Hello community,
here is the log from the commit of package octave-forge-doctest for
openSUSE:Factory checked in at 2017-01-29 10:33:28
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Comparing /work/SRC/openSUSE:Factory/octave-forge-doctest (Old)
and /work/SRC/openSUSE:Factory/.octave-forge-doctest.new (New)
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Package is "octave-forge-doctest"
Changes:
--------
---
/work/SRC/openSUSE:Factory/octave-forge-doctest/octave-forge-doctest.changes
2016-06-14 23:07:33.000000000 +0200
+++
/work/SRC/openSUSE:Factory/.octave-forge-doctest.new/octave-forge-doctest.changes
2017-02-03 17:51:42.897841757 +0100
@@ -1,0 +2,9 @@
+Thu Nov 24 14:09:56 UTC 2016 - [email protected]
+
+- Update to version 0.5.0
+ * SKIP_IF and other conditional directives can include small single-line
+ blocks of code.
+ * Recursion into subdirectories is now the default.
+ * In Texinfo mode, skip tests without output by default.
+
+-------------------------------------------------------------------
Old:
----
doctest-0.4.1.tar.gz
New:
----
doctest-0.5.0.tar.gz
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Other differences:
------------------
++++++ octave-forge-doctest.spec ++++++
--- /var/tmp/diff_new_pack.KQlygs/_old 2017-02-03 17:51:43.253791570 +0100
+++ /var/tmp/diff_new_pack.KQlygs/_new 2017-02-03 17:51:43.257791006 +0100
@@ -18,7 +18,7 @@
%define octpkg doctest
Name: octave-forge-%{octpkg}
-Version: 0.4.1
+Version: 0.5.0
Release: 0
Summary: Octave-Forge documentation tests
License: GPL-3.0+ and BSD-3-Clause
@@ -27,7 +27,7 @@
Source0:
http://downloads.sourceforge.net/octave/%{octpkg}-%{version}.tar.gz
BuildRequires: gcc-c++
BuildRequires: octave-devel
-Requires: octave-cli
+Requires: octave-cli >= 4.0.0
%description
The Octave-Forge Doctest package finds specially-formatted blocks of example
++++++ doctest-0.4.1.tar.gz -> doctest-0.5.0.tar.gz ++++++
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/doctest-0.4.1/DESCRIPTION
new/doctest-0.5.0/DESCRIPTION
--- old/doctest-0.4.1/DESCRIPTION 2016-01-04 19:13:29.000000000 +0100
+++ new/doctest-0.5.0/DESCRIPTION 2016-11-14 07:08:14.000000000 +0100
@@ -1,6 +1,6 @@
Name: doctest
-Version: 0.4.1
-Date: 2016-01-04
+Version: 0.5.0
+Date: 2016-11-13
Author: various authors
Maintainer: Colin B. Macdonald <[email protected]>, Michael Walter
<[email protected]>
Title: Documentation tests
@@ -9,5 +9,6 @@
the code and confirms the output is correct. This can be useful as part of
a testing framework or simply to ensure that documentation stays up-to-date
during software development.
+Depends: octave (>= 4.0.0)
Url: https://github.com/catch22/octave-doctest
-License: modified BSD
+License: BSD-3-Clause
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/doctest-0.4.1/NEWS new/doctest-0.5.0/NEWS
--- old/doctest-0.4.1/NEWS 2016-01-04 19:13:29.000000000 +0100
+++ new/doctest-0.5.0/NEWS 2016-11-14 07:08:14.000000000 +0100
@@ -1,3 +1,24 @@
+doctest 0.5.0 (2016-11-13)
+==========================
+
+ * SKIP_IF and other conditional directives can include small single-line
+ blocks of code. For example:
+
+ - "% doctest: +SKIP_IF(foo() && bar(42))"
+
+ * Recursion into subdirectories is now the default; pass "-nonrecursive"
+ for the previous default behaviour.
+
+ * In Texinfo mode, skip tests without output by default. This is an
+ experimental change to help test the GNU Octave project; the feature
+ might disappear without warning in a future version.
+
+ * Workaround regex bug on ARM architecture.
+
+ * Minimum supported Octave version is now 4.0.0.
