Difference between revisions of "Module:String"

From Timelines
Jump to: navigation, search
(add str.replace_plain)
m (1 revision imported: image crop)
 
(13 intermediate revisions by 5 users not shown)
Line 1: Line 1:
 +
--[[
 +
 +
This module is intended to provide access to basic string functions.
 +
 +
Most of the functions provided here can be invoked with named parameters,
 +
unnamed parameters, or a mixture.  If named parameters are used, Mediawiki will
 +
automatically remove any leading or trailing whitespace from the parameter.
 +
Depending on the intended use, it may be advantageous to either preserve or
 +
remove such whitespace.
 +
 +
Global options
 +
    ignore_errors: If set to 'true' or 1, any error condition will result in
 +
        an empty string being returned rather than an error message.
 +
 +
    error_category: If an error occurs, specifies the name of a category to
 +
        include with the error message.  The default category is
 +
        [Category:Errors reported by Module String].
 +
 +
    no_category: If set to 'true' or 1, no category will be added if an error
 +
        is generated.
 +
 +
Unit tests for this module are available at Module:String/tests.
 +
]]
 +
 
local str = {}
 
local str = {}
  
 +
--[[
 +
len
 +
 +
This function returns the length of the target string.
 +
 +
Usage:
 +
{{#invoke:String|len|target_string|}}
 +
OR
 +
{{#invoke:String|len|s=target_string}}
 +
 +
Parameters
 +
    s: The string whose length to report
 +
 +
If invoked using named parameters, Mediawiki will automatically remove any leading or
 +
trailing whitespace from the target string.
 +
]]
 
function str.len( frame )
 
function str.len( frame )
    return mw.ustring.len( frame.args.s )
+
local new_args = str._getParameters( frame.args, {'s'} )
 +
local s = new_args['s'] or ''
 +
return mw.ustring.len( s )
 
end
 
end
  
 +
--[[
 +
sub
 +
 +
This function returns a substring of the target string at specified indices.
 +
 +
Usage:
 +
{{#invoke:String|sub|target_string|start_index|end_index}}
 +
OR
 +
{{#invoke:String|sub|s=target_string|i=start_index|j=end_index}}
 +
 +
Parameters
 +
    s: The string to return a subset of
 +
    i: The fist index of the substring to return, defaults to 1.
 +
    j: The last index of the string to return, defaults to the last character.
 +
 +
The first character of the string is assigned an index of 1.  If either i or j
 +
is a negative value, it is interpreted the same as selecting a character by
 +
counting from the end of the string.  Hence, a value of -1 is the same as
 +
selecting the last character of the string.
 +
 +
If the requested indices are out of range for the given string, an error is
 +
reported.
 +
]]
 
function str.sub( frame )
 
function str.sub( frame )
    return mw.ustring.sub( frame.args.s, tonumber( frame.args.i ), tonumber( frame.args.j ) )
+
local new_args = str._getParameters( frame.args, { 's', 'i', 'j' } )
 +
local s = new_args['s'] or ''
 +
local i = tonumber( new_args['i'] ) or 1
 +
local j = tonumber( new_args['j'] ) or -1
 +
 
 +
local len = mw.ustring.len( s )
 +
 
 +
-- Convert negatives for range checking
 +
if i < 0 then
 +
i = len + i + 1
 +
end
 +
if j < 0 then
 +
j = len + j + 1
 +
end
 +
 
 +
if i > len or j > len or i < 1 or j < 1 then
 +
return str._error( 'String subset index out of range' )
 +
end
 +
if j < i then
 +
return str._error( 'String subset indices out of order' )
 +
end
 +
 
 +
return mw.ustring.sub( s, i, j )
 
end
 
end
  
 +
--[[
 +
This function implements that features of {{str sub old}} and is kept in order
 +
to maintain these older templates.
 +
]]
 
function str.sublength( frame )
 
function str.sublength( frame )
    local i = tonumber( frame.args.i ) or 0
+
local i = tonumber( frame.args.i ) or 0
    local len = tonumber( frame.args.len )
+
local len = tonumber( frame.args.len )
    return mw.ustring.sub( frame.args.s, i + 1, len and ( i + len ) )
+
return mw.ustring.sub( frame.args.s, i + 1, len and ( i + len ) )
 +
end
 +
 
 +
--[[
 +
_match
 +
 
 +
This function returns a substring from the source string that matches a
 +
specified pattern. It is exported for use in other modules
 +
 
 +
Usage:
 +
strmatch = require("Module:String")._match
 +
sresult = strmatch( s, pattern, start, match, plain, nomatch )
 +
 
 +
Parameters
 +
    s: The string to search
 +
    pattern: The pattern or string to find within the string
 +
    start: The index within the source string to start the search.  The first
 +
        character of the string has index 1.  Defaults to 1.
 +
    match: In some cases it may be possible to make multiple matches on a single
 +
        string.  This specifies which match to return, where the first match is
 +
        match= 1.  If a negative number is specified then a match is returned
 +
        counting from the last match.  Hence match = -1 is the same as requesting
 +
        the last match.  Defaults to 1.
 +
    plain: A flag indicating that the pattern should be understood as plain
 +
        text.  Defaults to false.
 +
    nomatch: If no match is found, output the "nomatch" value rather than an error.
 +
 
