ActionScript - Array.sortOn() For Non-English Data? - arrays

this is an array of objects that i want to alphabetize:
var streets:Array = new Array();
streets.push({name:"Édouard-Montpetit"});
streets.push({name:"Alexandre de Sève"});
streets.push({name:"Van Horne"});
streets.push({name:"Atwater"});
now i'll sort my array:
streets.sortOn("name", Array.CASEINSENSITIVE);
//Sorted
Alexandre de Sève
Atwater
Van Horne
Édouard-Montpetit
the accent above the E in Édouard-Montpetit, and any other first letter with a non-english accent is sorted after Z.
any ideas how i can sort this correctly? i do not have access to the named data.

I know this is late, but for anyone going through this answer, you could pass a Collator
object to the Array.sort() method. A simple example from the documentation:
var words:Array = new Array("coté", "côte");
var sorter:Collator = new Collator("fr-FR", CollatorMode.SORTING);
words.sort(sorter.compare);
trace(words);// côte,coté
Hope this helps

I don't think you can do it with sortOn, as there's no way to tell flash to use a particular collaction for sorting text (at least, not that I'm aware of).
However, you could use sort and a custom sort function.
In this sort function, basically you want to strip all accents and do a case insensitive comparation. Replacing diacritics is easy and after that, you can safely use < and > for comparing the strings. A sort function is called by sort with two of the items to be sorted at a time. It should return a negative number if the first passed items sorts first , a possitive number if the second comes first and 0 if they sort equal.
function sortText(obj1:Object,obj2:Object):int {
var a:String = replaceDiacritics(obj1.name);
var b:String = replaceDiacritics(obj2.name);
if(a < b) {
return -1;
} else if(b < a) {
return 1;
} else {
return 0;
}
}
function replaceDiacritics(str:String):String {
str = str.toLowerCase();
str = str.replace(/á/g,"a");
str = str.replace(/é/g,"e");
str = str.replace(/í/g,"i");
str = str.replace(/ó/g,"o");
str = str.replace(/ú/g,"u");
str = str.replace(/à/g,"a");
str = str.replace(/è/g,"e");
str = str.replace(/ì/g,"i");
str = str.replace(/ò/g,"o");
str = str.replace(/ù/g,"u");
return str;
}
streets.sort(sortText);
A couple of notes about this. I know this method won't work for Spanish, as you have ñ, which is considered a letter on its own (not a regular n with a funny mark) and comes after n and before o. So, it's not possible to just replace accents and do a < / > compare. I think this is not a problem in French, but I could be wrong (not sure how Ç / ç is considered for sort purposes, for instance). Also, note that I haven't replaced all possible diacritics, so you'd want to add circumflexes (^) and umlauts (¨) to replaceDiacritics as necessary.
Edit
For a table based approach you could try something like the following. Each letter is assigned a number that reflects the sort order. As long as you can assume that any letter will have an absolute sort order (that is, context won't change how this works, which is not the case on some languages), it should give you good results.
Out of laziness, I built the table with a loop and just did what was neccesary to put "Ñ" between "n" and "o". I'm not considering any diacritics for sorting purposes, so they have the same value than their unaccented counterpart. But you could change this table as neccesary. Also, this table probably should be hardcoded for the required locale, but this code is just to give you an idea of how you could do this, not a full implementation (and it's probably not entirely correct from a purist perspective, but I think it could do the job). Also, in case we find a character that is not mapped, I'm falling back to its code point to determine how it sorts.
var sortTable:Object = buildSortTable();
function buildSortTable():Object {
var sortTable:Object = {};
var char:String;
var offset:int = 0;
for(var i:int = 1; i < 256; i++) {
char = String.fromCharCode(i);
if(char == "Ñ" || char == "ñ") {
offset--;
continue;
}
sortTable[char] = i + offset;
if(char == "N") {
sortTable["Ñ"] = sortTable["N"] + 1;
offset++;
}
if(char == "n") {
sortTable["ñ"] = sortTable["n"] + 1;
offset++;
}
}
sortTable["Á"] = sortTable["À"] = sortTable["Ä"] = sortTable["Â"] = sortTable["A"];
sortTable["É"] = sortTable["È"] = sortTable["Ë"] = sortTable["Ê"] = sortTable["E"];
sortTable["Í"] = sortTable["Ì"] = sortTable["Ï"] = sortTable["Î"] = sortTable["I"];
sortTable["Ó"] = sortTable["Ò"] = sortTable["Ö"] = sortTable["Ô"] = sortTable["O"];
sortTable["Ú"] = sortTable["Ì"] = sortTable["Ü"] = sortTable["Û"] = sortTable["U"];
sortTable["á"] = sortTable["à"] = sortTable["ä"] = sortTable["â"] = sortTable["a"];
sortTable["é"] = sortTable["è"] = sortTable["ë"] = sortTable["ê"] = sortTable["e"];
sortTable["í"] = sortTable["ì"] = sortTable["ï"] = sortTable["î"] = sortTable["i"];
sortTable["ó"] = sortTable["ò"] = sortTable["ö"] = sortTable["ô"] = sortTable["o"];
sortTable["ú"] = sortTable["ù"] = sortTable["ü"] = sortTable["û"] = sortTable["u"];
return sortTable;
}
function sortText(obj1:Object,obj2:Object):int {
var a:String = obj1.name.toLowerCase();
var b:String = obj2.name.toLowerCase();
var len_a:int = a.length;
var len_b:int = b.length;
var char_a:String;
var char_b:String;
var val_a:Number;
var val_b:Number;
for(var i = 0; i < len_a && i < len_b; i++) {
char_a = a.charAt(i);
char_b = b.charAt(i);
val_a = sortTable[char_a];
val_b = sortTable[char_b];
// this is just in case we have a letter that we haven't mapped...
// let's fall back to using its code point
if(isNaN(val_a)) {
val_a = char_a.charCodeAt(0);
}
if(isNaN(val_b)) {
val_b = char_b.charCodeAt(0);
}
if(val_a < val_b) {
return -1;
} else if(val_a > val_b) {
return 1;
}
}
// both strings are equal so far; so the sorter one (if any) must sort first
if(len_a < len_b) {
return -1;
} else if(len_a > len_b) {
return 1;
} else {
return 0;
}
}

