std::ifstream fin(file);
if(!fin)
{
return false;
}
//TODO: Need to put a check for some validation - just one row that has GameLauncherDatabase
String line = "";
getline(fin, line, '\n');
if(line != "GameLauncherDatabase")
{
fin.close();
return false;
}
bool withinQuotes = false;
String word = "";
while(!fin.eof())
{
withinQuotes = false;
getline(fin, line, '\n');
m_hge->System_Log("%s\n",line);
m_hge->System_Log("line.size() = %i\n",line.length());
for(size_t i = 0; i < line.size(); i++)
{
if(line.substr(i,1) == "\"")
{
withinQuotes = !withinQuotes;
m_hge->System_Log("withinQuotes = %i\n",withinQuotes);
}
else if(line.substr(i,1) == "," && !withinQuotes)
{
m_hge->System_Log("%s\n", word);
word = "";
m_hge->System_Log("Comma\n");
}
else
{
word.append(line.substr(i,1));
//word.append("a");
m_hge->System_Log("word = %s\n", word);
}
}
m_hge->System_Log("%s\n", word);
//*wordIter = word;
}
fin.close();Just want to iterate through a string :(
Author
I've been looking at tutorials and String documentation, and I just get weird errors. I won't bother posting them because I'm guessing this must be simple enough and I'm going about it all wrong.
I read a line from a file (Yay I can do that)
Now all I want to do is go through one character at a time, copying each character to another string... basically I'm parsing it - yes it's tokenising basically but I'm reading a CSV, so I need to ignore commas between quotes. The logic of my code is fine - I just don't know the proper syntax for iterating through a string and copying one character to another. I keep getting errors...
What do I use? String stream? For or while loop? Built in C++ string functions or direct character access[?]
Here's what I've got so far. Currently it crashes.
Among other ways...
But you should really post the code you're trying to compile and the errors you get, otherwise its going to be hard to help you.
std::string a = /* ... */;std::string b;for(int i = 0; i < a.size(); ++i) b += a;But you should really post the code you're trying to compile and the errors you get, otherwise its going to be hard to help you.
My intuition would be:
std::string extractNext(std::string & from){ std::string::size_type end = from.find_first_of("\","); while (end != std::string::npos && from[end] == '\"') { if (end = from.find_first_of("\"",end+1) == std::string::npos) ; // throw an exception, because of unterminated quote end = from.find_first_of("\",",end+1); } std::string next = from.substr(0,end); from = from.substr(end); return next;}std::string text;while (text.size() > 0) vect.push_back(extractNext(text));
Author
Thanks guys - geez that was quick, I only just posted and was about to try and delete it to go and whittle down my problem properly by writing just a simple console app that does a mock up string parsing, without the other shit in there. I think I posted to early - 2am and getting frustrated I guess.
But those posts are probably enough info for me to go on. In any case I'm going to make a console app to test it out.
Thanks for the prompt replies. :)
But those posts are probably enough info for me to go on. In any case I'm going to make a console app to test it out.
Thanks for the prompt replies. :)
I would use standard C++ iterator, something like;
std::string str = "This is a string.";
std::string::const_iterator cursor = str.begin();
// move to next char and print it out
cursor++;
std::string letter = *cursor;
std::cout << letter << std::endl; // it should be "h"
And, I made a string tokenizer using C++ string class and iterator, it behaves like strtok() in standard C library. Check it out;
String Tokenizer
std::string str = "This is a string.";
std::string::const_iterator cursor = str.begin();
// move to next char and print it out
cursor++;
std::string letter = *cursor;
std::cout << letter << std::endl; // it should be "h"
And, I made a string tokenizer using C++ string class and iterator, it behaves like strtok() in standard C library. Check it out;
String Tokenizer
Author
Quote:Original post by jpetrie
Among other ways...std::string a = /* ... */;std::string b;for(int i = 0; i < a.size(); ++i) b += a;
Ah yes about that - I tried that earler, and VS 2003 gives me a compiler warning about the i < a.size(), saying it's a conversion between signed and unsigned. I've been taught to get rid of warnings - what do I do there?
