Graph Query Language (GQL, graphical Query Language) is a standard SQL while maintaining the International Working Group on the development and maintenance of a new language.
GQL owes much to the existing language, mainly inspired by Cypher (version now realize there are more than 10, including six commercial products), Oracle and SQL PGQL of itself. GQL project is the first since the SQL ISO / IEC international standard database language project.
In June this year, under the ISO / IEC Joint Technical Committee and many national standards bodies worldwide (in charge of customized IT standards) 1 GQL begin to vote on project proposals, seven countries have sent experts to participate in the four-year project. At the end of this week to vote, the proposal was adopted.
A total of ten countries cast in favor, including China, South Korea, Sweden, the United States, Germany, Britain, the Netherlands, Denmark, Kazakhstan, Canada and Finland. Another five countries choose to abstain, on the grounds that the lack of expertise to judge the proposal or comment.
Only Japan voted against it cited two reasons:
- Existing language has been able to achieve this kind of demand
- Specifically, to meet the rest of the SQL / Property Graph Query expansion of demand as well as the SQL standard
Why do we need a specific graphical query language?
Many suppliers, researchers and users agree that you can use non-relational "Graphic native" store and runtime model to develop graphical database. For example Neo4j industry-leading graphics and database platform new Redis Labs graphics products.
However, they also need a language similar to Cypher support the insertion and maintenance of data, not just data query. For the language to map the behavior of the center of it, SQL is unlikely to be a suitable model the desired language is capable of graphics as a query input, and outputs a chart, like SQL can read the table, and generate new reality table of results.
How SQL and GQL work?
Many proposals GQL support companies and national standards bodies do not think GQL and SQL are competitors, but by sharing basis and interoperability to complement each other. (Which refers to the formation of the core data type and manner of expression, and the concept of shared objects in the directory holding pattern as well as associated with the user / role session).
SQL / PgQ query is actually around a period of "proto-gql" son of SQL queries. Here is a sample query, when the General Assembly this year SIGMOD near the end, Oracle's database link Oskarvanrest as a reference Council, held in Amsterdam in July of this year (Link Database Benchmark Council) raised during the meeting.
MATCH keyword red part is the beginning of a fragment of pattern matching queries, the query is very similar written or PGQL Cypher query. You may notice that it is for the introduction of labels (such as the Creator IS Person in) the introduction of a host parameters or variables, as well as for. However, you can also use a colon in the label expression (if SQL parser engine is intelligently), and then the pre-existing "type" attribute maps query language similarity will be more obvious.
PgQ other parts of the query (black and green) to the Proto-GQL connect to a SQL SELECT statement. Columns clause flows through the tabular results of conventional SQL query. They focus only on SQL queries to interact with the graphics engine, GQL itself does not relate to this SQL "External Function Interface."
Generating SQL table, GQL FIG generate
SQL is a key aspect in a very different language and cypher. Cypher allows users to explore the structure of their data in the chart do not know what type of data will be returned in the case. It allows you to be a true graphical query, which concern not only value, but also includes the value aspects of shape, and match graphics mode defined subset of data elements. In other words, FIG query or subgraph FIG projected on one or more inputs for the calculation of FIG.
GQL will build openCypher Morpheus based on (it Cypher into the Apache Spark), in combination with G-CORE inspiration from LDBC, in order to provide users with a combination of map query language to support all those features that will make GQL conceptually equivalent to SQL.
More generally, GQL is a more modern language than SQL, it has a more structured type of system.
GQL project work will be SQL / GQL Standards Committee held later this month in Arusha, Tanzania beginning next meeting of the full 1 SC 32 / WG3 of ISO / IEC JTC on.
Still unable to determine the first version GQL can be achieved, but it's possible to develop a fairly complete draft before the second half of 2020.
Source: linkedin