
         while (my $f = readdir($dir)) {
             next unless $f =~ $re;
             log_info "Deleting $f ...";
             next if $dry_run;
             unlink $f;
         [200, "OK"];

    The above Perl function delete files, but if passed argument -dry_run => 1 (simulation mode), will not actually delete files, only display what files match the criteria and would have be deleted.

    Specifying a function as supporting dry_run means, among others:

    • If dry_run is requested, function will have no side effects

      It will behave like a pure function, and thus have the properties of a pure function.

  • pure

    Bool. Default: false. If set to true, specifies that function is "pure" and has no "side effects" (these are terms from functional programming / computer science). Having a side effect means changing something, somewhere (e.g. setting the value of a global variable, modifies its arguments, writing some data to disk, changing system date/time, etc.) Specifying a function as pure means, among others:

    • it can safely be inculded in transaction without recording in journal;

    • it can safely be included during dry run;

  • immutable

    Bool. Default is false. If set to true, specifies that function always returns the same result when given the same argument values. This enables optimization like memoization. An example of an immutable function is sub { $_[0]+$_[1] } where its results only depend on the arguments. Example of a mutable function would be rand() or read() that reads contents from a file.

  • idempotent

    Bool. Default is false. If set to true, specifies that function is idempotent. Idempotency means that repeated invocation of a function (each with the same arguments) will have the same effect as a single invocation. In other words, extra invocation will not have any effect.

    Some operations, like reading a database row or a file's content, is inherently idempotent (or to be exact nullipotent). Another example is setting or updating an entity to some specific value, or deleting some entity. Repeated invocation of the operation will still sets the entity to the same value, or still deletes the entity.

    Some other operations are inherently non-idempotent, for example sending an email. Repeated invocation will cause multiple emails to be sent.

    Yet some other operations are non-idempotent, but can be made idempotent simply by checking whether the target object(s) has (have) reached the final desired state, (optionally additionally also checking whether they are in the correct original state to begin with). For example, a function that renames a file can record the original file that was renamed (its MD5 checksum, size, or what not) or perhaps record the action in a history database or flag file, and refuse to rename again if the file to be renamed is not the original file.

  • check_arg

    Bool. Default is false. If set to true, specifies that function supports the action of checking only a single argument. Usually useful when doing form processing, where we want to check only a single form field (function argument). To check a single argument, one passes -action special argument with the value of check_arg and also passes the argument she wants to check. Function should check that argument and immediately return 200 status upon success, or 400 upon validation failure.

deps (function property)

Hash. This property specifies function's dependencies to various things. It is a hash of dep types and values. Some dep types are special: all, any, and none.

 deps => {
     all => [
         {DEPTYPE=>DEPVALUE, ...},
     any => [
         {DEPTYPE => DEPVALUE, ...},
     none => [
         {DEPTYPE => DEPVALUE, ...},

A dependency can be of any type: another function, environment variables, programs, OS software packages, etc. It is up to the dependency checker library to make use of this information.

For the dependencies to be declared as satisfied, all of the clauses must be satisfied.

Below is the list of defined dependency types. New dependency type may be defined by an extension.

  • env

    Str. Require that an environment variable exists and is true, where true is in the Perl sense (not an empty string or "0"; " " and "0.0" are both true). Example:

     env => 'HTTPS'
  • prog

    Str of defhash. Require that a program exists and is executable. If STR doesn't contain path separator character '/' it will be searched in PATH. Windows filesystem should also use Unix-style path, e.g. "C:/Program Files/Foo/Bar.exe".

     prog => 'rsync'   # any rsync found on PATH
     prog => '/bin/su' # won't accept any other su

    As an alternative to just a string, the dep value can also be a defhash. Currently known properties for the defhash: name (program name or path, str, required), path (program path, str, optional), min_version (minimum version, dotted number version str, optional).

  • pkg

    Str. Specify dependency on a Riap package. STR must be a valid Riap package URI string. Checker can check that requesting info on this URI succeeds and the type is indeed package. Example:

     pkg => '/Foo/'
  • func

    Str. Specify dependency on a Riap function. STR must be a valid Riap function URI string. Checker can check that requesting info on this URI succeeds and the type is indeed package. Example:

     pkg => '/Foo/somefunc'
     pkg => ''
  • code

    Code. Require that anonymous function returns a true value after called, where the notion of true depends on the host language. Example in Perl:

     code => sub {$>}  # i am not being run as root

    Example in Ruby:

     "code" => { Process.euid > 0 }  # i am not being run as root
  • tmp_dir

    Bool. If set to 1, specify that function requires temporary directory. Caller should provide path to this using special argument -tmp_dir.

  • trash_dir

    Bool. If set to 1, specify that function requires trash directory. Trash is not unlike a temporary directory. Caller should provide path to trash directory using special argument -trash_dir.

    Trash directory can be provided, e.g. by transaction manager (see Rinci::Transaction).

