Module:TableTools: Difference between revisions
fix cleanPattern function - some of the character classes were still being magic even when they were inside the set, so escape them all |
Implementing merger with Module:Table |
||
Line 17: | Line 17: | ||
local infinity = math.huge |
local infinity = math.huge |
||
local checkType = libraryUtil.checkType |
local checkType = libraryUtil.checkType |
||
local checkTypeMulti = libraryUtil.checkTypeMulti |
|||
--[[ |
--[[ |
||
Line 251: | Line 252: | ||
------------------------------------------------------------------------------------ |
------------------------------------------------------------------------------------ |
||
--]] |
--]] |
||
function p.size(t) |
function p.size(t) |
||
checkType('size', 1, t, 'table') |
checkType('size', 1, t, 'table') |
||
Line 258: | Line 260: | ||
end |
end |
||
return i |
return i |
||
end |
|||
local function defaultKeySort(item1, item2) |
|||
-- "number" < "string", so numbers will be sorted before strings. |
|||
local type1, type2 = type(item1), type(item2) |
|||
if type1 ~= type2 then |
|||
return type1 < type2 |
|||
else -- This will fail with table, boolean, function. |
|||
return item1 < item2 |
|||
end |
|||
end |
|||
--[[ |
|||
Returns a list of the keys in a table, sorted using either a default |
|||
comparison function or a custom keySort function. |
|||
]] |
|||
function p.keysToList(t, keySort, checked) |
|||
if not checked then |
|||
checkType('keysToList', 1, t, 'table') |
|||
checkTypeMulti('keysToList', 2, keySort, { 'function', 'boolean', 'nil' }) |
|||
end |
|||
local list = {} |
|||
local index = 1 |
|||
for key, value in pairs(t) do |
|||
list[index] = key |
|||
index = index + 1 |
|||
end |
|||
if keySort ~= false then |
|||
keySort = type(keySort) == 'function' and keySort or defaultKeySort |
|||
table.sort(list, keySort) |
|||
end |
|||
return list |
|||
end |
|||
--[[ |
|||
Iterates through a table, with the keys sorted using the keysToList function. |
|||
If there are only numerical keys, sparseIpairs is probably more efficient. |
|||
]] |
|||
function p.sortedPairs(t, keySort) |
|||
local check = _check('keysToList') |
|||
checkType('sortedPairs', 1, t, 'table') |
|||
checkType('sortedPairs', 2, keySort, 'function', true) |
|||
local list = p.keysToList(t, keySort, true) |
|||
local i = 0 |
|||
return function() |
|||
i = i + 1 |
|||
local key = list[i] |
|||
if key ~= nil then |
|||
return key, t[key] |
|||
else |
|||
return nil, nil |
|||
end |
|||
end |
|||
end |
|||
--[[ |
|||
Returns true if all keys in the table are consecutive integers starting at 1. |
|||
--]] |
|||
function p.isArray(t) |
|||
checkType("isArray", 1, t, "table") |
|||
local i = 0 |
|||
for k, v in pairs(t) do |
|||
i = i + 1 |
|||
if t[i] == nil then |
|||
return false |
|||
end |
|||
end |
|||
return true |
|||
end |
|||
-- { "a", "b", "c" } -> { a = 1, b = 2, c = 3 } |
|||
function p.invert(array) |
|||
checkType("invert", 1, array, "table") |
|||
local map = {} |
|||
for i, v in ipairs(array) do |
|||
map[v] = i |
|||
end |
|||
return map |
|||
end |
|||
--[[ |
|||
{ "a", "b", "c" } -> { ["a"] = true, ["b"] = true, ["c"] = true } |
|||
--]] |
|||
function p.listToSet(t) |
|||
checkType("listToSet", 1, t, "table") |
|||
local set = {} |
|||
for _, item in ipairs(t) do |
|||
set[item] = true |
|||
end |
|||
return set |
|||
end |
|||
--[[ |
|||
Recursive deep copy function. |
|||
Preserves identities of subtables. |
|||
]] |
|||
local function _deepCopy(orig, includeMetatable, already_seen) |
|||
-- Stores copies of tables indexed by the original table. |
|||
already_seen = already_seen or {} |
|||
local copy = already_seen[orig] |
|||
if copy ~= nil then |
|||
return copy |
|||
end |
|||
if type(orig) == 'table' then |
|||
copy = {} |
|||
for orig_key, orig_value in pairs(orig) do |
|||
copy[deepcopy(orig_key, includeMetatable, already_seen)] = deepcopy(orig_value, includeMetatable, already_seen) |
|||
end |
|||
