VMS Help
SCA, SCA Topics, Glossary
*Conan The Librarian (sorry for the slow response - running on an old VAX)
|
|
A file produced by a compiler which contains information that
describes the source code to SCA. It may contain one or more
analysis data modules.
A module containing all the information used by SCA for one
compilation unit.
One of the attributes of an occurrence. It tells you whether the
name of the occurrence is visible or hidden. You can instruct SCA
to show only occurrences with a particular appearance by using the
"occurrence=" attribute selection. For example, you can ask for
hidden occurrences by specifying "occurrence=hidden."
|
4 - associated_declaration
|
Any declaration which is not a primary declaration. Typically,
associated declarations provide information needed by the compiler
to refer to an object.
A way of limiting a result of a query to those occurrences which
match certain characteristics. You can select the following
attributes: NAME, SYMBOL_CLASS, OCCURRENCE, DOMAIN, or FILE_SPEC
attributes. In the query language, you use phrases like name=foo,
symbol=argument, occurrence=declaration, domain=module_specific,
and file="foo.c" to specify which attributes you want. You combine
these phrases with AND, OR and NOT operators.
|
6 - attribute_selection_expression
|
A query question which combines one or more attribute selections
(such as name=foo, symbol=routine) using AND, OR, NOT, and XOR.
Some examples are:
name=foo and symbol=routine and occurrence=primary
name=foo_module and occurrence=compilation_unit
Shows what procedures and functions are called from a subroutine,
and all subsequent calls in the call graph.
See the SCA_Topics CALLED_BY help topic.
See the SCA_Topics CALLING help topic.
|
10 - compilation_line_number
|
A line number generated by the compiler which starts at 1 for
the first file used during the compile and goes up by one for
each line read. If there is an include file, the compilation line
number is increased by one for each line in the include file. By
default, the line numbers in the query display produced by the
FIND or INSPECT commands are compilation line numbers.
A compilation unit is the smallest piece of source code that
can be separately compiled. For example, in FORTRAN, this is
a subroutine or function; in C, this is a single file. Some
languages allow you to compile more than one compilation unit
at once. Even if you compile more than one unit at a time, SCA
considers the units to be separate.
A check that INSPECT makes in which all occurrences of a symbol
are checked for consistency. For example, you can ensure that
all calls to a routine have the correct number and type of
arguments. In this type of check, each occurrence is compared
with a single occurrence selected by INSPECT to be the standard
against which all occurrences are checked. In this type of check,
each particular occurrence is either correct or incorrect.
See the SCA_Topics CONTAINED_BY help topic.
See the SCA_Topics CONTAINING help topic.
Tells the compiler about an object before the compiler uses it.
This can make the compiler create the object, or it can tell the
compiler about an object that was created elsewhere. A declaration
has a position in the source code with both a start and an end,
and can contain other declarations or references. A declaration
can be either a primary declaration or an associated declaration.
Tells you what symbol class a declaration is (subroutine,
function, variable). For example, both procedures and functions
(in Pascal terms) belong to the routine symbol class, but their
declaration classes are different. The declaration class tells
the user whether a declaration is a primary declaration or
an associated declaration. SCA uses the declaration class to
decide what to display as the occurrence class in the result of a
FIND command. The user can find the declaration class using the
callable interface.
|
17 - diagnostic_error_messages
|
The error messages that INSPECT produces. The query result from
an INSPECT command is like that of a FIND command, but with error
messages attached to each occurrence.
A line in the display resulting from a FIND or INSPECT command.
SCA numbers these lines if you use /DISPLAY=NUMBER. This feature
is most useful when you are saving the output in a file.
See expression.
One of the attributes of an occurrence. The expression attribute
tells you whether the occurrence is one that you the user placed
in the source code, or is one which the compiler created for
the user. An occurrence can be either explicit or implicit.
Explicit occurrences are those placed by the user in the source
code; implicit occurrences are those created by the compiler on
behalf of the user. For example, in the following FORTRAN program,
there are three references to i which are explicit. There is one
implicit declaration of i which is created by the FORTRAN compiler
when it sees the first reference to i.
program test
i = 0
i = i + 1
end
A data type used by the SCA callable interface. If you are using
the callable interface, see the on-line SCA guide to
platform specific information for more information.
See both appearance and hidden modules.
A module in one of the libraries in the current library list which
is also present in a library which is earlier in the library list.
The module in the earlier library is visible. The same module in
any later library is a hidden module, hidden by the first module.
See expression.
Uses the current cursor position for the query, when used with
a FIND or GOTO DECLARATION. If the cursor is in a query display,
SCA uses whatever occurrence on which it is positioned. If the
cursor is in a buffer, SCA uses the name on which the cursor is
positioned and the file the cursor is in to guide the query. SCA
relaxes the search criteria, if necessary, (for example, allowing
different versions of the source file) to find the indicated
occurrence.
The operation performed by the AND operator, which indicates
that SCA will accept any occurrence that matches both X and Y
as follows:
FIND X AND Y
Generic term usually referring to a physical library.
