Module pl.tablex

Extended operations on Lua tables.

See the Guide

Dependencies: pl.utils, pl.types

Functions

size (t) total number of elements in this table.
index_by (tbl, idx) return a list of all values in a table indexed by another list.
transform (fun, t, ...) apply a function to all values of a table, in-place.
range (start, finish[, step=1]) generate a table of all numbers in a range.
reduce (fun, t, memo) 'reduce' a list using a binary function.
index_map (t) create an index map from a list-like table.
makeset (t) create a set from a list-like table.
union (t1, t2) the union of two map-like tables.
intersection (t1, t2) the intersection of two map-like tables.
count_map (t, cmp) A table where the key/values are the values and value counts of the table.
set (t, val[, i1=1[, i2=#t]]) set an array range to a value.
new (n, val) create a new array of specified size with initial value.
clear (t, istart) clear out the contents of a table.
removevalues (t, i1, i2) remove a range of values from a table.
readonly (t) modifies a table to be read only.

Copying

update (t1, t2) copy a table into another, in-place.
copy (t) make a shallow copy of a table
deepcopy (t) make a deep copy of a table, recursively copying all the keys and fields.
icopy (dest, src[, idest=1[, isrc=1[, nsrc=#src]]]) copy an array into another one, clearing dest after idest+nsrc, if necessary.
move (dest, src[, idest=1[, isrc=1[, nsrc=#src]]]) copy an array into another one.
insertvalues (t[, position], values) insert values into a table.

Comparing

deepcompare (t1, t2[, ignore_mt[, eps]]) compare two values.
compare (t1, t2, cmp) compare two arrays using a predicate.
compare_no_order (t1, t2, cmp) compare two list-like tables using an optional predicate, without regard for element order.

Finding

find (t, val, idx) return the index of a value in a list.
rfind (t, val, idx) return the index of a value in a list, searching from the end.
find_if (t, cmp, arg) return the index (or key) of a value in a table using a comparison function.
search (t, value[, exclude]) find a value in a table by recursive search.

MappingAndFiltering

map (fun, t, ...) apply a function to all values of a table.
imap (fun, t, ...) apply a function to all values of a list.
map_named_method (name, t, ...) apply a named method to values from a table.
map2 (fun, t1, t2, ...) apply a function to values from two tables.
imap2 (fun, t1, t2, ...) apply a function to values from two arrays.
mapn (fun, ..., fun) Apply a function to a number of tables.
pairmap (fun, t, ...) call the function with the key and value pairs from a table.
filter (t, pred, arg) filter an array's values using a predicate function

Iterating

foreach (t, fun, ...) apply a function to all elements of a table.
foreachi (t, fun, ...) apply a function to all elements of a list-like table in order.
sort (t, f) return an iterator to a table sorted by its keys
sortv (t, f) return an iterator to a table sorted by its values

Extraction

keys (t) return all the keys of a table in arbitrary order.
values (t) return all the values of the table in arbitrary order
sub (t, first, last) Extract a range from a table, like 'string.sub'.

Merging

merge (t1, t2, dup) combine two tables, either as union or intersection.
difference (s1, s2, symm) a new table which is the difference of two tables.
zip (...) return a table where each element is a table of the ith values of an arbitrary number of tables.


Functions

size (t)
total number of elements in this table. Note that this is distinct from #t, which is the number of values in the array part; this value will always be greater or equal. The difference gives the size of the hash part, for practical purposes. Works for any object with a __pairs metamethod.

Parameters:

Returns:

    the size
index_by (tbl, idx)
return a list of all values in a table indexed by another list.

Parameters:

  • tbl table a table
  • idx array an index table (a list of keys)

Returns:

    a list-like table

Usage:

  • index_by({10,20,30,40},{2,4}) == {20,40}
  • index_by({one=1,two=2,three=3},{'one','three'}) == {1,3}
transform (fun, t, ...)
apply a function to all values of a table, in-place. Any extra arguments are passed to the function.

Parameters:

  • fun function A function that takes at least one argument
  • t table a table
  • ... extra arguments passed to fun

See also:

range (start, finish[, step=1])
generate a table of all numbers in a range. This is consistent with a numerical for loop.

Parameters:

  • start integer number
  • finish integer number
  • step integer make this negative for start < finish (default 1)
reduce (fun, t, memo)
'reduce' a list using a binary function.

Parameters:

  • fun function a function of two arguments
  • t array a list-like table
  • memo array optional initial memo value. Defaults to first value in table.