 +
For information on constructing Lua patterns, a form of [regular expression], see:
 +
 
 +
* http://www.lua.org/manual/5.1/manual.html#5.4.1
 +
* http://www.mediawiki.org/wiki/Extension:Scribunto/Lua_reference_manual#Patterns
 +
* http://www.mediawiki.org/wiki/Extension:Scribunto/Lua_reference_manual#Ustring_patterns
 +
 
 +
]]
 +
-- This sub-routine is exported for use in other modules
 +
function str._match( s, pattern, start, match_index, plain_flag, nomatch )
 +
if s == '' then
 +
return str._error( 'Target string is empty' )
 +
end
 +
if pattern == '' then
 +
return str._error( 'Pattern string is empty' )
 +
end
 +
start = tonumber(start) or 1
 +
if math.abs(start) < 1 or math.abs(start) > mw.ustring.len( s ) then
 +
return str._error( 'Requested start is out of range' )
 +
end
 +
if match_index == 0 then
 +
return str._error( 'Match index is out of range' )
 +
end
 +
if plain_flag then
 +
pattern = str._escapePattern( pattern )
 +
end
 +
 
 +
local result
 +
if match_index == 1 then
 +
-- Find first match is simple case
 +
result = mw.ustring.match( s, pattern, start )
 +
else
 +
if start > 1 then
 +
s = mw.ustring.sub( s, start )
 +
end
 +
 
 +
local iterator = mw.ustring.gmatch(s, pattern)
 +
if match_index > 0 then
 +
-- Forward search
 +
for w in iterator do
 +
match_index = match_index - 1
 +
if match_index == 0 then
 +
result = w
 +
break
 +
end
 +
end
 +
else
 +
-- Reverse search
 +
local result_table = {}
 +
local count = 1
 +
for w in iterator do
 +
result_table[count] = w
 +
count = count + 1
 +
end
 +
 
 +
result = result_table[ count + match_index ]
 +
end
 +
end
 +
 
 +
if result == nil then
 +
if nomatch == nil then
 +
return str._error( 'Match not found' )
 +
else
 +
return nomatch
 +
end
 +
else
 +
return result
 +
end
 
end
 
end
  
 +
--[[
 +
match
 +
 +
This function returns a substring from the source string that matches a
 +
specified pattern.
 +
 +
Usage:
 +
{{#invoke:String|match|source_string|pattern_string|start_index|match_number|plain_flag|nomatch_output}}
 +
OR
 +
{{#invoke:String|match|s=source_string|pattern=pattern_string|start=start_index
 +
    |match=match_number|plain=plain_flag|nomatch=nomatch_output}}
 +
 +
Parameters
 +
    s: The string to search
 +
    pattern: The pattern or string to find within the string
 +
    start: The index within the source string to start the search.  The first
 +
        character of the string has index 1.  Defaults to 1.
 +
    match: In some cases it may be possible to make multiple matches on a single
 +
        string.  This specifies which match to return, where the first match is
 +
        match= 1.  If a negative number is specified then a match is returned
 +
        counting from the last match.  Hence match = -1 is the same as requesting
 +
        the last match.  Defaults to 1.
 +
    plain: A flag indicating that the pattern should be understood as plain
 +
        text.  Defaults to false.
 +
    nomatch: If no match is found, output the "nomatch" value rather than an error.
 +
 +
If invoked using named parameters, Mediawiki will automatically remove any leading or
 +
trailing whitespace from each string.  In some circumstances this is desirable, in
 +
other cases one may want to preserve the whitespace.
 +
 +
If the match_number or start_index are out of range for the string being queried, then
 +
this function generates an error.  An error is also generated if no match is found.
 +
If one adds the parameter ignore_errors=true, then the error will be suppressed and
 +
an empty string will be returned on any failure.
 +
 +
For information on constructing Lua patterns, a form of [regular expression], see:
 +
 +
* http://www.lua.org/manual/5.1/manual.html#5.4.1
 +
* http://www.mediawiki.org/wiki/Extension:Scribunto/Lua_reference_manual#Patterns
 +
* http://www.mediawiki.org/wiki/Extension:Scribunto/Lua_reference_manual#Ustring_patterns
 +
 +
]]
 +
-- This is the entry point for #invoke:String|match
 
function str.match( frame )
 
function str.match( frame )
    return mw.ustring.match( frame.args.s, frame.args.pattern, tonumber( frame.args.i ) )
+
local new_args = str._getParameters( frame.args, {'s', 'pattern', 'start', 'match', 'plain', 'nomatch'} )
 +
local s = new_args['s'] or ''
 +
local start = tonumber( new_args['start'] ) or 1
 +
local plain_flag = str._getBoolean( new_args['plain'] or false )
 +
local pattern = new_args['pattern'] or ''
 +
local match_index = math.floor( tonumber(new_args['match']) or 1 )
 +
local nomatch = new_args['nomatch']
 +
 
 +
return str._match( s, pattern, start, match_index, plain_flag, nomatch )
 
end
 
end
  
--[====[
+
--[[
 
pos
 
pos
  
Line 34: Line 269:
  
 
If invoked using named parameters, Mediawiki will automatically remove any leading or
 
If invoked using named parameters, Mediawiki will automatically remove any leading or
trailing whitespace from the target string.  In some circumstances this is desirable, in  
+
trailing whitespace from the target string.  In some circumstances this is desirable, in
 
other cases one may want to preserve the whitespace.
 
other cases one may want to preserve the whitespace.
  
