I have the following hash in Ruby :
{
0 => {
:method=> "POST",
:path=> "/api/customer/191023",
:host=> "host.8",
:duration=> "1221"
},
1 => {
:method=> "GET",
:path=> "/api/customer/191023",
:host=> "host.8",
:duration=> "99"
},
2 => {
:method=> "POST",
:path=> "/api/customer/191023",
:host=> "host.10",
:duration=> "142"
},
3 => {
:method=> "POST",
:path=> "/api/customer/191023",
:host=> "host.8",
:duration=> "243"
}
4 => {
:method=> "POST",
:path=> "/api/customer/191023",
:host=> "host.10",
:duration=> "132"
}
}
I would like to do a simple search within these hashes to find the host with the highest frequency. For example, in the previous example, I should get host.8.
Thank you for your help,
M.
To find host value with highest frequency do:
hs = hash.values.group_by { |h| h[:host] =~ /host\.(\d+)/ && $1.to_i || 0 }.to_a
hs.reduce([-1,0]) { |sum,v| v[1].size > sum[1] && [ v[0], v[1].size ] || sum }.first
Description: [-1,0] is the default value for set for #reduce method, where -1 is a number (like in host.number), and 0 is a count of the number. So, when reduce encounters the number with size more than of passed sum, it replaces with the new value on next iteration.
Here's one way to do that.
Code
def max_host(hash)
hash.each_with_object(Hash.new(0)) { |(_,v),h| h[v[:host]] += 1 }
.max_by { |_,v| v }
.first
end
Example
Let's take the simplified example below. Note that I've changed, for example, :host = \"host.10\" to :host = "host.10", as the former is not a correct syntax. You could write the string as '\"host.10\" (=> "\\\"host.10\\\""), but I assume you simply want "host.10". The code is the same for both.
hash = {
0 => {
:method=>"POST",
:host =>"host.8"
},
1 => {
:method=>"GET",
:host =>"host.10"
},
2 => {
:method=>"POST",
:host =>"host.10"
}
}
max_host(hash)
#=> "host.10"
Explanation
For the example hash above,
enum = hash.each_with_object(Hash.new(0))
#=> #<Enumerator: {
# 0=>{:method=>"POST", :host=>"host.8"},
# 1=>{:method=>"GET", :host=>"host.10"},
# 2=>{:method=>"POST", :host=>"host.10"}}:each_with_object({})>
The enumerator will invoke the method Hash#each to pass each element of the enumerator into the block. We can see what those elements are by converting the enumerator to an array:
enum.to_a
#=> [[[0, {:method=>"POST", :host=>"host.8"}], {}],
# [[1, {:method=>"GET", :host=>"host.10"}], {}],
# [[2, {:method=>"POST", :host=>"host.10"}], {}]]
The empty hash shown in the first element is the initial value of the hash created by
Hash.new(0)
This creates a hash h with a default value of zero. By doing it this way, if h does not have a key k, h[k] will return the default value (0), but (important!) this does not change the hash.
The first value passed into the block is
[[0, {:method=>"POST", :host=>"host.8"}], {}]
This is then decomposed (or "disambiguated") into individual objects that are assigned to three block variables:
k => 0
v => {:method=>"POST", :host=>"host.8"}
h => Hash.new(0)
We then execute:
h[v[:host]] += 1
which is
h["host.8"] += 1
which is shorthand for
h["host.8"] = h["host.8"] + 1
[Aside: you may have noticed that in the code I show the block variables as |(_,v),h|, whereas above I refer to them above as |(k,v),h|. I could have used the latter, but since k is not reference in the block, I've chosen to replace it with a "placeholder" _. This ensures k won't be referenced and also tells any readers that I'm not using what would be the first block variable.]
