Module:string: Difference between revisions

no edit summary
No edit summary
No edit summary
Line 1: Line 1:
--[[
local str = {}


This module is intended to provide access to basic string functions.
-- Cannot include null byte.
local UTF8_char = "[\1-\127\194-\244][\128-\191]*"


Most of the functions provided here can be invoked with named parameters,
--[[
unnamed parameters, or a mixture.  If named parameters are used, Mediawiki will
ulen
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
Counts UTF-8 characters. Faster than mw.ustring.len.
    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
Assumes that the encoding is correct. Unlike mw.ustring.len, does not return nil
        include with the error message.  The default category is
if encoding is invalid.
        [Category:Errors reported by Module String].


    no_category: If set to 'true' or 1, no category will be added if an error
Does not count the bytes 192, 193, and 245-255. They are not used in UTF-8 and
        is generated.
will not occur if the string is valid. They are replaced with the replacement
 
character (U+FFFD) on MediaWiki pages.
Unit tests for this module are available at Module:String/tests.
--]]
]]
function str.ulen(text)
 
local _, length = string.gsub(text, UTF8_char, "")
local str = {}
return length
end


--[[
--[[
Line 31: Line 27:


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


Parameters
Parameters
Line 41: Line 37:
trailing whitespace from the target string.
trailing whitespace from the target string.
]]
]]
function str.len( frame )
function str.len(frame)
local new_args = str._getParameters( frame.args, {'s'} );
local new_args = str._getParameters(frame.args, { 's' });
local s = new_args['s'] or '';
local s = new_args['s'] or '';
return mw.ustring.len( s )
return mw.ustring.len(s)
end
end


Line 53: Line 49:


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


Parameters
Parameters
Line 61: Line 57:
     i: The fist index of the substring to return, defaults to 1.
     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.
     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
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
is a negative value, it is interpreted the same as selecting a character by
Line 70: Line 66:
reported.
reported.
]]
]]
function str.sub( frame )
function str.sub(frame)
local new_args = str._getParameters( frame.args, { 's', 'i', 'j' } );
local new_args = str._getParameters(frame.args, { 's', 'i', 'j' });
local s = new_args['s'] or '';
local s = new_args['s'] or '';
local i = tonumber( new_args['i'] ) or 1;
local i = tonumber(new_args['i']) or 1;
local j = tonumber( new_args['j'] ) or -1;
local j = tonumber(new_args['j']) or -1;
 
local len = mw.ustring.len( s );
local len = mw.ustring.len(s);
 
-- Convert negatives for range checking
-- Convert negatives for range checking
if i < 0 then
if i < 0 then
Line 85: Line 81:
j = len + j + 1;
j = len + j + 1;
end
end
 
if i > len or j > len or i < 1 or j < 1 then
if i > len or j > len or i < 1 or j < 1 then
return str._error( 'String subset index out of range' );
return str._error('String subset index out of range');
end
end
if j < i then
if j < i then
return str._error( 'String subset indices out of order' );
return str._error('String subset indices out of order');
end
end
 
return mw.ustring.sub( s, i, j )
return mw.ustring.sub(s, i, j)
end
end


Line 100: Line 96:
to maintain these older templates.
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
end


Line 113: Line 109:


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


Line 143: Line 139:
For information on constructing Lua patterns, a form of [regular expression], see:
For information on constructing Lua patterns, a form of [regular expression], see:


* https://www.lua.org/manual/5.1/manual.html#5.4.1
* http://www.lua.org/manual/5.1/manual.html#5.4.1
* https://www.mediawiki.org/wiki/Extension:Scribunto/Lua_reference_manual#Patterns
* http://www.mediawiki.org/wiki/Extension:Scribunto/Lua_reference_manual#Patterns
* https://www.mediawiki.org/wiki/Extension:Scribunto/Lua_reference_manual#Ustring_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(frame)
function str._match( s, pattern, start, match_index, plain, nomatch )
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'];
if s == '' then
if s == '' then
return str._error( 'Target string is empty' );
return str._error('Target string is empty');
end
end
if pattern == '' then
if pattern == '' then
return str._error( 'Pattern string is empty' );
return str._error('Pattern string is empty');
end
end
start = tonumber(start) or 1
if math.abs(start) < 1 or math.abs(start) > mw.ustring.len(s) then
if math.abs(start) < 1 or math.abs(start) > mw.ustring.len( s ) then
return str._error('Requested start is out of range');
return str._error( 'Requested start is out of range' );
end
end
if match_index == 0 then
if match_index == 0 then
return str._error( 'Match index is out of range' );
return str._error('Match index is out of range');
end
end
if plain_flag then
if plain_flag then
pattern = str._escapePattern( pattern );
pattern = str.pattern_escape(pattern);
end
end
 