Just make i an unsigned type; there's no reason it needs to allow negative numbers in that context.
Quote:Original post by Domarius
Ah yes about that - I tried that earler, and VS 2003 gives me a compiler warning about the i < a.size(), saying it's a conversion between signed and unsigned. I've been taught to get rid of warnings - what do I do there?
Use std::string::size_type instead of int.
Aside from what was said about actually finding a good way to do the iteration, I wanted to illustrate some general matters of technique.
(You might also want to consider more carefully exactly how the CSV spec works. It's surprisingly complex and I'm pretty sure you've missed a lot of corner cases.)
std::ifstream fin(file);if(!fin) { return false;}// Unless you made a typedef, "string" should be lowercase. :)std::string line = "";// Take advantage of default parameters. They have the values they do for a// reason, generally speaking.getline(fin, line);if (line != "GameLauncherDatabase") { // No need to close the file stream explicitly. Only do this when // a file *must* be released *before* the end of the variable's // lifetime, or when the stream object will be "reused" (generally a // bad idea anyway). return false;}bool withinQuotes = false;string word = "";// Don't use .eof() for loop conditions. Use the file read for the condition.// This way is idiomatic and avoids bugs - .eof() is only true after a read// has already failed. In practice, this typically means you process the// last line twice.while (getline(fin, line)) { withinQuotes = false; // I took out your logging stuff. Please don't emulate C stdio for // your logging functions. Emulate C++ streams instead: much cleaner. // If you really need the power of stdio's formatting, use boost::format // and gain typesafety (and the ability to work with user-defined // classes) at the same time. // In particular, "m_hge->System_Log("%s\n", line);" was probably itself // a likely cause of crashes. std::string instances are not char*s. for (size_t i = 0; i < line.size(); i++) { // Cache things that you'll use more than once, so that you // avoid repeating yourself and give a meaningful name to // important entities. std::string current = line.substr(i, 1); if (current == "\"") { withinQuotes = !withinQuotes; } else if (current == "," && !withinQuotes) { word = ""; } else { // Don't use .append() with strings, normally. Use // the operator overload instead; it reads better. word += current; } }}(You might also want to consider more carefully exactly how the CSV spec works. It's surprisingly complex and I'm pretty sure you've missed a lot of corner cases.)
Author
You guys have been really helpful considering the mess I posted - it was 2am and I was fed up, and I admit I could have done at least a little bit more towards reaching the answer.
But this is all awesome info and I am pretty sure I'll be able to get this done when I get back from work.
And to Zalman who went the extra step to help me with practice (the things I have not replied to you can be sure I've taken on board by default :) );
// Unless you made a typedef, "string" should be lowercase. :)
Yeah I did make a typedef :) ... our instructor (back when I used to study) did this so I have just been following suit, everything he does, till I know better.
// No need to close the file stream explicitly. Only do this when
// a file *must* be released *before* the end of the variable's
// lifetime, or when the stream object will be "reused" (generally a
// bad idea anyway).
Oh... I am just in the habit of cleaning up everything... if I delete all variables then I want to close all files too... is it bad practice?
// I took out your logging stuff. Please don't emulate C stdio for
// your logging functions. Emulate C++ streams instead: much cleaner.
// If you really need the power of stdio's formatting, use boost::format
// and gain typesafety (and the ability to work with user-defined
// classes) at the same time.
// In particular, "m_hge->System_Log("%s\n", line);" was probably itself
// a likely cause of crashes. std::string instances are not char*s.
Ah I am using the game engine HGE at hge.relishgames.com
I already posted there about how to get the logging working (it's a singleton)
// Cache things that you'll use more than once, so that you
// avoid repeating yourself and give a meaningful name to
// important entities.
Ah yes, thanks for the reminder, I know what you mean. Should be storing line.substr(i,1) as "currentChar" or something.
