Glossary

ABI: Application binary interface. The link between the MemCom library and the operating system.
API: Application programming interface. A collection of functions for interfacing MemCom with a specific programming language, such as C, C++, Fortran, Fortran95, Python.
Application: A program which makes use of the MemCom data manager.
ASCII Code: The American Standard Code for Information Interchange.
Database: A collection of structured data.
Database File: A collection of files stored under a directory, describing datasets.
Dataset: An entity of data recognized by the MemCom system by its name, dataset number, length etc.
Dataset Descriptor: A relational table attached to the dataset, storing user-defined attributes of a dataset.
Data Type: The internal structure of one data element or sequence of data elements recognized by MemCom.
Descriptor: See Dataset Descriptor.
Directory: A special file residing in a file system describing files and directories.
Dynamic Memory Manager: A system for dynamic run-time allocation of memory and in-core management of datasets.
File: An array of bytes ("byte stream") residing in a file system.
File offset: The byte address, counted from 0, pointing to data in a file.
Handle: Identifier of a database.
In-core Data Management: A mechanism that organizes MemCom datasets in the (virtual) memory. The Dynamic Memory Manager (DMM) manages datasets in-code.
Index: A table which describes MemCom datasets. It contains all information necessary for storing and retrieving a dataset. The index is transparent to the application
Label or Dataset Name: A character string that identifies a MemCom dataset.
Monitor: A program for interactively examining and editing of MemCom databases. Note that the Monitor is now replaced by the MemCom Python module.
Paged i/o: Data are exchanged in paged mode between the user or system buffers and the database.
Offset: Files: Byte of word (integer*4) offset with respect to the beginning of a file. Dataset: Offset counted in number of logical elements with respect to the beginning of the dataset (Fortran: Offsets start with 1, C: Offset starts with 0).
Record: A sequence of data, either (i) a MemCom dataset or subset or (ii) a record in the sense of operating system.
Relational Table: Database data type in the form of an associative array, also referred to as dictionary (Python) or map (C++).
Sparse Table: Database data type similar to the relational table.
Stream i/o: Data is transferred character-wise (byte-wise) directly to and from the data files. Stream input/output is not buffered within MemCom.
Sub-division number: The sub-set number in a dataset containing sub-sets. Note that sub-division numbers are contiguously numbered from 1 to the number of sub-sets ('rows') in a dataset.
Sub-set: A record in a dataset, all records being of the same size. A sub-set may be accessed directly without loading the whole dataset.
Word: The smallest storage unit of MemCom datasets (before version 7). The physical size of a word is 4 bytes. Obsolete from version 7+, the smallest storage unit now being a byte.
Word-addressable Mode: Data or MemCom data files are stored and retrieved by using the disk addresses. Obsolete from version 7+.