local result
local result
if match_index == 1 then
if match_index == 1 then
-- Find first match is simple case
-- Find first match is simple case
result = mw.ustring.match( s, pattern, start )
result = mw.ustring.match(s, pattern, start)
else
else
if start > 1 then
if start > 1 then
s = mw.ustring.sub( s, start );
s = mw.ustring.sub(s, start);
end
end
 
local iterator = mw.ustring.gmatch(s, pattern);
local iterator = mw.ustring.gmatch(s, pattern);
if match_index > 0 then
if match_index > 0 then
Line 194: Line 196:
count = count + 1;
count = count + 1;
end
end
 
result = result_table[ count + match_index ];
result = result_table[count + match_index];
end
end
end
end
 
if result == nil then
if result == nil then
if nomatch == nil then
if nomatch == nil then
return str._error( 'Match not found' );
return str._error('Match not found');
else
else
return nomatch;
return nomatch;
Line 208: Line 210:
return result;
return result;
end
end
end
-- 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, nomatch )
end
end


Line 228: Line 218:


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


Parameters
Parameters
Line 247: Line 237:
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 str._error( 'String index out of range' );
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


Line 273: Line 263:
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
if target_str == '' then
return 1;
return 1;
end
end
 
local start = mw.ustring.find( source_str, target_str, 1, true )
local start = mw.ustring.find(source_str, target_str, 1, true)
if start == nil then
if start == nil then
start = -1
start = -1
end
end
 
return start
return start
end
end
Line 297: Line 287:


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


Parameters
Parameters
Line 314: Line 304:
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 an empty string.  If either "source" or "target" are missing / empty, this
function also returns 0.
function also returns an empty string.


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 params = {
local source_str = new_args['source'] or '';
[1] = { required = true },
local pattern = new_args['target'] or '';
[2] = { required = true },
local start_pos = tonumber(new_args['start']) or 1;
[3] = { type = "number" },
local plain = new_args['plain'] or true;
[4] = { type = "boolean" },
 
}
if source_str == '' or pattern == '' then
return 0;
local args = require("Module:parameters").process(frame.args, params)
end
 
return mw.ustring.find(args[1], args[2], args[3], args[4])
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
end


Line 347: Line 329:


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


Line 360: Line 342:
         text and not as a Lua style regular expression, defaults to true
         text and not as a Lua style regular expression, defaults to true
]]
]]
function str.replace( frame )
function str.replace(frame)
local new_args = str._getParameters( frame.args, {'source', 'pattern', 'replace', 'count', 'plain' } );
local new_args = str._getParameters(frame.args, { 'source', 'pattern', 'replace', 'count', 'plain' });
local source_str = new_args['source'] or '';
local source_str = new_args['source'] or '';
local pattern = new_args['pattern'] or '';
local pattern = new_args['pattern'] or '';
local replace = new_args['replace'] or '';
local replace = new_args['replace'] or '';
local count = tonumber( new_args['count'] );
local count = tonumber(new_args['count']);
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 source_str;
return source_str;
end
end
plain = str._getBoolean( plain );
plain = str._getBoolean(plain);
 
if plain then
if plain then
pattern = str._escapePattern( pattern );
pattern = str.pattern_escape(pattern);
replace = mw.ustring.gsub( replace, "%%", "%%%%" ); --Only need to escape replacement sequences.
replace = mw.ustring.gsub(replace, "%%", "%%%%"); --Only need to escape replacement sequences.
end
end
 
local result;
local result;
result = mw.ustring.gsub(source_str, pattern, replace, count);
return result;
end