As h does not have a key "host.8", h["host.8"] to the right of = returns the default value:
h["host.8"] = 0 + 1
#=> 1
so now
h #=> {"host.8"=>1}
The second element passed into the block is
[[1, {:method=>"GET", :host=>"host.10"}], {"host.8"=>1}]
so the block variables become:
v => {:method=>"GET", :host=>"host.10"}
h => {"host.8"=>1}
Notice that the hash h has been updated. We execute
h[v[:host]] += 1
#=> h["host.10"] += 1
#=> h["host.10"] = h["host.10"] + 1
#=> h["host.10"] = 0 + 1
#=> 1
so now
h #=> {"host.8"=>1, "host.10"=>1}
Lastly, the block variables are assigned the values
v = {:method=>"POST", :host=>"host.10"}
h => {"host.8"=>1, "host.10"=>1}
so
h[v[:host]] += 1
#=> h["host.10"] += 1
#=> h["host.10"] = h["host.10"] + 1
#=> h["host.10"] = 1 + 1
#=> 2
h #=> {"host.8"=>1, "host.10"=>2}
and the value of h is returned by the method.
I'm having trouble displaying the contents of a table which contains nested tables (n-deep). I'd like to just dump it to std out or the console via a print statement or something quick and dirty but I can't figure out how. I'm looking for the rough equivalent that I'd get when printing an NSDictionary using gdb.
If the requirement is "quick and dirty"
I've found this one useful. Because of the recursion it can print nested tables too. It doesn't give the prettiest formatting in the output but for such a simple function it's hard to beat for debugging.
function dump(o)
if type(o) == 'table' then
local s = '{ '
for k,v in pairs(o) do
if type(k) ~= 'number' then k = '"'..k..'"' end
s = s .. '['..k..'] = ' .. dump(v) .. ','
end
return s .. '} '
else
return tostring(o)
end
end
e.g.
local people = {
{
name = "Fred",
address = "16 Long Street",
phone = "123456"
},
{
name = "Wilma",
address = "16 Long Street",
phone = "123456"
},
{
name = "Barney",
address = "17 Long Street",
phone = "123457"
}
}
print("People:", dump(people))
Produces the following output:
People: { [1] = { ["address"] = 16 Long Street,["phone"] =
123456,["name"] = Fred,} ,[2] = { ["address"] = 16 Long
Street,["phone"] = 123456,["name"] = Wilma,} ,[3] = { ["address"] = 17
Long Street,["phone"] = 123457,["name"] = Barney,} ,}
I know this question has already been marked as answered, but let me plug my own library here. It's called inspect.lua, and you can find it here:
https://github.com/kikito/inspect.lua
It's just a single file that you can require from any other file. It returns a function that transforms any Lua value into a human-readable string:
local inspect = require('inspect')
print(inspect({1,2,3})) -- {1, 2, 3}
print(inspect({a=1,b=2})
-- {
-- a = 1
-- b = 2
-- }
It indents subtables properly, and handles "recursive tables" (tables that contain references to themselves) correctly, so it doesn't get into infinite loops. It sorts values in a sensible way. It also prints metatable information.
Regards!
Feel free to browse the Lua Wiki on table serialization. It lists several ways on how to dump a table to the console.
You just have to choose which one suits you best. There are many ways to do it, but I usually end up using the one from Penlight:
> t = { a = { b = { c = "Hello world!", 1 }, 2, d = { 3 } } }
> require 'pl.pretty'.dump(t)
{
a = {
d = {
3
},
b = {
c = "Hello world!",
1
},
2
}
}
found this:
-- Print contents of `tbl`, with indentation.
-- `indent` sets the initial level of indentation.
function tprint (tbl, indent)
if not indent then indent = 0 end
for k, v in pairs(tbl) do
formatting = string.rep(" ", indent) .. k .. ": "
if type(v) == "table" then
print(formatting)
tprint(v, indent+1)
elseif type(v) == 'boolean' then
print(formatting .. tostring(v))
else
print(formatting .. v)
end
end
end
from here
https://gist.github.com/ripter/4270799
works pretty good for me...