Related

String Length changes if it is pulled from file

I'm coding in Game Maker Studio (Toby Fox used it for undertale) and I'm starting to try and work with files. Effectivly, I'm trying to build a custom "level" editor because the built in one for GMS lacks some features I want. So far I have set most of it up, but I still have to make it regurgitate the saved "levels" from the files. It's almost completly working, but there's one problem. Here's some info you should know ahead of time:
GMS uses really nonstandard syntax. I apologize ahead of time for that.
GMS is weird so arrays don't work with JSON formatting
as a result of No. 2, I have coded my own JSON formatting, which I reffer to as GSON (Gamemaker Studio Object Notation)
I am using GSON not only to store levels (and their respective components) but I also want to build a copy paste functionality, which I would use GSON for (copy the component and it just gives you a GSON string which is then interpreted back when you paste it, so that I can copy and paste not only the component type (i.e. a solid vs the player), but the variable values as well).
So, into the meat of it...
global.lineBreak = "\n"
global.sectionSign = "§"
{ // gsonStringifySelf
function gsonStringifySelf(varnamearray, exception = undefined){
/*
"exception" should be set to the variable you will use to store the returned string. if you decide to instead just upload it
directly to the file without a middleman variable, then leave the field empty. e.g.
gsonstring = gsonStringifySelf(varnamearray, "gsonstring")
varnamearray should be an array with the names of all the variables you want to save. I could hardwire it to use
variable_instance_get_names(self)
but this doesn't get instance_variables, so instead its best to have an input, especially so you can leave out
instance variables you don't care about. If you want to put in all the instance variables,you have to mannually add their names to the array, which could be done via the following:
var names = variable_instance_get_names(self) // (var means it is a strictly local variable, so it can't be accessed by any other objects)
var i = array_length(names) // array indexes start at 0 but array_length starts at 1, so I is refrencing the unset index that is closest to index 0
names[i++] = "id" // i++ returns i **then** increments it, causing it to refrence the correct array index, as mentioned above.
names[i++] = "visible"
names[i++] = "solid"
names[i++] = "persistent"
names[i++] = "depth"
names[i++] = "layer"
names[i++] = "alarm"
names[i++] = "toString"
names[i++] = "direction"
names[i++] = "friction"
names[i++] = "gravity"
names[i++] = "gravity_direction"
names[i++] = "hspeed"
names[i++] = "vspeed"
names[i++] = "speed"
names[i++] = "xstart"
names[i++] = "ystart"
names[i++] = "x"
names[i++] = "y"
names[i++] = "xprevious"
names[i++] = "yprevious"
names[i++] = "image_xscale"
names[i++] = "image_yscale"
strvars = gsonStringifySelf(names, strvars)
i++ returns the value of i, then increments it.
The gson formatting works as follows:
first it tells you what kind of object it is:
<objectname>{\n
(\n is the code for "new line")
then it adds the variables, which are formatted as follows:
<type>:<variablename>:<value>\n
then it puts "}\n" on the very end,
all together it looks like this:
<objectname>{\n
<type>:<variablename>:<value>\n
}\n
the types are as follows:
A = Array (list of entries)
B = Boolian (true/false)
R = Real (any and all numbers, yes this is unstandard)
S = String (Letters and Characters)
U = Undefined (Built in Variable "Undefined")
so these two are effectivly equivalent:
number = 20
"R:number:20"
Arrays are a special case:
A:<arrayname>:{,<type>:<value1>,<type>:<value2>...}\n
In this way multi dimentional arrays are instead shown as nested one dimensional arrays.
Multi dimensional arrays and nested arrays respectively look like this:
2 dimensional array that is 2x1 in size