already_seen[orig] = copy |
|||
if includeMetatable then |
|||
local mt = getmetatable(orig) |
|||
if mt ~= nil then |
|||
local mt_copy = deepcopy(mt, includeMetatable, already_seen) |
|||
setmetatable(copy, mt_copy) |
|||
already_seen[mt] = mt_copy |
|||
end |
|||
end |
|||
else -- number, string, boolean, etc |
|||
copy = orig |
|||
end |
|||
return copy |
|||
end |
|||
function p.deepCopy(orig, noMetatable, already_seen) |
|||
checkType("deepCopy", 3, already_seen, "table", true) |
|||
return _deepCopy(orig, not noMetatable, already_seen) |
|||
end |
|||
--[[ |
|||
Concatenates all values in the table that are indexed by a number, in order. |
|||
sparseConcat{ a, nil, c, d } => "acd" |
|||
sparseConcat{ nil, b, c, d } => "bcd" |
|||
]] |
|||
function p.sparseConcat(t, sep, i, j) |
|||
local list = {} |
|||
local list_i = 0 |
|||
for _, v in p.sparseIpairs(t) do |
|||
list_i = list_i + 1 |
|||
list[list_i] = v |
|||
end |
|||
return table.concat(list, sep, i, j) |
|||
end |
|||
--[[ |
|||
-- This returns the length of a table, or the first integer key n counting from |
|||
-- 1 such that t[n + 1] is nil. It is similar to the operator #, but may return |
|||
-- a different value when there are gaps in the array portion of the table. |
|||
-- Intended to be used on data loaded with mw.loadData. For other tables, use #. |
|||
--]] |
|||
function p.length(t) |
|||
local i = 0 |
|||
repeat |
|||
i = i + 1 |
|||
until t[i] == nil |
|||
return i - 1 |
|||
end |
|||
function p.inArray(arr, valueToFind) |
|||
checkType("inArray", 1, arr, "table") |
|||
-- if valueToFind is nil, error? |
|||
for _, v in ipairs(arr) do |
|||
if v == valueToFind then |
|||
return true |
|||
end |
|||
end |
|||
return false |
|||
end |
end |
||
Revision as of 21:28, 19 February 2019
This Lua module is used in system messages, and on approximately 6,250,000 pages, or roughly 10% of all pages. Changes to it can cause immediate changes to the Wikipedia user interface. To avoid major disruption and server load, any changes should be tested in the module's /sandbox or /testcases subpages, or in your own module sandbox. The tested changes can be added to this page in a single edit. Please discuss changes on the talk page before implementing them. |
This module is subject to page protection. It is a highly visible module in use by a very large number of pages, or is substituted very frequently. Because vandalism or mistakes would affect many pages, and even trivial editing might cause substantial load on the servers, it is protected from editing. |
This module depends on the following other modules: |
This module includes a number of functions for dealing with Lua tables. It is a meta-module, meant to be called from other Lua modules, and should not be called directly from #invoke.
Loading the module
To use any of the functions, first you must load the module.
local TableTools = require('Module:TableTools')
isPositiveInteger
TableTools.isPositiveInteger(value)
Returns true
if value
is a positive integer, and false
if not. Although it doesn't operate on tables, it is included here as it is useful for determining whether a given table key is in the array part or the hash part of a table.
isNan
TableTools.isNan(value)
Returns true
if value
is a NaN value, and false
if not. Although it doesn't operate on tables, it is included here as it is useful for determining whether a value can be a valid table key. (Lua will generate an error if a NaN value is used as a table key.)
shallowClone
TableTools.shallowClone(t)
Returns a clone of a table. The value returned is a new table, but all subtables and functions are shared. Metamethods are respected, but the returned table will have no metatable of its own. If you want to make a new table with no shared subtables and with metatables transferred, you can use mw.clone
instead. If you want to make a new table with no shared subtables and without metatables transferred, use deepCopy
with the noMetatable
option.
