This library is a lightweight modern wrapper around sqlite C api .
#include<iostream>
#include <sqlite_modern_cpp.h>
using namespace sqlite;
using namespace std;
int main() {
try {
// creates a database file 'dbfile.db' if it does not exists.
database db("dbfile.db");
// executes the query and creates a 'user' table
db <<
"create table if not exists user ("
" _id integer primary key autoincrement not null,"
" age int,"
" name text,"
" weight real"
");";
// inserts a new user record.
// binds the fields to '?' .
// note that only types allowed for bindings are :
// int ,long, long long, float, double
// string , u16string
// sqlite3 only supports utf8 and utf16 strings, you should use std::string for utf8 and std::u16string for utf16.
// note that u"my text" is a utf16 string literal of type char16_t * .
db << "insert into user (age,name,weight) values (?,?,?);"
<< 20
<< u"bob"
<< 83.25;
int age = 21;
float weight = 68.5;
string name = "jack";
db << u"insert into user (age,name,weight) values (?,?,?);" // utf16 query string
<< age
<< name
<< weight;
cout << "The new record got assigned id " << db.last_insert_rowid() << endl;
// slects from user table on a condition ( age > 18 ) and executes
// the lambda for each row returned .
db << "select age,name,weight from user where age > ? ;"
<< 18
>> [&](int age, string name, double weight) {
cout << age << ' ' << name << ' ' << weight << endl;
};
// selects the count(*) from user table
// note that you can extract a single culumn single row result only to : int,long,long,float,double,string,u16string
int count = 0;
db << "select count(*) from user" >> count;
cout << "cout : " << count << endl;
// you can also extract multiple column rows
db << "select age, name from user where _id=1;" >> tie(age, name);
cout << "Age = " << age << ", name = " << name << endl;
// this also works and the returned value will be automatically converted to string
string str_count;
db << "select count(*) from user" >> str_count;
cout << "scount : " << str_count << endl;
}
catch (exception& e) {
cout << e.what() << endl;
}
}
You can not execute multiple statements separated by semicolons in one go.
You can pass additional open flags to SQLite by using a config object:
sqlite_config config;
config.flags = OpenFlags::READONLY
database db("some_db", config);
int a;
// Now you can only read from db
auto ps = db << "select a from table where something = ? and anotherthing = ?" >> a;
config.flags = OpenFlags::READWRITE | OpenFlags::CREATE; // This is the default
config.encoding = Encoding::UTF16; // The encoding is respected only if you create a new database
database db2("some_db2", config);
// If some_db2 didn't exists before, it will be created with UTF-16 encoding.
It is possible to retain and reuse statments this will keep the query plan and in case of an complex query or many uses might increase the performance significantly.
database db(":memory:");
// if you use << on a sqlite::database you get a prepared statment back
// this will not be executed till it gets destroyed or you execute it explicitly
auto ps = db << "select a,b from table where something = ? and anotherthing = ?"; // get a prepared parsed and ready statment
// first if needed bind values to it
ps << 5;
int tmp = 8;
ps << tmp;
// now you can execute it with `operator>>` or `execute()`.
// If the statement was executed once it will not be executed again when it goes out of scope.
// But beware that it will execute on destruction if it wasn't executed!
ps >> [&](int a,int b){ ... };
// after a successfull execution the statment can be executed again, but the bound values are resetted.
// If you dont need the returned values you can execute it like this
ps++; // Does reset the bound values
// or like this
ps.execute(); // Does NOT reset the bound values, but we can reset them manually:
ps.reset();
// To disable the execution of a statment when it goes out of scope and wasn't used
ps.used(true); // or false if you want it to execute even if it was used
// Usage Example:
auto ps = db << "insert into complex_table_with_lots_of_indices values (?,?,?)";
int i = 0;
while( i < 100000 ){
ps << long_list[i++] << long_list[i++] << long_list[i++];
ps++; // Equal to: ps.execute(); ps.reset();
}
If you need the handle to the database connection to execute sqlite3 commands directly you can get a managed shared_ptr to it, so it will not close as long as you have a referenc to it.
Take this example on how to deal with a database backup using SQLITEs own functions in a safe and modern way.
try {
database backup("backup"); //Open the database file we want to backup to
auto con = db.connection(); // get a handle to the DB we want to backup in our scope
// this way we are sure the DB is open and ok while we backup
// Init Backup and make sure its freed on exit or exceptions!
auto state =
std::unique_ptr<sqlite3_backup,decltype(&sqlite3_backup_finish)>(
sqlite3_backup_init(backup.connection().get(), "main", con.get(), "main"),
sqlite3_backup_finish
);
if(state) {
int rc;
// Each iteration of this loop copies 500 database pages from database db to the backup database.
do {
rc = sqlite3_backup_step(state.get(), 500);
std::cout << "Remaining " << sqlite3_backup_remaining(state.get()) << "/" << sqlite3_backup_pagecount(state.get()) << "\n";
} while(rc == SQLITE_OK || rc == SQLITE_BUSY || rc == SQLITE_LOCKED);
}
} // Release allocated resources.
