Hello community,
here is the log from the commit of package octave-forge-io for openSUSE:Factory
checked in at 2016-07-09 09:25:10
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Comparing /work/SRC/openSUSE:Factory/octave-forge-io (Old)
and /work/SRC/openSUSE:Factory/.octave-forge-io.new (New)
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Package is "octave-forge-io"
Changes:
--------
--- /work/SRC/openSUSE:Factory/octave-forge-io/octave-forge-io.changes
2016-06-14 23:07:36.000000000 +0200
+++ /work/SRC/openSUSE:Factory/.octave-forge-io.new/octave-forge-io.changes
2016-07-09 09:25:28.000000000 +0200
@@ -1,0 +2,9 @@
+Fri Jul 8 10:02:20 UTC 2016 - dmitr...@opensuse.org
+
+- Update to version 2.4.2
+ * Changed several calls to (now deprecated) octave_config_info into
+ calls to __have_feature__.m
+ * Add support for complex numbers in read_namelist.m and write_namelist.m
+ * Bug fixes
+
+-------------------------------------------------------------------
Old:
----
io-2.4.1.tar.gz
New:
----
io-2.4.2.tar.gz
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Other differences:
------------------
++++++ octave-forge-io.spec ++++++
--- /var/tmp/diff_new_pack.dZrQra/_old 2016-07-09 09:25:29.000000000 +0200
+++ /var/tmp/diff_new_pack.dZrQra/_new 2016-07-09 09:25:29.000000000 +0200
@@ -18,7 +18,7 @@
%define octpkg io
Name: octave-forge-%{octpkg}
-Version: 2.4.1
+Version: 2.4.2
Release: 0
Summary: Input/Output in external formats from Octave
License: GPL-3.0+ and BSD-2-Clause
++++++ io-2.4.1.tar.gz -> io-2.4.2.tar.gz ++++++
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/io/DESCRIPTION new/io/DESCRIPTION
--- old/io/DESCRIPTION 2016-03-09 22:29:45.000000000 +0100
+++ new/io/DESCRIPTION 2016-07-04 21:44:38.000000000 +0200
@@ -1,13 +1,13 @@
Name: io
-Version: 2.4.1
-Date: 2015-03-09
+Version: 2.4.2
+Date: 2016-07-04
Author: various authors
Maintainer: Philip Nienhuis <prnienh...@users.sf.net>
Title: Input/Output
Description: Input/Output in external formats.
Categories: IO
Problems: Default initial Java memory probably too small, increase with
java.opts (see documentation). UNO support experimental.
-Depends: octave (>= 3.8.0), Octave (< 4.2.0)
+Depends: octave (>= 3.8.0), Octave (< 4.4.0)
Suggested: windows (>= 1.2.1); Java JRE (> 1.6)
Autoload: no
License: GPLv3+, simplified BSD
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/io/NEWS new/io/NEWS
--- old/io/NEWS 2016-03-10 22:02:24.000000000 +0100
+++ new/io/NEWS 2016-07-04 21:41:43.000000000 +0200
@@ -1,7 +1,27 @@
Summary of important user-visible changes for releases of the io package
===============================================================================
-io-2.4.1 Release Date: 2015-03-09 Release Manager: Philip Nienhuis
+io-2.4.2 Release Date: 2016-07-04 Release Manager: Philip Nienhuis
+===============================================================================
+
+*** Bug fixes:
+ - odsopen.m, xlsopen.m: do not allow writing to non-existent subdirs
+ - xlsread.m: return with error if file open failed, rather than trying
+ to proceed
+ - OCT interface: mention file creation errors verbosely rather than
+ silently
+ - OCT interface, OOXML: improved reading of cached values (but #47815)
+
+*** Code improvements:
+ - Swapped post-install.m hack to allow initialization of io package
+ for PKG_ADD/PKG_DEL directives; no more risk of affecting user space
+ - Changed several calls to (now deprecated) octave_config_info into
+ calls to __have_feature__.m
+ - Add support for complex numbers in read_namelist.m and write_namelist.m
+ (credits to Ryusuke Numata)
+
+===============================================================================
+io-2.4.1 Release Date: 2016-03-09 Release Manager: Philip Nienhuis
===============================================================================
*** Bug fixes:
@@ -11,20 +31,16 @@
provisional solution is to simply strip those characters from text
strings before writing them to file, or strip them before handing the
just read raw output cell array to Octave.
-
- Convert illegal XML characters ('"><&) to XML escape sequences and back.
-
- Delete temporary files when processing .xlsx files (bug #47378)
*** New features:
- New function tidyxml.m: simply strips characters < 32 or > 126 from
character arrays or cell arrays (including mixed cell arrays).
-
- test_spsh.m now shows a nice summary of test results for each interface.
*** Various changes
- Replace calls to deprecated sleep() with calls to pause().
-
- Adapt code to changes in octave_config_info in Octave development
version.
===============================================================================
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/io/PKG_ADD new/io/PKG_ADD
--- old/io/PKG_ADD 2016-03-09 22:13:29.000000000 +0100
+++ new/io/PKG_ADD 1970-01-01 01:00:00.000000000 +0100
@@ -1,114 +0,0 @@
-## PKG_ADD
-## add paths of io pkg java jars.
-
-## First check if Java support was built in anyway
-try
- HAVE_JAVA = octave_config_info ("features").JAVA;
-catch
- HAVE_JAVA = octave_config_info ("build_features").JAVA;
-end_try_catch
-if (HAVE_JAVA)
- ## OK, Java built-in / supported. Check environment var
- userdir = getenv ("OCTAVE_IO_JAVALIBS");
- if (ispc)
- homedir = getenv ("USERPROFILE");
- # (MinGW) assume jar files are in /lib/java
- libdir = octave_config_info ("libdir");
-# elseif (ismac)
-# ## Who knows where OSX keeps e.g., Apache POI stuff? if it does at all...
- elseif (isunix)
- homedir = tilde_expand ("~");
- ## On linux, spreadsheet .jars are often found somewhere in /usr/share/java
- libdir = "/usr/share";
- else
- ## Set libdir to "." to avoid searching in a root dir
- libdir = ".";
- endif
-
- ## Do we havea 64-bit indexing Octave at hand?
- ## Newer Octave-4.1.0+ has ENABLE_64, older Octave has USE_64_BIT_IDX_T
- try
- amd64y = strcmpi (octave_config_info ("USE_64_BIT_IDX_T"), "yes");
- catch
- amd64y = octave_config_info.ENABLE_64;
- end_try_catch
-
- ## Find LibreOffice or OpenOffice.org
- ooopath = '';
- ## Possible locations for OOo or LO.
- bnam = {"C:/Program Files (X86)", ...
- "C:/Program Files", ...
- "C:/Programs", ...
- "/opt", ...
- "/usr/lib"};
- if (amd64y)
- ## 64-bit indexing Octave won't work with 32-bit LibreOffice/OpenOffice.org
- bnam(1) = [];
- endif
- ii = 0;
- while (isempty (ooopath) && ii < numel (bnam))
- ooopath = glob ([ bnam{++ii} filesep "[Ll]ibre[Oo]ffice*"]);
- ## Watch out for uninstalled previous LO installations that just keep prefs
- if (! isempty (ooopath) && ! (exist ([ooopath{1} filesep "program"]) == 7))
- ooopath = '';
- endif
- endwhile
- ii = 0;
- while (isempty (ooopath) && ii < numel (bnam))
- ooopath = glob ([ bnam{++ii} filesep "[Oo]pen[Oo]ffice.org*"]);
- ## Watch out for uninstalled previous OOo installations that just keep
prefs
- if (! isempty (ooopath) && ! (exist ([ooopath{1} filesep "program"]) == 7))
- ooopath = '';
- endif
- endwhile
- ii = 0;
- while (isempty (ooopath) && ii < numel (bnam))
- ooopath = glob ([ bnam{++ii} filesep "ooo*"]);
- ## Watch out for uninstalled previous OOo installations that just keep
prefs
- if (! isempty (ooopath) && ! (exist ([ooopath{1} filesep "program"]) == 7))
- ooopath = '';
- endif
- endwhile
- if (! isempty (ooopath))
- ooopath = ooopath{:};
- else
- ooopath = '';
- endif
-
- ## One big try-catch to circumvent possible problems on Linux
- try
- if (! isempty (userdir))
- if (strcmpi (userdir, "no") || strcmpi (userdir, "false") || strcmpi
(userdir, "0"))
- ## Do not load Java class libs .jar files). First clean up, then return
- clear libdir spr_status userdir homedir bnam ooopath ii;
- return
- endif
- ## First allow some time for io package to be fully loaded
- pause (0.25);
- ## Check first for user-, then system supplied jars
- if (exist (userdir) == 7)
- ## Userdir is a subdir
- spr_status = chk_spreadsheet_support (userdir, 0, ooopath);
- endif
- ## Also try user's home directory
- elseif (isunix && ...