 
The first character has an index value of 1.
 
The first character has an index value of 1.
  
If one requests a negative value, this function will select a character by counting backwards  
+
If one requests a negative value, this function will select a character by counting backwards
 
from the end of the string.  In other words pos = -1 is the same as asking for the last character.
 
from the end of the string.  In other words pos = -1 is the same as asking for the last character.
  
 
A requested value of zero, or a value greater than the length of the string returns an error.
 
A requested value of zero, or a value greater than the length of the string returns an error.
]====]
+
]]
 
function str.pos( frame )
 
function str.pos( frame )
    local new_args = str._getParameters( frame.args, {'target', 'pos'} );
+
local new_args = str._getParameters( frame.args, {'target', 'pos'} )
    local target_str = new_args['target'] or '';
+
local target_str = new_args['target'] or ''
    local pos = tonumber( new_args['pos'] ) or 0;
+
local pos = tonumber( new_args['pos'] ) or 0
  
    if pos == 0 or math.abs(pos) > mw.ustring.len( target_str ) then
+
if pos == 0 or math.abs(pos) > mw.ustring.len( target_str ) then
        return '<strong class="error">String index out of range</strong>';
+
return str._error( 'String index out of range' )
    end  
+
end
   
+
 
    return mw.ustring.sub( target_str, pos, pos );
+
return mw.ustring.sub( target_str, pos, pos )
 
end
 
end
  
--[====[
+
--[[
 
str_find
 
str_find
  
 
This function duplicates the behavior of {{str_find}}, including all of its quirks.
 
This function duplicates the behavior of {{str_find}}, including all of its quirks.
This is provided in order to support existing templates, but is NOT RECOMMENDED for  
+
This is provided in order to support existing templates, but is NOT RECOMMENDED for
 
new code and templates.  New code is recommended to use the "find" function instead.
 
new code and templates.  New code is recommended to use the "find" function instead.
  
Line 69: Line 304:
 
value of "1", which is generally unexpected behavior, and must be accounted for
 
value of "1", which is generally unexpected behavior, and must be accounted for
 
separatetly.
 
separatetly.
]====]
+
]]
 
function str.str_find( frame )
 
function str.str_find( frame )
    local new_args = str._getParameters( frame.args, {'source', 'target'} );
+
local new_args = str._getParameters( frame.args, {'source', 'target'} )
    local source_str = new_args['source'] or '';
+
local source_str = new_args['source'] or ''
    local target_str = new_args['target'] or '';
+
local target_str = new_args['target'] or ''
 +
 
 +
if target_str == '' then
 +
return 1
 +
end
 +
 
 +
local start = mw.ustring.find( source_str, target_str, 1, true )
 +
if start == nil then
 +
start = -1
 +
end
  
    if target_str == '' then
+
return start
        return 1;
 
    end   
 
   
 
    local start = mw.ustring.find( source_str, target_str, 1, true )
 
    if start == nil then
 
        start = -1
 
    end
 
   
 
    return start
 
 
end
 
end
  
--[====[
+
--[[
 
find
 
find
  
Line 106: Line 341:
  
 
If invoked using named parameters, Mediawiki will automatically remove any leading or
 
If invoked using named parameters, Mediawiki will automatically remove any leading or
trailing whitespace from the parameter.  In some circumstances this is desirable, in  
+
trailing whitespace from the parameter.  In some circumstances this is desirable, in
 
other cases one may want to preserve the whitespace.
 
other cases one may want to preserve the whitespace.
  
This function returns the first index >= "start" where "target" can be found  
+
This function returns the first index >= "start" where "target" can be found
within "source".  Indices are 1-based.  If "target" is not found, then this  
+
within "source".  Indices are 1-based.  If "target" is not found, then this
 
function returns 0.  If either "source" or "target" are missing / empty, this
 
function returns 0.  If either "source" or "target" are missing / empty, this
 
function also returns 0.
 
function also returns 0.
  
 
This function should be safe for UTF-8 strings.
 
This function should be safe for UTF-8 strings.
]====]
+
]]
 
function str.find( frame )
 
function str.find( frame )
    local new_args = str._getParameters( frame.args, {'source', 'target', 'start', 'plain' } );
+
local new_args = str._getParameters( frame.args, {'source', 'target', 'start', 'plain' } )
    local source_str = new_args['source'] or '';
+
local source_str = new_args['source'] or ''
    local pattern = new_args['target'] or '';
+
local pattern = new_args['target'] or ''
    local start_pos = tonumber(new_args['start']) or 1;
+
local start_pos = tonumber(new_args['start']) or 1
    local plain = new_args['plain'] or true;
+
local plain = new_args['plain'] or true
       
+
 
    if source_str == '' or pattern == '' then
+
if source_str == '' or pattern == '' then
        return 0;
+
return 0
    end  
+
end
   