// Don't use .append() with strings, normally. Use
// the operator overload instead; it reads better.
word += current;
Heheh yeah I know - but you know what? I was tired and desperate and trying random things. Not good.
As for the CSV standard - yeah I realise it might break on some details, and if I really wanted to read a CSV file properly, I already downloaded an awesome pre-made class that someone made for free that apparently is very good.
But I have a very specific format of CSV for this app I'm making (my app cannot make use of any other format - must have string, string, int, int for each row), and while it would be more reliable to use that CSV thing I downloaded, I wouldn't be learning much, except how to use someone elses class. I wanted to take advantage of the situation and make something very simple as a matter of excersise. I even don't care if it crashes if it is forced to read in an innapropriate CSV - the user is never required to edit the CSV manually so they shouldn't be going there, and if anything goes wrong, the app is able to re-create the database automatically (it's based on a file structure that should be present).
Okay I'm using the CSV as sort of a database, and the entire thing is loaded into data structures in an array (an object for each item in the database, with public variables for each value, eg .name, .description, etc.), so the CSV is only read once at the start of the app, and the data is manipulated in memory after that. I'm sure my way will "work", but I'd be interested to hear your advice on the more "ideal" way to achieve such a thing.
Oh, and I'll remember your quote there about having char in my code ;)
But this is all awesome info and I am pretty sure I'll be able to get this done when I get back from work.
And to Zalman who went the extra step to help me with practice (the things I have not replied to you can be sure I've taken on board by default :) );
// Unless you made a typedef, "string" should be lowercase. :)
Yeah I did make a typedef :) ... our instructor (back when I used to study) did this so I have just been following suit, everything he does, till I know better.
// No need to close the file stream explicitly. Only do this when
// a file *must* be released *before* the end of the variable's
// lifetime, or when the stream object will be "reused" (generally a
// bad idea anyway).
Oh... I am just in the habit of cleaning up everything... if I delete all variables then I want to close all files too... is it bad practice?
// I took out your logging stuff. Please don't emulate C stdio for
// your logging functions. Emulate C++ streams instead: much cleaner.
// If you really need the power of stdio's formatting, use boost::format
// and gain typesafety (and the ability to work with user-defined
// classes) at the same time.
// In particular, "m_hge->System_Log("%s\n", line);" was probably itself
// a likely cause of crashes. std::string instances are not char*s.
Ah I am using the game engine HGE at hge.relishgames.com
I already posted there about how to get the logging working (it's a singleton)
// Cache things that you'll use more than once, so that you
// avoid repeating yourself and give a meaningful name to
// important entities.
Ah yes, thanks for the reminder, I know what you mean. Should be storing line.substr(i,1) as "currentChar" or something.
// Don't use .append() with strings, normally. Use
// the operator overload instead; it reads better.
word += current;
Heheh yeah I know - but you know what? I was tired and desperate and trying random things. Not good.
As for the CSV standard - yeah I realise it might break on some details, and if I really wanted to read a CSV file properly, I already downloaded an awesome pre-made class that someone made for free that apparently is very good.
But I have a very specific format of CSV for this app I'm making (my app cannot make use of any other format - must have string, string, int, int for each row), and while it would be more reliable to use that CSV thing I downloaded, I wouldn't be learning much, except how to use someone elses class. I wanted to take advantage of the situation and make something very simple as a matter of excersise. I even don't care if it crashes if it is forced to read in an innapropriate CSV - the user is never required to edit the CSV manually so they shouldn't be going there, and if anything goes wrong, the app is able to re-create the database automatically (it's based on a file structure that should be present).
Okay I'm using the CSV as sort of a database, and the entire thing is loaded into data structures in an array (an object for each item in the database, with public variables for each value, eg .name, .description, etc.), so the CSV is only read once at the start of the app, and the data is manipulated in memory after that. I'm sure my way will "work", but I'd be interested to hear your advice on the more "ideal" way to achieve such a thing.
Oh, and I'll remember your quote there about having char in my code ;)
This topic is closed to new replies.
Advertisement
Popular Topics
Advertisement