- ! (strcmpi (userdir, "no") || strcmpi (userdir, "false") || strcmpi
(userdir, "0")))
- ## On non-Windows systems, automatic loading of Java classes is opt-in
due to
- ## excessive search time (see bug #42044). Most of the delay is due to
searching
- ## for the Libre/OpenOffice.org jars
- clear libdir spr_status userdir homedir bnam ooopath ii HAVE_JAVA amd64y;
- return
- else
- ## Allow some time for io package to be fully loaded
- pause (0.25);
- endif
- ## Try <HOME>/java
- spr_status = chk_spreadsheet_support ([ homedir filesep "java" ], 0,
ooopath);
- ## Only then search for system-supplied jars. ooopath has been searched
- spr_status = chk_spreadsheet_support ([ libdir filesep "java" ], 0);
- catch
- warning ("(Automatic loading of spreadsheet I/O Java classlibs failed)\n");
- end_try_catch
-endif
-## Clean up
-clear libdir spr_status userdir homedir bnam ooopath ii HAVE_JAVA amd64y;
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/io/PKG_DEL new/io/PKG_DEL
--- old/io/PKG_DEL 2016-03-09 22:13:29.000000000 +0100
+++ new/io/PKG_DEL 1970-01-01 01:00:00.000000000 +0100
@@ -1,10 +0,0 @@
-## PKG_DEL
-##
-## All we need to do is try to remove all Java spreadsheet class libs loaded
-## by chk_spreadsheet_support.m from the javaclasspath
-
-try
- chk_spreadsheet_support ("", -1);
-catch
- warning ("Couldn't remove spreadsheet I/O javaclasspath entries while
unloading io pkg");
-end_try_catch
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/io/inst/__exit_io__.m new/io/inst/__exit_io__.m
--- old/io/inst/__exit_io__.m 1970-01-01 01:00:00.000000000 +0100
+++ new/io/inst/__exit_io__.m 2016-03-23 21:55:08.000000000 +0100
@@ -0,0 +1,36 @@
+## Copyright (C) 2016 Carnë Draug <carandr...@octave.org>
+## Copyright (C) 2011-2016 Philip Nienhuis
+##
+## This program is free software; you can redistribute it and/or
+## modify it under the terms of the GNU General Public License as
+## published by the Free Software Foundation; either version 3 of the
+## License, or (at your option) any later version.
+##
+## This program is distributed in the hope that it will be useful, but
+## WITHOUT ANY WARRANTY; without even the implied warranty of
+## MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+## General Public License for more details.
+##
+## You should have received a copy of the GNU General Public License
+## along with this program; if not, see
+## <http:##www.gnu.org/licenses/>.
+
+## -*- texinfo -*-
+## @deftypefn {} {} __exit_io_package__ ()
+## Undocumented internal function of io package.
+##
+## Remove io java jars loaded by io package functions from javaclasspath.
+##
+## @end deftypefn
+
+## PKG_DEL: __exit_io__ ()
+
+function __exit_io__ ()
+ ## All we need to do is try to remove all Java spreadsheet class libs loaded
+ ## by chk_spreadsheet_support.m from the javaclasspath
+ try
+ chk_spreadsheet_support ("", -1);
+ catch
+ warning ("Couldn't remove spreadsheet I/O javaclasspath entries while
unloading io pkg");
+ end_try_catch
+endfunction
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/io/inst/__init_io__.m new/io/inst/__init_io__.m
--- old/io/inst/__init_io__.m 1970-01-01 01:00:00.000000000 +0100
+++ new/io/inst/__init_io__.m 2016-06-28 22:23:05.000000000 +0200
@@ -0,0 +1,133 @@
+## Copyright (C) 2016 Carnë Draug <carandr...@octave.org>
+## Copyright (C) 2011-2016 Philip Nienhuis
+##
+## This program is free software; you can redistribute it and/or
+## modify it under the terms of the GNU General Public License as
+## published by the Free Software Foundation; either version 3 of the
+## License, or (at your option) any later version.
+##
+## This program is distributed in the hope that it will be useful, but
+## WITHOUT ANY WARRANTY; without even the implied warranty of
+## MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+## General Public License for more details.
+##
+## You should have received a copy of the GNU General Public License
+## along with this program; if not, see
+## <http:##www.gnu.org/licenses/>.
+
+## -*- texinfo -*-
+## @deftypefn {} {} __enter_io_package__ ()
+## Undocumented internal function of io package.
+##
+## Search io pkg java jars and add found ones to javaclasspath.
+##
+## @end deftypefn
+
+## PKG_ADD: __init_io__ ()
+
+function __init_io__ ()
+ ## First check if Java support was built in anyway
+ HAVE_JAVA = eval ("__octave_config_info__ ('build_features').JAVA", ...
+ "octave_config_info ('features').JAVA");
+ if (HAVE_JAVA)
+ ## OK, Java built-in / supported. Check environment var
+ userdir = getenv ("OCTAVE_IO_JAVALIBS");
+ if (ispc)
+ homedir = getenv ("USERPROFILE");
+ # (MinGW) assume jar files are in /lib/java
+ libdir = eval ('__octave_config_info__ ("libdir")', ...
+ 'octave_config_info ("libdir")');
+ # elseif (ismac)
+ # ## Who knows where OSX keeps e.g., Apache POI stuff? if it does at all...
+ elseif (isunix)
+ homedir = tilde_expand ("~");
+ ## On linux, spreadsheet .jars are often found somewhere in
/usr/share/java
+ libdir = "/usr/share";
+ else
+ ## Set libdir to "." to avoid searching in a root dir
+ libdir = ".";
+ endif
+
+ ## Do we havea 64-bit indexing Octave at hand?
+ ## Newer Octave-4.1.0+ has ENABLE_64, older Octave has USE_64_BIT_IDX_T
+ amd64y = eval ('__octave_config_info__.ENABLE_64', ...
+ 'strcmpi (octave_config_info ("USE_64_BIT_IDX_T"),
"yes")');
+
+ ## Find LibreOffice or OpenOffice.org
+ ooopath = '';
+ ## Possible locations for OOo or LO.
+ bnam = {"C:/Program Files (X86)", ...
+ "C:/Program Files", ...
+ "C:/Programs", ...
+ "/opt", ...
+ "/usr/lib"};
+ if (amd64y)
+ ## 64-bit indexing Octave won't work with 32-bit
LibreOffice/OpenOffice.org
+ bnam(1) = [];
+ endif
+ ii = 0;
+ while (isempty (ooopath) && ii < numel (bnam))
+ ooopath = glob ([ bnam{++ii} "/[Ll]ibre[Oo]ffice*"]);
+ ## Watch out for uninstalled previous LO installations that just keep
prefs
+ if (! isempty (ooopath) && ! (exist ([ooopath{1} filesep "program"]) ==
7))
+ ooopath = '';
+ endif
+ endwhile
+ ii = 0;
+ while (isempty (ooopath) && ii < numel (bnam))
+ ooopath = glob ([ bnam{++ii} "/[Oo]pen[Oo]ffice.org*"]);
+ ## Watch out for uninstalled previous OOo installations that just keep
prefs
+ if (! isempty (ooopath) && ! (exist ([ooopath{1} filesep "program"]) ==
7))
+ ooopath = '';
+ endif
+ endwhile
+ ii = 0;
+ while (isempty (ooopath) && ii < numel (bnam))
+ ooopath = glob ([ bnam{++ii} "/ooo*"]);
+ ## Watch out for uninstalled previous OOo installations that just keep
prefs
+ if (! isempty (ooopath) && ! (exist ([ooopath{1} filesep "program"]) ==
7))
+ ooopath = '';
+ endif
+ endwhile
+ if (! isempty (ooopath))
+ ooopath = ooopath{:};
+ else
+ ooopath = '';
+ endif
+
+ ## One big try-catch to circumvent possible problems on Linux
+ try
+ if (! isempty (userdir))
+ if (strcmpi (userdir, "no") || strcmpi (userdir, "false") || strcmpi
(userdir, "0"))
+ ## Do not load Java class libs .jar files). First clean up, then
return
+ clear libdir spr_status userdir homedir bnam ooopath ii;
+ return
+ endif
+ ## First allow some time for io package to be fully loaded
+ pause (0.25);
+ ## Check first for user-, then system supplied jars
+ if (exist (userdir) == 7)
+ ## Userdir is a subdir
+ spr_status = chk_spreadsheet_support (userdir, 0, ooopath);
+ endif
+ ## Also try user's home directory
+ elseif (isunix && ...
