Sophie: ftnchek-3.1.2-2mdk i586

ftnchek-3.1.2-2mdk.i586.rpm

<HTML>
<HEAD>
<TITLE>Option: style</TITLE>
</HEAD>
<BODY bgcolor=white>
<A HREF="toc.html">Table of Contents</A><P>
<P>Previous: <A HREF="source.html">-source</A><HR><P>
<H2>Option: <font color=#FF0080>style</font></H2><P>
<A NAME="style"></A><DT><B><A HREF="style.html">-style</A>=</B><I>list</I>

<DD>
<DL COMPACT><DT><DD>
Provides extra-picky warnings about obsolescent or old-fashioned
programming constructions.  This option is helpful for efforts to follow a
modern programming style.  (Most of the things complained about under
this option are forbidden in the <B>F</B> subset language.)  By
default, all warnings are turned off.
<P>

The <I>list</I> consists of keywords
separated by commas or colons.
There are three special keywords: <B>all</B> to turn on all the
options, <B>none</B> to turn them all
off, and <B>help</B> to print the list of all the keywords with a brief
explanation of each.
If <I>list</I> is omitted, <B><A HREF="style.html">-style</A></B> is equivalent to
<B><A HREF="style.html">-style</A>=all</B>, and <B>-nostyle</B> is equivalent to
<B><A HREF="style.html">-style</A>=none</B>.
The warning keywords with their meanings are as follows:
<DL COMPACT>
<DT><B>block-if</B>:

<DD>
Complain about arithmetic IF statement.  Accept block IF or logical IF
(which controls a single statement).
<DT><B>construct-name</B>:

<DD>
Complain about unnamed block constructs: IF,
DO, and SELECT CASE.  Note that if a construct name is present on the
opening statement of a construct, then it is required to be present on
all other component statements (ELSE, END IF, etc.) of the construct.
In that case a missing construct name on those statements generates a
syntax error regardless of this option.  The purpose of this option is
to warn if the construct completely lacks the optional name.
<DT><B>distinct-do</B>:

<DD>
Complain if two DO loops share a common terminator statement.
<DT><B>do-construct</B>:

<DD>
Complain if terminator of a DO loop is anything other than an END DO
or CONTINUE statement.  This is the requirement in order for the loop to
meet the Fortran 90 definition of a do-construct.
<DT><B>do-enddo</B>:

<DD>
Complain if terminator of a DO loop is anything other than an END DO
statement.  (This option overrides the <B>do-construct</B> option,
being even stricter.)
<DT><B>end-name</B>:

<DD>
Complain about the absence of the subprogram name on structured END
statements.
<DT><B>format-stmt</B>:

<DD>
Complain about the presence of FORMAT statements.  Only the FORMAT
statements themselves are flagged, not the references to them in I/O lists.
<DT><B>goto</B>:

<DD>
Complain about the presence of unconditional, computed or assigned
GOTO statements.  Also complain about alternate returns (but not about
labels as subprogram arguments).
<DT><B>labeled-stmt</B>:

<DD>
Complain about the presence of labels (numbers) on statements other
than FORMAT statements.  (Since FORMAT statements are arguably
convenient and not readily abused, complaints about them are
controlled by the separate <B>format-stmt</B> keyword.)
<DT><B>program-stmt</B>:

<DD>
Complain about the absence of a PROGRAM statement at the head of the
main program.
<DT><B>structured-end</B>:

<DD>
Complain about the use of a plain END statement to end a subprogram,
rather than a structured END statement (END PROGRAM, END SUBROUTINE,
END FUNCTION, or END BLOCK DATA).
</DL>
<P>

See also: <B><A HREF="f77.html">-f77</A></B>, <B><A HREF="f90.html">-f90</A></B>, <B><A HREF="f95.html">-f95</A></B>, <B><A HREF="pretty.html">-pretty</A></B>,
<B><A HREF="portability.html">-portability</A></B>.
</DL>

<P>
<P><HR><P>Next: <A HREF="symtab.html">-symtab</A>
</BODY></HTML>