Most pure lua print table functions I've seen have a problem with deep recursion
and tend to cause a stack overflow when going too deep. This print
table function that I've written does not have this problem. It should also be capable of handling really large tables due to the way it handles concatenation. In my personal usage of this function, it outputted 63k lines to file in about a second.
The output also keeps lua syntax and the script can easily be modified
for simple persistent storage by writing the output to file if modified to allow
only number, boolean, string and table data types to be formatted.
function print_table(node)
local cache, stack, output = {},{},{}
local depth = 1
local output_str = "{\n"
while true do
local size = 0
for k,v in pairs(node) do
size = size + 1
end
local cur_index = 1
for k,v in pairs(node) do
if (cache[node] == nil) or (cur_index >= cache[node]) then
if (string.find(output_str,"}",output_str:len())) then
output_str = output_str .. ",\n"
elseif not (string.find(output_str,"\n",output_str:len())) then
output_str = output_str .. "\n"
end
-- This is necessary for working with HUGE tables otherwise we run out of memory using concat on huge strings
table.insert(output,output_str)
output_str = ""
local key
if (type(k) == "number" or type(k) == "boolean") then
key = "["..tostring(k).."]"
else
key = "['"..tostring(k).."']"
end
if (type(v) == "number" or type(v) == "boolean") then
output_str = output_str .. string.rep('\t',depth) .. key .. " = "..tostring(v)
elseif (type(v) == "table") then
output_str = output_str .. string.rep('\t',depth) .. key .. " = {\n"
table.insert(stack,node)
table.insert(stack,v)
cache[node] = cur_index+1
break
else
output_str = output_str .. string.rep('\t',depth) .. key .. " = '"..tostring(v).."'"
end
if (cur_index == size) then
output_str = output_str .. "\n" .. string.rep('\t',depth-1) .. "}"
else
output_str = output_str .. ","
end
else
-- close the table
if (cur_index == size) then
output_str = output_str .. "\n" .. string.rep('\t',depth-1) .. "}"
end
end
cur_index = cur_index + 1
end
if (size == 0) then
output_str = output_str .. "\n" .. string.rep('\t',depth-1) .. "}"
end
if (#stack > 0) then
node = stack[#stack]
stack[#stack] = nil
depth = cache[node] == nil and depth + 1 or depth - 1
else
break
end
end
-- This is necessary for working with HUGE tables otherwise we run out of memory using concat on huge strings
table.insert(output,output_str)
output_str = table.concat(output)
print(output_str)
end
Here is an example:
local t = {
["abe"] = {1,2,3,4,5},
"string1",
50,
["depth1"] = { ["depth2"] = { ["depth3"] = { ["depth4"] = { ["depth5"] = { ["depth6"] = { ["depth7"]= { ["depth8"] = { ["depth9"] = { ["depth10"] = {1000}, 900}, 800},700},600},500}, 400 }, 300}, 200}, 100},
["ted"] = {true,false,"some text"},
"string2",
[function() return end] = function() return end,
75
}
print_table(t)
Output:
{
[1] = 'string1',
[2] = 50,
[3] = 'string2',
[4] = 75,
['abe'] = {
[1] = 1,
[2] = 2,
[3] = 3,
[4] = 4,
[5] = 5
},
['function: 06472B70'] = 'function: 06472A98',
['depth1'] = {
[1] = 100,
['depth2'] = {
[1] = 200,
['depth3'] = {
[1] = 300,
['depth4'] = {
[1] = 400,
['depth5'] = {
[1] = 500,
['depth6'] = {
[1] = 600,
['depth7'] = {
[1] = 700,
['depth8'] = {
[1] = 800,
['depth9'] = {
[1] = 900,
['depth10'] = {
[1] = 1000
}
}
}
}
}
}
}
}
}
},
['ted'] = {
[1] = true,
[2] = false,
[3] = 'some text'
}
}
As previously mentioned, you have to write it.