A:<arrayname>:{A:{<type>:<value>},A:{<type>:<value>}}\n
2 length 1 dimensional array, with a 2 length array nested inside
A:<arrayname>:{A:{<type>:<value>,<type>:<value>}, <type>:<value>}\n
I mention this because in a case like this:
a = [
10,
5,
false
]
b = [
a,
"nope"
]
there will be no connection between b[0] and a. Instead, b[0] will be only store the values, and as such changing b[0][1] will not affect the value of a[1]
note that in scenarios like this:
array[0] = 1
array[2] = "hello"
array[1] will also be part of the string, so it will look like this:
A:array:{R:1;R:0;S:hello}\n
This is because of a quirk with GMS. If you set a location in an array "past" an undefined location, then the undefined location is set to 0 when it is accessed.
*/
static lineBreak = "\n"
static sectionSign = "§"
var str = sectionSign + object_get_name(object_index) + "{" + lineBreak // Make the first line = "§<objectname>{\n"
var str = global.sectionSign + object_get_name(object_index) + "{" + global.lineBreak // Make the first line = "§<objectname>{\n"
var names = varnamearray
for(var i = 0; i < array_length(names); i++){
if(names[i] == string(exception)){
continue
}
var tempstr = ""
var r = variable_instance_get(self, names[i])
if(is_array(r)){
tempstr = "A:" + names[i] + ":"
tempstr += string(gsonStringifyArray(r)) + global.lineBreak
} else {
tempstr = gsonValueTypeNotate(r, names[i]) + global.lineBreak
}
str += tempstr
}
str += "}" + global.lineBreak
return str
}
function gsonValueTypeNotate(val, name){
if(name == undefined){
var a = ""
name = ""
} else {
var a = ":"
}
if(is_bool(val)){
var str = "B:" + name + a + string(val)
} else if(is_numeric(val)){
var str = "R:" + name + a + string(val)
} else if(is_string(val)){
var str = "S:" + name + a + val
} else if(is_undefined(val)){
var str = "U:" + name + a + string(val)
}
return str
}
function gsonStringifyArray(array){
var str = "{" + global.sectionSign + ","
var len = array_length(array)
var len1 = len - 1
for(var i = 0; i < len1; i++){
if(is_array(array[i])){
str += "A:" + gsonStringifyArray(array[i]) + global.sectionSign + ","
continue
} else {
str += gsonValueTypeNotate(array[i], undefined) + global.sectionSign + ","
}
}
if(is_array(array[i])){
str += gsonStringifyArray(array[i])
} else {
str += gsonValueTypeNotate(array[i], undefined)
}
str += global.sectionSign + ",}"
return str
}
}
{ // gsonParse
function gsonParseObject(str){
var array
array = stringLineify(str)
var a1 = array[0]
var a2 = string_length(a1)
var a = string_copy(array[0], 2, string_length(array[0]) - 3)
var b = instance_create_depth(0, 0, 0, asset_get_index(a))
var c
for(var i = 1; i < array_length(array) - 1; i++){
if(string_copy(array[i], 1, 1) = "A"){
for(var j = 3; string_copy(array[i], j, 1) != ":"; j++){
}
c[0] = string_copy(array[i], 3, j - 3)
c[1] = parseArrayValue(array[i])
variable_instance_set(b, c[0], c[1])
} else {
c = getVarNameValue(array[i])
if(string_copy(array[i], 1, 1) = "B"){
variable_instance_set(b, c[0], bool(c[1]))
} else if(string_copy(array[i], 1, 1) = "R"){
variable_instance_set(b, c[0], real(c[1]))
} else if(string_copy(array[i], 1, 1) = "S"){
variable_instance_set(b, c[0], c[1])
} else if(string_copy(array[i], 1, 1) = "U"){
variable_instance_set(b, c[0], undefined)
}
}
}
}
function getVarNameValue(str){
for(var i = 3; string_copy(str, i, 1) != ":"; i++){
}
var a
a[0] = string_copy(str, 3, i - 3)
var ind = ++i
for(; string_copy(str, i, 1) != global.lineBreak; i++){
}
a[1] = string_copy(str, ind, i - ind)
return a
}
function parseArrayValue(str){
var array
array[0] = ""
var l
var b = 0
var c = 1
for(var i = 2; i <= string_length(str); i++){
var l = string_copy(str, i, 1)
if(l == global.sectionSign){
i += 2
var l = string_copy(str, i, 1)
if(l == "A"){
c = i + 2
for(var j = c; string_copy(str, j, 1) != "}"; ++j){
}
show_debug_message(string_copy(str, c, ++j))
array[b++] = parseArrayValue(string_copy(str, c, j - c))
i = j - 1
