Vector reserve & iterator

oops, I left out a line...that's why it's not making sense.
should be this:

// insert the first file
todoVector.push_back(ModelFile(masterModelName));
for(todoVector_itr=todoVector.begin(); todoVector_itr!=todoVector.end(); ++todoVector_itr){
// process the current file
process(*todoVector_itr);
// add each file referenced in the current file to list of files to examine.
for(name_itr=todoVector_itr->vectorINCLUDE.begin();name_itr!=todoVector_itr->vectorINCLUDE.end();++name_itr){
todoVector.push_back(ModelFile(*name_itr));
}
}

Ok, this is how it is supposed to work.
Initially, I push one ModelFile on to the vector.
I then go into the first for loop to process all the files in the todovector. There is at least one file to process.

Each ModelFile obj has a vectorINCLUDE to keep track of all the files referenced in the current file. Process() adds to vectorInclud any filenames mentioned in the current file.

After a file is processed, I need to add all the files in the vectorINCLUDE to the todovector. VectorInclude is a string vector so that's why the second for loop is also pushing Modelfile obj.

So this is where the todoVector_itr iterator invalidating occurs. If I don't specify a reserve size, the first cycle of the 2nd for loop processes fine. But the second cycle will then cause an error because the outer loop iterator was setup before the second push_back of the inner for loop. This 2nd push somehow invalidates the iterator.

my fix came from here: http://www.sgi.com/tech/stl/Vector.html

A vector's iterators are invalidated when its memory is reallocated. Additionally, inserting or deleting an element in the middle of a vector invalidates all iterators that point to elements following the insertion or deletion point. It follows that you can prevent a vector's iterators from being invalidated if you use reserve() to preallocate as much memory as the vector will ever use, and if all insertions and deletions are at the vector's end.

I suppose I could get around this problem and brute force it with a while less then todovector.size type of loop. But this way, I learn a little more. :)

In general you want to use iterators rather the a while loop with counter right?

ok, suppose I have the following 3 files.
filein.txt:

123
345
fileb.txt
324
filec.txt
12
fileb.txt:

123465
34565
324789
12
filec.txt:

123.123
345.2
324.6
12.9

The ModelFile object has a vector<strings> vectorINCLUDE.

I don't know beforehand how many files there are. I need to process the first file to know exactly how many. In this example, I should have three ModelFile obj. (filein, fileb, filec).

todoVector is a global vector<ModelFile>. As each text file is processed, any filename (eg. fileb.txt) called out in the file being processed (filein.txt) needs to be added to the global todoVector so that its contents can be examined later. Since I read a string from the files, that's why I have to push ModelFile(somestring).

so my code starts with:
ModelFile filea("filein.txt") and push that in to the global todovector.
I then setup the global vector iterator (with only one item initially).
after process() command is done, filein.vectorINCLUDE then will have two items (fileb.txt & filec.txt)
this is where the second loop comes in. I need to turn the two string names into ModelFIle obj and push those on to the global todovector. This is because after I'm done with reading filein.txt, I need to then read fileb.txt and filec.txt and so on...

in the case of fileb, and filec, there are no filename strings so nothing needs to be added the the global vector.

does this make more sense?