Here is my humble version: (super basic one)
function tprint (t, s)
for k, v in pairs(t) do
local kfmt = '["' .. tostring(k) ..'"]'
if type(k) ~= 'string' then
kfmt = '[' .. k .. ']'
end
local vfmt = '"'.. tostring(v) ..'"'
if type(v) == 'table' then
tprint(v, (s or '')..kfmt)
else
if type(v) ~= 'string' then
vfmt = tostring(v)
end
print(type(t)..(s or '')..kfmt..' = '..vfmt)
end
end
end
example:
local mytbl = { ['1']="a", 2, 3, b="c", t={d=1} }
tprint(mytbl)
output (Lua 5.0):
table[1] = 2
table[2] = 3
table["1"] = "a"
table["t"]["d"] = 1
table["b"] = "c"
I use my own function to print the contents of a table but not sure how well it translates to your environment:
---A helper function to print a table's contents.
---#param tbl table #The table to print.
---#param depth number #The depth of sub-tables to traverse through and print.
---#param n number #Do NOT manually set this. This controls formatting through recursion.
function PrintTable(tbl, depth, n)
n = n or 0;
depth = depth or 5;
if (depth == 0) then
print(string.rep(' ', n).."...");
return;
end
if (n == 0) then
print(" ");
end
for key, value in pairs(tbl) do
if (key and type(key) == "number" or type(key) == "string") then
key = string.format("[\"%s\"]", key);
if (type(value) == "table") then
if (next(value)) then
print(string.rep(' ', n)..key.." = {");
PrintTable(value, depth - 1, n + 4);
print(string.rep(' ', n).."},");
else
print(string.rep(' ', n)..key.." = {},");
end
else
if (type(value) == "string") then
value = string.format("\"%s\"", value);
else
value = tostring(value);
end
print(string.rep(' ', n)..key.." = "..value..",");
end
end
end
if (n == 0) then
print(" ");
end
end
The simplest way, with circular reference handling and all:
function dump(t, indent, done)
done = done or {}
indent = indent or 0
done[t] = true
for key, value in pairs(t) do
print(string.rep("\t", indent))
if type(value) == "table" and not done[value] then
done[value] = true
print(key, ":\n")
dump(value, indent + 2, done)
done[value] = nil
else
print(key, "\t=\t", value, "\n")
end
end
end
There are 2 solutions that I want to mention: a quick&dirty one, and another which properly escapes all keys and values but is bigger
Simple & fast solution (use only on "safe" inputs):
local function format_any_value(obj, buffer)
local _type = type(obj)
if _type == "table" then
buffer[#buffer + 1] = '{"'
for key, value in next, obj, nil do
buffer[#buffer + 1] = tostring(key) .. '":'
format_any_value(value, buffer)
buffer[#buffer + 1] = ',"'
end
buffer[#buffer] = '}' -- note the overwrite
elseif _type == "string" then
buffer[#buffer + 1] = '"' .. obj .. '"'
elseif _type == "boolean" or _type == "number" then
buffer[#buffer + 1] = tostring(obj)
else
buffer[#buffer + 1] = '"???' .. _type .. '???"'
