Thread: Identify and count word sequences within a text file

It depends what you mean by sequences, if there is an easily identified fixed string to search for then std library objects like std::string iteslf will probably suffice, as the member functions like find_first_of replace, etc can be used.

Like if you just want to find how many times 'the' occurs
then you could just use getline() on your file stream, and parse each line as a temp string using find() for your text

If it is likely to be more jumbled up and complex searching you might consider a regular expression library, or even a fuzzy matching algorithm

well a simplistic (and expensive) approach is just brute force, you decide how many words form a sequence, say 3. Then probably buffer the file twice, step to word one in buffer1, get a three word sequence from theat start point and parse all through buffer2 every time a match for buffer1 sequence is found save the result for it. Then step to word two in buffer1, get three words from that start point, check it against buffer2, save results..then step to word 3 in buffer1 blah blah etc etc. Punctuation will have to be dealt with, included or discarded. If you have enough time you could do the whole process starting with single word, then two word seqwuence, then three, four, five, etc.
But like i say this is very simplistic.