+
 
    if type( plain ) == 'string' then
+
plain = str._getBoolean( plain )
        plain = plain:lower();
+
 
        if plain == 'false' or plain == 'no' or plain == '0' then
+
local start = mw.ustring.find( source_str, pattern, start_pos, plain )
            plain = false;
+
if start == nil then
        else
+
start = 0
            plain = true;
+
end
        end   
 
    end
 
  
    local start = mw.ustring.find( source_str, pattern, start_pos, plain )
+
return start
    if start == nil then
 
        start = 0
 
    end
 
   
 
    return start
 
 
end
 
end
  
--[====[
+
--[[
replace_plain
+
replace
  
 
This function allows one to replace a target string or pattern within another
 
This function allows one to replace a target string or pattern within another
Line 151: Line 379:
  
 
Usage:
 
Usage:
{{#invoke:String|replace_plain|source_str|pattern_string|replace_string|firstonlyflag}}
+
{{#invoke:String|replace|source_str|pattern_string|replace_string|replacement_count|plain_flag}}
 
OR
 
OR
{{#invoke:String|replace_plain|source=source_str|pattern=pattern_str|replace=replace_string|firstonly=firstonlyflag}}
+
{{#invoke:String|replace|source=source_string|pattern=pattern_string|replace=replace_string|
 +
  count=replacement_count|plain=plain_flag}}
  
 
Parameters
 
Parameters
 
     source: The string to search
 
     source: The string to search
     patten: The string or pattern to find within source
+
     pattern: The string or pattern to find within source
 
     replace: The replacement text
 
     replace: The replacement text
     firstonly: Boolean flag indicating that only the first occurence found should be replaced
+
     count: The number of occurences to replace, defaults to all.
]====]
+
    plain: Boolean flag indicating that pattern should be understood as plain
function str.replace_plain( frame )
+
        text and not as a Lua style regular expression, defaults to true
    local new_args = str._getParameters( frame.args, {'source', 'pattern', 'replace', 'firstonly' } );
+
]]
    local source_str = new_args['source'] or '';
+
function str.replace( frame )
    local pattern = new_args['pattern'] or '';
+
local new_args = str._getParameters( frame.args, {'source', 'pattern', 'replace', 'count', 'plain' } )
    local replace = new_args['replace'] or '';
+
local source_str = new_args['source'] or ''
    local firstonly = new_args['firstonly'] or '';
+
local pattern = new_args['pattern'] or ''
    firstonly = firstonly:lower();
+
local replace = new_args['replace'] or ''
       
+
local count = tonumber( new_args['count'] )
    if source_str == '' or pattern == '' then
+
local plain = new_args['plain'] or true
        return source_str;
+
 
    end  
+
if source_str == '' or pattern == '' then
 +
return source_str
 +
end
 +
plain = str._getBoolean( plain )
 +
 
 +
if plain then
 +
pattern = str._escapePattern( pattern )
 +
replace = mw.ustring.gsub( replace, "%%", "%%%%" ) --Only need to escape replacement sequences.
 +
end
  
    local pattern_plain = mw.ustring.gsub(pattern, '%%', '%%%%');
+
local result
    local replace_plain = mw.ustring.gsub(replace, '%%', '%%%%');
 
  
    if firstonly == 'true' or firstonly == 'yes' or firstonly == '1' then
+
if count ~= nil then
        local result = gsub( source_str, pattern_plain, replace_plain, 1 );
+
result = mw.ustring.gsub( source_str, pattern, replace, count )
    else
+
else
        local result = gsub( source_str, pattern_plain, replace_plain, n );
+
result = mw.ustring.gsub( source_str, pattern, replace )
    end
+
end
 +
 
 +
return result
 +
end
  
     return result;
+
--[[
 +
     simple function to pipe string.rep to templates.
 +
]]
 +
function str.rep( frame )
 +
local repetitions = tonumber( frame.args[2] )
 +
if not repetitions then
 +
return str._error( 'function rep expects a number as second parameter, received "' .. ( frame.args[2] or '' ) .. '"' )
 +
end
 +
return string.rep( frame.args[1] or '', repetitions )
 
end
 
end
  
--[====[
+
--[[
 +
escapePattern
 +
 
 +
This function escapes special characters from a Lua string pattern. See [1]
 +
for details on how patterns work.
 +
 
 +
[1] https://www.mediawiki.org/wiki/Extension:Scribunto/Lua_reference_manual#Patterns
 +
 
 +
Usage:
 +
{{#invoke:String|escapePattern|pattern_string}}
 +
 
 +
Parameters
 +
    pattern_string: The pattern string to escape.
 +
]]
 +
function str.escapePattern( frame )
 +
local pattern_str = frame.args[1]
 +
if not pattern_str then
 +
return str._error( 'No pattern string specified' )
 +
end
 +
local result = str._escapePattern( pattern_str )
 +
return result
 +
end
 +
 
 +
--[[
 +
count
 +
This function counts the number of occurrences of one string in another.
 +
]]
 +
function str.count(frame)
 +
local args = str._getParameters(frame.args, {'source', 'pattern', 'plain'})
 +
local source = args.source or ''
 +
local pattern = args.pattern or ''
 +
local plain = str._getBoolean(args.plain or true)
 +
if plain then
 +
pattern = str._escapePattern(pattern)
 +
end
 +
local _, count = mw.ustring.gsub(source, pattern, '')
 +
return count
 +
end
 +
 
