Return to Question

Based on Edward answer, I'm doing rewriting and getting rid of pointers. However, here I've got a problem coz of reference needs to be initialized at declaration. ShopStockFound is initialized later and it already contains data from next item while previous items still needs to be processed somehow.

 // {{{ ProcessShopStocks
void ProcessShopStocks(ShopStocksMap& shopStocks, const ShopsMap& shops,
 FILE* out) {
 std::ifstream file(SHOP_STOCKS_FILE);
 std::string line;
 ShopStockStruct *shopStock= NULL;
 AvailabilityStruct *availability;
 int numShopStocks = 0;
 char *itemID;
 char *itemIDprev = (char *)"";
 int shopID;
 int amount;
 std::vector<std::string> items;
 std::getline(file, line); // skip header
 while (std::getline(file, line)) {
 items = split(line, (char) '\t');
 if (items.size() < 3)
 continue;
 itemID = strdup(items[0].c_str());
 if (strcmp(itemIDprev, itemID) != 0) {
 // item changed - post process previous item
 if (shopStock != NULL)
 PostProcessItem(*shopStock, shops, out);
 }
 shopID = atoi(items[1].c_str());
 amount = atoi(items[2].c_str());
 /*
 * replaced - see below
 auto shopStockFound = shopStocks.find(itemID);
 if (shopStockFound == shopStocks.end()) {
 // not found, so we need to create a new one
 shopStock = new ShopStockStruct();
 shopStock->itemID = itemID;
 // and store it, we'll add from value a bit later
 shopStocks[itemID] = *shopStock;
 }
 else {
 // found group already, get a pointer to it for adding nGroupFrom
 // value later
 shopStock = &(shopStockFound->second);
 }
 // in any case (found or not), we've to add GroupFrom value
 // to do: create struct, and store it in map
 availability = new AvailabilityStruct();
 availability->shopID = shopID;
 availability->amount = amount;
 availability->totalAmount = 0;
 availability->totalAmount10k = 0;
 (shopStock->availability)[shopID] = *availability;
 */
 auto& shopStockFound = shopStocks[itemID] = { .itemID = itemID };
 shopStockFound.availability[shopID] = { 
 .shopID = shopID,
 .amount = amount,
 .totalAmount = 0,
 .totalAmount10k = 0
 };
 shopStock = &shopStockFound;
 numShopStocks++;
 }
 // process the last item if exists 
 if (shopStock != NULL)
 PostProcessItem(*shopStock, shops, out);
 printf("Total %d ShopStocks loaded.\n", numShopStocks);
}
// }}}

BTW, after replacing some of the pointers with references the performance improved from 1.1 to 0.4 seconds. However, after modifying this function it goes back and even become worst (1.2 seconds now).

Based on Edward answer, I'm doing rewriting and getting rid of pointers. However, here I've got a problem coz of reference needs to be initialized at declaration. ShopStockFound is initialized later and it already contains data from next item while previous items still needs to be processed somehow.

 // {{{ ProcessShopStocks
void ProcessShopStocks(ShopStocksMap& shopStocks, const ShopsMap& shops,
 FILE* out) {
 std::ifstream file(SHOP_STOCKS_FILE);
 std::string line;
 ShopStockStruct *shopStock= NULL;
 AvailabilityStruct *availability;
 int numShopStocks = 0;
 char *itemID;
 char *itemIDprev = (char *)"";
 int shopID;
 int amount;
 std::vector<std::string> items;
 std::getline(file, line); // skip header
 while (std::getline(file, line)) {
 items = split(line, (char) '\t');
 if (items.size() < 3)
 continue;
 itemID = strdup(items[0].c_str());
 if (strcmp(itemIDprev, itemID) != 0) {
 // item changed - post process previous item
 if (shopStock != NULL)
 PostProcessItem(*shopStock, shops, out);
 }
 shopID = atoi(items[1].c_str());
 amount = atoi(items[2].c_str());
 /*
 * replaced - see below
 auto shopStockFound = shopStocks.find(itemID);
 if (shopStockFound == shopStocks.end()) {
 // not found, so we need to create a new one
 shopStock = new ShopStockStruct();
 shopStock->itemID = itemID;
 // and store it, we'll add from value a bit later
 shopStocks[itemID] = *shopStock;
 }
 else {
 // found group already, get a pointer to it for adding nGroupFrom
 // value later
 shopStock = &(shopStockFound->second);
 }
 // in any case (found or not), we've to add GroupFrom value
 // to do: create struct, and store it in map
 availability = new AvailabilityStruct();
 availability->shopID = shopID;
 availability->amount = amount;
 availability->totalAmount = 0;
 availability->totalAmount10k = 0;
 (shopStock->availability)[shopID] = *availability;
 */
 auto& shopStockFound = shopStocks[itemID] = { .itemID = itemID };
 shopStockFound.availability[shopID] = { 
 .shopID = shopID,
 .amount = amount,
 .totalAmount = 0,
 .totalAmount10k = 0
 };
 shopStock = &shopStockFound;
 numShopStocks++;
 }
 // process the last item if exists 
 if (shopStock != NULL)
 PostProcessItem(*shopStock, shops, out);
 printf("Total %d ShopStocks loaded.\n", numShopStocks);
}
// }}}

BTW, after replacing some of the pointers with references the performance improved from 1.1 to 0.4 seconds. However, after modifying this function it goes back and even become worst (1.2 seconds now).

Additionally, all shops within the same group also share their effective inventory without needing a special line in the sgm.csv file.