History files
The history files are introduced according to the SEP
(Stanford Exploration Project) history files, although there
are some differences, especially in running the programs.
The history files are designed to contain both the data
and the information how to execute the programs.
The following items are recognized in the history files:
- #Comments:
- Everything from a # (hash mark) to the end of line is assumed
to be a comment and is ignored. If a line contains a hash mark,
its content is thus interpreted just between the
beginning of the line and the hash mark.
- =Data:
- Each = (equal sign) immediately preceded by a string is interpreted
as the PARAMETER=VALUE couple specifying the value of the parameter
named PARAMETER. For example, N1=27 means that the value of the
parameter named N1 is 27. This convention is identical
to the SEP (Stanford Exploration Project) parameter files.
For more details refer to Fortran file
sep.for.
- :Programs:
- Each : (colon) immediately preceded by a string is interpreted
as the instruction to execute a program. The string between the
last space (or the beginning of the line if there are no spaces
to the left of :) and the colon is the name of the program.
If the name contains substring ".pl", it is assumed to represent
a Perl script. Otherwise, it is assumed to represent executable
program, compiled from the Fortran source.
Everything to the right from the colon is assumed to represent
the DATA for the program. This convention differs from
the SEP (Stanford Exploration Project) history files.
If the DATA is blank, it defaults to the name of the corresponding
history file in apostrophes, followed by a space and a slash.
This default is useful for Fortran programs, not for Perl scripts.
Data for Perl scripts are assumed to be read from the command line.
The perl scripts are thus executed by a command like
perl program.pl DATA
Data for executable programs are assumed to be read from the
console, within a single READ statement.
The executable programs are thus executed by a command like
program
and the DATA are submitted from the standard input.
No other items are interpreted in this version.
All text containing neither equals nor colons has thus the effect
of comments.
If the DATA for a program (i.e., string right to the colon)
contain a substring identical to the name of the history file,
the program is assumed to read the history file.
In such a case, only the PARAMETER=VALUE couples between the
beginning of the history file and the "colon" instruction to execute
the program are assumed to apply to the program.
If there is more PARAMETER=VALUE couples for the same parameter
in the history file, the last couple before the "colon" of the currently
executed program takes preference.
This convention enables to modify the data for the program
by entering additional PARAMETER=VALUE couples and additional
"colon" instructions to execute the programs within the same
history file.
When executing the history file, it is copied line by line
to the output history file. When a "colon" instruction to execute
the program is encountered, the program is executed with the
name of the output history file substituted for the name of
the input history file in order to hide the PARAMETER=VALUE couples
between the "colon" instruction and end of file to the program.
The history files may be executed by Perl script
go.pl. For example, history file
'file.h' may be executed by command
perl go.pl file.h
generating the output history file named 'file.out'.
If you wish to name the output history file, e.g., 'new.out',
the history file may be executed by command
perl go.pl file.h new.out
Perl interpreter should be available at
www.perl.org
Example of the history file: corfun.h.
Examples of papers devoted to the description of history files
(ordered chronologically):
- Klimes, L. (1998):
Comparison of ray-matrix and finite-difference methods
in a simple 1-D model.
In: Seismic Waves in Complex 3-D Structures, Report 7, pp. 169-180,
Dep. Geophys., Charles Univ., Prague.
- Bulant, P. & Klimes, L. (1998):
Computations in the model composed during the 1998 consortium meeting.
In: Seismic Waves in Complex 3-D Structures, Report 7, pp. 33-55,
Dep. Geophys., Charles Univ., Prague.
- Bucha, V. & Klimes, L. (1999):
Finite differences above the MODEL package.
In: Seismic Waves in Complex 3-D Structures, Report 8, pp. 171-192,
Dep. Geophys., Charles Univ., Prague.
- Klimes, L.(2000):
Smoothing the Marmousi model for Gaussian-packet migrations.
In: Seismic Waves in Complex 3-D Structures, Report 10, pp. 63-74,
Dep. Geophys., Charles Univ., Prague.
- Bucha, V. (2001):
Displaying 3-D seismic models through the VRML and GOCAD.
In: Seismic Waves in Complex 3-D Structures, Report 11, pp. 337-355,
Dep. Geophys., Charles Univ., Prague.
Alphabetical list of input parameters of all programs.