.proto file structure
message repetedMSG
{
required string data = 1;
}
message mainMSG
{
required repetedMSG_id = 1;
repeated repetedMSG rptMSG = 2;
}
I have one mainMSG and in it too many (suppose 10) repetedMSG are present.
Now i want to delete any particular repetedMSG (suppose 5th repetedMSG )from mainMSG. For this i tried 3 ways but none of them worked.
for (int j = 0; j<mainMSG->repetedMSG_size(); j++){
repetedMSG reptMsg = mainMsg->mutable_repetedMSG(j);
if (QString::fromStdString(reptMsg->data).compare("deleteMe") == 0){
*First tried way:-* reptMsg->Clear();
*Second tried Way:-* delete reptMsg;
*Third tried way:-* reptMsg->clear_formula_name();
break;
}
}
I get run-time error when i serialize the mainMSG for writing to a file i.e. when execute this line
mainMSG.SerializeToOstream (std::fstream output("C:/A/test1", std::ios::out | std::ios::trunc | std::ios::binary)) here i get run-time error
You can use RepeatedPtrField::DeleteSubrange() for this. However, be careful about using this in a loop -- people commonly write code like this which is O(n^2):
// BAD CODE! O(n^2)!
for (int i = 0; i < message.foo_size(); i++) {
if (should_filter(message.foo(i))) {
message.mutable_foo()->DeleteSubrange(i, 1);
--i;
}
}
Instead, if you plan to remove multiple elements, do something like this:
// Move all filtered elements to the end of the list.
int keep = 0; // number to keep
for (int i = 0; i < message.foo_size(); i++) {
if (should_filter(message.foo(i))) {
// Skip.
} else {
if (keep < i) {
message.mutable_foo()->SwapElements(i, keep)
}
++keep;
}
}
// Remove the filtered elements.
message.mutable_foo()->DeleteSubrange(keep, message.foo_size() - keep);
Related
I am working on a compiler and one aspect currently is how to wait for interpolated variable names to be resolved. So I am wondering how to take a nested interpolated variable string and build some sort of simple data model/schema for unwrapping the evaluated string so to speak. Let me demonstrate.
Say we have a string like this:
foo{a{x}-{y}}-{baz{one}-{two}}-foo{c}
That has 1, 2, and 3 levels of nested interpolations in it. So essentially it should resolve something like this:
wait for x, y, one, two, and c to resolve.
when both x and y resolve, then resolve a{x}-{y} immediately.
when both one and two resolve, resolve baz{one}-{two}.
when a{x}-{y}, baz{one}-{two}, and c all resolve, then finally resolve the whole expression.
I am shaky on my understanding of the logic flow for handling something like this, wondering if you could help solidify/clarify the general algorithm (high level pseudocode or something like that). Mainly just looking for how I would structure the data model and algorithm so as to progressively evaluate when the pieces are ready.
I'm starting out trying and it's not clear what to do next:
{
dependencies: [
{
path: [x]
},
{
path: [y]
}
],
parent: {
dependency: a{x}-{y} // interpolated term
parent: {
dependencies: [
{
}
]
}
}
}
Some sort of tree is probably necessary, but I am having trouble figuring out what it might look like, wondering if you could shed some light on that with some pseudocode (or JavaScript even).
watch the leaf nodes at first
then, when the children of a node are completed, propagate upward to resolving the next parent node. This would mean once x and y are done, it could resolve a{x}-{y}, but then wait until the other nodes are ready before doing the final top-level evaluation.
You can just simulate it by sending "events" to the system theoretically, like:
ready('y')
ready('c')
ready('x')
ready('a{x}-{y}')
function ready(variable) {
if ()
}
...actually that may not work, not sure how to handle the interpolated nodes in a hacky way like that. But even a high level description of how to solve this would be helpful.
export type SiteDependencyObserverParentType = {
observer: SiteDependencyObserverType
remaining: number
}
export type SiteDependencyObserverType = {
children: Array<SiteDependencyObserverType>
node: LinkNodeType
parent?: SiteDependencyObserverParentType
path: Array<string>
}
(What I'm currently thinking, some TypeScript)
Here is an approach in JavaScript:
Parse the input string to create a Node instance for each {} term, and create parent-child dependencies between the nodes.