+ ! (strcmpi (userdir, "no") || strcmpi (userdir, "false") || strcmpi
(userdir, "0")))
+ ## On non-Windows systems, automatic loading of Java classes is opt-in
due to
+ ## excessive search time (see bug #42044). Most of the delay is due to
searching
+ ## for the Libre/OpenOffice.org jars
+ clear libdir spr_status userdir homedir bnam ooopath ii HAVE_JAVA
amd64y;
+ return
+ else
+ ## Allow some time for io package to be fully loaded
+ pause (0.25);
+ endif
+ ## Try <HOME>/java
+ spr_status = chk_spreadsheet_support ([ homedir filesep "java" ], 0,
ooopath);
+ ## Only then search for system-supplied jars. ooopath has been searched
+ spr_status = chk_spreadsheet_support ([ libdir filesep "java" ], 0);
+ catch
+ warning ("(Automatic loading of spreadsheet I/O Java classlibs
failed)\n");
+ end_try_catch
+ endif
+endfunction
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/io/inst/chk_spreadsheet_support.m
new/io/inst/chk_spreadsheet_support.m
--- old/io/inst/chk_spreadsheet_support.m 2016-03-09 22:13:29.000000000
+0100
+++ new/io/inst/chk_spreadsheet_support.m 2016-03-23 08:02:56.000000000
+0100
@@ -144,12 +144,7 @@
dbug = 0;
endif
- ## Provisions for octave_config_info changes in Octvae-4.1.0+
- try
- HAVE_JAVA = octave_config_info ("features").JAVA;
- catch
- HAVE_JAVA = octave_config_info ("build_features").JAVA;
- end_try_catch
+ HAVE_JAVA = __have_feature__ ("JAVA");
if (dbug < 0 && HAVE_JAVA)
## Remove loaded Java class libs from the javaclasspath
if (dbug < -2 && numel (loaded_jars))
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/io/inst/odsopen.m new/io/inst/odsopen.m
--- old/io/inst/odsopen.m 2016-03-09 22:13:29.000000000 +0100
+++ new/io/inst/odsopen.m 2016-04-23 12:07:47.000000000 +0200
@@ -1,4 +1,4 @@
-## Copyright (C) 2009-2015 Philip Nienhuis
+## Copyright (C) 2009-2016 Philip Nienhuis
##
## This program is free software; you can redistribute it and/or modify
## it under the terms of the GNU General Public License as published by
@@ -96,6 +96,26 @@
error ("No return argument specified!\n usage: ODS = odsopen (ODSfile,
[Rw]).\n");
endif
+ if (! (islogical (rw) || isnumeric (rw)))
+ error ("odsopen.m: numerical or logical value expected for arg ## 2
(readwrite)\n")
+ endif
+
+ if (ischar (filename))
+ [pth, fnam, ext] = fileparts (filename);
+ if (isempty (fnam))
+ error ("odsopen.m: no filename or empty filename specified");
+ endif
+ if (rw && ! isempty (pth))
+ apth = make_absolute_filename (pth);
+ if (exist (apth) != 7)
+ error ("odsopen.m: cannot write into non-existent directory:\n'%s'\n",
...
+ apth);
+ endif
+ endif
+ else
+ error ("odsopen.m: filename expected for argument #1");
+ endif
+
if (! isempty (reqinterface))
intfmsg = "requested";
if (! (ischar (reqinterface) || iscell (reqinterface)))
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/io/inst/private/__OCT_spsh_close__.m
new/io/inst/private/__OCT_spsh_close__.m
--- old/io/inst/private/__OCT_spsh_close__.m 2016-03-09 22:13:29.000000000
+0100
+++ new/io/inst/private/__OCT_spsh_close__.m 2016-04-23 12:22:01.000000000
+0200
@@ -1,4 +1,4 @@
-## Copyright (C) 2013-2015 Philip Nienhuis
+## Copyright (C) 2013-2016 Philip Nienhuis
##
## This program is free software; you can redistribute it and/or modify
## it under the terms of the GNU General Public License as published by
@@ -43,9 +43,11 @@
pause (0.05);
## Zip tmp directory into .ods or .xlsx and copy it over original file
try
- [stts, op] = system (sprintf ("zip -q -r %s *.* .", filename));
+ [stts, op] = system (sprintf ('zip -q -r "%s" *.* .', filename));
pause (0.20);
- xls.changed = 0;
+ if (stts == 0)
+ xls.changed = 0;
+ endif
catch
printf ("odsclose: could not zip ods contents in %s to %s\n", ...
xls.workbook, filename);
@@ -55,13 +57,13 @@
elseif (strcmpi (xls.filename(end-8:end), ".gnumeric"))
## gnumeric files are gzipped
try
- status = system (sprintf ("gzip -c -S=gnumeric %s > %s", ...
+ status = system (sprintf ('gzip -c -S=gnumeric %s > "%s"', ...
xls.workbook, filename));
if (! status)
## Delete temporary file
unlink (xls.workbook);
+ xls.changed = 0;
endif
- xls.changed = 0;
catch
status = 1;
end_try_catch
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/io/inst/private/__OCT_spsh_open__.m
new/io/inst/private/__OCT_spsh_open__.m
--- old/io/inst/private/__OCT_spsh_open__.m 2016-03-10 21:59:32.000000000
+0100
+++ new/io/inst/private/__OCT_spsh_open__.m 2016-04-03 18:08:30.000000000
+0200
@@ -50,7 +50,7 @@
opwd = pwd;
if (ftype == 5)
## Gnumeric xml files are gzipped
- system (sprintf ("gzip -d -c -S=gnumeric %s > %s", templ, tmpdir));
+ system (sprintf ('gzip -d -c -S=gnumeric "%s" > %s', templ, tmpdir));
fid = fopen (tmpdir, 'r');
xml = fread (fid, "char=>char").';
## Close file handle, don't delete file
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/io/inst/private/__OCT_xlsx2oct__.m
new/io/inst/private/__OCT_xlsx2oct__.m
--- old/io/inst/private/__OCT_xlsx2oct__.m 2016-03-09 22:13:29.000000000
+0100
+++ new/io/inst/private/__OCT_xlsx2oct__.m 2016-05-03 22:21:09.000000000
+0200
@@ -1,5 +1,5 @@
## Copyright (C) 2013,2014 Markus Bergholz
-## Parts Copyright (C) 2013-2015 Philip Nienhuis
+## Parts Copyright (C) 2013-2016 Philip Nienhuis
##
## This program is free software; you can redistribute it and/or modify
## it under the terms of the GNU General Public License as published by
@@ -44,7 +44,8 @@
warning ("xls2oct: illegal worksheet number (%d) - worksheet #1 assumed",
wsh);
endif
- ## Prepare to open requested worksheet file in subdir xl/ . Note: Win
accepts forward slashes
+ ## Prepare to open requested worksheet file in subdir xl/ .
+ ## Note: Windows accepts forward slashes
rawsheet = fopen (sprintf ('%s/xl/worksheets/sheet%d.xml', xls.workbook,
wsh));
if (rawsheet <= 0)
# Try to open sheet from r:id in worksheets.rels.xml
@@ -63,220 +64,237 @@
strings = fread (fid, "char=>char").';
fclose (fid);
catch
- ## No sharedStrings.xml; implies no "fixed" strings (computed strings
can still be there)
+ ## No sharedStrings.xml; implies no "fixed" strings (computed strings can
+ ## still be present)
strings = "";
end_try_catch
endif
rstatus = 0;
- raw = {};
-
- ## 'val' are the actual values. 'valraw' are the corresponding(!) cell
positions (e.g. B3).
- if (isempty (raw))
- ## General note for tuning: '"([^"]*)"' (w/o single quotes) could be
faster than '"(.*?)"'
- ##
(http://stackoverflow.com/questions/2503413/regular-expression-to-stop-at-first-match
comment #7)
-
- ## Below are loads of nested IFs. They're needed to catch empty previous
results, even empty sheets
+ ## General note for tuning: '"([^"]*)"' (w/o single quotes) could be faster
+ ## than '"(.*?)"'
+ ##
(http://stackoverflow.com/questions/2503413/regular-expression-to-stop-at-first-match
comment #7)
+
+ ## As to requested subranges: it's complicated to extract just part of a
sheet;
+ ## either way the entire sheet would need to be scanned for cell addresses
+ ## before one can know what part of the sheet XML the requested range lives.
+ ## In addition the endpoint cells of that rangemay not exist in the sheet XML
+ ## (e.g., if they're empty).