+
+
+
doctest 0.4.1 (2016-01-04)
==========================
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/doctest-0.4.1/inst/doctest.m
new/doctest-0.5.0/inst/doctest.m
--- old/doctest-0.4.1/inst/doctest.m 2016-01-04 19:13:29.000000000 +0100
+++ new/doctest-0.5.0/inst/doctest.m 2016-11-14 07:08:14.000000000 +0100
@@ -1,5 +1,5 @@
%% Copyright (c) 2010 Thomas Grenfell Smith
-%% Copyright (c) 2011, 2013-2015 Michael Walter
+%% Copyright (c) 2011, 2013-2016 Michael Walter
%% Copyright (c) 2015-2016 Colin B. Macdonald
%%
%% Redistribution and use in source and binary forms, with or without
@@ -30,12 +30,10 @@
%% -*- texinfo -*-
%% @documentencoding UTF-8
-%% @deftypefn {Function File} {} doctest @var{target}
-%% @deftypefnx {Function File} {} doctest @var{target} -recursive
-%% @deftypefnx {Function File} {} doctest @var{target} -DIRECTIVE
-%% @deftypefnx {Function File} {} doctest @var{target} +DIRECTIVE
-%% @deftypefnx {Function File} {@var{success} =} doctest (@var{target},
@dots{})
-%% @deftypefnx {Function File} {[@var{numpass}, @var{numtests}, @var{summary]}
=} doctest (@dots{})
+%% @deftypefn {} {} doctest @var{target}
+%% @deftypefnx {} {} doctest @var{target} -nonrecursive
+%% @deftypefnx {} {@var{success} =} doctest (@var{target}, @dots{})
+%% @deftypefnx {} {[@var{numpass}, @var{numtests}, @var{summary}] =} doctest
(@dots{})
%% Run examples embedded in documentation.
%%
%% Doctest finds and runs code found in @var{target}, which can be a:
@@ -43,8 +41,7 @@
%% @item function;
%% @item class;
%% @item Texinfo file;
-%% @item directory/folder (pass @code{-recursive} to descend
-%% into subfolders);
+%% @item directory/folder (pass @code{-nonrecursive} to skip subfolders);
%% @item cell array of such items.
%% @end itemize
%% When called with a single return value, return whether all tests have
@@ -171,8 +168,8 @@
%% @end example
%%
%% Doctest provides the default flags @code{DOCTEST_OCTAVE} and
-%% @code{DOCTEST_MATLAB}, but you can access arbitrary variables and
-%% (nullary) functions.
+%% @code{DOCTEST_MATLAB}, but you can call functions and access arbitrary
+%% variables (including those defined by previous tests).
%%
%%
%% By default, all adjacent white space is collapsed into a single space
@@ -194,29 +191,35 @@
%%
%% To disable the @code{...} wildcard, use the @code{-ELLIPSIS} directive.
%%
-%% The default directives can be overridden on the command line using, for
-%% example, @code{doctest target -NORMALIZE_WHITESPACE +ELLIPSIS}. Note that
-%% directives local to a test still take precident over these.
-%%
-%%
%% @strong{Diary Style}
%% When the m-file contains plaintext documentation, doctest finds tests
%% by searching for lines that begin with @code{>>}. It then finds the
%% expected output by searching for the next @code{>>} or two blank lines.
%%
%% @strong{Octave/Texinfo Style}
-%% If your m-file contains Texinfo markup, then doctest finds code inside
-%% @code{@@example @dots{} @@end example} blocks. Some comments:
+%% If your m-file contains Texinfo markup, then doctest finds code in
+%% @code{@@example @dots{} @@end example} blocks. Note:
%% @itemize
%% @item The two-blank-lines convention is not required.
-%% @item The use of @code{>>} is not required as Octave documentation
-%% conventionally indicates output with @code{@@print} and
-%% @code{@@result}. Ambiguities are resolving by assuming output
+%% @item The use of @code{>>} is neither required nor recommended as Octave
+%% documentation conventionally indicates output with @code{@@result@{@}}
+%% and @code{@@print@{@}}. Ambiguities are resolving by assuming output
%% is indented further than input.
-%% @item You are free to use diary-style doctests inside
-%% @code{@@example} blocks.
%% @end itemize
%%
+%% A typical Texinfo-style doctest looks like:
+%% @example
+%% a = 5;
+%% b = a + 1
+%% @result{} b = 6
+%% disp("hello\nthere")
+%% @print{} hello
+%% @print{} there
+%% @end example
+%%
+%% The two styles are not mutually exclusive: this documentation is written
+%% in Texinfo using a hybrid approach.