Collect the leaf Nodes of this tree as the tree is being constructed: group these leaf nodes by their identifier. Note that the same identifier could occur multiple times in the input string, leading to multiple Nodes. If a variable x is resolved, then all Nodes with that name (the group) will be resolved.
Each node has a resolve method to set its final value
Each node has a notify method that any of its child nodes can call in order to notify it that the child has been resolved with a value. This may (or may not yet) lead to a cascading call of resolve.
In a demo, a timer is set up that at every tick will resolve a randomly picked variable to some number
I think that in your example, foo, and a might be functions that need to be called, but I didn't elaborate on that, and just considered them as literal text that does not need further treatment. It should not be difficult to extend the algorithm with such function-calling features.
class Node {
constructor(parent) {
this.source = ""; // The slice of the input string that maps to this node
this.texts = []; // Literal text that's not part of interpolation
this.children = []; // Node instances corresponding to interpolation
this.parent = parent; // Link to parent that should get notified when this node resolves
this.value = undefined; // Not yet resolved
}
isResolved() {
return this.value !== undefined;
}
resolve(value) {
if (this.isResolved()) return; // A node is not allowed to resolve twice: ignore
console.log(`Resolving "${this.source}" to "${value}"`);
this.value = value;
if (this.parent) this.parent.notify();
}
notify() {
// Check if all dependencies have been resolved
let value = "";
for (let i = 0; i < this.children.length; i++) {
const child = this.children[i];
if (!child.isResolved()) { // Not ready yet
console.log(`"${this.source}" is getting notified, but not all dependecies are ready yet`);
return;
}
value += this.texts[i] + child.value;
}
console.log(`"${this.source}" is getting notified, and all dependecies are ready:`);
this.resolve(value + this.texts.at(-1));
}
}
function makeTree(s) {
const leaves = {}; // nodes keyed by atomic names (like "x" "y" in the example)
const tokens = s.split(/([{}])/);
let i = 0; // Index in s
function dfs(parent=null) {
const node = new Node(parent);
const start = i;
while (tokens.length) {
const token = tokens.shift();
i += token.length;
if (token == "}") break;
if (token == "{") {
node.children.push(dfs(node));
} else {
node.texts.push(token);
}
}
node.source = s.slice(start, i - (tokens.length ? 1 : 0));
if (node.children.length == 0) { // It's a leaf
const label = node.texts[0];
leaves[label] ??= []; // Define as empty array if not yet defined
leaves[label].push(node);
}
return node;
}
dfs();
return leaves;
}
// ------------------- DEMO --------------------
let s = "foo{a{x}-{y}}-{baz{one}-{two}}-foo{c}";
const leaves = makeTree(s);
// Create a random order in which to resolve the atomic variables:
function shuffle(array) {
for (var i = array.length - 1; i > 0; i--) {
var j = Math.floor(Math.random() * (i + 1));
[array[j], array[i]] = [array[i], array[j]];
}
return array;
}
const names = shuffle(Object.keys(leaves));
// Use a timer to resolve the variables one by one in the given random order
let index = 0;
function resolveRandomVariable() {
if (index >= names.length) return; // all done
console.log("\n---------------- timer tick --------------");
const name = names[index++];
console.log(`Variable ${name} gets a value: "${index}". Calling resolve() on the connected node instance(s):`);
for (const node of leaves[name]) node.resolve(index);
setTimeout(resolveRandomVariable, 1000);
}
setTimeout(resolveRandomVariable, 1000);
your idea of building a dependency tree it's really likeable.
Anyway I tryed to find a solution as simplest possible.
Even if it already works, there are many optimizations possible, take this just as proof of concept.
The background idea it's produce a List of Strings which you can read in order where each element it's what you need to solve progressively. Each element might be mandatory to solve something that come next in the List, hence for the overall expression. Once you solved all the chunks you have all pieces to solve your original expression.