Returns:

    the result of the function

Usage:

    reduce('+',{1,2,3,4}) == 10
index_map (t)
create an index map from a list-like table. The original values become keys, and the associated values are the indices into the original list.

Parameters:

  • t array a list-like table

Returns:

    a map-like table
makeset (t)
create a set from a list-like table. A set is a table where the original values become keys, and the associated values are all true.

Parameters:

  • t array a list-like table

Returns:

    a set (a map-like table)
union (t1, t2)
the union of two map-like tables. If there are duplicate keys, the second table wins.

Parameters:

Returns:

    tab

See also:

intersection (t1, t2)
the intersection of two map-like tables.

Parameters:

Returns:

    tab

See also:

count_map (t, cmp)
A table where the key/values are the values and value counts of the table.

Parameters:

  • t array a list-like table
  • cmp function a function that defines equality (otherwise uses ==)

Returns:

    a map-like table

See also:

set (t, val[, i1=1[, i2=#t]])
set an array range to a value. If it's a function we use the result of applying it to the indices.

Parameters:

  • t array a list-like table
  • val a value
  • i1 integer start range (default 1)
  • i2 integer end range (default #t)
new (n, val)
create a new array of specified size with initial value.

Parameters:

  • n integer size
  • val initial value (can be nil, but don't expect # to work!)

Returns:

    the table
clear (t, istart)
clear out the contents of a table.

Parameters:

  • t array a list
  • istart optional start position
removevalues (t, i1, i2)
remove a range of values from a table. End of range may be negative.

Parameters:

  • t array a list-like table
  • i1 integer start index
  • i2 integer end index

Returns:

    the table
readonly (t)
modifies a table to be read only. This only offers weak protection. Tables can still be modified with table.insert and rawset.

NOTE: for Lua 5.1 length, pairs and ipairs will not work, since the equivalent metamethods are only available in Lua 5.2 and newer.

Parameters:

Returns:

    the table read only (a proxy).

Copying

update (t1, t2)
copy a table into another, in-place.

Parameters:

  • t1 table destination table
  • t2 table source (actually any iterable object)

Returns:

    first table
copy (t)
make a shallow copy of a table

Parameters:

  • t table an iterable source

Returns:

    new table
deepcopy (t)
make a deep copy of a table, recursively copying all the keys and fields. This supports cycles in tables; cycles will be reproduced in the copy. This will also set the copied table's metatable to that of the original.

Parameters:

Returns:

    new table
icopy (dest, src[, idest=1[, isrc=1[, nsrc=#src]]])
copy an array into another one, clearing dest after idest+nsrc, if necessary.

Parameters:

  • dest array a list-like table
  • src array a list-like table
  • idest integer where to start copying values into destination (default 1)
  • isrc integer where to start copying values from source (default 1)
  • nsrc integer number of elements to copy from source (default #src)
move (dest, src[, idest=1[, isrc=1[, nsrc=#src]]])
copy an array into another one.

Parameters:

  • dest array a list-like table
  • src array a list-like table
  • idest integer where to start copying values into destination (default 1)
  • isrc integer where to start copying values from source (default 1)
  • nsrc integer number of elements to copy from source (default #src)
insertvalues (t[, position], values)
insert values into a table. similar to table.insert but inserts values from given table values, not the object itself, into table t at position pos.

Parameters:

  • t array the list
  • position integer (default is at end) (optional)
  • values array

Comparing

deepcompare (t1, t2[, ignore_mt[, eps]])
compare two values. if they are tables, then compare their keys and fields recursively.

Parameters:

  • t1 A value
  • t2 A value
  • ignore_mt boolean if true, ignore __eq metamethod (default false) (optional)
  • eps number if defined, then used for any number comparisons (optional)

Returns:

    true or false
compare (t1, t2, cmp)
compare two arrays using a predicate.

Parameters:

  • t1 array an array
  • t2 array an array
  • cmp function A comparison function; bool = cmp(t1_value, t2_value)

Returns:

    true or false

Usage:

    assert(tablex.compare({ 1, 2, 3 }, { 1, 2, 3 }, "=="))

assert(tablex.compare(
   {1,2,3, hello = "world"},  -- fields are not compared!
   {1,2,3}, function(v1, v2) return v1 == v2 end)
compare_no_order (t1, t2, cmp)
compare two list-like tables using an optional predicate, without regard for element order.