 +
--[[
 +
endswith
 +
This function determines whether a string ends with another string.
 +
]]
 +
function str.endswith(frame)
 +
local args = str._getParameters(frame.args, {'source', 'pattern'})
 +
local source = args.source or ''
 +
local pattern = args.pattern or ''
 +
if pattern == '' then
 +
-- All strings end with the empty string.
 +
return "yes"
 +
end
 +
if mw.ustring.sub(source, -mw.ustring.len(pattern), -1) == pattern then
 +
return "yes"
 +
else
 +
return ""
 +
end
 +
end
 +
 
 +
--[[
 +
join
 +
 
 +
Join all non empty arguments together; the first argument is the separator.
 +
Usage:
 +
{{#invoke:String|join|sep|one|two|three}}
 +
]]
 +
function str.join(frame)
 +
local args = {}
 +
local sep
 +
for _, v in ipairs( frame.args ) do
 +
if sep then
 +
if v ~= '' then
 +
table.insert(args, v)
 +
end
 +
else
 +
sep = v
 +
end
 +
end
 +
return table.concat( args, sep or '' )
 +
end
 +
 
 +
--[[
 
Helper function that populates the argument list given that user may need to use a mix of
 
Helper function that populates the argument list given that user may need to use a mix of
 
named and unnamed parameters.  This is relevant because named parameters are not
 
named and unnamed parameters.  This is relevant because named parameters are not
 
identical to unnamed parameters due to string trimming, and when dealing with strings
 
identical to unnamed parameters due to string trimming, and when dealing with strings
 
we sometimes want to either preserve or remove that whitespace depending on the application.
 
we sometimes want to either preserve or remove that whitespace depending on the application.
]====]
+
]]
 
function str._getParameters( frame_args, arg_list )
 
function str._getParameters( frame_args, arg_list )
    local new_args = {};
+
local new_args = {}
    local index = 1;
+
local index = 1
    local value;
+
local value
   
+
 
    for i,arg in ipairs( arg_list ) do
+
for _, arg in ipairs( arg_list ) do
        value = frame_args[arg]
+
value = frame_args[arg]
        if value == nil then
+
if value == nil then
            value = frame_args[index];
+
value = frame_args[index]
            index = index + 1;
+
index = index + 1
        end
+
end
        new_args[arg] = value;
+
new_args[arg] = value
    end
+
end
   
+
 
    return new_args;
+
return new_args
end      
+
end
 +
 
 +
--[[
 +
Helper function to handle error messages.
 +
]]
 +
function str._error( error_str )
 +
local frame = mw.getCurrentFrame()
 +
local error_category = frame.args.error_category or 'Errors reported by Module String'
 +
local ignore_errors = frame.args.ignore_errors or false
 +
local no_category = frame.args.no_category or false
 +
 
 +
if str._getBoolean(ignore_errors) then
 +
return ''
 +
end
 +
 
 +
local error_str = '<strong class="error">String Module Error: ' .. error_str .. '</strong>'
 +
if error_category ~= '' and not str._getBoolean( no_category ) then
 +
error_str = '[[Category:' .. error_category .. ']]' .. error_str
 +
end
 +
 
 +
return error_str
 +
end
 +
 
 +
--[[
 +
Helper Function to interpret boolean strings
 +
]]
 +
function str._getBoolean( boolean_str )
 +
local boolean_value
 +
 
 +
if type( boolean_str ) == 'string' then
 +
boolean_str = boolean_str:lower()
 +
if boolean_str == 'false' or boolean_str == 'no' or boolean_str == '0'
 +
or boolean_str == '' then
 +
boolean_value = false
 +
else
 +
boolean_value = true
 +
end
 +
elseif type( boolean_str ) == 'boolean' then
 +
boolean_value = boolean_str
 +
else
 +
error( 'No boolean value found' )
 +
end
 +
return boolean_value
 +
end
 +
 
 +
--[[
 +
Helper function that escapes all pattern characters so that they will be treated
 +
as plain text.
 +
]]
 +
function str._escapePattern( pattern_str )
 +
return mw.ustring.gsub( pattern_str, "([%(%)%.%%%+%-%*%?%[%^%$%]])", "%%%1" )
 +
end
  
 
return str
 
return str

Latest revision as of 11:39, 12 September 2020

Documentation for this module may be created at Module:String/doc

--[[

This module is intended to provide access to basic string functions.

Most of the functions provided here can be invoked with named parameters,
unnamed parameters, or a mixture.  If named parameters are used, Mediawiki will
automatically remove any leading or trailing whitespace from the parameter.
Depending on the intended use, it may be advantageous to either preserve or
remove such whitespace.

Global options
    ignore_errors: If set to 'true' or 1, any error condition will result in
        an empty string being returned rather than an error message.

    error_category: If an error occurs, specifies the name of a category to
        include with the error message.  The default category is
        [Category:Errors reported by Module String].

    no_category: If set to 'true' or 1, no category will be added if an error
        is generated.

Unit tests for this module are available at Module:String/tests.
]]

local str = {}

--[[
len

This function returns the length of the target string.