It's written in Java, I hope it's understandable.
import java.util.ArrayList;
import java.util.Collections;
import java.util.List;
import java.util.Objects;
public class StackOverflow {
public static void main(String[] args) {
String exp = "foo{a{x}-{y}}-{baz{one}-{two}}-foo{c}";
List<String> chunks = expToChunks(exp);
//it just reverse the order of the list
Collections.reverse(chunks);
System.out.println(chunks);
//output -> [c, two, one, baz{one}-{two}, y, x, a{x}-{y}]
}
public static List<String> expToChunks(String exp) {
List<String> chunks = new ArrayList<>();
//this first piece just find the first inner open parenthesys and its relative close parenthesys
int begin = exp.indexOf("{") + 1;
int numberOfParenthesys = 1;
int end = -1;
for(int i = begin; i < exp.length(); i++) {
char c = exp.charAt(i);
if (c == '{') numberOfParenthesys ++;
if (c == '}') numberOfParenthesys --;
if (numberOfParenthesys == 0) {
end = i;
break;
}
}
//this if put an end to recursive calls
if(begin > 0 && begin < exp.length() && end > 0) {
//add the chunk to the final list
String substring = exp.substring(begin, end);
chunks.add(substring);
//remove from the starting expression the already considered chunk
String newExp = exp.replace("{" + substring + "}", "");
//recursive call for inner element on the chunk found
chunks.addAll(Objects.requireNonNull(expToChunks(substring)));
//calculate other chunks on the remained expression
chunks.addAll(Objects.requireNonNull(expToChunks(newExp)));
}
return chunks;
}
}
Some details on the code:
The following piece find the begin and the end index of the first outer chunk of expression. The background idea is: in a valid expression the number of open parenthesys must be equal to the number of closing parenthesys. The count of open(+1) and close(-1) parenthesys can't ever be negative.
So using that simple loop once I find the count of parenthesys to be 0, I also found the first chunk of the expression.
int begin = exp.indexOf("{") + 1;
int numberOfParenthesys = 1;
int end = -1;
for(int i = begin; i < exp.length(); i++) {
char c = exp.charAt(i);
if (c == '{') numberOfParenthesys ++;
if (c == '}') numberOfParenthesys --;
if (numberOfParenthesys == 0) {
end = i;
break;
}
}
The if condition provide validation on the begin and end indexes and stop the recursive call when no more chunks can be found on the remained expression.
if(begin > 0 && begin < exp.length() && end > 0) {
...
}
How to get the result of EXPECT_EXIT() in gtest?
I want to do something according to the results of EXPECT_EXIT. For example:
for(int i = 0; i < 1000; i++) {
for(int j = 0; j < 10; j++) {
auto &result = EXPECT_EXIT(Function(i, inputs[j]));
if (result) {
do something;
break; //jump out of the loop
}
}
}
From your comment:
But once it failed, I need record something and skip rest of the input
check for this function
Short answer:
To stop execution at first mismatch you need to use ASSERT version instead of EXPECT, in your case ASSERT_EXIT.
To output some extra data for failed check you can just use stream operator of check:
ASSERT_EXIT(Function(i, j), testing::ExitedWithCode(0), "") << "Some custom data: " << i << "; " << j;
If you want to do something more on fail - you can use ::testing::Test::HasFailure() function.
Long suggestion:
One of the properties of good unittest is simplicity. That implies absence of any loops or branches. Setup logic (if any) should be moved to fixtures. Also unittests must be executed in isolation.
By switching your test to parametric one you would cover that. Check this:
void Function(int i, int j)
{
if (j != 3)
exit(0);
}
class FunTest : public testing::TestWithParam<std::tuple<int, int>> { };
TEST_P(FunTest, Should_Exit)
{
const auto i = std::get<0>(GetParam());
const auto j = std::get<1>(GetParam());
EXPECT_EXIT(Function(i, j), testing::ExitedWithCode(0), "");
}
INSTANTIATE_TEST_SUITE_P(,
FunTest,
testing::Combine(
testing::Range(1, 20),
testing::Range(1, 10)),
[](const testing::TestParamInfo<std::tuple<int, int>>& info) -> std::string {
const auto i = std::get<0>(info.param);
const auto j = std::get<1>(info.param);
return std::to_string(i) + '_' + std::to_string(j);
});
This approach would help with your problem: you can stop execution by using --gtest_break_on_failure and you can control naming of test cases by implementing custom printer for specific TestParamInfo (lambda in example).
It would also guarantee isolation and provide certain level of simplicity (all complex stuff provided by framework, imho test code is minimalistic and simple).
Also this would allow execution of specific test using gtest_filter
(e.g. --gtest_filter=FunTest.Should_Exit/4_3).
sometime based on some condition it may want to jump (or move forward) a few steps inside the for loop,
how to do it is kolin?
a simplified use case:
val datArray = arrayOf(1, 2, 3......)