  • all

    Array of deps. A "meta" type that allows several dependencies to be joined together in a logical-AND fashion. All dependency hashes must be satisfied. For example, to declare a dependency to several programs and an environment variable:

     all => [
         {prog => 'rsync'},
         {prog => 'tar'},
         {env  => 'FORCE'},
  • any

    Array of deps. Like all, but specify a logical-OR relationship. Any one of the dependencies will suffice. For example, to specify requirement to alternative Perl modules:

     or => [
         {perl_module => 'HTTP::Daemon'},
         {perl_module => 'HTTP::Daemon::SSL'},
  • none

    Array of deps. Specify that none of the dependencies must be satisfied for this type to be satisfied. Example, to specify that the function not run under SUDO or by root:

     none => [
         {env  => 'SUDO_USER'   },
         {code => sub {$> != 0} },

    Note that the above is not equivalent to below:

     none => [
         {env => 'SUDO_USER', code => sub {$> != 0} },

    which means that if none or only one of 'env'/'code' is satisfied, the whole dependency becomes a success (since it is negated by 'none'). Probably not what you want.

If you add a new language-specific dependency type, please prefix it with the language code, e.g. perl_module, perl_func, ruby_gem, python_egg. These dependency types have also been defined by some existing tools: deb (dependency to a Debian package), rpm (dependency to an RPM package), js_url (loading a remote JavaScript script URL), file (existence of a), perl_run_func (running a Perl subroutine and getting a successful enveloped result). Some of these might be declared as part of the core dependency types in the future.


What is the difference between summary or description in the Sah schema and arg specification?


     args => {
         src => {
             summary => "Source path",
             description => "...",
             schema => ["str*", {
                 summary => "...",
                 description => "...",
         dest => {
             summary => "Target path",
             description => "...",
             schema => ["str*", {
                 summary => "...",
                 description => "...",

As you can see, each argument has a summary and description, but the schema for each argument also has a summary and description schema clauses. What is the difference and which should be put into which?

The argument specification's summary (and description) describe the argument itself, in this example it says that src means "The source path" and dest means "The target path". The argument schema's summary (and description) describe the data type and valid values. In this example it could say, e.g., "a Unix-path string with a maximum length of 255 characters". In fact, src and dest are probably of the same type ("Unix path") and can share schema.

     args => {
         src => {
             schema => "unix_path",
         dest => {
             schema => "unix_path",

What is the difference between setting req=>1 in the argument specification and req=>1 in schema?


 # Note: remember that in Sah, str* is equivalent to [str => {req=>1}]
 args => {
     a => {         schema=>"str"  },
     b => {         schema=>"str*" },
     c => { req=>1, schema=>"str"  },
     d => { req=>1, schema=>"str*" },

In particular look at b and c. b is not a required argument (no req=>1 in the argument spec) but if it is specified, than it cannot be undef/null (since the schema says [str=>{req=>1}], a.k.a "str*"). On the other hand, c is a required argument (req=>1 in the argument spec) but you can specify undef/null as the value. The following are valid:

 func(c=>undef, d=>1);

But the following are not:

 func(b=>1, d=>1);  # c is not specified
 func(b=>undef, c=>1, d=>1);  # b has undef value
 func(b=>1, c=>1, d=>undef);  # d has undef value

Should I add a new metadata property, or add a new feature name to the features property, or add a new dependency type to the deps property?

If your property describes a dependency to something, it should definitely be a new dependency type. If your property only describes what the function can do and does not include any wrapper code, then it probably goes into features. Otherwise, it should probably become a new metadata property.

For example, if you want to declare that your function can only be run under a certain moon phase (e.g. full moon), it should definitely go as a new dependency type, so it becomes: deps => { moon_phase => 'full' }.

Another example, reverse is a feature name, because it just states that if we pass -reverse => 1 special argument to a reversible function, it can do a reverse operation. It doesn't include any wrapper code, all functionality is realized by the function itself. On the other hand, timeout is a metadata property because it involves adding adding some wrapping code (a timeout mechanism, e.g. an eval() block and alarm() in Perl).


Please visit the project's homepage at


Source repository is at


Related specifications: Sah, HTTP/1.1 (RFC 2068)



perlancar <>


To contribute, you can send patches by email/via RT, or send pull requests on GitHub.

Most of the time, you don't need to build the distribution yourself. You can simply modify the code, then test via:

 % prove -l

If you want to build the distribution (e.g. to try to install it locally on your system), you can install Dist::Zilla, Dist::Zilla::PluginBundle::Author::PERLANCAR, Pod::Weaver::PluginBundle::Author::PERLANCAR, and sometimes one or two other Dist::Zilla- and/or Pod::Weaver plugins. Any additional steps required beyond that are considered a bug and can be reported to me.


This software is copyright (c) 2022, 2021, 2020, 2019, 2018, 2017, 2016, 2015, 2014, 2013, 2012 by perlancar <>.

This is free software; you can redistribute it and/or modify it under the same terms as the Perl 5 programming language system itself.


Please report any bugs or feature requests on the bugtracker website

When submitting a bug or request, please include a test-file or a patch to an existing test-file that illustrates the bug or desired feature.