+ ## So we read *all* data and in the end just return the requested rectangle.
+
+ ## 'val' are the actual values. 'valraw' are the corresponding(!) cell
+ ## positions (e.g. B3). They're extracted separately to avoid (rare cases
of)
+ ## out-of-phase regexp results that would screw up reshape().
+
+ ## Below are loads of nested IFs. They're needed to catch empty previous
+ ## results, even empty sheets.
+
+ ## In below regexps, we ignore "cm" and "ph" tags immediately after <c and
+ ## a "vm" tag immediately after <t> tag. As soon as we hit them in the wild
+ ## these can be added (at the cost of speed performance).
+
+ ## 1. Get pure numbers, including booleans, double and boolean formula
+ ## results, from cells w/o 's=""' tag
+ val = cell2mat (regexp (rawdata, '<c r="\w+"(?:
t="[bn]+")?>(?:<f.+?(?:</f>|/>))?<v(?:.*?)>(.*?)</v>', "tokens"));
+ if (! isempty (val))
+ valraw = cell2mat (regexp (rawdata, '<c r="(\w+)"(?:
t="[bn]+")?>(?:<f.+?(?:</f>|/>))?<v(?:.*?)>.*?</v>', "tokens"));
+ endif
- ## 1. Get pure numbers, including booleans, double and boolean formula
results, from cells w/o 's=""' tag
- val = cell2mat (regexp (rawdata, '<c r="\w+"(?:
t="[bn]+")?>(?:<f.+?(?:</f>|/>))?<v>(.*?)</v>', "tokens"));
+ ## If val is still empty, try another regexpression (PRN: will this ever
+ ## work? haven't seen such cells)
+ if (numel (val) == 0)
+ val = cell2mat (regexp (rawdata, '<c r="\w+" s="\d+"(?!
t="s")><v(?:.*?)>(.*?)</v>', "tokens"));
if (! isempty (val))
- valraw = cell2mat (regexp (rawdata, '<c r="(\w+)"(?:
t="[bn]+")?>(?:<f.+?(?:</f>|/>))?<v>.*?</v>', "tokens"));
+ valraw = cell2mat (regexp (rawdata, '<c r="(\w+)" s="\d+"(?!
t="s")><v(?:.*?)>.*?</v>', "tokens"));
endif
+ endif
- ## If val is still empty, try another regexpression (PRN: will this ever
work? haven't seen such cells)
- if (numel (val) == 0)
- val = cell2mat (regexp (rawdata, '<c r="\w+" s="\d+"(?!
t="s")><v>(.*?)</v>', "tokens"));
+ ## If 'val' exist, check if there are tagged s="NUMBERS" values
+ if (numel (regexp (rawdata, ' s="', "once")) > 0)
+ ## Time/date values. Exclude formulas (having <f> </f> of <f /> tags),
+ ## strings ('t="s"') and error results ('t="e"')
+ valp = cell2mat (regexp (rawdata, '<c r="\w+" s="\d+"(?:
t="[^se]*")?><v(?:.*?)>(.*?)</v>', "tokens"));
+ if (! isempty (valp))
+ valrawp = cell2mat(regexp (rawdata, '<c r="(\w+)" s="\d+"(?:
t="[^se]*")?><v(?:.*?)>.*?</v>', "tokens"));
if (! isempty (val))
- valraw = cell2mat (regexp (rawdata, '<c r="(\w+)" s="\d+"(?!
t="s")><v>.*?</v>', "tokens"));
+ val = [val valp];
+ valraw = [valraw valrawp];
+ else
+ val = valp;
+ valraw = valrawp;
+ clear valp valrawp ;
endif
endif
+ endif
+ ## Turn strings into numbers
+ if (! isempty (val))
+ val = num2cell (str2double (val));
+ endif
- ## If 'val' exist, check if there are tagged s="NUMBERS" values
- if (numel (regexp (rawdata, ' s="', "once")) > 0)
- ## Time/date values. Exclude formulas (having <f> </f> of <f /> tags),
- ## strings ('t="s"') and error results ('t="e"')
- valp = cell2mat (regexp (rawdata, '<c r="\w+" s="\d+"(?:
t="[^se]*")?><v>(.*?)</v>', "tokens"));
- if (! isempty (valp))
- valrawp = cell2mat(regexp (rawdata, '<c r="(\w+)" s="\d+"(?:
t="[^se]*")?><v>.*?</v>', "tokens"));
- if (! isempty (val))
- val = [val valp];
- valraw = [valraw valrawp];
- else
- val = valp;
- valraw = valrawp;
- clear valp valrawp ;
- endif
+ ## 2. String / text formulas (cached results are in this sheet; fixed strings
+ ## in <sharedStrings.xml>)
+ ## 2.A Formulas
+ if (spsh_opts.formulas_as_text)
+
+ ## Get formulas themselves as text strings. Formulas have no 't="s"'
attribute. Keep starting '>' for next line
+ valf1 = cell2mat (regexp (rawdata, '<c r="\w+"(?: s="\d+")?(?:
t="\w+")?><f(?: .*?)*(>.*?)</f>(?:<v(?:.*?)>.*?</v>)?', "tokens"));
+ if (! isempty (valf1))
+ valf1 = regexprep (valf1, '^>', '=');
+ ## Pretty text output
+ valf1 = strrep (valf1, """, '"');
+ valf1 = strrep (valf1, "<", "<");
+ valf1 = strrep (valf1, ">", ">");
+ valf1 = strrep (valf1, "&", "&");
+ valrawf1 = cell2mat(regexp (rawdata, '<c r="(\w+)"(?: s="\d+")?(?:
t="\w+")?><f(?: .*?)*>.*?</f>(?:<v(?:.*?)>.*?</v>)?', "tokens"));
+ if (isempty (val))
+ val = valf1;
+ else
+ ## Formulas start with '=' so:
+ val = [val valf1];
+ valraw = [valraw valrawf1];
endif
endif
- ## Turn strings into numbers
- if (! isempty (val))
- val = num2cell (str2double (val));
+ clear valf1 valrawf1 ;
+ else
+ ## Get (cached) formula results. Watch out! as soon as a "t" attibute
equals "b" or is missing it is a number
+ ## First the non-numeric formula results
+ valf2 = cell2mat (regexp (rawdata, '<c r="\w+" s="\d+"
t="(?:[^sb]?|str)">(?:<f.+?(?:</f>|/>))<v(?:.*?)>(.*?)</v>', "tokens"));
+ if (! isempty (valf2))
+ ## Pretty text output
+ valf2 = strrep (valf2, """, '"');
+ valf2 = strrep (valf2, "<", "<");
+ valf2 = strrep (valf2, ">", ">");
+ valf2 = strrep (valf2, "&", "&");
+ valrawf2 = cell2mat(regexp (rawdata, '<c r="(\w+)" s="\d+"
t="(?:[^sb]?|str)">(?:<f.+?(?:</f>|/>))<v(?:.*?)>.*?</v>', "tokens"));
+ if (isempty (val))
+ val = valf2;
+ valraw = valrawf2;
+ else
+ val = [val valf2];
+ valraw = [valraw valrawf2];
+ endif
endif
+ clear valf2 valrawf2 ;
+ ## Next the numeric formula results. These need additional conversion
+ valf3 = cell2mat (regexp (rawdata, '<c r="\w+" s="\d+"(?:
t="b")?>(?:<f.+?(?:</f>|/>))<v(?:.*?)>(.*?)</v>', "tokens"));
+ if (! isempty (valf3))
+ valrawf3 = cell2mat(regexp (rawdata, '<c r="(\w+)" s="\d+"(?:
t="b")?>(?:<f.+?(?:</f>|/>))<v(?:.*?)>.*?</v>', "tokens"));
+ if (isempty (val))
+ val = num2cell(str2double (valf3));
+ valraw = valrawf3;
+ else
+ val = [val num2cell(str2double (valf3))];
+ valraw = [valraw valrawf3];
+ endif
+ endif
+ clear valf3 valrawf3 ;
+ endif
- ## 2. String / text formulas (cached results are in this sheet; fixed
strings in <sharedStrings.xml>)
- ## Formulas
- if (spsh_opts.formulas_as_text)
-
- ## Get formulas themselves as text strings. Formulas have no 't="s"'
attribute. Keep starting '>' for next line
- valf1 = cell2mat (regexp (rawdata, '<c r="\w+"(?: s="\d+")?(?:
t="\w+")?><f(?: .*?)*(>.*?)</f>(?:<v>.*?</v>)?', "tokens"));
- if (! isempty (valf1))
- valf1 = regexprep (valf1, '^>', '=');
- ## Pretty text output
- valf1 = strrep (valf1, """, '"');
- valf1 = strrep (valf1, "<", "<");
- valf1 = strrep (valf1, ">", ">");
- valf1 = strrep (valf1, "&", "&");
- valrawf1 = cell2mat(regexp (rawdata, '<c r="(\w+)"(?: s="\d+")?(?:
t="\w+")?><f(?: .*?)*>.*?</f>(?:<v>.*?</v>)?', "tokens"));
- if (isempty (val))
- val = valf1;
- else
- ## Formulas start with '=' so:
- val = [val valf1];
- valraw = [valraw valrawf1];
+ ## 2.B. Strings
+ if (! isempty (strings))
+ ## Extract string values. May be much more than present in current sheet
+ strings = regexp (strings, '<si[^>]*>.*?</si>', "match");
+ if (! isempty (strings))
+ for n = 1:columns(strings)
+ ctext{1,n} = cell2mat (cell2mat (regexp (strings{1,n},
'<t[^>]*>(.*?)</t>', "tokens")));
+ end
+ ## Pointers into sharedStrings.xml. "Hard" (fixed) strings have 't="s"'
attribute
+ ## For reasons known only to M$ those pointers are zero-based, so:
+ vals = str2double (cell2mat (regexp (rawdata, '<c r="\w+"(?: s="\d+")?