+%%
%% @seealso{test}
%% @end deftypefn
@@ -238,7 +241,7 @@
end
% input parsing for options and directives
-recursive = false;
+recursive = true;
directives = doctest_default_directives();
for i = 1:(nargin-1)
assert(ischar(varargin{i}))
@@ -246,10 +249,18 @@
directive = varargin{i}(2:end);
switch directive
case 'recursive'
+ % weakly deprecated, not mentioned in help text
assert(strcmp(pm, '-'))
recursive = true;
+ case 'nonrecursive'
+ assert(strcmp(pm, '-'))
+ recursive = false;
otherwise
assert(strcmp(pm, '+') || strcmp(pm, '-'))
+ warning('Doctest:deprecated', ...
+ ['Support for specifying directives on the command line is
deprecated\n' ...
+ ' and will be removed in a future version (for
discussion, see\n' ...
+ '
https://github.com/catch22/octave-doctest/issues/127).']);
enable = strcmp(varargin{i}(1), '+');
directives = doctest_default_directives(directives, directive, enable);
end
@@ -262,7 +273,7 @@
[color_ok, color_err, color_warn, reset] = doctest_colors(fid);
% print banner
-fprintf(fid, 'Doctest v0.4.1: this is Free Software without warranty, see
source.\n\n');
+fprintf(fid, 'Doctest v0.5.0: this is Free Software without warranty, see
source.\n\n');
summary = struct();
@@ -275,7 +286,7 @@
for i=1:numel(what)
- summary = doctest_collect(what{i}, directives, summary, recursive, fid);
+ summary = doctest_collect(what{i}, directives, summary, recursive, 0, fid);
end
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/doctest-0.4.1/inst/private/doctest_collect.m
new/doctest-0.5.0/inst/private/doctest_collect.m
--- old/doctest-0.4.1/inst/private/doctest_collect.m 2016-01-04
19:13:29.000000000 +0100
+++ new/doctest-0.5.0/inst/private/doctest_collect.m 2016-11-14
07:08:14.000000000 +0100
@@ -1,4 +1,4 @@
-function summary = doctest_collect(what, directives, summary, recursive, fid)
+function summary = doctest_collect(what, directives, summary, recursive,
depth, fid)
% Find and run doctests.
%
% The parameter WHAT is the name of a class, directory, function or filename:
@@ -70,9 +70,20 @@
% Deal with directories
if (strcmp(type, 'dir'))
- %if (~ strcmp(what, '.'))
- % fprintf(fid, 'Descending into directory "%s"\n', what);
- %end
+ if (strcmp(what, '.'))
+ if (depth == 0)
+ % cheap hack to not indent when calling "doctest ."
+ depth = -1;
+ end
+ else
+ spaces = repmat(' ', 1, 2*depth);
+ if (strcmp(what(end), filesep()))
+ slashchar = '';
+ else
+ slashchar = filesep();
+ end
+ fprintf(fid, '%s%s%s\n', spaces, what, slashchar);
+ end
oldcwd = chdir(what);
files = dir('.');
for i=1:numel(files)
@@ -97,7 +108,7 @@
continue
end
end
- summary = doctest_collect(f, directives, summary, recursive, fid);
+ summary = doctest_collect(f, directives, summary, recursive, depth + 1,
fid);
end
chdir(oldcwd);
return
@@ -110,21 +121,26 @@
% TARGETS(i).link Hyperlink to test for use in Matlab.
% TARGETS(i).docstring Associated docstring.
% TARGETS(i).error: Contains error string if extraction failed.
+% TARGETS(i).depth How "deep" in a recursive traversal are we
if strcmp(type, 'function')
- targets = collect_targets_function(what);
+ target = collect_targets_function(what);
+ target.depth = depth;
+ targets = [target];
elseif strcmp(type, 'class')
- targets = collect_targets_class(what);
+ targets = collect_targets_class(what, depth);
elseif strcmp(type, 'texinfo')
target = struct();
target.name = what;
target.link = '';
+ target.depth = depth;
[target.docstring, target.error] = parse_texinfo(fileread(what));
targets = [target];
else
target = struct();
target.name = what;
target.link = '';
+ target.depth = depth;
target.docstring = '';
target.error = 'Function or class not found.';
targets = [target];
@@ -141,7 +157,9 @@
for i=1:numel(targets)
% run doctests for target and update statistics
target = targets(i);
- fprintf(fid, '%s %s ', target.name, repmat('.', 1, 55 - numel(target.name)));
+ spaces = repmat(' ', 1, 2*target.depth);
+ dots = repmat('.', 1, 55 - numel(target.name) - 2*target.depth);
+ fprintf(fid, '%s%s %s ', spaces, target.name, dots);
% extraction error?