A list of one or more physical libraries which compose the virtual
library. The order of the physical libraries is important. A
module found in one physical library in the library list hides
the same module in other physical libraries that come later in the
library list.
For example, suppose you have three libraries. Library 1 contains
module A. Library 2 contains modules A, B, and C. Library 3
contains modules C and D. SCA uses module A from Library 1,
modules B and C from library 2, and module D from library 3.
Refers to the position of a single physical library in a library
list. This can be used with any command that refers to a library
already in the library list. For example, the command SET
NOLIBRARY 1 removes the first library from the library list, and
the command LOAD/LIBRARY=2 FOO.ANA loads the file FOO.ANA into the
second library on the library list.
Represents a single compilation unit. You can list the modules
with the SHOW MODULE command. Each module shown represents a
single compilation unit - the smallest piece of source that can
be separately compiled. Even if several of these are compiled at
once (which is common in languages such as FORTRAN and BASIC),
each compilation unit appears separately.
In the query language, SYMBOL=MODULE specifies a certain type
of symbol. This is not the same as a compilation unit. You can
have modules which are not compilation units, and compilation
units which are not modules. How a module is defined varies from
language to language.
A string of characters used to identify symbols in your source
code. Legal characters for a name are defined by whatever language
you use. Each name has zero or more characters. Any character
may appear in a name. Special characters that appear in a name
must be quoted using double quotes. You do not need to quote the
following: $, _, *, %, &, -, alphanumeric characters.
|
32 - name_selection_expression
|
Selects occurrences with a particular name. For example, you can
specify the following:
FIND name=xyz
You can use wildcards in the name expression. Unlike other
atttributes, you can omit "name=" from the expression and only
specify the following:
FIND xyz
|
33 - nonstructured_relationship_expression
|
A relationship query (such as CONTAINED_BY, CALLED_BY, CALLING)
which requests that structure information be excluded. You can
specify RESULT=NOSTRUCTURE, RESULT=BEGIN, or RESULT=END as one
of the parameters to the relationship function. If you specify
RESULT=BEGIN or RESULT=END, this is displayed like any query
which does not have a relationship function. If you specify
RESULT=NOSTRUCTURE, this is displayed like any query which does
not have a relationship function, but has all the occurrences that
would be present if the structure were present.
A query result that contains occurrences, but does not contain any
relationships between occurrences. These are produced by queries
that do not involve relationship functions, or queries that do
involve relationship functions but specify RESULT=NOSTRUCTURE,
RESULT=BEGIN, or RESULT=END.
An occurrence is any instance of an entity in your program. An
entity can be any language construct, such as a variable, a
routine, or a constant. To further clarify this, consider the
following code fragment (not written in a particular language):
1 MODULE myprog;
2
3 VAR i,j;
4
5 ROUTINE foo()
6 BEGIN
7 i = 5;
8 j = i;
9 END;
10 END;
The code contains four entities: myprog, foo, i, j. There is one
occurrence each of the module myprog, and the routine foo. The
variable i, however, has three occurrences, and the variable j has
two.
A check that INSPECT performs on a single occurrence. For example,
INSPECT can check whether an occurrence is an implicit declaration
of a variable without having to look at any other occurrences.
One of the attributes of an occurrence that tells you what kind of
occurrence it is. The occurrence class indicates if an occurrence
is a declaration, a reference, or another class. It also indicates
what kind of declaration or reference it is. You can instruct SCA
to show you only occurrences with a particular occurrence class
using the OCCURRENCE= attribute selection. For example, you can
ask for write references using OCCURRENCE=WRITE.
|
38 - occurrence_selection_expression
|
The expression containing the occurrence class for each occurrence
you want to find.
A single directory containing an SCA database. The directory
should not contain anything else. SCA always locks an entire
physical library when it accesses it. When you are reading a
physical library (for example, with a FIND command), other users
are not allowed to write to the physical library, but other users
are allowed to read the physical library. When you are writing
to a physical library (for example, using LOAD), no other user is
allowed to read or write to the physical library.
Any declaration which affects how a particular object, such as a
routine or a variable, is implemented.
The first library in the library list. Commands which change the
SCA library, such as LOAD and REORGANIZE, apply to the first
library, unless you use the /LIBRARY qualifier to instruct SCA
to use another library.
The question you ask SCA together with the information you receive
from SCA. The question uses the SCA query language with the FIND
or INSPECT commands. The answer is called a query result. You use
the FIND or INSPECT commands to display the query result. You can
also use the query result as part of the question for a subsequent
query.
A query result is the information you get when you evaluate a
query, using the FIND or INSPECT commands. It consists of a set of
occurrences, and relationships between occurrences.
The use of some object in the source code. For example: X = X + 1
In this example, there are two references to the variable X. One
(to the left of the =) is a write reference; the other is a read
reference. A reference has a position in the source code, but it
is a single point and cannot contain anything.