t="s"><v(?:.*?)>(\d+)</v>', "tokens"))) + 1;
+ if (! isempty (vals) && isfinite (vals))
+ ## Get actual values. Watch out for empty strings
+ vals = ctext(vals);
+ ids = cellfun (@isempty, vals);
+ if (any (ids))
+ vals {find (ids)} = "";
endif
- endif
- clear valf1 valrawf1 ;
- else
- ## Get (cached) formula results. Watch out! as soon as a "t" attibute
equals "b" or is missing it is a number
- ## First the non-numeric formula results
- valf2 = cell2mat (regexp (rawdata, '<c r="\w+" s="\d+"
t="(?:[^sb]?|str)">(?:<f.+?(?:</f>|/>))<v>(.*?)</v>', "tokens"));
- if (! isempty (valf2))
## Pretty text output
- valf2 = strrep (valf2, """, '"');
- valf2 = strrep (valf2, "<", "<");
- valf2 = strrep (valf2, ">", ">");
- valf2 = strrep (valf2, "&", "&");
- valrawf2 = cell2mat(regexp (rawdata, '<c r="(\w+)" s="\d+"
t="(?:[^sb]?|str)">(?:<f.+?(?:</f>|/>))<v>.*?</v>', "tokens"));
- if (isempty (val))
- val = valf2;
- valraw = valrawf2;
- else
- val = [val valf2];
- valraw = [valraw valrawf2];
- endif
- endif
- clear valf2 valrawf2 ;
- ## Next the numeric formula results. These need additional conversion
- valf3 = cell2mat (regexp (rawdata, '<c r="\w+" s="\d+"(?:
t="b")?>(?:<f.+?(?:</f>|/>))<v>(.*?)</v>', "tokens"));
- if (! isempty (valf3))
- valrawf3 = cell2mat(regexp (rawdata, '<c r="(\w+)" s="\d+"(?:
t="b")?>(?:<f.+?(?:</f>|/>))<v>.*?</v>', "tokens"));
+ vals = strrep (vals, """, '"');
+ vals = strrep (vals, "<", "<");
+ vals = strrep (vals, ">", ">");
+ vals = strrep (vals, "&", "&");
+ ## Cell addresses
+ valraws = cell2mat (regexp (rawdata, '<c r="(\w+)"(?: s="\d+")?
t="s"><v(?:.*?)>\d+</v>', "tokens"));
if (isempty (val))
- val = num2cell(str2double (valf3));
- valraw = valrawf3;
+ val = vals;
+ valraw = valraws;
else
- val = [val num2cell(str2double (valf3))];
- valraw = [valraw valrawf3];
+ val = [val vals];
+ valraw = [valraw valraws];
endif
endif
- clear valf3 valrawf3 ;
endif
+ clear vals valraws ;
+ endif
- ## 3. Strings
- if (! isempty (strings))
- ## Extract string values. May be much more than present in current sheet
- strings = regexp (strings, '<si[^>]*>.*?</si>', "match");
- if (! isempty (strings))
- for n = 1:columns(strings)
- ctext{1,n} = cell2mat (cell2mat (regexp (strings{1,n},
'<t[^>]*>(.*?)</t>', "tokens")));
- end
- ## Pointers into sharedStrings.xml. "Hard" (fixed) strings have
't="s"' attribute
- ## For reasons known only to M$ those pointers are zero-based, so:
- vals = str2double (cell2mat (regexp (rawdata, '<c r="\w+"(?: s="\d+")?
t="s"><v>(\d+)</v>', "tokens"))) + 1;
- if (! isempty (vals) && isfinite (vals))
- ## Get actual values. Watch out for empty strings
- vals = ctext(vals);
- ids = cellfun (@isempty, vals);
- if (any (ids))
- vals {find (ids)} = "";
- endif
- ## Pretty text output
- vals = strrep (vals, """, '"');
- vals = strrep (vals, "<", "<");
- vals = strrep (vals, ">", ">");
- vals = strrep (vals, "&", "&");
- ## Cell addresses
- valraws = cell2mat (regexp (rawdata, '<c r="(\w+)"(?: s="\d+")?
t="s"><v>\d+</v>', "tokens"));
- if (isempty (val))
- val = vals;
- valraw = valraws;
- else
- val = [val vals];
- valraw = [valraw valraws];
- endif
- endif
- endif
- clear vals valraws ;
- endif
-
- ## If val is empty, sheet is empty
- if (isempty (val))
- xls.limits = [];
- raw = {};
- return
- endif
-
- ## 4. Prepare
- ## Get the row number (currently supported from 1 to 999999)
- vi.row = str2double (cell2mat (regexp (valraw,
'(\d+|\d+\d+|\d+\d+\d+|\d+\d+\d+\d+|\d+\d+\d+\d+\+d|\d+\d+\d+\d+\d+\d+)?',
"match"))')';
-
- ## Get the column character (A to ZZZ) (that are more than 18k supported
columns atm)
- vi.alph = cell2mat (regexp (valraw,
'([A-Za-z]+|[A-Za-z]+[A-Za-z]+|[A-Za-z]+[A-Za-z]+[A-Za-z]+)?', "match"));
-
- ## Free memory; might be useful while reading huge files
- clear valraw ;
-
- ## If missed formular indices
- idx.all = cell2mat (regexp (rawdata, '<c r="(\w+)"[^>]*><f', "tokens"));
- if (0 < numel (idx.all))
- idx.num = str2double (cell2mat (regexp (idx.all,
'(\d+|\d+\d+|\d+\d+\d+|\d+\d+\d+\d+|\d+\d+\d+\d+\+d|\d+\d+\d+\d+\d+\d+)?',
"match"))')';
- idx.alph = cell2mat (regexp (idx.all,
'([A-Za-z]+|[A-Za-z]+[A-Za-z]+|[A-Za-z]+[A-Za-z]+[A-Za-z]+)?', "match"));
- idx.alph = double (cell2mat (cellfun (@col2num, vi.alph,
"UniformOutput", 0)));
- else
- ## To prevent warnings or errors while calculating corresponding NaN
matrix
- idx.num = [];
- idx.alph = [];
- end
- ## Transform column character to column number
- ## A -> 1; C -> 3, AB -> 28 ...
