string: A Row of Characters That Knows Its Own Length

Maria types her name into a C program that has char name[10]. "Maria" fits. Her full name, "Maria Lopez", does not, and the extra letters spill into memory the array does not own. Every strlen then walks the letters again to count them. In C++ none of that is your job: string name; grows to fit, and name.size() already knows the length. This lesson shows the string, and one honest surprise about Maria's name in Bangla.

The problem: text that will not sit still

In the C track, text was a char array ending in '\0'. You chose its length before the program ran. You copied with strcpy, joined with strcat and measured with strlen, which counts the characters one by one every time you call it. A string is the C++ answer: a row of characters that grows by itself and keeps its own length. Its full name is std::string, and it lives in the header <string>.

Here is the same job twice. Read a first name and a last name, join them with a space, and print the result and its length. First the C way, which still compiles as C++.

#include <cstdio>
#include <cstring>

int main() {
    char first[20], last[20], full[41];
    scanf("%19s %19s", first, last);
    strcpy(full, first);
    strcat(full, " ");
    strcat(full, last);
    printf("%s has %zu characters\n", full, strlen(full));
    return 0;
}

Maria Lopez has 11 characters

Now the C++ way, on the same input.

#include <iostream>
#include <string>
using namespace std;

int main() {
    string first, last;
    cin >> first >> last;
    string full = first + " " + last;
    cout << full << " has " << full.size() << " characters\n";
    return 0;
}

Maria Lopez has 11 characters

That output is for the input Maria Lopez. The C version needs three array sizes, a %19s to stop an overflow, and two library calls to join. The C++ version has no size anywhere. So a string is text that you never size by hand.

A string is a row of characters, and a little more

Inside, a string keeps its characters side by side in one block of memory, like a C array or a vector from Module 2. A character here is one char, one byte. The string also keeps two numbers: its size, how many characters it holds, and its capacity, how many fit before it needs a bigger block.

One more thing sits in the block. After the last character, the library keeps a '\0', the same end marker C strings use. It is not counted in the size. It is there so the string can hand its text to a C function in one step (lesson 02 shows c_str()).

So a string is a vector of characters that also keeps an end marker, which is why name.size() is instant where strlen(name) was a walk.

The syntax you need on day one

#include <string>

string s;                     an empty string, ""
string s = "Maria";           a copy of this text
string s(5, '*');             5 copies of one character, "*****"
s.size()   s.length()         how many characters (bytes) it holds
s.empty()                     true when it holds none
s[i]                          the character at index i, 0 to s.size() - 1
s + t      s += t             a new joined string; add t to the end of s
s == t                        true when the two texts are the same
cin >> s;                     read one word, up to a space or a line break
getline(cin, s);              read a whole line, spaces included

So a string is declared like any variable, and everything you do to it is an operator or a call with a dot.

Four ways to make one

You will make strings in four shapes. This program makes one of each and prints it between brackets, so an empty one still shows.

#include <iostream>
#include <string>
using namespace std;

int main() {
    string empty_one;
    string name = "Maria";
    string stars(5, '*');
    string copy = name;
    copy[0] = 'm';

    cout << "[" << empty_one << "] size " << empty_one.size() << '\n';
    cout << "[" << name << "] size " << name.size() << '\n';
    cout << "[" << stars << "] size " << stars.size() << '\n';
    cout << "[" << copy << "] size " << copy.size() << '\n';
    return 0;
}

[] size 0
[Maria] size 5
[*****] size 5
[maria] size 5

The table reads the four lines back.

The last line shows something C never did with =. Changing copy[0] left name alone. In C, char* copy = name; makes a second name for the same letters. So = on strings copies the text; it does not share it.

Reading a word, and reading a line

cin >> s skips any spaces, then reads one word: characters up to the next space, tab or line break. getline(cin, s) reads everything up to the end of the line, spaces included, and throws the line break away. Module 1, lesson 02, met both. Here they are on one line of input.

