#gStore

gStore2.0

Overview

gStore is a graph-based RDF data management system (or what is commonly called a “triple store”) that maintains the graph structure of the original RDF data. Its data model is a labeled, directed multiedge graph, where each vertex corresponds to a subject or an object. We also represent a given SPARQL query by a query graph Q. Query processing involves finding subgraph matches of Q over the RDF graph G. gStore incorporates an index over the RDF graph (called VS-tree) to speed up query processing. VS-tree is a heightbalanced tree with a number of associated pruning techniques to speed up subgraph matching.

Install Steps

System Requirement: 64-bit linux server with GCC, make, readline installed. We have tested on linux server with CentOS 6.2 x86_64 and CentOS 6.6 x86_64. The version of GCC should be 4.4.7 or later.

You can install gStore2.0 in two commands.

1. run

# make lib_antlr

to compile and link the library we need.

2. run

# make

to compile the gStore code and build executable "gload", "gquery", "gserver", "gclient".

Usage

gStore2.0 currently includes four executables.

####1. gload gload is used to build a new database from a RDF triple format file.

# ./gload db_name rdf_triple_file_name

For example, we build a database from LUBM_10.n3 which can be found in example folder.

[root@master Gstore]# ./gload db_LUBM10 ./example/LUBM_10.n3 
2015年05月21日 星期四 20时58分21秒  -0.484698 seconds
gload...
argc: 3 DB_store:db_LUBM10      RDF_data: ./example/LUBM_10.n3  
begin encode RDF from : ./example/LUBM_10.n3 ...

####2. gquery gquery is used to query an exsisting database with SPARQL files.

./gquery db_name (use ./gquery --help for detail)

The program shows a command prompt("gsql>"): you can type in a command here, use help to see information of all commands.

For sparql command, input a file name which can be interpreted as a single SPARQL query.

When the program finish answering the query, it shows the command prompt again.

gStore2.0 only support simple “select” queries now.

We also take lubm_10.nt as an example.

[root@master Gstore]# ./gquery lubm_10.nt/
gquery...
argc: 2 DB_store:lubm_10.nt/
loadTree...
LRUCache initial...
LRUCache initial finish
finish loadCache
finish loadEntityID2FileLineMap
open KVstore
finish load
finish loading
Type `help` for information of all commands
gsql>sparql /media/wip/common/data/LUBM/query/q0.sql
... ...
Total time used: 4ms.
final result is : 
<http://www.Department0.University0.edu/FullProfessor0>
<http://www.Department1.University0.edu/FullProfessor0>
<http://www.Department2.University0.edu/FullProfessor0>
<http://www.Department3.University0.edu/FullProfessor0>
<http://www.Department4.University0.edu/FullProfessor0>
<http://www.Department5.University0.edu/FullProfessor0>
<http://www.Department6.University0.edu/FullProfessor0>
<http://www.Department7.University0.edu/FullProfessor0>
<http://www.Department8.University0.edu/FullProfessor0>
<http://www.Department9.University0.edu/FullProfessor0>
<http://www.Department10.University0.edu/FullProfessor0>
<http://www.Department11.University0.edu/FullProfessor0>
<http://www.Department12.University0.edu/FullProfessor0>
<http://www.Department13.University0.edu/FullProfessor0>
<http://www.Department14.University0.edu/FullProfessor0>

Notice:

"[empty result]" will be printed if no answer, and there is an empty line after all results.

Readline lib is used, so you can use arrow key to see command history, and use arrow key to move and modify your entire command.

What is more, path completion is supported for utility. (not built-in command completion)

####3. gserver gserver is a daemon. It should be launched first when accessing gStore by gclient or API. It communicates with client through socket.

[root@master Gstore]# ./gserver 
port=3305
Wait for input...

You can also assign a custom port for listening.

[root@master Gstore]# ./gserver 3307
port=3307
Wait for input...

####4. gclient gclient is designed as a client to send commands and receive feedbacks.

[root@master Gstore]# ./gclient 
ip=127.0.0.1 port=3305
->

You can also assign gserver's ip and port.

[root@centos74 Gstore]# ./gclient 172.31.19.15 3307
ip=172.31.19.15 port=3307
->

We can use these following commands now:

->import db_name rdf_triple_file_name; build a database from RDF triple file.

->load db_name; load an exsisting database.

->unload db_name; unload database, but will not delete it on disk, you can load it next time.

->query SPARQL; query the current database with a SPARQL query.

->show databases; show the current database's name.

->insert db_name rdf_triple_file_name; insert a handful of new RDF triples to the orginal database.

Note that each command ends with ";".

####5. API We provide JAVA and C++ API for accessing gStore now. Please refer to example codes in api/cpp/example and api/java/example.

####6. gtest We provide a test tool to generate structural logs for datasets. Please type ./gtest --help in the working directory for detail.

(change paths in the main/gtest.cpp if needed)

You should place the datasets and queries this way:

DIR/WatDiv/database/*.nt

DIR/WatDiv/query/*.sql

And the output will be sorted into there logs: load.log/(for database loading time and size), time.log/(for query time) and result.log/(for all query results).

All logs produced by this program are in TSV format, you can load them into Calc/Excel directly.

System Update Log

####Oct 20, 2015.

We add a gtest tool for utility, you can use it to query several datasets with their own queries.

In addition, gquery console is improved. Readline lib is used for input instead of fgets, and the gquery console can support commands history, modifying command and commands completion now.

What is more, we found and fix a bug in Database/(a pointer for debugging log is not set to NULL after fclose operation, so if you close one databse and open another, the system will fail entirely because the system think that the debugging log is still open)

####Sep 25, 2015.

We implement the version of BTree, and replace the old one.

After testing on DBpedia, LUBM, and WatDiv benchmark, we conclude that the new BTree performs more efficient than the old version. For the same triple file, the new version spends shorter time on executing gload command.

Besides, the new version can handle the long literal objects ocasion efficiently, while triples whose object's length exceeds 4096 bytes result in frequent inefficent split operations on the old version BTree.

####Feb 2, 2015

We modify the RDF parser and SPARQL parser.

Under the new RDF parser, we also redesign the encode strategy, which reduces RDF file scanning times.

Now we can parse the standard SPARQL v1.1 grammar correctly, and can support basic graph pattern(BGP) SPARQL queries written by this standard grammar.

####Dec 11, 2014.

We add API for C/CPP and JAVA.

####Nov 20, 2014.

We share our gStore2.0 code as an open-source project under BSD license on github.