Parameters:

  • t1 array a list-like table
  • t2 array a list-like table
  • cmp A comparison function (may be nil)

Finding

find (t, val, idx)
return the index of a value in a list. Like string.find, there is an optional index to start searching, which can be negative.

Parameters:

  • t array A list-like table
  • val A value
  • idx integer index to start; -1 means last element,etc (default 1)

Returns:

    index of value or nil if not found

Usage:

  • find({10,20,30},20) == 2
  • find({'a','b','a','c'},'a',2) == 3
rfind (t, val, idx)
return the index of a value in a list, searching from the end. Like string.find, there is an optional index to start searching, which can be negative.

Parameters:

  • t array A list-like table
  • val A value
  • idx index to start; -1 means last element,etc (default #t)

Returns:

    index of value or nil if not found

Usage:

    rfind({10,10,10},10) == 3
find_if (t, cmp, arg)
return the index (or key) of a value in a table using a comparison function.

NOTE: the 2nd return value of this function, the value returned by the comparison function, has a limitation that it cannot be false. Because if it is, then it indicates the comparison failed, and the function will continue the search. See examples.

Parameters:

  • t table A table
  • cmp function A comparison function
  • arg an optional second argument to the function

Returns:

  1. index of value, or nil if not found
  2. value returned by comparison function (cannot be false!)

Usage:

    -- using an operator
local lst = { "Rudolph", true, false, 15 }
local idx, cmp_result = tablex.rfind(lst, "==", "Rudolph")
assert(idx == 1)
assert(cmp_result == true)

local idx, cmp_result = tablex.rfind(lst, "==", false)
assert(idx == 3)
assert(cmp_result == true)       -- looking up 'false' works!

-- using a function returning the value looked up
local cmp = function(v1, v2) return v1 == v2 and v2 end
local idx, cmp_result = tablex.rfind(lst, cmp, "Rudolph")
assert(idx == 1)
assert(cmp_result == "Rudolph")  -- the value is returned

-- NOTE: this fails, since 'false' cannot be returned!
local idx, cmp_result = tablex.rfind(lst, cmp, false)
assert(idx == nil)               -- looking up 'false' failed!
assert(cmp_result == nil)
search (t, value[, exclude])
find a value in a table by recursive search.

Parameters:

  • t table the table
  • value the value
  • exclude array any tables to avoid searching (optional)

Returns:

    a fieldspec, e.g. 'a.b' or 'math.sin'

Usage:

    search(_G,math.sin,{package.path}) == 'math.sin'

MappingAndFiltering

map (fun, t, ...)
apply a function to all values of a table. This returns a table of the results. Any extra arguments are passed to the function.

Parameters:

  • fun function A function that takes at least one argument
  • t table A table
  • ... optional arguments

Usage:

    map(function(v) return v*v end, {10,20,30,fred=2}) is {100,400,900,fred=4}
imap (fun, t, ...)
apply a function to all values of a list. This returns a table of the results. Any extra arguments are passed to the function.

Parameters:

  • fun function A function that takes at least one argument
  • t array a table (applies to array part)
  • ... optional arguments

Returns:

    a list-like table

Usage:

    imap(function(v) return v*v end, {10,20,30,fred=2}) is {100,400,900}
map_named_method (name, t, ...)
apply a named method to values from a table.

Parameters:

  • name string the method name
  • t array a list-like table
  • ... any extra arguments to the method

Returns:

    a List with the results of the method (1st result only)

Usage:

    local Car = {}
Car.__index = Car
function Car.new(car)
  return setmetatable(car or {}, Car)
end
Car.speed = 0
function Car:faster(increase)
  self.speed = self.speed + increase
  return self.speed
end

local ferrari = Car.new{ name = "Ferrari" }
local lamborghini = Car.new{ name = "Lamborghini", speed = 50 }
local cars = { ferrari, lamborghini }

assert(ferrari.speed == 0)
assert(lamborghini.speed == 50)
tablex.map_named_method("faster", cars, 10)
assert(ferrari.speed == 10)
assert(lamborghini.speed == 60)
map2 (fun, t1, t2, ...)
apply a function to values from two tables.

Parameters:

  • fun function a function of at least two arguments
  • t1 table a table
  • t2 table a table
  • ... extra arguments

Returns:

    a table

Usage:

    map2('+',{1,2,3,m=4},{10,20,30,m=40}) is {11,22,23,m=44}
imap2 (fun, t1, t2, ...)
apply a function to values from two arrays. The result will be the length of the shortest array.