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;
function str.gsub(frame)
local params = {
[1] = { required = true, allow_empty = true, allow_whitespace = true},
[2] = { required = true, allow_empty = true, allow_whitespace = true},
[3] = { required = true, allow_empty = true, allow_whitespace = true},
[4] = { type = "number" },
}
local args = require("Module:parameters").process(frame.args, params)
return (mw.ustring.gsub(args[1], args[2], args[3], args[4]))
end
end


--[[
--[[
Line 393: Line 386:
]]
]]


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


--[[
function str.lower(frame)
escapePattern
local text = frame.args[1] or ''
 
return mw.ustring.lower(text)
This function escapes special characters from a Lua string pattern. See [1]
end
for details on how patterns work.


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


--[[
--[[
Line 430: Line 407:
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
Line 443: Line 420:
new_args[arg] = value;
new_args[arg] = value;
end
end
 
return new_args;
return new_args;
end
end
Line 450: Line 427:
Helper function to handle error messages.
Helper function to handle error messages.
]]
]]
function str._error( error_str )
function str._error(error_str)
local frame = mw.getCurrentFrame();
local frame = mw.getCurrentFrame();
local error_category = frame.args.error_category or 'Errors reported by Module String';
local error_category = frame.args.error_category or 'Errors reported by Module String';
local ignore_errors = frame.args.ignore_errors or false;
local ignore_errors = frame.args.ignore_errors or false;
local no_category = frame.args.no_category or false;
local no_category = frame.args.no_category or false;
 
if str._getBoolean(ignore_errors) then
if str._getBoolean(ignore_errors) then
return '';
return '';
end
end
 
local error_str = '<strong class="error">String Module Error: ' .. error_str .. '</strong>';
local error_str = '<strong class="error">String Module Error: ' .. error_str .. '</strong>';
if error_category ~= '' and not str._getBoolean( no_category ) then
if error_category ~= '' and not str._getBoolean(no_category) then
error_str = '[[Category:' .. error_category .. ']]' .. error_str;
error_str = '[[Category:' .. error_category .. ']]' .. error_str;
end
end
 
return error_str;
return error_str;
end
end
Line 471: Line 448:
Helper Function to interpret boolean strings
Helper Function to interpret boolean strings
]]
]]
function str._getBoolean( boolean_str )
function str._getBoolean(boolean_str)
local boolean_value;
local boolean_value;
 
if type( boolean_str ) == 'string' then
if type(boolean_str) == 'string' then
boolean_str = boolean_str:lower();
boolean_str = boolean_str:lower();
if boolean_str == 'false' or boolean_str == 'no' or boolean_str == '0'
if boolean_str == 'false' or boolean_str == 'no' or boolean_str == '0'
Line 482: Line 459:
boolean_value = true;
boolean_value = true;
end
end
elseif type( boolean_str ) == 'boolean' then
elseif type(boolean_str) == 'boolean' then
boolean_value = boolean_str;
boolean_value = boolean_str;
else
else
error( 'No boolean value found' );
error('No boolean value found');
end
end
return boolean_value
return boolean_value
Line 491: Line 468:


--[[
--[[
Helper function that escapes all pattern characters so that they will be treated
Helper function that escapes all pattern characters – ().%+-*?[^$] – so that they will be treated
as plain text.
as plain text.
]]
]]
function str._escapePattern( pattern_str )
function str.pattern_escape(pattern_str)
return mw.ustring.gsub( pattern_str, "([%(%)%.%%%+%-%*%?%[%^%$%]])", "%%%1" );
local invoked = false
if type(pattern_str) == "table" then
if pattern_str.args then
local frame = pattern_str
invoked = true
if frame.args[1] then
pattern_str = frame.args[1]
else
pattern_str = frame:getParent().args[1]
end
else
error("First argument to pattern_escape should be a string, a number, or a frame object.")
end
elseif not (type(pattern_str) == "string" or type(pattern_str) == "number") then
error("First argument to pattern_escape should be a string or a number.")
end
if invoked then
local escaped = mw.ustring.gsub(pattern_str, "([%(%)%.%%%+%-%*%?%[%^%$%]])", "%%%1")
return escaped
else
return mw.ustring.gsub(pattern_str, "([%(%)%.%%%+%-%*%?%[%^%$%]])", "%%%1");
end
end
 