You can use transactions with begin;
, commit;
and rollback;
commands.
db << "begin;"; // begin a transaction ...
db << "insert into user (age,name,weight) values (?,?,?);"
<< 20
<< u"bob"
<< 83.25f;
db << "insert into user (age,name,weight) values (?,?,?);" // utf16 string
<< 21
<< u"jack"
<< 68.5;
db << "commit;"; // commit all the changes.
db << "begin;"; // begin another transaction ....
db << "insert into user (age,name,weight) values (?,?,?);" // utf16 string
<< 19
<< u"chirs"
<< 82.7;
db << "rollback;"; // cancel this transaction ...
Use std::vector<T>
to store and retrieve blob data.
T
could be char,short,int,long,long long, float or double
.
db << "CREATE TABLE person (name TEXT, numbers BLOB);";
db << "INSERT INTO person VALUES (?, ?)" << "bob" << vector<int> { 1, 2, 3, 4};
db << "INSERT INTO person VALUES (?, ?)" << "sara" << vector<double> { 1.0, 2.0, 3.0, 4.0};
vector<int> numbers_bob;
db << "SELECT numbers from person where name = ?;" << "bob" >> numbers_bob;
db << "SELECT numbers from person where name = ?;" << "sara" >> [](vector<double> numbers_sara){
for(auto e : numbers_sara) cout << e << ' '; cout << endl;
};
If you have databases where some rows may be null, you can use std::unique_ptr<T>
to retain the NULL values between C++ variables and the database.
db << "CREATE TABLE tbl (id integer,age integer, name string, img blob);";
db << "INSERT INTO tbl VALUES (?, ?, ?, ?);" << 1 << 24 << "bob" << vector<int> { 1, 2 , 3};
unique_ptr<string> ptr_null; // you can even bind empty unique_ptr<T>
db << "INSERT INTO tbl VALUES (?, ?, ?, ?);" << 2 << nullptr << ptr_null << nullptr;
db << "select age,name,img from tbl where id = 1"
>> [](unique_ptr<int> age_p, unique_ptr<string> name_p, unique_ptr<vector<int>> img_p) {
if(age_p == nullptr || name_p == nullptr || img_p == nullptr) {
cerr << "ERROR: values should not be null" << std::endl;
}
cout << "age:" << *age_p << " name:" << *name_p << " img:";
for(auto i : *img_p) cout << i << ","; cout << endl;
};
db << "select age,name,img from tbl where id = 2"
>> [](unique_ptr<int> age_p, unique_ptr<string> name_p, unique_ptr<vector<int>> img_p) {
if(age_p != nullptr || name_p != nullptr || img_p != nullptr) {
cerr << "ERROR: values should be nullptr" << std::endl;
exit(EXIT_FAILURE);
}
cout << "OK all three values are nullptr" << endl;
};
We have native support for SQLCipher.
If you want to use encrypted databases, include the sqlite_moder_cpp/sqlcipher.h
header.
Then create a sqlcipher_database
instead.
#include<iostream>
#include <sqlite_modern_cpp/sqlcipher.h>
using namespace sqlite;
using namespace std;
int main() {
try {
// creates a database file 'dbfile.db' if it does not exists with password 'secret'
sqlcipher_config config;
config.key = secret;
sqlcipher_database db("dbfile.db", config);
// executes the query and creates a 'user' table
db <<
"create table if not exists user ("
" _id integer primary key autoincrement not null,"
" age int,"
" name text,"
" weight real"
");";
// More queries ...
db.rekey("new_secret"); // Change the password of the already encrypted database.
// Even more queries ..
}
catch (exception& e) { cout << e.what() << endl; }
}
You can use std::optional<T>
as an alternative for std::unique_ptr<T>
to work with NULL values.
#include <sqlite_modern_cpp.h>
struct User {
long long _id;
std::optional<int> age;
std::optional<string> name;
std::optional<real> weight;
};
int main() {
User user;
user.name = "bob";
// Same database as above
database db("dbfile.db");
// Here, age and weight will be inserted as NULL in the database.
db << "insert into user (age,name,weight) values (?,?,?);"
<< user.age
<< user.name
<< user.weight;
user._id = db.last_insert_rowid();
// Here, the User instance will retain the NULL value(s) from the database.
db << "select _id,age,name,weight from user where age > ? ;"
<< 18
>> [&](long long id,
std::optional<int> age,
std::optional<string> name
std::optional<real> weight) {
cout << "id=" << _id
<< " age = " << (age ? to_string(*age) ? string("NULL"))
<< " name = " << (name ? *name : string("NULL"))
<< " weight = " << (weight ? to_string(*weight) : string(NULL))
<< endl;
};
}
If you do not have C++17 support, you can use boost optional instead by defining _MODERN_SQLITE_BOOST_OPTIONAL_SUPPORT
before importing the sqlite_modern_cpp
header.