#include <iostream>
#include <string>
using namespace std;

int main() {
    string word, rest;
    cin >> word;
    getline(cin, rest);
    cout << "word: [" << word << "]\n";
    cout << "rest: [" << rest << "]\n";
    cout << "rest has " << rest.size() << " characters\n";
    return 0;
}

word: [Maria]
rest: [ Lopez likes tea]
rest has 16 characters

That output is for the input Maria Lopez likes tea. >> stopped at the first space and left it in the input. getline then took the rest of the line, starting with that space. This is also the trap Module 1 showed: after cin >> n, a getline reads the empty rest of the number's line. So read words with >>, lines with getline, and finish a line before you switch.

Indexing, and Bob's last index

The characters are numbered from 0, like an array's boxes. A string of size n has indexes 0 to n - 1, and s[i] is one char you can read or change. Bob prints the character codes of "Maria", one per index. He writes <= to be sure he reaches the last one.

#include <iostream>
#include <string>
using namespace std;

int main() {
    string name = "Maria";
    for (size_t i = 0; i <= name.size(); i++) {
        cout << "[" << (int)name[i] << "]";
    }
    cout << '\n';
    return 0;
}

[77][97][114][105][97][0]

Five codes are the letters: 77 is 'M' and 97 is 'a', from the ASCII table. The sixth, 0, is name[5], which is name[name.size()]. A vector would have read past its block there, as Module 2 showed. A string has its end marker in that slot, and the standard says s[s.size()] reads it. So Bob's loop prints a 0 he did not want, with no crash and no message.

One step further is real trouble. name[name.size() + 1] is outside the text, and nothing checks it. On Compiler Explorer, at the Playground's flags, name[name.size() + 5] read 9 in one run and -15 in the next. Those are leftover bytes, so any run may give anything. So write i < s.size(), and remember that the slot at s.size() is the end marker, not a character.

size() is the same unsigned type as a vector's, so s.size() - 1 on an empty string is 18446744073709551615, exactly as Module 2, lesson 01 showed. Check empty() before you subtract.

A string is bytes: Maria's name in Bangla

Maria asks why her name has a different length when she types it in Bangla. The answer is the most honest line in this module: a string holds bytes, and size() counts bytes. A byte is 8 bits, a number from 0 to 255. Every ASCII character, the English letters, digits and punctuation, is one byte. A Bangla letter is not.

The Playground stores text in UTF-8, the encoding almost every web page and Linux system uses. UTF-8 gives every character in every language a number, its code point, written like U+09AE. Then it writes that number as one to four bytes. ASCII code points take one byte. Each Bangla code point takes three.

#include <iostream>
#include <string>
using namespace std;

int main() {
    string en = "Maria";
    string bn = "মারিয়া";
    cout << en << ": size " << en.size() << '\n';
    cout << bn << ": size " << bn.size() << '\n';
    cout << "bytes:";
    for (unsigned char c : bn) {
        cout << ' ' << (int)c;
    }
    cout << '\n';
    return 0;
}

Maria: size 5
মারিয়া: size 21
bytes: 224 166 174 224 166 190 224 166 176 224 166 191 224 166 175 224 166 188 224 166 190

Five letters, size 5. The Bangla name looks like three letters, মা, রি and য়া, and its size is 21. It is seven code points: three letters, three vowel signs and one nukta dot, each written as three bytes. Every byte is above 127, so none of them is an ASCII character on its own. The loop uses unsigned char so the bytes print as 0 to 255; a plain char on the Playground is signed and would print them as negative numbers. Step through the name below, one code point at a time.

So three numbers describe one Bangla word: the bytes size() counts, the code points, and the letters a reader sees. For English text they are all the same number, which is why the problems in this module use ASCII input. For Bangla, s[0] is one byte of a letter, never the whole letter.