Usage:
{{#invoke:String|len|target_string|}}
OR
{{#invoke:String|len|s=target_string}}

Parameters
    s: The string whose length to report

If invoked using named parameters, Mediawiki will automatically remove any leading or
trailing whitespace from the target string.
]]
function str.len( frame )
	local new_args = str._getParameters( frame.args, {'s'} )
	local s = new_args['s'] or ''
	return mw.ustring.len( s )
end

--[[
sub

This function returns a substring of the target string at specified indices.

Usage:
{{#invoke:String|sub|target_string|start_index|end_index}}
OR
{{#invoke:String|sub|s=target_string|i=start_index|j=end_index}}

Parameters
    s: The string to return a subset of
    i: The fist index of the substring to return, defaults to 1.
    j: The last index of the string to return, defaults to the last character.

The first character of the string is assigned an index of 1.  If either i or j
is a negative value, it is interpreted the same as selecting a character by
counting from the end of the string.  Hence, a value of -1 is the same as
selecting the last character of the string.

If the requested indices are out of range for the given string, an error is
reported.
]]
function str.sub( frame )
	local new_args = str._getParameters( frame.args, { 's', 'i', 'j' } )
	local s = new_args['s'] or ''
	local i = tonumber( new_args['i'] ) or 1
	local j = tonumber( new_args['j'] ) or -1

	local len = mw.ustring.len( s )

	-- Convert negatives for range checking
	if i < 0 then
		i = len + i + 1
	end
	if j < 0 then
		j = len + j + 1
	end

	if i > len or j > len or i < 1 or j < 1 then
		return str._error( 'String subset index out of range' )
	end
	if j < i then
		return str._error( 'String subset indices out of order' )
	end

	return mw.ustring.sub( s, i, j )
end

--[[
This function implements that features of {{str sub old}} and is kept in order
to maintain these older templates.
]]
function str.sublength( frame )
	local i = tonumber( frame.args.i ) or 0
	local len = tonumber( frame.args.len )
	return mw.ustring.sub( frame.args.s, i + 1, len and ( i + len ) )
end

--[[
_match

This function returns a substring from the source string that matches a
specified pattern. It is exported for use in other modules

Usage:
strmatch = require("Module:String")._match
sresult = strmatch( s, pattern, start, match, plain, nomatch )

Parameters
    s: The string to search
    pattern: The pattern or string to find within the string
    start: The index within the source string to start the search.  The first
        character of the string has index 1.  Defaults to 1.
    match: In some cases it may be possible to make multiple matches on a single
        string.  This specifies which match to return, where the first match is
        match= 1.  If a negative number is specified then a match is returned
        counting from the last match.  Hence match = -1 is the same as requesting
        the last match.  Defaults to 1.
    plain: A flag indicating that the pattern should be understood as plain
        text.  Defaults to false.
    nomatch: If no match is found, output the "nomatch" value rather than an error.

For information on constructing Lua patterns, a form of [regular expression], see:

* http://www.lua.org/manual/5.1/manual.html#5.4.1
* http://www.mediawiki.org/wiki/Extension:Scribunto/Lua_reference_manual#Patterns
* http://www.mediawiki.org/wiki/Extension:Scribunto/Lua_reference_manual#Ustring_patterns

]]
-- This sub-routine is exported for use in other modules
function str._match( s, pattern, start, match_index, plain_flag, nomatch )
	if s == '' then
		return str._error( 'Target string is empty' )
	end
	if pattern == '' then
		return str._error( 'Pattern string is empty' )
	end
	start = tonumber(start) or 1
	if math.abs(start) < 1 or math.abs(start) > mw.ustring.len( s ) then
		return str._error( 'Requested start is out of range' )
	end
	if match_index == 0 then
		return str._error( 'Match index is out of range' )
	end
	if plain_flag then
		pattern = str._escapePattern( pattern )
	end

	local result
	if match_index == 1 then
		-- Find first match is simple case
		result = mw.ustring.match( s, pattern, start )
	else
		if start > 1 then
			s = mw.ustring.sub( s, start )
		end

		local iterator = mw.ustring.gmatch(s, pattern)
		if match_index > 0 then
			-- Forward search
			for w in iterator do
				match_index = match_index - 1
				if match_index == 0 then
					result = w
					break
				end
			end
		else
			-- Reverse search
			local result_table = {}
			local count = 1
			for w in iterator do
				result_table[count] = w
				count = count + 1
			end

			result = result_table[ count + match_index ]
		end
	end

	if result == nil then
		if nomatch == nil then
			return str._error( 'Match not found' )
		else
			return nomatch
		end
	else
		return result
	end
end

--[[
match

This function returns a substring from the source string that matches a
specified pattern.

Usage:
{{#invoke:String|match|source_string|pattern_string|start_index|match_number|plain_flag|nomatch_output}}
OR
{{#invoke:String|match|s=source_string|pattern=pattern_string|start=start_index
    |match=match_number|plain=plain_flag|nomatch=nomatch_output}}

Parameters
    s: The string to search
    pattern: The pattern or string to find within the string
    start: The index within the source string to start the search.  The first
        character of the string has index 1.  Defaults to 1.
    match: In some cases it may be possible to make multiple matches on a single
        string.  This specifies which match to return, where the first match is
        match= 1.  If a negative number is specified then a match is returned
        counting from the last match.  Hence match = -1 is the same as requesting
        the last match.  Defaults to 1.
    plain: A flag indicating that the pattern should be understood as plain
        text.  Defaults to false.
    nomatch: If no match is found, output the "nomatch" value rather than an error.