removeDuplicates
TableTools.removeDuplicates(t)
Removes duplicate values from an array. This function is only designed to work with standard arrays: keys that are not positive integers are ignored, as are all values after the first nil
value. (For arrays containing nil
values, you can use compressSparseArray
first.) The function tries to preserve the order of the array: the earliest non-unique value is kept, and all subsequent duplicate values are removed. For example, for the table {5, 4, 4, 3, 4, 2, 2, 1}
removeDuplicates
will return {5, 4, 3, 2, 1}
.
numKeys
TableTools.numKeys(t)
Takes a table t
and returns an array containing the numbers of any positive integer keys that have non-nil values, sorted in numerical order. For example, for the table {'foo', nil, 'bar', 'baz', a = 'b'}
, numKeys
will return {1, 3, 4}
.
affixNums
TableTools.affixNums(t, prefix, suffix)
Takes a table t
and returns an array containing the numbers of keys with the optional prefix prefix
and the optional suffix suffix
. For example, for the table {a1 = 'foo', a3 = 'bar', a6 = 'baz'}
and the prefix 'a'
, affixNums
will return {1, 3, 6}
. All characters in prefix
and suffix
are interpreted literally.
numData
TableTools.numData(t, compress)
Given a table with keys like "foo1"
, "bar1"
, "foo2"
, and "baz2"
, returns a table of subtables in the format { [1] = {foo = 'text', bar = 'text'}, [2] = {foo = 'text', baz = 'text'} }
. Keys that don't end with an integer are stored in a subtable named "other"
. The compress option compresses the table so that it can be iterated over with ipairs
.
compressSparseArray
TableTools.compressSparseArray(t)
Takes an array t
with one or more nil values, and removes the nil values while preserving the order, so that the array can be safely traversed with ipairs
. Any keys that are not positive integers are removed. For example, for the table {1, nil, foo = 'bar', 3, 2}
, compressSparseArray
will return {1, 3, 2}
.
sparseIpairs
TableTools.sparseIpairs(t)
This is an iterator function for traversing a sparse array t
. It is similar to ipairs
, but will continue to iterate until the highest numerical key, whereas ipairs
may stop after the first nil
value. Any keys that are not positive integers are ignored.
Usually sparseIpairs
is used in a generic for
loop.
for i, v in TableTools.sparseIpairs(t) do
-- code block
end
Note that sparseIpairs
uses the pairs
function in its implementation. Although some table keys appear to be ignored, all table keys are accessed when it is run.
size
TableTools.size(t)
Finds the size of a key/value pair table (associative array). For example, for {foo = 'foo', bar = 'bar'}
, size
will return 2
. The function will also work on arrays, but for arrays it is more efficient to use the #
operator. Note that to find the size, this function uses the pairs
function to iterate through all of the keys.
keysToList
TableTools.keysToList(t, keySort, checked)
Returns a list of the keys in a table, sorted using either a default comparison function or a custom keySort
function, which follows the same rules as the comp
function supplied to table.sort
. If keySort
is false
, no sorting is done. Set checked
to true
to skip the internal type checking.
sortedPairs
TableTools.sortedPairs(t, keySort)
Iterates through a table, with the keys sorted using the keysToList
function. If there are only numerical keys, sparseIpairs
is probably more efficient.
isArray
TableTools.isArray(value)
Returns true
if value
is a table and all keys are consecutive integers starting at 1.
isArrayLike
TableTools.isArrayLike(value)
Returns true
if value
is iterable and all keys are consecutive integers starting at 1.
invert
TableTools.invert(arr)
Transposes the keys and values in an array. For example, invert{ "a", "b", "c" }
yields { a=1, b=2, c=3 }
.
listToSet
TableTools.listToSet(arr)
Creates a set from the array part of the table arr
. Indexing the set by any of the values of the array returns true
. For example, listToSet{ "a", "b", "c" }
yields { a=true, b=true, c=true }
.
deepCopy
TableTools.deepCopy(orig, noMetatable, alreadySeen)
Creates a copy of the table orig
. As with mw.clone
, all values that are not functions are duplicated and the identity of tables is preserved. If noMetatable
is true
, then the metatable (if any) is not copied. Can copy tables loaded with mw.loadData
.