if target.error
@@ -210,7 +228,7 @@
end
-function targets = collect_targets_class(what)
+function targets = collect_targets_class(what, depth)
if (strcmp(what(1), '@'))
% Octave methods('@foo') gives java error, Matlab just says "No methods"
what = what(2:end);
@@ -224,6 +242,7 @@
else
target.link = sprintf('<a
href="matlab:editorservices.openAndGoToLine(''%s'', 1);">%s</a>', which(what),
what);
end
+ target.depth = depth;
[target.docstring, target.error] = extract_docstring(target.name);
targets = target;
@@ -238,6 +257,7 @@
target.name = sprintf('%s.%s', what, meths{i});
target.link = sprintf('<a
href="matlab:editorservices.openAndGoToFunction(''%s'', ''%s'');">%s</a>',
which(what), meths{i}, target.name);
end
+ target.depth = depth;
[target.docstring, target.error] = extract_docstring(target.name);
targets = [targets; target];
end
@@ -265,7 +285,7 @@
end
else
format
- warning('Unexpected format in that file/function');
+ warning('Doctest:unexpected-format', 'Unexpected format in that
file/function');
error = '';
end
else
@@ -299,20 +319,21 @@
% Mark the occurrence of “@example” and “@end example” to be able to find
% example blocks after conversion from texi to plain text. Also consider
% indentation, so we can later correctly unindent the example's content.
+ % Note: uses “@example” instead of “$2” to avoid ARM-specific bug #130.
str = regexprep (str, ...
- '^([ \t]*)(@example)(.*)$', ...
- [ '$1$2$3\n', ... % retain original line
+ '^([ \t]*)(\@example)(.*)$', ...
+ [ '$1\@example$3\n', ... % retain original line
'$1###### EXAMPLE START ######'], ...
'lineanchors', 'dotexceptnewline', 'emptymatch');
str = regexprep (str, ...
- '^([ \t]*)(@end example)(.*)$', ...
+ '^([ \t]*)(\@end example)(.*)$', ...
[ '$1###### EXAMPLE STOP ######\n', ...
- '$1$2$3'], ... % retain original line
+ '$1\@end example$3'], ... % retain original line
'lineanchors', 'dotexceptnewline', 'emptymatch');
% special comments "@c doctest: cmd" are translated
% FIXME the expression would also match @@c doctest: ...
- re = [ '(?:@c(?:omment)?\s' ... % @c or @comment, ?: means no token
+ re = [ '(?:\@c(?:omment)?\s' ... % @c or @comment, ?: means no token
'|#|%)\s*' ... % or one of #,%
'(doctest:\s*.*)' ]; % want the doctest token
str = regexprep (str, re, '% $1', 'dotexceptnewline');
@@ -321,7 +342,7 @@
% the __makeinfo__ function exists in Octave only.
[str, err] = eval('__makeinfo__ (str, ''plain text'')');
if (err ~= 0)
- error = '__makeinfo__ returned with error code'
+ error = sprintf('__makeinfo__ returned error code %d', err);
return
end
@@ -363,17 +384,30 @@
return
end
- % split into lines
- L = strsplit (T{i}, '\n');
-
if (regexp (T{i}, '^\s*>>', 'once'))
% First nonblank line starts with '>>': assume diary style. However,
% we strip @result and @print macros (TODO: perhaps unwisely?)
+ L = strsplit (T{i}, '\n');
L = regexprep (L, '^(\s*)(?:⇒|=>|⊣|-\|)', '$1', 'once', 'lineanchors');
T{i} = strjoin (L, '\n');
continue
end
+
+ % Hack: the @example block is commonly mis-used to store non-examples such
as
+ % diagrams or math. Delete an example block that has no indicated output.
+ % (Hard to leave for "later" as we don't keep track of @example blocks.)