If invoked using named parameters, Mediawiki will automatically remove any leading or
trailing whitespace from each string.  In some circumstances this is desirable, in
other cases one may want to preserve the whitespace.

If the match_number or start_index are out of range for the string being queried, then
this function generates an error.  An error is also generated if no match is found.
If one adds the parameter ignore_errors=true, then the error will be suppressed and
an empty string will be returned on any failure.

For information on constructing Lua patterns, a form of [regular expression], see:

* http://www.lua.org/manual/5.1/manual.html#5.4.1
* http://www.mediawiki.org/wiki/Extension:Scribunto/Lua_reference_manual#Patterns
* http://www.mediawiki.org/wiki/Extension:Scribunto/Lua_reference_manual#Ustring_patterns

]]
-- This is the entry point for #invoke:String|match
function str.match( frame )
	local new_args = str._getParameters( frame.args, {'s', 'pattern', 'start', 'match', 'plain', 'nomatch'} )
	local s = new_args['s'] or ''
	local start = tonumber( new_args['start'] ) or 1
	local plain_flag = str._getBoolean( new_args['plain'] or false )
	local pattern = new_args['pattern'] or ''
	local match_index = math.floor( tonumber(new_args['match']) or 1 )
	local nomatch = new_args['nomatch']

	return str._match( s, pattern, start, match_index, plain_flag, nomatch )
end

--[[
pos

This function returns a single character from the target string at position pos.

Usage:
{{#invoke:String|pos|target_string|index_value}}
OR
{{#invoke:String|pos|target=target_string|pos=index_value}}

Parameters
    target: The string to search
    pos: The index for the character to return

If invoked using named parameters, Mediawiki will automatically remove any leading or
trailing whitespace from the target string.  In some circumstances this is desirable, in
other cases one may want to preserve the whitespace.

The first character has an index value of 1.

If one requests a negative value, this function will select a character by counting backwards
from the end of the string.  In other words pos = -1 is the same as asking for the last character.

A requested value of zero, or a value greater than the length of the string returns an error.
]]
function str.pos( frame )
	local new_args = str._getParameters( frame.args, {'target', 'pos'} )
	local target_str = new_args['target'] or ''
	local pos = tonumber( new_args['pos'] ) or 0

	if pos == 0 or math.abs(pos) > mw.ustring.len( target_str ) then
		return str._error( 'String index out of range' )
	end

	return mw.ustring.sub( target_str, pos, pos )
end

--[[
str_find

This function duplicates the behavior of {{str_find}}, including all of its quirks.
This is provided in order to support existing templates, but is NOT RECOMMENDED for
new code and templates.  New code is recommended to use the "find" function instead.

Returns the first index in "source" that is a match to "target".  Indexing is 1-based,
and the function returns -1 if the "target" string is not present in "source".

Important Note: If the "target" string is empty / missing, this function returns a
value of "1", which is generally unexpected behavior, and must be accounted for
separatetly.
]]
function str.str_find( frame )
	local new_args = str._getParameters( frame.args, {'source', 'target'} )
	local source_str = new_args['source'] or ''
	local target_str = new_args['target'] or ''

	if target_str == '' then
		return 1
	end

	local start = mw.ustring.find( source_str, target_str, 1, true )
	if start == nil then
		start = -1
	end

	return start
end

--[[
find

This function allows one to search for a target string or pattern within another
string.

Usage:
{{#invoke:String|find|source_str|target_string|start_index|plain_flag}}
OR
{{#invoke:String|find|source=source_str|target=target_str|start=start_index|plain=plain_flag}}

Parameters
    source: The string to search
    target: The string or pattern to find within source
    start: The index within the source string to start the search, defaults to 1
    plain: Boolean flag indicating that target should be understood as plain
        text and not as a Lua style regular expression, defaults to true

If invoked using named parameters, Mediawiki will automatically remove any leading or
trailing whitespace from the parameter.  In some circumstances this is desirable, in
other cases one may want to preserve the whitespace.

This function returns the first index >= "start" where "target" can be found
within "source".  Indices are 1-based.  If "target" is not found, then this
function returns 0.  If either "source" or "target" are missing / empty, this
function also returns 0.

This function should be safe for UTF-8 strings.
]]
function str.find( frame )
	local new_args = str._getParameters( frame.args, {'source', 'target', 'start', 'plain' } )
	local source_str = new_args['source'] or ''
	local pattern = new_args['target'] or ''
	local start_pos = tonumber(new_args['start']) or 1
	local plain = new_args['plain'] or true

	if source_str == '' or pattern == '' then
		return 0
	end

	plain = str._getBoolean( plain )

	local start = mw.ustring.find( source_str, pattern, start_pos, plain )
	if start == nil then
		start = 0
	end

	return start
end

--[[
replace

This function allows one to replace a target string or pattern within another
string.