Parameters:

  • fun function a function of at least two arguments
  • t1 array a list-like table
  • t2 array a list-like table
  • ... extra arguments

Usage:

    imap2('+',{1,2,3,m=4},{10,20,30,m=40}) is {11,22,23}
mapn (fun, ..., fun)
Apply a function to a number of tables. A more general version of map The result is a table containing the result of applying that function to the ith value of each table. Length of output list is the minimum length of all the lists

Parameters:

  • fun A function that takes as many arguments as there are tables
  • ... table n tables
  • fun A function that takes as many arguments as there are tables

Usage:

  • mapn(function(x,y,z) return x+y+z end, {1,2,3},{10,20,30},{100,200,300}) is {111,222,333}
  • mapn(math.max, {1,20,300},{10,2,3},{100,200,100}) is    {100,200,300}
pairmap (fun, t, ...)
call the function with the key and value pairs from a table. The function can return a value and a key (note the order!). If both are not nil, then this pair is inserted into the result: if the key already exists, we convert the value for that key into a table and append into it. If only value is not nil, then it is appended to the result.

Parameters:

  • fun function A function which will be passed each key and value as arguments, plus any extra arguments to pairmap.
  • t table A table
  • ... optional arguments

Usage:

  • pairmap(function(k,v) return v end,{fred=10,bonzo=20}) is {10,20} _or_ {20,10}
  • pairmap(function(k,v) return {k,v},k end,{one=1,two=2}) is {one={'one',1},two={'two',2}}
filter (t, pred, arg)
filter an array's values using a predicate function

Parameters:

  • t array a list-like table
  • pred function a boolean function
  • arg optional argument to be passed as second argument of the predicate

Iterating

foreach (t, fun, ...)
apply a function to all elements of a table. The arguments to the function will be the value, the key and finally any extra arguments passed to this function. Note that the Lua 5.0 function table.foreach passed the key first.

Parameters:

  • t table a table
  • fun function a function on the elements; function(value, key, ...)
  • ... extra arguments passed to fun

See also:

foreachi (t, fun, ...)
apply a function to all elements of a list-like table in order. The arguments to the function will be the value, the index and finally any extra arguments passed to this function

Parameters:

  • t array a table
  • fun function a function with at least one argument
  • ... optional arguments
sort (t, f)
return an iterator to a table sorted by its keys

Parameters:

  • t table the table
  • f function an optional comparison function (f(x,y) is true if x < y)

Returns:

    an iterator to traverse elements sorted by the keys

Usage:

    for k,v in tablex.sort(t) do print(k,v) end
sortv (t, f)
return an iterator to a table sorted by its values

Parameters:

  • t table the table
  • f function an optional comparison function (f(x,y) is true if x < y)

Returns:

    an iterator to traverse elements sorted by the values

Usage:

    for k,v in tablex.sortv(t) do print(k,v) end

Extraction

keys (t)
return all the keys of a table in arbitrary order.

Parameters:

  • t table A list-like table where the values are the keys of the input table
values (t)
return all the values of the table in arbitrary order

Parameters:

  • t table A list-like table where the values are the values of the input table
sub (t, first, last)
Extract a range from a table, like 'string.sub'. If first or last are negative then they are relative to the end of the list eg. sub(t,-2) gives last 2 entries in a list, and sub(t,-4,-2) gives from -4th to -2nd

Parameters:

  • t array a list-like table
  • first integer An index
  • last integer An index

Returns:

    a new List

Merging

merge (t1, t2, dup)
combine two tables, either as union or intersection. Corresponds to set operations for sets () but more general. Not particularly useful for list-like tables.

Parameters:

  • t1 table a table
  • t2 table a table
  • dup boolean true for a union, false for an intersection.

See also:

Usage:

  • merge({alice=23,fred=34},{bob=25,fred=34}) is {fred=34}
  • merge({alice=23,fred=34},{bob=25,fred=34},true) is {bob=25,fred=34,alice=23}
difference (s1, s2, symm)
a new table which is the difference of two tables. With sets (where the values are all true) this is set difference and symmetric difference depending on the third parameter.

Parameters:

  • s1 table a map-like table or set
  • s2 table a map-like table or set
  • symm boolean symmetric difference (default false)

Returns:

    a map-like table or set
zip (...)
return a table where each element is a table of the ith values of an arbitrary number of tables. It is equivalent to a matrix transpose.

Parameters:

  • ... array arrays to be zipped

Usage:

    zip({10,20,30},{100,200,300}) is {{10,100},{20,200},{30,300}}
generated by LDoc 1.5.0