+ R1 = regexp (T{i}, '^\s*(⇒|=>|⊣|-\|)', 'lineanchors');
+ R2 = regexp (T{i}, '(doctest:\s+-TEXINFO_SKIP_BLOCKS_WO_OUTPUT)');
+ T{i} = regexprep (T{i}, '(doctest:\s+-TEXINFO_SKIP_BLOCKS_WO_OUTPUT)', '');
+ if (isempty (R1) && isempty (R2))
+ T{i} = '';
+ continue
+ end
+
+ % split into lines
+ L = strsplit (T{i}, '\n');
+
% Categorize input and output lines in the example using
% @result and @print macros. Everything else, including comment lines and
% empty lines, is categorized as input (for now).
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/doctest-0.4.1/inst/private/doctest_compare.m
new/doctest-0.5.0/inst/private/doctest_compare.m
--- old/doctest-0.4.1/inst/private/doctest_compare.m 2016-01-04
19:13:29.000000000 +0100
+++ new/doctest-0.5.0/inst/private/doctest_compare.m 2016-11-14
07:08:14.000000000 +0100
@@ -1,10 +1,13 @@
function match = doctest_compare(want, got, normalize_whitespace, ellipsis)
-% Matches two strings together. They should be identical, except:
+%DOCTEST_COMPARE check if two strings match
%
+% Returns true if string GOT matches the template string WANT. Basically
+% WANT and GOT should be identical, except:
+%
+% * whitespace at the start/end of each line is trimmed;
% * multiple spaces are collapsed (if NORMALIZE_WHITESPACE is true);
-% * the first one can contain '...', which matches anything in the
-% second (if ELLIPSIS is true);
-% * they might match after putting "ans = " on the first;
+% * WANT can have "..."; matches anything in GOT (if ELLIPSIS is true);
+% * WANT can omit "ans = ";
% * various other nonsense of unknown current relevance.
%
@@ -16,32 +19,38 @@
% WHY do they need backspaces? huh.
got = regexprep(got, '.\x08', '');
-% collapse multiple spaces to one
-if normalize_whitespace
- want = strtrim(regexprep(want, '\s+', ' '));
- got = strtrim(regexprep(got, '\s+', ' '));
-else
- want = strtrim(strtrim_lines_discard_empties(want));
- got = strtrim(strtrim_lines_discard_empties(got));
-end
+ want = strtrim(want);
+ got = strtrim(got);
-if isempty(got) && (isempty(want) || (ellipsis && strcmp(want, '...')))
+ if (isempty(got) && (isempty(want) || (ellipsis && strcmp(want, '...'))))
match = 1;
return
-end
+ end
-want_re = regexptranslate('escape', want);
-if ellipsis
- want_re = regexprep(want_re, '(\\\.){3}', '.*');
-end
+ want = regexptranslate('escape', want);
+ if normalize_whitespace
+ % collapse multiple spaces, then have each match many
+ if is_octave && compare_versions (OCTAVE_VERSION, '4.1', '<')
+ want = regexprep(want, '\s+', '\s+');
+ else
+ want = regexprep(want, '\s+', '\\s\+');
+ end
+ else
+ want = strtrim_lines_discard_empties(want);
+ got = strtrim_lines_discard_empties(got);
+ end
+
+ if ellipsis
+ want = regexprep(want, '(\\\.){3}', '.*');
+ end
-% allow "ans = " to be missing
-want_re = ['^(ans\s*=\s*)?' want_re '$'];
+ % allow "ans = " to be missing
+ want = ['^(ans\s*=\s*)?' want '$'];
-result = regexp(got, want_re, 'once');
+ result = regexp(got, want, 'once');
-match = ~ isempty(result);
+ match = ~ isempty(result);
end
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/doctest-0.4.1/inst/private/doctest_run.m
new/doctest-0.5.0/inst/private/doctest_run.m
--- old/doctest-0.4.1/inst/private/doctest_run.m 2016-01-04
19:13:29.000000000 +0100
+++ new/doctest-0.5.0/inst/private/doctest_run.m 2016-11-14
07:08:14.000000000 +0100
@@ -48,12 +48,18 @@
tests(i).xfail = {};
% find and process directives
- directive_matches = regexp(tests(i).source,
'(?:#|%)\s*doctest:\s+([(\+|\-)][\w]+)(\([\w]+\))?', 'tokens');
+ re = [ ...