If the optional library is not available, the experimental/optional one will be used instead.
Note: boost support is deprecated and will be removed in future versions.
If your columns may have flexible types, you can use C++17's std::variant
to extract the value.
db << "CREATE TABLE tbl (id integer, data);";
db << "INSERT INTO tbl VALUES (?, ?);" << 1 << vector<int> { 1, 2, 3};
db << "INSERT INTO tbl VALUES (?, ?);" << 2 << 2.5;
db << "select data from tbl where id = 1"
>> [](std::variant<vector<int>, double> data) {
if(data.index() != 1) {
cerr << "ERROR: we expected a blob" << std::endl;
}
for(auto i : get<vector<int>>(data)) cout << i << ","; cout << endl;
};
db << "select data from tbl where id = 2"
>> [](std::variant<vector<int>, double> data) {
if(data.index() != 2) {
cerr << "ERROR: we expected a real number" << std::endl;
}
cout << get<double>(data) << endl;
};
If you read a specific type and this type does not match the actual type in the SQlite database, yor data will be converted.
This does not happen if you use a variant
.
If the variant
does an alternative of the same value type, an mismatch
exception will be thrown.
The value types are NULL, integer, real number, text and BLOB.
To support all possible values, you can use variant<nullptr_t, sqlite_int64, double, string, vector<char>
.
On error, the library throws an error class indicating the type of error. The error classes are derived from the SQLITE3 error names, so if the error code is SQLITE_CONSTRAINT, the error class thrown is sqlite::errors::constraint. SQLite3 extended error names are supported too. So there is e.g. a class sqlite::errors::constraint_primarykey derived from sqlite::errors::constraint. Note that all errors are derived from sqlite::sqlite_exception and that itself is derived from std::runtime_exception.
sqlite::sqlite_exception has a get_code()
member function to get the SQLITE3 error code or get_extended_code()
to get the extended error code.
Additionally you can use get_sql()
to see the SQL statement leading to the error.
database db(":memory:");
db << "create table person (id integer primary key not null, name text);";
try {
db << "insert into person (id, name) values (?,?)" << 1 << "jack";
// inserting again to produce error
db << "insert into person (id, name) values (?,?)" << 1 << "jack";
}
/* if you are trying to catch all sqlite related exceptions
* make sure to catch them by reference */
catch (sqlite_exception& e) {
cerr << e.get_code() << ": " << e.what() << " during "
<< e.get_sql() << endl;
}
/* you can catch specific exceptions as well,
catch(sqlite::errors::constraint e) { } */
/* and even more specific exceptions
catch(sqlite::errors::constraint_primarykey e) { } */
You can also register a error logging function with sqlite::error_log
.
The <sqlite_modern_cpp/log.h>
header has to be included to make this function available.
The call to sqlite::error_log
has to be the first call to any sqlite_modern_cpp
function by your program.
error_log(
[&](sqlite_exception& e) {
cerr << e.get_code() << ": " << e.what() << endl;
},
[&](errors::misuse& e) {
/* You can behave differently to specific errors */
}
);
database db(":memory:");
db << "create table person (id integer primary key not null, name text);";
try {
db << "insert into person (id, name) values (?,?)" << 1 << "jack";
// inserting again to produce error
db << "insert into person (id, name) values (?,?)" << 1 << "jack";
}
catch (sqlite_exception& e) {}
To extend SQLite with custom functions, you just implement them in C++:
database db(":memory:");
db.define("tgamma", [](double i) {return std::tgamma(i);});
db << "CREATE TABLE numbers (number INTEGER);";
for(auto i=0; i!=10; ++i)
db << "INSERT INTO numbers VALUES (?);" << i;
db << "SELECT number, tgamma(number+1) FROM numbers;" >> [](double number, double factorial) {
cout << number << "! = " << factorial << '\n';
};
Just Make sure you are using the full path of your database file :
sqlite::database db("/data/data/com.your.package/dbfile.db")
.
The usual way works for installing:
./configure && make && sudo make install
Note, there's nothing to make, so you there's no need to run configure and you can simply point your compiler at the hdr/ directory.
See breaking changes documented in each Release.
Pull requests are welcome 😉
- AUR Arch Linux
- maintainer Nissar Chababy
- Nuget (TODO nuget.org)
- Conan (TODO conan.io)
- vcpkg
MIT license - http://www.opensource.org/licenses/mit-license.php