end
end
Usage:
local function format_as_json(obj)
if obj == nil then return "null" else
local buffer = {}
format_any_value(obj, buffer)
return table.concat(buffer)
end
end
local function print_as_json(obj)
print(_format_as_json(obj))
end
print_as_json {1, 2, 3}
print_as_json(nil)
print_as_json("string")
print_as_json {[1] = 1, [2] = 2, three = { { true } }, four = "four"}
Correct solution with key/value escaping
Small library that I wrote in pure Lua for this specific use-case: https://github.com/vn971/fast_json_encode
Or specifically this 1 file that includes both a formatter and a printer: https://github.com/vn971/fast_json_encode/blob/master/json_format.lua
You have to code it yourself I'm afraid. I wrote this, and it may be of some use to you
function printtable(table, indent)
indent = indent or 0;
local keys = {};
for k in pairs(table) do
keys[#keys+1] = k;
table.sort(keys, function(a, b)
local ta, tb = type(a), type(b);
if (ta ~= tb) then
return ta < tb;
else
return a < b;
end
end);
end
print(string.rep(' ', indent)..'{');
indent = indent + 1;
for k, v in pairs(table) do
local key = k;
if (type(key) == 'string') then
if not (string.match(key, '^[A-Za-z_][0-9A-Za-z_]*$')) then
key = "['"..key.."']";
end
elseif (type(key) == 'number') then
key = "["..key.."]";
end
if (type(v) == 'table') then
if (next(v)) then
printf("%s%s =", string.rep(' ', indent), tostring(key));
printtable(v, indent);
else
printf("%s%s = {},", string.rep(' ', indent), tostring(key));
end
elseif (type(v) == 'string') then
printf("%s%s = %s,", string.rep(' ', indent), tostring(key), "'"..v.."'");
else
printf("%s%s = %s,", string.rep(' ', indent), tostring(key), tostring(v));
end
end
indent = indent - 1;
print(string.rep(' ', indent)..'}');
end
The table.tostring metehod of metalua is actually very complete. It deals with nested tables, the indentation level is changeable, ...
See https://github.com/fab13n/metalua/blob/master/src/lib/metalua/table2.lua
This is my version that supports excluding tables and userdata
-- Lua Table View by Elertan
table.print = function(t, exclusions)
local nests = 0
if not exclusions then exclusions = {} end
local recurse = function(t, recurse, exclusions)
indent = function()
for i = 1, nests do
io.write(" ")
end
end
local excluded = function(key)
for k,v in pairs(exclusions) do
if v == key then
return true
end
end
return false
end
local isFirst = true
for k,v in pairs(t) do
if isFirst then
indent()
print("|")
isFirst = false
end
if type(v) == "table" and not excluded(k) then
indent()
print("|-> "..k..": "..type(v))
nests = nests + 1
recurse(v, recurse, exclusions)
elseif excluded(k) then
indent()
print("|-> "..k..": "..type(v))
elseif type(v) == "userdata" or type(v) == "function" then
indent()
print("|-> "..k..": "..type(v))
elseif type(v) == "string" then
indent()
print("|-> "..k..": ".."\""..v.."\"")
else
indent()
print("|-> "..k..": "..v)
end
end
nests = nests - 1
end
nests = 0
print("### START TABLE ###")
for k,v in pairs(t) do
print("root")
if type(v) == "table" then
print("|-> "..k..": "..type(v))
nests = nests + 1
recurse(v, recurse, exclusions)
elseif type(v) == "userdata" or type(v) == "function" then
print("|-> "..k..": "..type(v))
elseif type(v) == "string" then
print("|-> "..k..": ".."\""..v.."\"")
else
print("|-> "..k..": "..v)
end
end
print("### END TABLE ###")
end
This is an example
t = {
location = {
x = 10,
y = 20
},
size = {
width = 100000000,
height = 1000,
},
name = "Sidney",
test = {
hi = "lol",
},
anotherone = {
1,
2,
3
}
}
table.print(t, { "test" })
Prints:
### START TABLE ###
root
|-> size: table
|
|-> height: 1000
|-> width: 100000000
root
|-> location: table
|
|-> y: 20
|-> x: 10
root
|-> anotherone: table
|
|-> 1: 1
|-> 2: 2
|-> 3: 3
root
|-> test: table
|
|-> hi: "lol"
root
|-> name: "Sidney"
### END TABLE ###
Notice that the root doesn't remove exclusions
Made this version to print tables with identation. Can probably be extended to work recursively.
