Hi there! You may be interested in a sister project, IonDB - A key-value store for embedded devices.
LittleD is a relational database for microprocessor devices, using as little as 1kB of RAM for most queries. LittleD supports SELECT-FROM-WHERE syntax, including inner joins, projections, and selections over arbitrary expressions. LittleD also has support for CREATE TABLE and INSERT statements. Only integers and fixed-width strings are supported at this time. In the future, query pre-compilation, network transmission, network relational operators, aggregation, and improved indexing are hoped to be implemented.
If you are interested in the design of LittleD, you might want to check out my published research: here or here.
Sure!
#include "Littled/dbparser/dbparser.h"
#define BYTES_LEN 400
int main(void)
{
char memseg[BYTES_LEN];
db_query_mm_t mm;
db_op_base_t* root;
db_tuple_t tuple;
init_query_mm(&mm, memseg, BYTES_LEN);
parse("CREATE TABLE sensors (id int, temp int);", &mm);
init_query_mm(&mm, memseg, BYTES_LEN);
parse("INSERT INTO sensors VALUES (1, 221);", &mm);
init_query_mm(&mm, memseg, BYTES_LEN);
parse("INSERT INTO sensors VALUES (2, 89884);", &mm);
init_query_mm(&mm, memseg, BYTES_LEN);
parse("INSERT INTO sensors VALUES (3, 112);", &mm);
init_query_mm(&mm, memseg, BYTES_LEN);
parse("INSERT INTO sensors VALUES (4, 455);", &mm);
init_query_mm(&mm, memseg, BYTES_LEN);
parse("INSERT INTO sensors VALUES (5, 3313);", &mm);
init_query_mm(&mm, memseg, BYTES_LEN);
parse("INSERT INTO sensors VALUES (6, 11);", &mm);
init_query_mm(&mm, memseg, BYTES_LEN);
parse("INSERT INTO sensors VALUES (7, 99996);", &mm);
init_query_mm(&mm, memseg, BYTES_LEN);
parse("INSERT INTO sensors VALUES (8, 6565);", &mm);
init_query_mm(&mm, memseg, BYTES_LEN);
parse("INSERT INTO sensors VALUES (9, 6565);", &mm);
init_query_mm(&mm, memseg, BYTES_LEN);
root = parse("SELECT * FROM sensors;", &mm);
if (root == NULL)
{
printf("NULL root\n");
}
else
{
init_tuple(&tuple, root->header->tuple_size, root->header->num_attr, &mm);
while (next(root, &tuple, &mm) == 1)
{
int id = getintbyname(&tuple, "id", root->header);
int sensor_val = getintbyname(&tuple, "temp", root->header);;
printf("sensor val: %i (%i)\n", sensor_val, id);
}
}
return 0;
}
LittleD provides a SQL frontend to manage data under the relational model. This project targets devices of about the same capabilities as the Arduino Mega2560. In a nutshell, the system supports the creation of tables, data insertion, and SELECT-FROM-WHERE syntax. Error messages are supported but not mandatory. A simple configuration header file dictates which features are compiled in so the database's codespace footprint can be controlled according to application needs. All code except that needed to interface with Arduinos is written in C.
This project is part of ongoing research being conducted at the University of British Columbia's Okanagan Campus under the direction of Dr. Ramon Lawrence. The work has been supported by Dr. Lawrence's Distributed Data Lab, the Irving K. Barber School of Arts and Sciences, the Natural Sciences and Engineering Research Council of Canada, along with a list of people too numerous to name.
To simply build the database locally, in your shell of choice navigate to the repository root and execute the command
make
To build the unit tests, issue the command
make tests
and navigate to the bin/tests directory within the repository's root directory. You may run all tests by
./runalltests
Single tests are included in the same directory. The documentation can be generated using
make docs
and reside in the doc/doxygen directory. Note that doxygen must be installed for the documentation to be generated. Finally,
make clean
deletes all files created at compile time.
LittleD can be managed using an Eclipse Makefile project. Either the packaged C/C++ Eclipse IDE or the installable CDT package must be used.
First make sure that the repository is cloned. In Eclipse click File -> Import.... Within the Select window, choose under C/C++ the Existing Code as Makefile Project option and click Next. Give your Eclipse Project a name and browse to the root of the project repository to use the code. Choose the correct compiler for your system and then click Finish.
As noted above, there are a small number of make targets that can be used with the project's Makefile. By right clicking the project in the Package Explorer and selecting Make Targets -> Create.... Leaving all options at the their default, enter the name of the target to add (for instance, tests) and then click OK. You can then run the Make target by right clicking on the project in the Project Explorer and choosing Make Targets -> Build.... Select the target from the list that appears and click Build. Finally, you can run individual files with main methods by opening them and click the green Run button at the top.
Absolutely! Contributions are welcome! Please make pull requests off the master branch.