Similar to mw.clone
, but mw.clone
cannot copy tables loaded with mw.loadData
and does not allow metatables not to be copied.
sparseConcat
TableTools.sparseConcat(t, sep, i, j)
Concatenates all values in the table that are indexed by a positive integer, in order. For example, sparseConcat{ "a", nil, "c", "d" }
yields "acd"
and sparseConcat{ nil, "b", "c", "d" }
yields "bcd"
.
length
TableTools.length(t, prefix)
Finds the length of an array or of a quasi-array with keys with an optional prefix
such as "data1", "data2", etc. It uses an exponential search algorithm to find the length, so as to use as few table lookups as possible.
This algorithm is useful for arrays that use metatables (e.g. frame.args) and for quasi-arrays. For normal arrays, just use the # operator, as it is implemented in C and will be quicker.
inArray
TableTools.inArray(arr, valueToFind)
Returns true
if valueToFind
is a member of the array arr
, and false
otherwise.
merge
TableTools.merge(...)
Given the arrays, returns an array containing the elements of each input array in sequence.
extend
TableTools.extend(arr1, arr2)
Extends the first array in place by appending all elements from the second array.
See also
- {{#invoke:params}}
--[[
------------------------------------------------------------------------------------
-- TableTools --
-- --
-- This module includes a number of functions for dealing with Lua tables. --
-- It is a meta-module, meant to be called from other Lua modules, and should --
-- not be called directly from #invoke. --
------------------------------------------------------------------------------------
--]]
local libraryUtil = require('libraryUtil')
local p = {}
-- Define often-used variables and functions.
local floor = math.floor
local infinity = math.huge
local checkType = libraryUtil.checkType
local checkTypeMulti = libraryUtil.checkTypeMulti
--[[
------------------------------------------------------------------------------------
-- isPositiveInteger
--
-- This function returns true if the given value is a positive integer, and false
-- if not. Although it doesn't operate on tables, it is included here as it is
-- useful for determining whether a given table key is in the array part or the
-- hash part of a table.
------------------------------------------------------------------------------------
--]]
function p.isPositiveInteger(v)
if type(v) == 'number' and v >= 1 and floor(v) == v and v < infinity then
return true
else
return false
end
end
--[[
------------------------------------------------------------------------------------
-- isNan
--
-- This function returns true if the given number is a NaN value, and false
-- if not. Although it doesn't operate on tables, it is included here as it is
-- useful for determining whether a value can be a valid table key. Lua will
-- generate an error if a NaN is used as a table key.
------------------------------------------------------------------------------------
--]]
function p.isNan(v)
if type(v) == 'number' and tostring(v) == '-nan' then
return true
else
return false
end
end
--[[
------------------------------------------------------------------------------------
-- shallowClone
--
-- This returns a clone of a table. The value returned is a new table, but all
-- subtables and functions are shared. Metamethods are respected, but the returned
-- table will have no metatable of its own.
------------------------------------------------------------------------------------
--]]
function p.shallowClone(t)
local ret = {}
for k, v in pairs(t) do
ret[k] = v
end
return ret
end
--[[
------------------------------------------------------------------------------------
-- removeDuplicates
--
-- This removes duplicate values from an array. Non-positive-integer keys are
-- ignored. The earliest value is kept, and all subsequent duplicate values are
-- removed, but otherwise the array order is unchanged.
------------------------------------------------------------------------------------
--]]
function p.removeDuplicates(t)
checkType('removeDuplicates', 1, t, 'table')
local isNan = p.isNan
local ret, exists = {}, {}
for i, v in ipairs(t) do
if isNan(v) then
-- NaNs can't be table keys, and they are also unique, so we don't need to check existence.
ret[#ret + 1] = v
else
if not exists[v] then
ret[#ret + 1] = v
exists[v] = true
end
end
end
return ret
end
--[[
------------------------------------------------------------------------------------
-- numKeys
--
-- This takes a table and returns an array containing the numbers of any numerical
-- keys that have non-nil values, sorted in numerical order.