An association between two different occurrences. For example, the
CALLED_BY relationship in SCA associates a primary declaration of
a procedure with call references to other procedures and functions
in that procedure. A relationship has both a source and a target.
A relationship also has a relationship type. Relationships go in
both directions. For example, the CALLED_BY relationship is the
inverse of the CALLING relationship.
The kind of relationship between two occurrences. For example, a
CALLED_BY relationship between a declaration of a routine FOO and
a reference to a routine BAR shows that routine FOO calls routine
BAR. The relationship types that are valid for SCA are: CALLED_BY,
CALLING, CONTAINED_BY, CONTAINING, TYPED_BY, and TYPING.
The occurrences that result from each query.
The analysis of a software system performed by looking at the
source code. This is in contrast to dynamic analysis, which is
analysis of the software while it is running.
SCA has some commands which do static analysis. These commands are
the INSPECT command, which does consistency checking, the REPORT
command, that can generate reports about modules, imported and
exported objects, and so forth, and some uses of the FIND command,
to generate call graphs and type graphs.
|
49 - structured_relationship_expression
|
A query that uses a relationship function (such as CONTAINED_
BY, CALLED_BY, or TYPED) which asks for structure information.
Structure information shows the relationships between occurrences
found as well as the occurrences found. Structure information is
provided by default.
A query result which has both occurrences and relationships
between occurrences. These are produced by queries which involve
relationship functions.
Any object in a program. For example, a FUNCTION, a VARIABLE, a
CONSTANT, or any of the entities with which a programmer typically
deals.
A symbol has occurrences. For example, the declaration of
a variable is an occurrence, and uses of the variable are
occurrences. SCA determines which occurrences belong to which
symbols using the rules of the language you are using. For
example, you may have two different variables named INDEX in
separate subroutines. According to the rules of your language,
these are usually different variables, so they are different
symbols for SCA.
It does not matter whether all occurrences of a symbol are in
a single compilation unit, or spread over several compilation
units. All the occurrences still belong to the same symbol. For
example, you may have a subroutine SUB1 in one module, and calls
to that subroutine in several other modules. These all appear as
occurrences of the same symbol, named SUB1.
The programmer and SCA should have the same definition of what
constitutes a unique item. SCA's term for a unique item is symbol.
A check that INSPECT performs on all occurrences of a symbol.
For example, INSPECT can ensure that there are both read and
write references to a variable. In this type of check, no single
occurrence is either correct or incorrect. If there are problems,
the problems are with the symbol as a whole.
An attribute selection that identifies the type of symbol. Tells
you whether the symbol is a variable, constant, or some other
class. You can use the FIND command to find only symbols with
a particular symbol class. For example, you can specify "FIND
symbol=argument." You can abbreviate both "symbol" and "argument".
|
54 - symbol_class_selection_expression
|
The expression containing the symbol class for each symbol you
want to find.
A set of occurrences and relationships that describes a complex
data type. For example, a declaration of a record consists of a
record and some record components. Each record component has a
type, which may be another record, a pointer to the same record,
a basic data type such as integer, and so forth. In SCA, the type
graph connects all these together, with relationships connecting
the record to its components and the components to their types.
See the SCA_Topics TYPED_BY help topic.
See the SCA_Topics TYPING help topic.
The operation performed by the OR operator, which indicates that
SCA will accept any occurrence that matches either X or Y as
follows:
FIND X OR Y
A library that allows you to split your SCA library into pieces.
Each piece is called a physical library. SCA works the same way
whether your virtual library has only one physical library or
several physical libraries.
Some of the reasons for using more than one physical library are
as follows:
o Physical libraries can be placed on more than one disk, if
there is not enough room on a single disk.
o Analysis date files can be loaded into more than one physical
library at the same time, to make LOAD run faster.
o A small personal physical library can be used to keep track of
your personal changes without affecting other users.
o A separate physical library can be used for each major
component in your system.
A single virtual library is a list of one or more physical
libraries. The order is important. A module in one physical
library hides the same module in physical libraries later on in
the list. This list of physical libraries is called a library
list.
See both appearance and visible modules.
Modules that SCA can examine when performing a FIND or INSPECT
command. The current library list tells you what modules are
visible. All modules in the first library in the library list
are visible. Modules in the second library which are not in the
first library are visible. Modules in the third library which are
not in the first or second libraries are visible. Any module which
is not visible is hidden.
Wildcards are used to match more than one name at once. There
are two wildcards in SCA: the asterisk (*) and percent (%). For
example, in the name expression A*, the wildcard * will match any
number of characters, so this would match A, AB, ABC, AXYZ, and
so forth. The * wildcard matches any number of characters, and %
wildcard matches just one character. If you do not want SCA to use
* or % as a wildcard, you use an ampersand (&) to quote it. For
example, to find the single name *, you would use &*.
[legal]
[privacy]
[GNU]
[policy]
[netiquette]
[sponsors]
[FAQ]
Polarhome, production since 1999.
Member of Polarhome portal.