Overview of Database Systems (third edition)
A basic textbook of the database, published by the higher Education press. In order to deepen the basic knowledge of the database
Understand, look over again, to further grasp the knowledge of the database to lay the groundwork for learning.
The book is divided into four parts: Foundation, design, System, new technology. According to the order in the book, the following records the Reading
's Notes:
First Basic article
1: Introduction
1.1: Overview
The composition of database system: Database, Database management system, application system, database administrator and user.
Development of Database technology: Manual Management--file system--database system.
Database System features: To achieve the overall data structure, data sharing, low redundancy, easy expansion, data physics
Independent and data logic independent, data by the DBMS unified management.
1.2: Model
Data Model: By data structure (static), retrieval/update (dynamic), constraint conditions. The
main data models are: Hierarchical model, mesh model, relational model, object-oriented model.
Conceptual Model: By entity, attribute, code, domain, entity, entity set, contact composition, the method is E-R diagram.
1.3: Structure
Level three mode: mode (public Data View for all users), external mode (data user view), internal mode (storage mode
). The
three-level model constitutes two mappings to guarantee logical independence and physical independence.
External mode/pattern mapping: Keeping the external mode unchanged makes the logic independence of data program, called the logical independence of data.
Mode/internal mode image: Keep the mode unchanged, so that the external mode depends on the program is not changed, called the physical independence of data.
1.4: Compose
hardware platform and database, software, personnel.
personnel include: DBA (Database administrator), System Analyst/Database Designer, application programmer, user.
1.5: Research
Research area is the database management system software, database design, database theory.
2: Relational database
2.1: Overview:
relationship refers to a two-dimensional table, relational operations are set operations, the need to meet entity integrity, Referential integrity, user-defined
integrity.
2.2: Definition:
Relationships can have three types: Basic table, query table, view chart.
R (U, D, Dom, F): u--Property name collection d--properties corresponding to the domain dom--property to the domain image
f--property to the data dependency collection.
2.3: Integrity
Entity Integrity-The primary code in the table guarantees the uniqueness of the row data.
Referential Integrity--the main code and the Outer Code Association.
user-defined integrity-user-defined semantic conditions.
2.4: Relational algebra
traditional set operations are: Sum, difference, intersection, and generalized Cartesian product.
relational operations: selecting, projecting, connecting, excluding.
2.5: Relational calculus
tuple relational calculus language: ALPHA.
Domain Relational Calculus language: QBE.
3:sql
SQL too familiar, not much to say, just look at embedded SQL:
3.1:sql Communication with the main language
SQL Communications Area: Stores execution state through SQLCA data structures.
Master variables: parameters are passed by program variables.
cursors: Used to store multiple records, in combination with both of these methods. The
cursors are primarily defined by declare, open, fetch read, close closed. You can also use current to
Update or delete.
3.2: Dynamic SQL
Dynamic SQL here more refers to the knowledge of programming, not to SQL itself.
4: Relational Systems and query optimization
4.1: Definition of a relational system
non-demanding relational systems must support a complete relational model, The minimum definition is:
support relational database (table structure), support selection, projection, and connection.
Classification: Form system, minimum relational system, complete relation system, whole relation system. The 12 principles of the
total relationship system, slightly.
4.2: Query optimization
optimization is generally divided into four steps:
A: Turn the query into a syntax tree.
B: Use the equivalence principle to change the syntax tree to a standard form.
C: Select the optimal algorithm.
D: Generate a final query plan.
General Merchandise adoption of cost-based optimization algorithms:
Total Cost = I/o cost + CPU cost + memory cost
4.3: Principles of Optimization
A: Select operations first.
B: Preprocessing connections: such as sorting.
C: Projections and selections can be made at the same time.
D: Combines projection and binocular operations to reduce scanning.
E: Combine the selection with the Cartesian product as a concatenation operation.
F: Find common subexpression.
4.4: Use the equivalence change rule optimization of relational algebra.
5: Relational data theory
This section is more or less complex, but the most important is the paradigm definition:
First paradigm: relationships, As a two-dimensional table, each component must not be divided.
2NF: On a 1NF basis, a non-primary property full function relies on code.
3NF: Each non-primary attribute relies on code and does not pass on code.
BCNF: All primary attributes are fully functional dependent on each code that does not contain it, and no property full function relies on
any set of properties that are not code. Alas, it's too complicated.
mode decomposition: not clear.