} else if(l == "}"){
break
} else {
for(var j = i; string_copy(str, j, 1) != global.sectionSign and j <= string_length(str); ++j){
var l = string_copy(str, j, 1)
}
array[b++] = getArrValue(string_copy(str, i, j - i))
i = j - 1
}
}
}
return array
}
function getArrValue(str){
var v = string_copy(str, 3, string_length(str))
if(string_copy(str, 1, 1) = "B"){
return bool(v)
} else if(string_copy(str, 1, 1) = "R"){
return real(v)
} else if(string_copy(str, 1, 1) = "S"){
return string(v)
} else if(string_copy(str, 1, 1) = "U"){
return undefined
}
}
}
{ // Save GSON to file
function gsonRoomUnload(file){
global.roomUnload = file
with(all){
event_user(0)
}
event_user(0)
}
function saveString(fname, str){
var b = file_text_open_append(fname)
file_text_write_string(b, str)
file_text_close(b)
}
}
{ // Load GSON from file
function gsonRoomLoad(fname){
if(file_exists(fname)){
var file = file_text_open_read(fname)
var a = ""
var b = ""
for(var i = 0; ; i++){
b = file_text_readln(file)
a += b
if(string_copy(b, 1, 1) == "}"){
//gsonParseObject(a)
a = ""
}
if(string_copy(b, 2, 1) == ""){
break
}
}
return true
} else {
return false
}
}
}
function stringLineify(str){
var strings
strings[0] = ""
var a
a[1] = 0
a[2] = ""
for(var i = 0; a[2] != false; i++){
a = stringLine(str, a[1])
strings[i] = a[0]
}
return strings
}
function stringLine(str, index){
var carriage = "\n"
for(var i = 1; ; i++){
var b = string_copy(str, index + i, 1)
if(b == carriage){
return [string_copy(str, index, i), index + i + 1, true]
} else if(index + i >= string_length(str)){
return [string_copy(str, index, i), index + i + 1, false]
}
}
}
There are a couple of functions you'll notice aren't defined. Those functions are built into GMS. I've linked their documentation at the bottom of the question.
Okay, so here's my problem:
If I run gsonStringifySelf() then take the string it returns and put it in gsonParseObject() then this line:
string_copy(array[0], 2, string_length(array[0]) - 3)
Is correct. Remember, what this does is take this example string:
"§<type>{\n"
and turn it into this:
"<type>"
string_copy(array[0], 2, string_length(array[0]) - 3)
copies from the second location in the string (I will be using "|" to inducate where the code is selecting in the string)
"§|<type>{\n"
and string_length returns the length of the string. so its trying to copy within the "|" in the string
"§|<type>{\n|"
But it is actually trying to copy 1 past the final location of the string, because we skipped the first location of the string but told it to copy the length of the string. Thus, when we tell it how many spaces to copy, we must subtract 3. We subtract 1 so it is only copying up to the final point in the string, and another 2 so that "{\n" isn't included in the final string. (\n is newline, and is considered one character even though it is represented with 2)
so base string:
"§<type>{\n"
copy from second point
"§|<type>{\n"
copy to the end of the string +1
"§|<type>{\n|"
subtract one from how many we are copying to shave the extra space off the end
"§|<type>{\n|"
and then subtract another 2 from how many we are copying to remove "{\n"
"§|<type>|{\n"
and return it:
"<type>"
Here's the thing. If I run gsonStringifySelf() and put it in a file then when I pull it down from the file and put that into gsonParseObject(), instead of this being correct:
string_copy(array[0], 2, string_length(array[0]) - 3)
this is correct:
string_copy(array[0], 2, string_length(array[0]) - 4)
So somehow, in all the file beeswax, an extra character is being added on, and I can't figure out where.
Thanks in advance for the help.
bool(n)
https://manual.yoyogames.com/GameMaker_Language/GML_Reference/Variable_Functions/bool.htm
real(n)
https://manual.yoyogames.com/GameMaker_Language/GML_Reference/Strings/real.htm
is_bool(n)
https://manual.yoyogames.com/GameMaker_Language/GML_Reference/Variable_Functions/is_bool.htm
is_numeric(n)