------------------------------------------------------------------------------------
--]]
function p.numKeys(t)
checkType('numKeys', 1, t, 'table')
local isPositiveInteger = p.isPositiveInteger
local nums = {}
for k, v in pairs(t) do
if isPositiveInteger(k) then
nums[#nums + 1] = k
end
end
table.sort(nums)
return nums
end
--[[
------------------------------------------------------------------------------------
-- affixNums
--
-- This takes a table and returns an array containing the numbers of keys with the
-- specified prefix and suffix. For example, for the table
-- {a1 = 'foo', a3 = 'bar', a6 = 'baz'} and the prefix "a", affixNums will
-- return {1, 3, 6}.
------------------------------------------------------------------------------------
--]]
function p.affixNums(t, prefix, suffix)
checkType('affixNums', 1, t, 'table')
checkType('affixNums', 2, prefix, 'string', true)
checkType('affixNums', 3, suffix, 'string', true)
local function cleanPattern(s)
-- Cleans a pattern so that the magic characters ()%.[]*+-?^$ are interpreted literally.
s = s:gsub('([%(%)%%%.%[%]%*%+%-%?%^%$])', '%%%1')
return s
end
prefix = prefix or ''
suffix = suffix or ''
prefix = cleanPattern(prefix)
suffix = cleanPattern(suffix)
local pattern = '^' .. prefix .. '([1-9]%d*)' .. suffix .. '$'
local nums = {}
for k, v in pairs(t) do
if type(k) == 'string' then
local num = mw.ustring.match(k, pattern)
if num then
nums[#nums + 1] = tonumber(num)
end
end
end
table.sort(nums)
return nums
end
--[[
------------------------------------------------------------------------------------
-- numData
--
-- Given a table with keys like ("foo1", "bar1", "foo2", "baz2"), returns a table
-- of subtables in the format
-- { [1] = {foo = 'text', bar = 'text'}, [2] = {foo = 'text', baz = 'text'} }
-- Keys that don't end with an integer are stored in a subtable named "other".
-- The compress option compresses the table so that it can be iterated over with
-- ipairs.
------------------------------------------------------------------------------------
--]]
function p.numData(t, compress)
checkType('numData', 1, t, 'table')
checkType('numData', 2, compress, 'boolean', true)
local ret = {}
for k, v in pairs(t) do
local prefix, num = mw.ustring.match(tostring(k), '^([^0-9]*)([1-9][0-9]*)$')
if num then
num = tonumber(num)
local subtable = ret[num] or {}
if prefix == '' then
-- Positional parameters match the blank string; put them at the start of the subtable instead.
prefix = 1
end
subtable[prefix] = v
ret[num] = subtable
else
local subtable = ret.other or {}
subtable[k] = v
ret.other = subtable
end
end
if compress then
local other = ret.other
ret = p.compressSparseArray(ret)
ret.other = other
end
return ret
end
--[[
------------------------------------------------------------------------------------
-- compressSparseArray
--
-- This takes an array with one or more nil values, and removes the nil values
-- while preserving the order, so that the array can be safely traversed with
-- ipairs.
------------------------------------------------------------------------------------
--]]
function p.compressSparseArray(t)
checkType('compressSparseArray', 1, t, 'table')
local ret = {}
local nums = p.numKeys(t)
for _, num in ipairs(nums) do
ret[#ret + 1] = t[num]
end
return ret
end
--[[
------------------------------------------------------------------------------------
-- sparseIpairs
--
-- This is an iterator for sparse arrays. It can be used like ipairs, but can
-- handle nil values.
------------------------------------------------------------------------------------
--]]
function p.sparseIpairs(t)
checkType('sparseIpairs', 1, t, 'table')
local nums = p.numKeys(t)
local i = 0
local lim = #nums
return function ()
i = i + 1
if i <= lim then
local key = nums[i]
return key, t[key]
else
return nil, nil
end
end
end
--[[
------------------------------------------------------------------------------------
-- size
--
-- This returns the size of a key/value pair table. It will also work on arrays,
-- but for arrays it is more efficient to use the # operator.