function str.count(text, pattern, plain)
if not (type(text) == "string" or type(text) == "number") then
error('The first argument to the function "count" must be a string or a number, not a ' .. type(text) .. '.')
end
if not (type(pattern) == "string" or type(pattern) == "number") then
error('The first argument to the function "count" must be a string or a number, not a ' .. type(text) .. '.')
end
if plain then
pattern = str.pattern_escape(pattern)
end
local _, count = mw.ustring.gsub(text, pattern, "")
return count
end
 
function str.plain_gsub(text, pattern, replacement)
local invoked = false
if type(text) == "table" then
invoked = true
if text.args then
local frame = text
local params = {
[1] = {},
[2] = {},
[3] = { allow_empty = true },
}
local args = require("Module:parameters").process(frame.args, params)
text = args[1]
pattern = args[2]
replacement = args[3]
else
error("If the first argument to plain_gsub is a table, it should be a frame object.")
end
else
if not (type(pattern) == "string" or type(pattern) == "number") then
error("The second argument to plain_gsub should be a string or a number.")
end
if not (type(replacement) == "string" or type(replacement) == "number") then
error("The third argument to plain_gsub should be a string or a number.")
end
end
pattern = str.pattern_escape(pattern)
if invoked then
text = mw.ustring.gsub(text, pattern, replacement)
return text
else
return mw.ustring.gsub(text, pattern, replacement)
end
end
 
function str.matchToArray(text, pattern)
local invoked = false
if type(text) == "table" then
invoked = true
if text.args then
local frame = text
local params = {
[1] = { required = true },
[2] = { required = true },
}
local args = require("Module:parameters").process(frame.args, params)
text = args[1]
pattern = args[2]
else
error("If the first argument to matchToArray is a table, it should be a frame object.")
end
else
if not (type(pattern) == "string" or type(pattern) == "number") then
error("The second argument to matchToArray should be a string or a number.")
end
end
local matches = {}
local i = 0
for match in mw.ustring.gmatch(text, pattern) do
i = i + 1
matches[i] = match
end
if i > 0 then
if invoked then
return table.concat(matches, ", ")
else
return matches
end
else
if invoked then
return ""
else
return nil
end
end
end
 
--[=[
Similar to gmatch, but it returns the count of the match in addition to the
list of captures, something like ipairs().
If the pattern doesn't contain any captures, the whole match is returned.
Invoke thus:
for i, whole_match in require("Module:string").imatch(text, pattern) do
[ do something with i and whole_match ]
end
or
for i, capture1[, capture2[, capture3[, ...]]] in require("Module:string").imatch(text, pattern) do
[ do something with i and capture1 ]
end
For example, this code
for i, whole_match in require("Module:string").imatch("a b c", "[a-z]") do
mw.log(i, whole_match)
end
will log
1 a
2 b
3 c
]=]
function str.imatch(text, pattern, pos, plain, use_basic_Lua_function)
local i = 0
pos = pos or 0
if not string.find(pattern, "%b()") then
pattern = "(" .. pattern .. ")"
end
local find = use_basic_Lua_function and string.find or mw.ustring.find
return function()
i = i + 1
local return_values = { find(text, pattern, pos, plain) }
local j = return_values[2]
if return_values[3] then
pos = j + 1
-- Skip the first two returned values, which are the indices of the
-- whole match.
return i, unpack(return_values, 3)
end
end
end
 
function str.escapebytes(s)
return (string.gsub(s,
'.',
function(char)
return ('\\%03d'):format(string.byte(char))
end))
end
 
function str.URIdecode(frame)
return mw.uri.decode(frame.args[1], frame.args[2] or "PATH")
end
end


return str
return str