Joining with +, and comparing with ==

+ makes a new string from two others, and += adds to the end of an existing one. == asks whether two strings hold the same text. In C, == on two char arrays compared their addresses, and you needed strcmp. This program shows both behaviours side by side.

#include <iostream>
#include <string>
using namespace std;

int main() {
    char a[] = "tea";
    char b[] = "tea";
    cout << "C arrays equal? " << (a == b) << '\n';

    string x = "tea";
    string y = "te";
    y += 'a';
    cout << "strings equal? " << (x == y) << '\n';

    string order = x + " and " + y;
    cout << order << '\n';
    return 0;
}

C arrays equal? 0
strings equal? 1
tea and tea

The two arrays hold the same letters but live at two addresses, so a == b is false. The strings compare their characters, so x == y is true even though y was built in two steps. += took a single char, 'a', as easily as a whole string. So == means "same text" for strings, which is what you meant all along.

One rule about +: a string must be on one side of each join. x + " and " + y works because it starts from a string, and each step gives a string back. Two quoted texts are not strings; they are C arrays, so "Maria" + " " + last fails at every command line. GCC 12 says error: invalid operands of types 'const char [6]' and 'const char [2]' to binary 'operator+'. Start from a string, or write string("Maria").

#include <iostream>
#include <string>
using namespace std;

int main() {
    string dish;
    cin >> dish;
    cout << dish << " has " << dish.size() << " letters\n";
    cout << "first " << dish[0] << ", last " << dish[dish.size() - 1] << '\n';
    return 0;
}

biryani has 7 letters
first b, last i

#include <iostream>
#include <string>
using namespace std;

int main() {
    string line, longest;
    while (getline(cin, line)) {
        cout << line << " (" << line.size() << ")\n";
        if (line.size() > longest.size()) {
            longest = line;
        }
    }
    cout << "longest: " << longest << '\n';
    return 0;
}

Maria Lopez (11)
Bob (3)
Kenji Watanabe (14)
longest: Kenji Watanabe

#include <iostream>
#include <string>
using namespace std;

int main() {
    string pw;
    while (getline(cin, pw)) {
        bool digit = false, capital = false;
        for (char c : pw) {
            if (c >= '0' && c <= '9') {
                digit = true;
            }
            if (c >= 'A' && c <= 'Z') {
                capital = true;
            }
        }
        bool ok = pw.size() >= 8 && digit && capital;
        cout << "[" << pw << "] " << (ok ? "strong" : "weak") << '\n';
    }
    return 0;
}

[tea4two] weak
[Tea4two!] strong
[] weak
[Biryani2026] strong

string name;
cin >> name;
cout << "Hello, " << name << "!\n";

string name = "Maria";
printf("%s\n", name);

string s;
s[0] = 'M';
cout << "[" << s << "] size " << s.size() << '\n';

#include <iostream>
#include <string>
using namespace std;

int main() {
    string kenji(3, 'a');
    string bob = "3" + 1;
    cout << "[" << kenji << "] [" << bob << "]\n";
    return 0;
}

#include <iostream>
#include <string>
using namespace std;

int main()
{
    ios::sync_with_stdio(false);
    cin.tie(nullptr);

    string word;
    while (cin >> word) {
        // Keep the longest word seen so far.
        // Replace it only when this word is strictly longer.
    }

    // Print the longest word, a space, and its length.

    return 0;
}

#include <iostream>
#include <string>
using namespace std;

int main() {
    string name;
    getline(cin, name);

    // Make a string of stars as long as the framed line,
    // then print the three lines.

    return 0;
}

#include <iostream>
#include <string>
using namespace std;

int main() {
    string line;
    getline(cin, line);

    // Count the bytes that start a code point.
    // Hint: look at the bytes of the Bangla name in this lesson.

    return 0;
}

Next, Bob builds a long line with s = s + c and waits. Lesson 02 puts a cost on every string operation and shows why.