+ '[#%]\s*doctest:\s+' ... % e.g., "# doctest: "
+ '([\+\-]\w+)' ... % token for cmd, e.g., "+XSKIP_IF"
+ '(\s*\(' ... % token for code, starting with "("
+ '[^#%\n]+' ... % no newlines, no comments in code
+ '\))?']; % ")" of code, at most one code arg
+ directive_matches = regexp(tests(i).source, re, 'tokens');
for j = 1:length(directive_matches)
directive = directive_matches{j}{1};
if (strcmp('+SKIP_IF', directive) || strcmp('+SKIP_UNLESS', directive) ||
strcmp('+XFAIL_IF', directive) || strcmp('+XFAIL_UNLESS', directive))
if length(directive_matches{j}) == 2
- condition = directive_matches{j}{2}(2:end - 1);
+ condition = directive_matches{j}{2};
else
error('doctest: syntax error, expected %s(varname)', directive);
end
@@ -76,7 +82,7 @@
elseif strcmp('+XFAIL_UNLESS', directive)
tests(i).xfail{end + 1} = sprintf('~(%s)', condition);
else
- error('doctest: unexpected directive %s', directive);
+ warning('Doctest:unexpected-directive', 'doctest: ignoring unexpected
directive %s', directive);
end
end
end
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/doctest-0.4.1/octave-doctest.metainfo.xml
new/doctest-0.5.0/octave-doctest.metainfo.xml
--- old/doctest-0.4.1/octave-doctest.metainfo.xml 2016-01-04
19:13:29.000000000 +0100
+++ new/doctest-0.5.0/octave-doctest.metainfo.xml 2016-11-14
07:08:14.000000000 +0100
@@ -1,11 +1,35 @@
<?xml version="1.0" encoding="UTF-8"?>
+<!--
+Copyright 2016 Colin B. Macdonald
+
+Copying and distribution of this file, with or without modification,
+are permitted in any medium without royalty provided the copyright
+notice and this notice are preserved. This file is offered as-is,
+without any warranty.
+-->
<component type="addon">
<id>octave-doctest</id>
<extends>www.octave.org-octave.desktop</extends>
<name>Doctest</name>
<summary>Finds and tests example code within documentation</summary>
+ <description>
+ <p>Finds specially-formatted blocks of example code within
+ documentation files. Then execute that code and confirm the
+ output is correct. This can be useful as part of a testing
+ framework or simply to ensure that documentation stays up-to-date
+ during software development.</p>
+ </description>
+ <keywords>
+ <keyword>development</keyword>
+ <keyword>software testing</keyword>
+ <keyword>testing</keyword>
+ <keyword>documentation</keyword>
+ </keywords>
+ <translation/>
<url type="homepage">http://octave.sourceforge.net/doctest</url>
<url
type="bugtracker">https://github.com/catch22/octave-doctest/issues/new</url>
- <metadata_license>CC0-1.0</metadata_license>
+ <metadata_license>FSFAP</metadata_license>-->
<project_license>BSD-3-Clause</project_license>
+ <developer_name>Octave-Forge Community</developer_name>
+ <update_contact>[email protected]</update_contact>
</component>
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/doctest-0.4.1/test/test_angle_brackets.m
new/doctest-0.5.0/test/test_angle_brackets.m
--- old/doctest-0.4.1/test/test_angle_brackets.m 2016-01-04
19:13:29.000000000 +0100
+++ new/doctest-0.5.0/test/test_angle_brackets.m 2016-11-14
07:08:14.000000000 +0100
@@ -1,9 +1,13 @@
function s = test_angle_brackets()
% https://savannah.gnu.org/bugs/?45084 (Fixed in Octave 4.0)
%
-% >> s = test_angle_brackets()
+% Fails on 3.8
+% >> oct38 = DOCTEST_OCTAVE && compare_versions(OCTAVE_VERSION, '4.0.0', '<');
+%
+%
+% >> s = test_angle_brackets() % doctest: +XFAIL_IF(oct38)
% s = I <3 U
-% >> s = '<p>I heart you</p>'
+% >> s = '<p>I heart you</p>' % doctest: +XFAIL_IF(oct38)
% s = <p>I heart you</p>
s = 'I <3 U';
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/doctest-0.4.1/test/test_comments_with_directives.m
new/doctest-0.5.0/test/test_comments_with_directives.m
--- old/doctest-0.4.1/test/test_comments_with_directives.m 1970-01-01
01:00:00.000000000 +0100
+++ new/doctest-0.5.0/test/test_comments_with_directives.m 2016-11-14
07:08:14.000000000 +0100
@@ -0,0 +1,11 @@
+function test_comments_with_directives()
+% >> a = 6 %doctest: +XFAIL % comment (with
parenthetical)
+% b = 5
+%
+%
+% >> a = 7 % doctest: +XFAIL_IF (true) % comment
+% b = 5
+%
+%
+% >> a = 8 % doctest: +XFAIL_IF (false) % comment (with
parenthetical)
+% a = 8
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/doctest-0.4.1/test/test_ellipsis_match_empty.m
new/doctest-0.5.0/test/test_ellipsis_match_empty.m
--- old/doctest-0.4.1/test/test_ellipsis_match_empty.m 1970-01-01
01:00:00.000000000 +0100
+++ new/doctest-0.5.0/test/test_ellipsis_match_empty.m 2016-11-14
07:08:14.000000000 +0100
@@ -0,0 +1,20 @@
+function test_ellipsis_match_empty()
+% Ellipses should match empty string:
+% >> disp('abcdef')
+% abc...def
+%
+%
+% empty at ends:
+% >> disp('def')
+% ...def
+%
+% >> disp('def')
+% def...