https://manual.yoyogames.com/GameMaker_Language/GML_Reference/Variable_Functions/is_numeric.htm
is_string(n)
https://manual.yoyogames.com/GameMaker_Language/GML_Reference/Variable_Functions/is_string.htm
is_undefined(n)
https://manual.yoyogames.com/GameMaker_Language/GML_Reference/Variable_Functions/is_string.htm
asset_get_index(str)
https://manual.yoyogames.com/GameMaker_Language/GML_Reference/Asset_Management/Assets_And_Tags/asset_get_index.htm
string_copy(str, index, count)
https://manual.yoyogames.com/GameMaker_Language/GML_Reference/Strings/string_copy.htm
string_length(str)
https://manual.yoyogames.com/GameMaker_Language/GML_Reference/Strings/string_length.htm
with(id)
https://manual.yoyogames.com/GameMaker_Language/GML_Overview/Language_Features/with.htm
event_user(int)
https://manual.yoyogames.com/GameMaker_Language/GML_Reference/Asset_Management/Objects/Object_Events/event_user.htm
instance_create_depth()
https://manual.yoyogames.com/GameMaker_Language/GML_Reference/Asset_Management/Instances/instance_create_depth.htm
variable_instance_get()
https://manual.yoyogames.com/GameMaker_Language/GML_Reference/Variable_Functions/variable_instance_get.htm
variable_instance_set()
https://manual.yoyogames.com/GameMaker_Language/GML_Reference/Variable_Functions/variable_instance_set.htm
file_exists(fname)
https://manual.yoyogames.com/GameMaker_Language/GML_Reference/File_Handling/File_System/file_exists.htm
file_text_open_append(fname)
https://manual.yoyogames.com/GameMaker_Language/GML_Reference/File_Handling/Text_Files/file_text_open_append.htm
file_text_write_string(file)
https://manual.yoyogames.com/GameMaker_Language/GML_Reference/File_Handling/Text_Files/file_text_write_string.htm
file_text_open_read(fname)
https://manual.yoyogames.com/GameMaker_Language/GML_Reference/File_Handling/Text_Files/file_text_open_read.htm
file_text_readln(file)
https://manual.yoyogames.com/GameMaker_Language/GML_Reference/File_Handling/Text_Files/file_text_readln.htm
file_text_close(file)
https://manual.yoyogames.com/GameMaker_Language/GML_Reference/File_Handling/Text_Files/file_text_close.htm
Ive tried most everything I can think of. The extra Character is an invisible character, and because of that, itis also invisible when I try to check the value with the debugger. I could bodge it, but if I can i'd rather avoid that, since it'll just cause me problems later anyway.
So after a bit of testing, I've realized that with just how long it takes to load a single instance, it's not worth using. Presumably I would use this for loading the areas during gameplay as well, but it takes well over 10 seconds to load a single instance, and a couple of minutes for more than 4 objects. I'll just have to avoid using arrays. Thanks! (bit annoying that I spent so much time on it for no reason tho...)
GMS is weird so arrays don't work with JSON formatting
The not-so-recently-added json_stringify does. The following
var thing = {
an_int: 1,
a_float: 1.5,
a_bool: true,
a_string: "hi",
an_array: [1, 2, "oh"],
a_struct: { x: 1, y: 2, name: "me", arr: [3, 4, undefined] },
};
show_debug_message(json_stringify(thing));
would output
{ "a_struct": { "x": 1.0, "y": 2.0, "name": "me", "arr": [ 3.0, 4.0, null ] }, "an_int": 1.0, "a_float": 1.5, "a_bool": true, "a_string": "hi", "an_array": [ 1.0, 2.0, "oh" ] }
With help of variable_instance_get_names/variable_struct_get_names, you could assemble a struct with all of the instance's variables (and desired built-in ones), encode that, and upon decoding write them back in.
As for your encoder-decoder, a few implementation caveats plague it:
GameMaker strings are immutable, meaning that adding two strings together generally allocates a new, third one. This makes your encoding slower than it could have been.
Writing to a buffer with kind=buffer_grow is a good way to limit memory re-allocations (as it will double the size whenever it runs out of space).