- vi.col = double (cell2mat (cellfun (@col2num, vi.alph, "UniformOutput",
0)));
-
- ## Find data rectangle limits
- idx.mincol = min ([idx.alph vi.col]);
- idx.minrow = min ([idx.num vi.row]);
- idx.maxrow = max ([idx.num vi.row]);
- idx.maxcol = max ([idx.alph vi.col]);
-
- ## Convey limits of data rectangle to xls2oct. Must be done here
- xls.limits = [idx.mincol, idx.maxcol; idx.minrow, idx.maxrow];
-
- ## column adjustment when first number or formula don't begin in the first
column
- if (1 < idx.mincol)
- vi.col = vi.col - (idx.mincol - 1);
- endif
- ## row adjustment when first number or formular don't begin in the first
row
- if (1 < idx.minrow)
- vi.row = vi.row - (idx.minrow - 1);
- endif
- ## Initialize output cell array
- raw = cell (idx.maxrow - idx.minrow + 1, idx.maxcol - idx.mincol + 1);
-
- ## get logical indices for 'val' from 'valraw' positions in NaN matrix
- vi.idx = sub2ind (size (raw), (vi.row), (vi.col));
- ## set values to the corresponding indizes in final cell matrix
- raw(vi.idx) = val;
+ ## If val is empty, sheet is empty
+ if (isempty (val))
+ xls.limits = [];
+ raw = {};
+ return
+ endif
+ ## 3. Prepare for assigning extracted values to output cell array
+ ## Get the row numbers (currently supported from 1 to 999999)
+ vi.row = str2double (cell2mat (regexp (valraw,
'(\d+|\d+\d+|\d+\d+\d+|\d+\d+\d+\d+|\d+\d+\d+\d+\+d|\d+\d+\d+\d+\d+\d+)?',
"match"))')';
+
+ ## Get the column characters (A to ZZZ) (that are more than 18k supported
cols)
+ vi.alph = cell2mat (regexp (valraw,
'([A-Za-z]+|[A-Za-z]+[A-Za-z]+|[A-Za-z]+[A-Za-z]+[A-Za-z]+)?', "match"));
+
+ ## Free memory; might be useful while reading huge files
+ clear valraw ;
+
+ ## If missed formular indices
+ idx.all = cell2mat (regexp (rawdata, '<c r="(\w+)"[^>]*><f', "tokens"));
+ if (0 < numel (idx.all))
+ idx.num = str2double (cell2mat (regexp (idx.all,
'(\d+|\d+\d+|\d+\d+\d+|\d+\d+\d+\d+|\d+\d+\d+\d+\+d|\d+\d+\d+\d+\d+\d+)?',
"match"))')';
+ idx.alph = cell2mat (regexp (idx.all,
'([A-Za-z]+|[A-Za-z]+[A-Za-z]+|[A-Za-z]+[A-Za-z]+[A-Za-z]+)?', "match"));
+ idx.alph = double (cell2mat (cellfun (@col2num, vi.alph, "UniformOutput",
0)));
+ else
+ ## To prevent warnings or errors while calculating corresponding NaN matrix
+ idx.num = [];
+ idx.alph = [];
+ end
+ ## Transform column character to column number
+ ## A -> 1; C -> 3, AB -> 28 ...
+ vi.col = double (cell2mat (cellfun (@col2num, vi.alph, "UniformOutput", 0)));
+
+ ## Find data rectangle limits
+ idx.mincol = min ([idx.alph vi.col]);
+ idx.minrow = min ([idx.num vi.row]);
+ idx.maxrow = max ([idx.num vi.row]);
+ idx.maxcol = max ([idx.alph vi.col]);
+
+ ## Convey limits of data rectangle to xls2oct. Must be done here as first
start
+ xls.limits = [idx.mincol, idx.maxcol; idx.minrow, idx.maxrow];
+
+ ## column adjustment when first number or formula don't begin in first column
+ if (1 < idx.mincol)
+ vi.col = vi.col - (idx.mincol - 1);
endif
+ ## row adjustment when first number or formular don't begin in first row
+ if (1 < idx.minrow)
+ vi.row = vi.row - (idx.minrow - 1);
+ endif
+ ## Initialize output cell array
+ raw = cell (idx.maxrow - idx.minrow + 1, idx.maxcol - idx.mincol + 1);
+
+ ## get logical indices for 'val' from 'valraw' positions in NaN matrix
+ vi.idx = sub2ind (size (raw), (vi.row), (vi.col));
+ ## set values to the corresponding indizes in final cell matrix
+ raw(vi.idx) = val;
- ## FIXME maybe reading parts of the data can be done faster above by better
regexps
+ ## Process requested cell range argument
if (! isempty (crange))
- ## We'll do it the easy way: just read all data, then return only the
requested part
+ ## Extract only the requested cell rectangle (see comments higher up)
[~, nr, nc, tr, lc] = parse_sp_range (crange);
- xls.limits = [max(idx.mincol, lc), min(idx.maxcol, lc+nc-1);
max(idx.minrow, tr), min(idx.maxrow, tr+nr-1)];
- ## Correct spreadsheet locations for lower right shift or raw
+ xls.limits = [max(idx.mincol, lc), min(idx.maxcol, lc+nc-1); ...
+ max(idx.minrow, tr), min(idx.maxrow, tr+nr-1)];
+ ## Correct spreadsheet locations for lower right shift of raw
rc = idx.minrow - 1;
cc = idx.mincol - 1;
- raw = raw(xls.limits(2, 1)-rc : xls.limits(2, 2)-rc, xls.limits(1, 1)-cc :
xls.limits(1, 2)-cc);
+ raw = raw(xls.limits(2, 1)-rc : xls.limits(2, 2)-rc, ...
+ xls.limits(1, 1)-cc : xls.limits(1, 2)-cc);
endif
if (! isempty (val))
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/io/inst/private/__UNO_spsh_open__.m
new/io/inst/private/__UNO_spsh_open__.m
--- old/io/inst/private/__UNO_spsh_open__.m 2016-03-09 22:13:29.000000000
+0100
+++ new/io/inst/private/__UNO_spsh_open__.m 2016-03-22 22:52:23.000000000
+0100
@@ -102,11 +102,10 @@
catch
amd64y = octave_config_info.ENABLE_64;
end_try_catch
- if (amd64y || ...
- strncmpi (octave_config_info("canonical_host_type"), "x86_64", 6))
+ if (amd64y || ! isempty (strfind (lower (computer ("arch")), "x86_64")))
printf ("Maybe you have 32-bit Libre/OpenOffice installed?\n");
printf ("64-bit Octave requires 64-bit L.O. / OOo.\n\n");
- elseif (strncmpi (octave_config_info("canonical_host_type"), "i686", 4))
+ elseif (! isempty (strfind (lower (computer ("arch")), "i686")))
printf ("Maybe you have 64-bit Libre/OpenOffice installed?\n");
printf ("32-bit Octave requires 32-bit L.O. / OOo.\n\n");
endif
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/io/inst/private/__chk_java_sprt__.m
new/io/inst/private/__chk_java_sprt__.m
--- old/io/inst/private/__chk_java_sprt__.m 2016-03-09 22:13:29.000000000
+0100
+++ new/io/inst/private/__chk_java_sprt__.m 2016-03-23 08:08:24.000000000
+0100
@@ -23,11 +23,7 @@
jcp = {};
tmp1 = 0;
- try
- has_java = octave_config_info.features.JAVA;
- catch
- has_java = octave_config_info.build_features.JAVA;
- end_try_catch
+ has_java = __have_feature__ ("JAVA");
if (! has_java)
## No Java support built in => any further checks are moot
return
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/io/inst/private/getodsinterfaces.m
new/io/inst/private/getodsinterfaces.m
--- old/io/inst/private/getodsinterfaces.m 2016-03-09 22:13:29.000000000
+0100
+++ new/io/inst/private/getodsinterfaces.m 2016-03-22 22:36:40.000000000
+0100
@@ -46,11 +46,7 @@
persistent uno_1st_time = 0;
if (isempty (has_java))
- try
- has_java = octave_config_info.features.JAVA;
- catch
- has_java = octave_config_info.build_features.JAVA;
- end_try_catch
+ has_java = __have_feature__ ("JAVA");
endif
if (isempty (odsinterfaces.OTK) && isempty (odsinterfaces.JOD) ...
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/io/inst/private/getxlsinterfaces.m
new/io/inst/private/getxlsinterfaces.m
--- old/io/inst/private/getxlsinterfaces.m 2016-03-09 22:13:29.000000000
+0100
+++ new/io/inst/private/getxlsinterfaces.m 2016-03-22 22:37:07.000000000
+0100
@@ -49,11 +49,7 @@
persistent uno_1st_time = 0;
if (isempty (has_java))
- try
- has_java = octave_config_info.features.JAVA;
- catch
- has_java = octave_config_info.build_features.JAVA;
- end_try_catch
+ has_java = __have_feature__ ("JAVA");
endif
if (isempty (xlsinterfaces.COM) && isempty (xlsinterfaces.POI) ...