function printtable(table, indent)
print(tostring(table) .. '\n')
for index, value in pairs(table) do
print(' ' .. tostring(index) .. ' : ' .. tostring(value) .. '\n')
end
end
--~ print a table
function printTable(list, i)
local listString = ''
--~ begin of the list so write the {
if not i then
listString = listString .. '{'
end
i = i or 1
local element = list[i]
--~ it may be the end of the list
if not element then
return listString .. '}'
end
--~ if the element is a list too call it recursively
if(type(element) == 'table') then
listString = listString .. printTable(element)
else
listString = listString .. element
end
return listString .. ', ' .. printTable(list, i + 1)
end
local table = {1, 2, 3, 4, 5, {'a', 'b'}, {'G', 'F'}}
print(printTable(table))
Hi man, I wrote a siple code that do this in pure Lua, it has a bug (write a coma after the last element of the list) but how i wrote it quickly as a prototype I will let it to you adapt it to your needs.
Adding another version. This one tries to iterate over userdata as well.
function inspect(o,indent)
if indent == nil then indent = 0 end
local indent_str = string.rep(" ", indent)
local output_it = function(str)
print(indent_str..str)
end
local length = 0
local fu = function(k, v)
length = length + 1
if type(v) == "userdata" or type(v) == 'table' then
output_it(indent_str.."["..k.."]")
inspect(v, indent+1)
else
output_it(indent_str.."["..k.."] "..tostring(v))
end
end
local loop_pairs = function()
for k,v in pairs(o) do fu(k,v) end
end
local loop_metatable_pairs = function()
for k,v in pairs(getmetatable(o)) do fu(k,v) end
end
if not pcall(loop_pairs) and not pcall(loop_metatable_pairs) then
output_it(indent_str.."[[??]]")
else
if length == 0 then
output_it(indent_str.."{}")
end
end
end
Convert to json and then print.
local json = require('cjson')
json_string = json.encode(this_table)
print (json_string)
simple example of dump a table in lua
i suggest using serpent.lua
local function parser(value, indent, subcategory)
local indent = indent or 2
local response = '(\n'
local subcategory = type(subcategory) == 'number' and subcategory or indent
for key, value in pairs(value) do
if type(value) == 'table' then
value = parser(value, indent, subcategory + indent)
elseif type(value) == 'string' then
value = '\''.. value .. '\''
elseif type(value) ~= 'number' then
value = tostring(value)
end
if type(tonumber(key)) == 'number' then
key = '[' .. key .. ']'
elseif not key:match('^([A-Za-z_][A-Za-z0-9_]*)$') then
key = '[\'' .. key .. '\']'
end
response = response .. string.rep(' ', subcategory) .. key .. ' = ' .. value .. ',\n'
end
return response .. string.rep(' ', subcategory - indent) .. ')'
end
example
response = parser{1,2,3, {ok = 10, {}}}
print(response)
result
(
[1] = 1,
[2] = 2,
[3] = 3,
[4] = (
[1] = (),
ok = 10
)
)
here's my little snippet for that:
--- Dump value of a variable in a formatted string
--
--- #param o table Dumpable object
--- #param tbs string|nil Tabulation string, ' ' by default
--- #param tb number|nil Initial tabulation level, 0 by default
--- #return string
local function dump(o, tbs, tb)
tb = tb or 0
tbs = tbs or ' '
if type(o) == 'table' then
local s = '{'
if (next(o)) then s = s .. '\n' else return s .. '}' end
tb = tb + 1
for k,v in pairs(o) do
if type(k) ~= 'number' then k = '"' .. k .. '"' end
s = s .. tbs:rep(tb) .. '[' .. k .. '] = ' .. dump(v, tbs, tb)
s = s .. ',\n'
end
tb = tb - 1
return s .. tbs:rep(tb) .. '}'
else
return tostring(o)
end
end
I have humbly modified a bit Alundaio code:
-- by Alundaio
-- KK modified 11/28/2019
function dump_table_to_string(node, tree, indentation)
local cache, stack, output = {},{},{}
local depth = 1
if type(node) ~= "table" then
return "only table type is supported, got " .. type(node)
end
if nil == indentation then indentation = 1 end
local NEW_LINE = "\n"
local TAB_CHAR = " "
if nil == tree then
NEW_LINE = "\n"
elseif not tree then
NEW_LINE = ""
TAB_CHAR = ""
end
local output_str = "{" .. NEW_LINE
while true do
local size = 0
for k,v in pairs(node) do
size = size + 1
end
local cur_index = 1
for k,v in pairs(node) do
if (cache[node] == nil) or (cur_index >= cache[node]) then
if (string.find(output_str,"}",output_str:len())) then
output_str = output_str .. "," .. NEW_LINE
elseif not (string.find(output_str,NEW_LINE,output_str:len())) then
output_str = output_str .. NEW_LINE
end
-- This is necessary for working with HUGE tables otherwise we run out of memory using concat on huge strings
table.insert(output,output_str)
output_str = ""
local key
if (type(k) == "number" or type(k) == "boolean") then
key = "["..tostring(k).."]"