GameMaker strings are encoded as UTF-8, meaning that string_char_at isn't as simple as adding an index to the string pointer. This makes your decoding slower than it could have been.
Reading from a buffer is a common way around this, but take care - since you'd be storing a UTF-8 string in the buffer, a character will not necessarily be a single byte. Fortunately, you can usually completely ignore this detail so long as you read/write a string as a whole.
You are storing strings verbatim, which probably means that it's going to catch fire the moment one of the strings contains a } or other delimiter.
You are doing loops searching for characters often instead of calling string_pos[_ext] to let the engine do this [quicker].
Although technically recursive, its reliance on grabbing a substring with an encoded value instead of having a read position makes the process of reading nested values far messier than it should be (if I were to guess, it fails after misattributing one of the closing })
But also, you know, if it's your format, who said that it has to be a string? You can work with bytes instead, sparing yourself of a number of problems at once - no need for delimiters when you are writing/reading data in the same order.
A simple buffer-based encoder-decoder fits in just a little over 50 lines of code:
enum BinType { Undefined, Bool, Float, Int, String, Array, Struct };
function buffer_write_value(_buf, _val) {
if (is_real(_val)) {
buffer_write(_buf, buffer_u8, BinType.Float);
buffer_write(_buf, buffer_f64, _val);
} else if (is_bool(_val)) {
buffer_write(_buf, buffer_u8, BinType.Bool);
buffer_write(_buf, buffer_bool, _val);
} else if (is_numeric(_val)) {
buffer_write(_buf, buffer_u8, BinType.Int);
buffer_write(_buf, buffer_u64, _val); // u64 writes signed int64s fine
} else if (is_string(_val)) {
buffer_write(_buf, buffer_u8, BinType.String);
buffer_write(_buf, buffer_string, _val);
} else if (is_struct(_val)) {
buffer_write(_buf, buffer_u8, BinType.Struct);
var _names = variable_struct_get_names(_val);
var _count = array_length(_names);
buffer_write(_buf, buffer_u32, _count);
for (var i = 0; i < _count; i++) {
var _name = _names[i];
buffer_write(_buf, buffer_string, _name);
buffer_write_value(_buf, _val[$ _name]);
}
} else if (is_array(_val)) {
buffer_write(_buf, buffer_u8, BinType.Array);
var _count = array_length(_val);
buffer_write(_buf, buffer_u32, _count);
for (var i = 0; i < _count; i++) buffer_write_value(_buf, _val[i]);
} else buffer_write(_buf, buffer_u8, BinType.Undefined);
}
function buffer_read_value(_buf) {
switch (buffer_read(_buf, buffer_u8)) {
case BinType.Bool: return buffer_read(_buf, buffer_bool);
case BinType.Float: return buffer_read(_buf, buffer_f64);
case BinType.Int: return buffer_read(_buf, buffer_u64);
case BinType.String: return buffer_read(_buf, buffer_string);
case BinType.Array:
var _count = buffer_read(_buf, buffer_u32);
var _arr = array_create(_count);
for (var i = 0; i < _count; i++) _arr[i] = buffer_read_value(_buf);
return _arr;
case BinType.Struct:
var _struct = {};
repeat (buffer_read(_buf, buffer_u32)) {
var _name = buffer_read(_buf, buffer_string);
_struct[$ _name] = buffer_read_value(_buf);
}
return _struct;
default: return undefined;
}
}
and will successfully process the nested struct from the beginning of my answer if you write a value to a buffer, rewind it, and read it back. Add a little logic to detect instances or special cases, and you'll have a solution fit to your specific problems.
Further reading:
Haxe's JsonPrinter is a great example of a recursive JSON encoder.
SNAP is a GameMaker library with a number of encoders-decoders (following the aforementioned buffer principles, often non-recursive) that is both an example of good code and a solution to problems that you might be thinking of having.
Debugger (for stepping through and figuring out where you go wrong)
Profiler (for figuring out what's slow)
YYC (for performance)