Usage:
{{#invoke:String|replace|source_str|pattern_string|replace_string|replacement_count|plain_flag}}
OR
{{#invoke:String|replace|source=source_string|pattern=pattern_string|replace=replace_string|
   count=replacement_count|plain=plain_flag}}

Parameters
    source: The string to search
    pattern: The string or pattern to find within source
    replace: The replacement text
    count: The number of occurences to replace, defaults to all.
    plain: Boolean flag indicating that pattern should be understood as plain
        text and not as a Lua style regular expression, defaults to true
]]
function str.replace( frame )
	local new_args = str._getParameters( frame.args, {'source', 'pattern', 'replace', 'count', 'plain' } )
	local source_str = new_args['source'] or ''
	local pattern = new_args['pattern'] or ''
	local replace = new_args['replace'] or ''
	local count = tonumber( new_args['count'] )
	local plain = new_args['plain'] or true

	if source_str == '' or pattern == '' then
		return source_str
	end
	plain = str._getBoolean( plain )

	if plain then
		pattern = str._escapePattern( pattern )
		replace = mw.ustring.gsub( replace, "%%", "%%%%" ) --Only need to escape replacement sequences.
	end

	local result

	if count ~= nil then
		result = mw.ustring.gsub( source_str, pattern, replace, count )
	else
		result = mw.ustring.gsub( source_str, pattern, replace )
	end

	return result
end

--[[
    simple function to pipe string.rep to templates.
]]
function str.rep( frame )
	local repetitions = tonumber( frame.args[2] )
	if not repetitions then
		return str._error( 'function rep expects a number as second parameter, received "' .. ( frame.args[2] or '' ) .. '"' )
	end
	return string.rep( frame.args[1] or '', repetitions )
end

--[[
escapePattern

This function escapes special characters from a Lua string pattern. See [1]
for details on how patterns work.

[1] https://www.mediawiki.org/wiki/Extension:Scribunto/Lua_reference_manual#Patterns

Usage:
{{#invoke:String|escapePattern|pattern_string}}

Parameters
    pattern_string: The pattern string to escape.
]]
function str.escapePattern( frame )
	local pattern_str = frame.args[1]
	if not pattern_str then
		return str._error( 'No pattern string specified' )
	end
	local result = str._escapePattern( pattern_str )
	return result
end

--[[
count
This function counts the number of occurrences of one string in another.
]]
function str.count(frame)
	local args = str._getParameters(frame.args, {'source', 'pattern', 'plain'})
	local source = args.source or ''
	local pattern = args.pattern or ''
	local plain = str._getBoolean(args.plain or true)
	if plain then
		pattern = str._escapePattern(pattern)
	end
	local _, count = mw.ustring.gsub(source, pattern, '')
	return count
end

--[[
endswith
This function determines whether a string ends with another string.
]]
function str.endswith(frame)
	local args = str._getParameters(frame.args, {'source', 'pattern'})
	local source = args.source or ''
	local pattern = args.pattern or ''
	if pattern == '' then
		-- All strings end with the empty string.
		return "yes"
	end
	if mw.ustring.sub(source, -mw.ustring.len(pattern), -1) == pattern then
		return "yes"
	else
		return ""
	end
end

--[[
join

Join all non empty arguments together; the first argument is the separator.
Usage:
{{#invoke:String|join|sep|one|two|three}}
]]
function str.join(frame)
	local args = {}
	local sep
	for _, v in ipairs( frame.args ) do
		if sep then
			if v ~= '' then
				table.insert(args, v)
			end
		else
			sep = v
		end
	end
	return table.concat( args, sep or '' )
end

--[[
Helper function that populates the argument list given that user may need to use a mix of
named and unnamed parameters.  This is relevant because named parameters are not
identical to unnamed parameters due to string trimming, and when dealing with strings
we sometimes want to either preserve or remove that whitespace depending on the application.
]]
function str._getParameters( frame_args, arg_list )
	local new_args = {}
	local index = 1
	local value

	for _, arg in ipairs( arg_list ) do
		value = frame_args[arg]
		if value == nil then
			value = frame_args[index]
			index = index + 1
		end
		new_args[arg] = value
	end

	return new_args
end

--[[
Helper function to handle error messages.
]]
function str._error( error_str )
	local frame = mw.getCurrentFrame()
	local error_category = frame.args.error_category or 'Errors reported by Module String'
	local ignore_errors = frame.args.ignore_errors or false
	local no_category = frame.args.no_category or false

	if str._getBoolean(ignore_errors) then
		return ''
	end

	local error_str = '<strong class="error">String Module Error: ' .. error_str .. '</strong>'
	if error_category ~= '' and not str._getBoolean( no_category ) then
		error_str = '[[Category:' .. error_category .. ']]' .. error_str
	end

	return error_str
end

--[[
Helper Function to interpret boolean strings
]]
function str._getBoolean( boolean_str )
	local boolean_value

	if type( boolean_str ) == 'string' then
		boolean_str = boolean_str:lower()
		if boolean_str == 'false' or boolean_str == 'no' or boolean_str == '0'
				or boolean_str == '' then
			boolean_value = false
		else
			boolean_value = true
		end
	elseif type( boolean_str ) == 'boolean' then
		boolean_value = boolean_str
	else
		error( 'No boolean value found' )
	end
	return boolean_value
end

--[[
Helper function that escapes all pattern characters so that they will be treated
as plain text.
]]
function str._escapePattern( pattern_str )
	return mw.ustring.gsub( pattern_str, "([%(%)%.%%%+%-%*%?%[%^%$%]])", "%%%1" )
end

return str