/**
* start from the index to process some data, return how many data has been
consumed
*/
fun processData_1(startIndex: Int) : Int {
// process dataArray starting from the index of startIndex
// return the data it has processed
}
fun processData_2(startIndex: Int) : Int {
// process dataArray starting from the index of startIndex
// return the data it has processed
}
in Java it could be:
for (int i=0; i<datArray.lenght-1; i++) {
int processed = processData_1(i);
i += processed; // jump a few steps for those have been processed, then start 2nd process
if (i<datArray.lenght-1) {
processed = processData_2(i);
i += processed;
}
}
How to do it in kotlin?
for(i in array.indices){
val processed = processData(i);
// todo
}
With while:
var i = 0
while (i < datArray.length - 1) {
var processed = processData_1(i)
i += processed // jump a few steps for those have been processed, then start 2nd process
if (i < datArray.length - 1) {
processed = processData_2(i)
i += processed
}
i++
}
You can do that with continue as stated in the Kotlin docs here: https://kotlinlang.org/docs/reference/returns.html
Example:
val names = arrayOf("james", "john", "jim", "jacob", "johan")
for (name in names) {
if(name.length <= 4) continue
println(name)
}
This would only print names longer than 4 characters (as it skips names with a length of 4 and below)
Edit: this only skips one iteration at a time. So if you want to skip multiple, you could store the process state somewhere else and check the status for each iteration.
So, I'm trying to make a game that requires randomly colored pictureboxes. I've been trying to make the random color generator, but I'm running into an issue that I can't explain.
When this code runs (inside of Form1_Load event):
for(int i=0; i<6, i++)
{
DateTime moment = DateTime::Now;
Random^RNG=gcnew Random(moment.Millisecond);
color[i]=RNG->Next(16);
if(color[i]<=9)
{
colorStr[i]=color[i].ToString();
}
else if(color[i]==10)
{
colorStr[i]="A";
}
else if(color[i]==11)
{
colorStr[i]="B";
}
else if(color[i]==12)
{
colorStr[i]="C";
}
else if(color[i]==13)
{
colorStr[i]="D";
}
else if(color[i]==14)
{
colorStr[i]="E";
}
else if(color[i]==15)
{
colorStr[i]="F";
}
FullColor+=colorStr[i]; //FullColor was initialized with a value of "#";
}
this->textBox1->Text=FullColor;
this->Player->BackColor = System::Drawing::ColorTranslator::FromHTML(FullColor);
The textbox displays either all the same number (i.e. #000000), or the first number will be unique, but the other five will be equal to each other (i.e. #A22222).
Random generator should not be re-created every time. Try to do it once, before the loop:
Random^RNG=gcnew Random(moment.Millisecond);
for(int i=0; i<6, i++)
{
....
(In your case, it seems that the moment.Millisecond is the same for sequential calls. But even if it would be different, the generator is not supposed to be re-created.)
Instead of the loop, you may consider the following code:
Random^ RNG = gcnew Random(); // somewhere at the beginning
....
int color = RNG->Next(0x1000000);
String^ colorStr = color.ToString("X6");
I have an ini file with information like:
slExpsave_0 = 0.23;
slExpsave_1 = 0.40;
slExpsave_2 = -0.85;
...
I need to use them in a loop;
...
for(var i=0; i<3; i++)
{
slExpArray[i].value = slExpsave_[i];
}
But it doesn't work.
What is the reason for this?
I'm guessing this is Javascript, so you should use an array if you want to access them within a loop:
slExpsave_ = [0.23, 0.40, -0.85];
Does this help?
Let me explain my problem more clearly.
As I did it manually, It worked as I wanted:
slExpsave_0 = 0.23;
slExpsave_1 = 0.40;
slExpsave_2 = -0.85;
slExpArray[0].value = slExpsave_0; // output is 0.23;
slExpArray[1].value = slExpsave_1; // output is 0.40;
slExpArray[2].value = slExpsave_2; // output is -0.85;
when I tried to do it automatically by put in the for loop:
the first one I tried:
for(var i=0; i<3; i++)
{
slExpArray[i].value = "slExpsave_" + i;
}
// output is slExpsave_0, slExpsave_1, slExpsave_2
the second was:
for(var i=0; i<3; i++)
{
slExpArray[i].value = slExpsave_[i];
}
// doesn't work.