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/io/inst/read_namelist.m new/io/inst/read_namelist.m
--- old/io/inst/read_namelist.m 2016-03-09 22:13:29.000000000 +0100
+++ new/io/inst/read_namelist.m 2016-06-19 23:15:53.000000000 +0200
@@ -1,4 +1,4 @@
-## Copyright (C) Darien Pardinas Diaz <darien.pardinas-d...@monash.edu>
+## Copyright (C) 2013 Darien Pardinas Diaz <darien.pardinas-d...@monash.edu>
##
## This program is free software; you can redistribute it and/or modify it
under
## the terms of the GNU General Public License as published by the Free
Software
@@ -30,12 +30,12 @@
## |-- NMLST_M--|--VAR2
## |...
## |--VARNm
-##
+##
## Note: The function can read multidimensional variables as well. The
## function assumes that there is no more than one namelist section per
## line. At this time there is no syntax checking functionality so the
## function will crash in case of errors.
-##
+##
## Example:
## NMLST = read_namelist ("OPTIONS.nam");
## NMLST.NAM_FRAC.XUNIF_NATURE = 0.1;
@@ -65,13 +65,13 @@
fid = fopen (filename, "r");
c = 0;
lines = cell(1);
-
+
## Read all the text lines in namelist file
while (! feof (fid))
line = fgetl (fid);
## Remove comments if any on the line
idx = find (line == "!");
- if ~isempty (idx),
+ if (! isempty (idx))
line = line (1:idx(1) - 1);
end
if (! isempty (line)),
@@ -79,8 +79,8 @@
lines{c} = line; ## FIXME each time appending to a cell array is slow
end
end
- fclose(fid);
-
+ fclose (fid);
+
ii = 0;
while (ii < c);
## Find a record
@@ -98,11 +98,15 @@
namelst = line;
line = []; ##TDuell 31/03/2013 Provisional fix L.102 PRN
1apr2013
endif
- nmlst_bdy = [];
- if (! isempty (line)); idx = strfind (line, "/"); endif
+ nmlst_bdy = "";
+ if (! isempty (line));
+ idx = strfind (line, "/");
+ else
+ line = "";
+ endif
## Get the variable specification section
while (isempty (idx) && ii < c)
- nmlst_bdy = [ nmlst_bdy " " line ];
+ nmlst_bdy = [ nmlst_bdy " " line ];
++ii;
line = lines{ii};
idx = strfind (line, "/");
@@ -111,7 +115,7 @@
nmlst_bdy = [ nmlst_bdy " " line ];
endif
## Parse current namelist (set of variables)
- S.(namelst) = parse_namelist (nmlst_bdy);
+ S.(namelst) = parse_namelist (nmlst_bdy);
endif
endwhile
@@ -128,60 +132,81 @@
strng = regexprep (strng, '\.false\.', "F", "ignorecase");
strng = regexprep (strng, '\.t\.', "T", "ignorecase");
strng = regexprep (strng, '\.f\.', "F", "ignorecase");
-
- ## Make evaluable the (:) expression in MATLAB if any
+
+ ## Make evaluable the (:) expression in Octave if any
strng = regexprep (strng, '\(:\)', "(1,:)");
[strng, islit] = parse_literal_strings ([strng " "]);
-
+
## Find the position of all the "="
- eq_idx = find (strng == "=");
+ eq_idx = find (strng == "="); ## PRN 19Jun2016 use strfind() ?
nvars = length (eq_idx);
-
+
arg_start = eq_idx + 1;
arg_end = zeros (size (eq_idx));
vars = cell (nvars, 1);
S = struct;
-
+
## Loop through every variable
for kk = 1:nvars,
ii = eq_idx(kk) - 1;
## Move to the left and discard blank spaces
- while (strng(ii) == " "); --ii; endwhile
+ while (strng(ii) == " ")
+ --ii;
+ endwhile
## Now we are over the variable name or closing parentesis
jj = ii;
if (strng(ii) == ")"),
- while (strng(ii) ~= "("); --ii; endwhile
+ while (strng(ii) != "(")
+ --ii;
+ endwhile
--ii;
## Move to the left and discard any possible blank spaces
- while (strng(ii) == " "); --ii; endwhile
+ while (strng(ii) == " ")
+ --ii;
+ endwhile
endif
-
+
## Now we are over the last character of the variable name
- while (strng(ii) ~= " "); --ii; endwhile
-
+ while (strng(ii) != " ")
+ --ii;
+ endwhile
+
if (kk > 1);
arg_end(kk - 1) = ii;
endif
vars{kk} = [ "S." strng(ii + 1: jj) ];
endfor
-
+
arg_end(end) = length (strng);
-
+
## This variables are used in the eval function to evaluate True/False,
## so don't remove it!
T = ".true.";
F = ".false.";
## Loop through every variable guess variable type
- for kk = 1:nvars
+ for kk = 1:nvars
arg = strng(arg_start(kk):arg_end(kk));
arglit = islit(arg_start(kk):arg_end(kk))';
-
+
+ # complex numbers
+ if (! any (arglit))
+ bra = strfind (arg, "(");
+ ckt = strfind (arg, ")");
+ if (! isempty (bra))
+ list = [];
+ for i=1:length(bra)
+ list = [ list, eval([ "complex", arg(bra(i):ckt(i)) ])];
+ endfor
+ arg = num2str(list);
+ endif
+ endif
+
## Remove commas in non literal string...
- commas = (! arglit && arg == ',');
+ commas = (! arglit && arg == ",");
if (any (commas))
arg(commas) = " ";
endif
-
+
if (any (arglit))
## We are parsing a variable that is literal string
arg = [ "{" arg "};"];
@@ -192,23 +217,22 @@
## We are parsing a numerical array
arg = [ "[" arg "];"];
endif
- ## Eval the modified syntax in Matlab
+ ## Eval the modified syntax in Octave
eval ([vars{kk} " = " arg]);
endfor
endfunction
-
-## Parse the literal declarations of strings and change to Matlab syntax
+## Parse the literal declarations of strings and change to Octave syntax
function [strng, is_lit] = parse_literal_strings (strng)
len = length (strng);
- add_squote = []; ## Positions to add a scape single quote on syntax
+ add_squote = []; ## Positions to add a scape single quote on syntax
rem_dquote = []; ## Positions to remove a double quote scape on syntax
ii = 1;
while (ii < len)
- if strng(ii) == "'", ## Opening string with single quote...
+ if (strng(ii) == "'") ## Opening string with single quote...
++ii;
- while ((ii < len && strng(ii) ~= "'") || strcmp (strng(ii:ii+1),
''''''))
+ while ((ii < len && strng(ii) != "'") || strcmp (strng(ii:ii+1), ''''''))
++ii;
if strcmp (strng(ii-1:ii), ''''''),
++ii;
@@ -218,9 +242,9 @@
if (strng(ii) == '"') ## Opening string with double quote...
strng(ii) = "'"; ## Change to single quote
++ii;
- while (strng(ii) ~= '"' || strcmp (strng(ii:i+1),'""') && ii < len)
+ while (strng(ii) != '"' || strcmp (strng(ii:i+1),'""') && ii < len)
## Check for a possible single quote here
- if strng(ii) == "'",
+ if (strng(ii) == "'")
add_squote = [ add_squote ii ];
endif
++ii;
@@ -239,8 +263,8 @@
for ii = 1:length(rem_dquote);
strng = [ strng(1:rem_dquote(ii)-1) strng(rem_squote(ii)+1:end) ];
endfor
-
- ## Now everything should be in Matlab string syntax
+
+ ## Now everything should be in Octave string syntax
## Classify syntax as literal or regular expression
ii = 1;
len = length (strng);
@@ -249,9 +273,9 @@
if (strng(ii) == "'") ## Opening string with single quote...
is_lit(ii) = 1;
++ii;
- while ((ii < len && strng(ii) ~= "'") || strcmp(strng(ii:ii+1), "''"))
+ while ((ii < len && strng(ii) != "'") || strcmp (strng(ii:ii+1), "''"))
is_lit(ii) = 1;
- ii = ii + 1;
+ ++ii;
if (strcmp (strng(ii-1:ii), '''''')),
is_lit(ii) = 1;
++ii;
@@ -261,5 +285,22 @@
endif
++ii;
endwhile
-
+
endfunction
+
+
+## Read complex (data by Ryusuke Numata)
+%!test
+%! fn = tempname ();
+%! fid = fopen (fn, "w");
+%! fprintf (fid, "&test\n");
+%! fprintf (fid, " z = (1.1,2.2), (0.,1.), (1.,0.)\n");
+%! fprintf (fid, " y = (9,8)\n");
+%! fprintf (fid, " a = 1. 2. 3.\n");
+%! fprintf (fid, [' c = "(test)"' "\n" '\\' "\n"]);
+%! fclose (fid);
+%! nm = read_namelist (fn);
+%! unlink (fn);
+%! assert (nm.test.z, [1.1+2.2i, 1i, 1], eps);
+%! assert (nm.test.y, 9+8i, eps);
+%! assert (nm.test.a, [1 2 3], eps);
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/io/inst/test_spsh.m new/io/inst/test_spsh.m
--- old/io/inst/test_spsh.m 2016-03-10 22:54:05.000000000 +0100
+++ new/io/inst/test_spsh.m 2016-06-28 21:41:05.000000000 +0200
@@ -141,7 +141,7 @@
endfor
endfor
printf ("\n + = correct result returned\n");
- printf ( " 0 = partly correct (e.g., double rather than logical)\n");
+ printf ( " o = partly correct (e.g., double rather than logical)\n");
printf ( " - = erroneous or no result.\n");
printf ("\n- End of test_spsh -\n");
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/io/inst/write_namelist.m new/io/inst/write_namelist.m