Replace user in a protected range by script (Google Sheet)

here my specific case:
I have some range protected in google sheets
I need to replace some specific Editor if is editor of those range (var Editor2Replace and Editor2Add are emails)
Logically I tried to, for each sheet:
Cycle (FOR) of all the protected range (counter p)
For each protected range catch current editors and have it in array
Of the Editors read the email ==> this is what generate the mistake
Cycle (FOR) all the editors looking if someone of those is == Editor2Replace (that
is an email)
Here the code, but something is logically wrong, I doubt in what is an array and what not..
var Protections = sheet.getProtections(SpreadsheetApp.ProtectionType.RANGE);
for (var p = 0; p < Protections.length; p++) {
var Protection_Desc = Protections[p].getDescription();
var Protection_Editors = [];
var Protection_Editors = [Protections[p].getEditors()];
for (var r = 0; r < Protection_Editors.length; r++){
var Protection_Email[r] = [Protection_Editors[r].getEmail()];
if (Protection_Idontknow == Editor2Replace){
Protections[p].addEditor = Editor2Add;
Protections[p].removeEditor = Editor2Replace;
var Protection_Range = Protections[p].getRange();
var Protection_Row = Protection_Range.getRow();
var Owner1 = sheet.getRange(Protection_Row,5).getValue();
var Owner2 = sheet.getRange(Protection_Row,6).getValue();
if (Owner1 == Editor2Replace){
sheet.getRange(Protection_Row,5).setValue(Editor2Add);
}
if (Owner2 == Editor2Replace){
sheet.getRange(Protection_Row,6).setValue(Editor2Add);
}
}
}
Many thanks for hepling
There were a lot of issues in your script and I will enumerate them one by one. Also, I was able to replace a user in the protected sheet by modifying your script.
Issues:
Duplicate declaration
var Protection_Editors = [];
var Protection_Editors = [Protections[p].getEditors()];
Storing the returned value (array) in another array (which should not be done in your issue, it doesn't help you with anything)
var Protection_Editors = [Protections[p].getEditors()];
...
var Protection_Email[r] = [Protection_Editors[r].getEmail()];
Newly declared variable having an index (which I don't understand why)
var Protection_Email[r] = [Protection_Editors[r].getEmail()];
Variable not declared Protection_Idontknow
if (Protection_Idontknow == Editor2Replace){
Incorrect usage of methods addEditor and removeEditor
Protections[p].addEditor = Editor2Add;
Protections[p].removeEditor = Editor2Replace;
Below code should fix those issues (added some comments):
Code:
var Protections = sheet.getProtections(SpreadsheetApp.ProtectionType.RANGE);
for (var p = 0; p < Protections.length; p++) {
var Protection_Desc = Protections[p].getDescription();
// returned value of getEditors is already an array, return as is
var Protection_Editors = Protections[p].getEditors();
for (var r = 0; r < Protection_Editors.length; r++) {
var Protection_Email = Protection_Editors[r].getEmail();
// compare current email with the one you want to replace
if (Protection_Email == Editor2Replace) {
// add new and remove the one to replace
Protections[p].addEditor(Editor2Add);
Protections[p].removeEditor(Editor2Replace);
}
}
}
Note:
I have removed anything that were unrelated to the replacement of editors.
Reference:
Protection