------------------------------------------------------------------------------------
--]]
function p.size(t)
checkType('size', 1, t, 'table')
local i = 0
for k in pairs(t) do
i = i + 1
end
return i
end
local function defaultKeySort(item1, item2)
-- "number" < "string", so numbers will be sorted before strings.
local type1, type2 = type(item1), type(item2)
if type1 ~= type2 then
return type1 < type2
else -- This will fail with table, boolean, function.
return item1 < item2
end
end
--[[
Returns a list of the keys in a table, sorted using either a default
comparison function or a custom keySort function.
]]
function p.keysToList(t, keySort, checked)
if not checked then
checkType('keysToList', 1, t, 'table')
checkTypeMulti('keysToList', 2, keySort, { 'function', 'boolean', 'nil' })
end
local list = {}
local index = 1
for key, value in pairs(t) do
list[index] = key
index = index + 1
end
if keySort ~= false then
keySort = type(keySort) == 'function' and keySort or defaultKeySort
table.sort(list, keySort)
end
return list
end
--[[
Iterates through a table, with the keys sorted using the keysToList function.
If there are only numerical keys, sparseIpairs is probably more efficient.
]]
function p.sortedPairs(t, keySort)
local check = _check('keysToList')
checkType('sortedPairs', 1, t, 'table')
checkType('sortedPairs', 2, keySort, 'function', true)
local list = p.keysToList(t, keySort, true)
local i = 0
return function()
i = i + 1
local key = list[i]
if key ~= nil then
return key, t[key]
else
return nil, nil
end
end
end
--[[
Returns true if all keys in the table are consecutive integers starting at 1.
--]]
function p.isArray(t)
checkType("isArray", 1, t, "table")
local i = 0
for k, v in pairs(t) do
i = i + 1
if t[i] == nil then
return false
end
end
return true
end
-- { "a", "b", "c" } -> { a = 1, b = 2, c = 3 }
function p.invert(array)
checkType("invert", 1, array, "table")
local map = {}
for i, v in ipairs(array) do
map[v] = i
end
return map
end
--[[
{ "a", "b", "c" } -> { ["a"] = true, ["b"] = true, ["c"] = true }
--]]
function p.listToSet(t)
checkType("listToSet", 1, t, "table")
local set = {}
for _, item in ipairs(t) do
set[item] = true
end
return set
end
--[[
Recursive deep copy function.
Preserves identities of subtables.
]]
local function _deepCopy(orig, includeMetatable, already_seen)
-- Stores copies of tables indexed by the original table.
already_seen = already_seen or {}
local copy = already_seen[orig]
if copy ~= nil then
return copy
end
if type(orig) == 'table' then
copy = {}
for orig_key, orig_value in pairs(orig) do
copy[deepcopy(orig_key, includeMetatable, already_seen)] = deepcopy(orig_value, includeMetatable, already_seen)
end
already_seen[orig] = copy
if includeMetatable then
local mt = getmetatable(orig)
if mt ~= nil then
local mt_copy = deepcopy(mt, includeMetatable, already_seen)
setmetatable(copy, mt_copy)
already_seen[mt] = mt_copy
end
end
else -- number, string, boolean, etc
copy = orig
end
return copy
end
function p.deepCopy(orig, noMetatable, already_seen)
checkType("deepCopy", 3, already_seen, "table", true)
return _deepCopy(orig, not noMetatable, already_seen)
end
--[[
Concatenates all values in the table that are indexed by a number, in order.
sparseConcat{ a, nil, c, d } => "acd"
sparseConcat{ nil, b, c, d } => "bcd"
]]
function p.sparseConcat(t, sep, i, j)
local list = {}
local list_i = 0
for _, v in p.sparseIpairs(t) do
list_i = list_i + 1
list[list_i] = v
end
return table.concat(list, sep, i, j)
end
--[[
-- This returns the length of a table, or the first integer key n counting from
-- 1 such that t[n + 1] is nil. It is similar to the operator #, but may return
-- a different value when there are gaps in the array portion of the table.
-- Intended to be used on data loaded with mw.loadData. For other tables, use #.
--]]
function p.length(t)
local i = 0
repeat
i = i + 1
until t[i] == nil
return i - 1
end
function p.inArray(arr, valueToFind)
checkType("inArray", 1, arr, "table")
-- if valueToFind is nil, error?
for _, v in ipairs(arr) do
if v == valueToFind then
return true
end
end
return false
end
return p