else
key = "['"..tostring(k).."']"
end
if (type(v) == "number" or type(v) == "boolean") then
output_str = output_str .. string.rep(TAB_CHAR,depth*indentation) .. key .. " = "..tostring(v)
elseif (type(v) == "table") then
output_str = output_str .. string.rep(TAB_CHAR,depth*indentation) .. key .. " = {" .. NEW_LINE
table.insert(stack,node)
table.insert(stack,v)
cache[node] = cur_index+1
break
else
output_str = output_str .. string.rep(TAB_CHAR,depth*indentation) .. key .. " = '"..tostring(v).."'"
end
if (cur_index == size) then
output_str = output_str .. NEW_LINE .. string.rep(TAB_CHAR,(depth-1)*indentation) .. "}"
else
output_str = output_str .. ","
end
else
-- close the table
if (cur_index == size) then
output_str = output_str .. NEW_LINE .. string.rep(TAB_CHAR,(depth-1)*indentation) .. "}"
end
end
cur_index = cur_index + 1
end
if (size == 0) then
output_str = output_str .. NEW_LINE .. string.rep(TAB_CHAR,(depth-1)*indentation) .. "}"
end
if (#stack > 0) then
node = stack[#stack]
stack[#stack] = nil
depth = cache[node] == nil and depth + 1 or depth - 1
else
break
end
end
-- This is necessary for working with HUGE tables otherwise we run out of memory using concat on huge strings
table.insert(output,output_str)
output_str = table.concat(output)
return output_str
end
then:
print(dump_table_to_string("AA", true,3))
print(dump_table_to_string({"AA","BB"}, true,3))
print(dump_table_to_string({"AA","BB"}))
print(dump_table_to_string({"AA","BB"},false))
print(dump_table_to_string({"AA","BB",{22,33}},true,2))
gives:
only table type is supported, got string
{
[1] = 'AA',
[2] = 'BB'
}
{
[1] = 'AA',
[2] = 'BB'
}
{[1] = 'AA',[2] = 'BB'}
{
[1] = 'AA',
[2] = 'BB',
[3] = {
[1] = 22,
[2] = 33
}
}
Now the function print can print the (flat) tables!
oprint = print -- origin print
print = function (...)
if type(...) == "table" then
local str = ''
local amount = 0
for i,v in pairs(...) do
amount=amount+1
local pre = type(i) == "string" and i.."=" or ""
str = str .. pre..tostring(v) .. "\t"
end
oprint('#'..amount..':', str)
else
oprint(...)
end
end
For example:
print ({x=7, y=9, w=11, h="height", 7, 8, 9})
prints:
#7: 7 8 9 y=9 x=7 h=height w=11
The same way it can be just new function tostring:
otostring = tostring -- origin tostring
tostring = function (...)
if type(...) == "table" then
local str = '{'
for i,v in pairs(...) do
local pre = type(i) == "string" and i.."=" or ""
str = str .. pre..tostring(v) .. ", "
end
str = str:sub(1, -3)
return str..'}'
else
return otostring(...)
end
end