can I check a regular expression on one array element?

I am trying to test if a specific element in an array is a range of characters, from a-z lowercase. what am I doing wrong? I am very new to coding (1 month in) and I am probably trying to do stuff thats too hard for me.
var array ["a","b","c"];
var myRegularExpression = /[a-z]/;
if (myRegularExpression.test(array[index])) {
//do stuff
}
To give you a working example as #Tushar mentioned:
var arr = ["a","b","c","123"];
var myRegularExpression = new RegExp("^[a-z]+$");
var matchCount = 0;
for (var index = 0; index < arr.length; index++) {
if (myRegularExpression.test(arr[index])) {
//do stuff
matchCount += 1;
}
}
document.getElementById("result").innerText = matchCount;
Number of elements matching the regex "^[a-z]+$":
<div id="result"></div>

AS3 - Fast search through array with strings

I need help. I am making something like dictionary (but you are filling it by yourself). I need fast search for words in it. I need to use Object or Array (because Dictionary is not supported by JSON. There is option to save file). I have this code, but I am afraid it is not so optimized to search fast when there will be a lot of words in array. Please help.
public function Search (string:String,section:String = Wordbook.NEWW):int
{
var str:String = string.toUpperCase();
for (i = 0; i < NewWords.length; i++)
{
if (NewWords[i].toUpperCase.indexOf(str) > -1)
{
return i;
}
}
return -1;//If not found
}
And example how it should work: (SearchTxt - text field, user should type here word he need to find; WB - Wordbook class; WB.NewWords & WB.NewWordsT - Arrays in this class)
var index:int = WB.Search(SearchTxt.text,Wordbook.NEWW);
if(index>-1){
WordTxt.text = WB.NewWords[index];
TranslationTxt.text = WB.NewWordsT[index];
} else {
dispatchEvent(new EventWithMessage(EventWithMessage.ERROR,{error:"No match!"}));
}
There are two solutions here, depending on your use-cases.
Solution 1: Leave it alone!
If the biggest dictionary you're going to build is only a hundred words or so, performance will be fine with a linear search.
Solution 2: Implement a binary search algorithm
If you do need to deal with large dictionaries of thousands of words, a binary search algorithm will offer improved performance. For this, you'll need to guarantee that your search array is sorted.
Something like this should do the trick:
public function Search (string:String,section:Array = NewWords, offset:int = 0):int
{
if (section.length == 0) {
return -1;
}
var str:String = string.toUpperCase();
var firstCharCode:int = str.charCodeAt(0);
var middleWord:String = section[section.length / 2];
var middleWordCharCode:int = middleWord.charCodeAt(0);
if (middleWord.substr(0,str.length) == str) {
return (section.length / 2) + offset;
} else {
var comparison:int = str.localeCompare(middleWord);
if (comparison < 0) {
return Search(string, section.splice(0, section.length / 2), offset);
} else { // then comparison > 0
var newOffset:int = offset + section.length / 2;
return Search(string, section.splice(section.length / 2, section.length-1), newOffset);
}
}
}

AS3: How to check if a value already exists in the Array before adding with FOR loop?

I believe is something simple but obviously not simple enough :). Any ideas how to check if a value already exists in the Array before adding the value using FOR loop?
I have this so far and it doesn't work as I want to because the Array can contain duplicate values!
var n:int = 5;
var cnt:int;
for (var i = 0; i < n; i++)
{
cnt = randomThief();
for (var a = 0; a < loto5.length; a++)
{
if (loto5[i] == cnt)
{
loto5[i] = cnt;
}
}
}
You can use the indexOf() method of the Array class to check if the value exists like this :
var index:int = loto5.indexOf(cnt);
indexOf() returns a -1, if the value doesn't exist. Here is an example of how to do a check :
if (loto5.indexOf(cnt) >= 0)
{
// do something
}
for (var a = 0; a < loto5.length; a++)
{
cnt = randomThief();
if (loto5.indexOf(cnt) == -1) //if cnt isn't in array do ...
{
trace (cnt+" is not in Array");
loto5[a] = cnt;
}
}
Works, simple and beauty :)

Resources