--- old/io/inst/write_namelist.m 2016-03-09 22:13:29.000000000 +0100
+++ new/io/inst/write_namelist.m 2016-06-19 23:31:02.000000000 +0200
@@ -31,11 +31,11 @@
## |...
## |--VARNm
##
-## Notes: Only supports variables of type:
+## Notes: Only supports variables of type:
## Scalars, vectors and 2D numeric arrays (integers and floating points)
## Scalars and 1D boolean arrays specified as '.true.' and '.false.' strings
## Single and 1D arrays of strings
-##
+##
## Example:
## NMLST = read_namelist ("OPTIONS.nam");
## NMLST.NAM_FRAC.XUNIF_NATURE = 0.1;
@@ -52,21 +52,21 @@
fid = fopen (filename, "w");
name_lists = fieldnames (S);
n_name_lists = length(name_lists);
-
+
for ii = 1:n_name_lists,
## Write individual namelist records
fprintf (fid, "&%s\n", name_lists{ii});
rcrds = S.(name_lists{ii});
-
+
rcrds_name = fieldnames(rcrds);
n_rcrds = length(rcrds_name);
-
+
for jj = 1:n_rcrds,
var = rcrds.(rcrds_name{jj});
## Find variable type...
if (iscell (var)),
fprintf (fid, " %s =", rcrds_name{jj});
- if (strcmp (var{1}, ".true.") || strcmp (var{1},"'.false.")),
+ if (strcmp (var{1}, ".true.") || strcmp (var{1}, "'.false.")),
for kk = 1:length (var),
fprintf (fid, " %s,", var{kk});
endfor
@@ -81,10 +81,16 @@
if (r == 1 || c == 1)
## Variable is a scalar or vector
fprintf (fid, " %s =", rcrds_name{jj});
- fprintf (fid, " %g,", var);
+ if (iscomplex (var))
+ for i=1:length(var)
+ fprintf (fid, " (%g,%g),", real(var(i)), imag(var(i)));
+ endfor
+ else
+ fprintf (fid, " %g,", var);
+ endif
fprintf (fid, "%s\n", "");
else
- ## Varible is a two dimensional array
+ ## Variable is a two dimensional array
for kk = 1:r,
fprintf (fid, " %s(%i,:) =", rcrds_name{jj}, kk);
fprintf (fid, " %g,", var(kk,:));
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/io/inst/xlsopen.m new/io/inst/xlsopen.m
--- old/io/inst/xlsopen.m 2016-03-09 22:13:29.000000000 +0100
+++ new/io/inst/xlsopen.m 2016-04-23 12:07:41.000000000 +0200
@@ -1,4 +1,4 @@
-## Copyright (C) 2009-2015 Philip Nienhuis
+## Copyright (C) 2009-2016 Philip Nienhuis
##
## This program is free software; you can redistribute it and/or modify
## it under the terms of the GNU General Public License as published by
@@ -105,6 +105,22 @@
error ("xlsopen.m: numerical or logical value expected for arg ## 2
(readwrite)\n")
endif
+ if (ischar (filename))
+ [pth, fnam, ext] = fileparts (filename);
+ if (isempty (fnam))
+ error ("xlsopen.m: no filename or empty filename specified");
+ endif
+ if (xwrite && ! isempty (pth))
+ apth = make_absolute_filename (pth);
+ if (exist (apth) != 7)
+ error ("xlsopen.m: cannot write into non-existent directory:\n'%s'\n",
...
+ apth);
+ endif
+ endif
+ else
+ error ("xlsopen.m: filename expected for argument #1");
+ endif
+
if (! isempty (reqinterface))
intfmsg = "requested";
if (! (ischar (reqinterface) || iscell (reqinterface)))
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/io/inst/xlsread.m new/io/inst/xlsread.m
--- old/io/inst/xlsread.m 2016-03-09 22:13:29.000000000 +0100
+++ new/io/inst/xlsread.m 2016-04-23 12:07:33.000000000 +0200
@@ -1,4 +1,4 @@
-## Copyright (C) 2009-2015 by Philip Nienhuis
+## Copyright (C) 2009-2016 by Philip Nienhuis
##
## This program is free software; you can redistribute it and/or modify it
under
## the terms of the GNU General Public License as published by the Free
Software
@@ -197,19 +197,19 @@
xls = xlsopen (fn, 0, reqintf);
if (! isempty (xls))
xls_ok = 1;
- endif
-
- ## Get data from spreadsheet file & return handle
- [rawarr, xls, rstatus] = xls2oct (xls, wsh, datrange);
-
- ## Save some results before xls is wiped
- rawlimits = xls.limits;
- xtype = xls.xtype;
-
- if (rstatus)
- [numarr, txtarr, lims] = parsecell (rawarr, rawlimits);
- else
- rawarr = {}; numarr = []; txtarr = {};
+
+ ## Get data from spreadsheet file & return handle
+ [rawarr, xls, rstatus] = xls2oct (xls, wsh, datrange);
+
+ ## Save some results before xls is wiped
+ rawlimits = xls.limits;
+ xtype = xls.xtype;
+
+ if (rstatus)
+ [numarr, txtarr, lims] = parsecell (rawarr, rawlimits);
+ else
+ rawarr = {}; numarr = []; txtarr = {};
+ endif
endif
unwind_protect_cleanup
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/io/inst/xmlread.m new/io/inst/xmlread.m
--- old/io/inst/xmlread.m 2016-03-09 22:13:29.000000000 +0100
+++ new/io/inst/xmlread.m 2016-03-22 22:32:02.000000000 +0100
@@ -48,12 +48,7 @@
## This is a Java based function
if (isempty (java_ok))
- try
- HAVE_JAVA = octave_config_info.features.JAVA;
- catch
- HAVE_JAVA = octave_config_info.build_features.JAVA;
- end_try_catch
- if (! HAVE_JAVA)
+ if (! __have_features__ ("JAVA"))
error ("xmlread: Octave was built without Java support, exiting");
elseif (! usejava ("jvm"))
error ("xmlread: no Java JRE or JDK detected, exiting");
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/io/inst/xmlwrite.m new/io/inst/xmlwrite.m
--- old/io/inst/xmlwrite.m 2016-03-09 22:13:29.000000000 +0100
+++ new/io/inst/xmlwrite.m 2016-03-22 22:31:39.000000000 +0100
@@ -51,12 +51,7 @@
## Check Java support and support files
if (isempty (java_ok))
- try
- HAVE_JAVA = octave_config_info.features.JAVA;
- catch
- HAVE_JAVA = octave_config_info.build_features.JAVA;
- end_try_catch
- if (! HAVE_JAVA)
+ if (! __have_features__ ("JAVA"))
error ("xmlwrite: Octave was built without Java support, exiting");
elseif (! usejava ("jvm"))
error ("xmlwrite: no Java JRE or JDK detected, exiting");
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/io/post_install.m new/io/post_install.m
--- old/io/post_install.m 2016-03-09 22:13:29.000000000 +0100
+++ new/io/post_install.m 1970-01-01 01:00:00.000000000 +0100
@@ -1,24 +0,0 @@
-## (C) 2011,2012 Nitzan Arani
-## (C) 2014 Philip Nienhuis <prnienhuis at users.sf.net>
-## This function moves the PKG_ADD & PKG_DEL scripts from the .oct directory
of the io
-## package to the function script file after installation. The reason is that
PKG_ADD
-## calls chk_spreadsheet_support before the function subdir has been added to
Octave's
-## search path.
-
-## Changes:
-## 2014-01-08 Changed to move, rather than remove, PKG_ADD & PKG_DEL
-
-function post_install (desc)
-
- ## Try to move PKG_ADD & PKG_DEL to the io function script subdir to avoid
- ## PKG_ADD invoking chk_spreadsheet_support before the function dir is in
- ## the load path during loading of io pkg (esp. during installation).
-
- files = cstrcat (desc.archprefix, filesep, octave_config_info
("canonical_host_type"),
- "-", octave_config_info ("api_version"), filesep, "PKG_???");
- [status, msg] = movefile (files, desc.dir, 'f');
- if (! status)
- warning ("post_install.m: unable to move PKG_ADD & PKG_DEL to script dir:
%s", msg);
- endif
-
-endfunction