+%
+%
+% Empty and whitespace:
+% >> disp('abc def')
+% abc ...def
+%
+% >> disp('abc def')
+% abc... def
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/doctest-0.4.1/test/test_ellipsis_match_whitespace.m
new/doctest-0.5.0/test/test_ellipsis_match_whitespace.m
--- old/doctest-0.4.1/test/test_ellipsis_match_whitespace.m 1970-01-01
01:00:00.000000000 +0100
+++ new/doctest-0.5.0/test/test_ellipsis_match_whitespace.m 2016-11-14
07:08:14.000000000 +0100
@@ -0,0 +1,36 @@
+function test_ellipsis_match_whitespace()
+% Whitespace in middle
+% >> disp('abc def')
+% abc ... def
+% >> disp('abc def')
+% abc ... def
+% >> disp('abc def')
+% abc...def
+%
+%
+% Should fail: expects something surrounded by whitespace
+% >> disp('abc def') % doctest: +XFAIL
+% abc ... def
+%
+%
+% This is ok, because there are two whitespaces in input
+% >> disp('abc def')
+% abc ... def
+%
+%
+% Currently, ellipses will match empty the string but we trim begin/end of
+% lines, so these probably fail because there is nothing to match the space
+% after/before the "...". Probably ok to change this behaviour.
+% >> disp(' def') % doctest: +XFAIL
+% ... def
+%
+% >> disp('def ') % doctest: +XFAIL
+% def ...
+%
+%
+% However, these are ok:
+% >> disp('def')
+% ...def
+%
+% >> disp('def')
+% def...
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/doctest-0.4.1/test/test_multi_directives.m
new/doctest-0.5.0/test/test_multi_directives.m
--- old/doctest-0.4.1/test/test_multi_directives.m 1970-01-01
01:00:00.000000000 +0100
+++ new/doctest-0.5.0/test/test_multi_directives.m 2016-11-14
07:08:14.000000000 +0100
@@ -0,0 +1,36 @@
+function test_multi_directives()
+% Dummy:
+% >> z = 0;
+%
+%
+% Will fail if we do not skip
+% >> s = 'a b' % doctest: +SKIP_IF (1 > 0) % doctest:
-NORMALIZE_WHITESPACE
+% .. z = 1;
+% s = a b
+%
+%
+% and we did skip
+% >> z
+% z = 0
+%
+%
+%
+% Will fail if we do not skip
+% >> s = 'a b' % my comment % doctest: -NORMALIZE_WHITESPACE %
doctest: +SKIP_IF (true) % other comment
+% .. z = 2;
+% s = a b
+%
+%
+% and we did skip:
+% >> z
+% z = 0
+%
+%
+%
+% >> s = 'a ...23 b' % doctest: -ELLIPSIS % doctest:
+NORMALIZE_WHITESPACE % doctest: +XFAIL
+% .. z = 3;
+% s = a ... b
+%
+%
+% >> z
+% z = 3
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/doctest-0.4.1/test/test_skip_wo_output.texinfo
new/doctest-0.5.0/test/test_skip_wo_output.texinfo
--- old/doctest-0.4.1/test/test_skip_wo_output.texinfo 1970-01-01
01:00:00.000000000 +0100
+++ new/doctest-0.5.0/test/test_skip_wo_output.texinfo 2016-11-14
07:08:14.000000000 +0100
@@ -0,0 +1,61 @@
+In the Octave community, the @code{@@example} block is commonly used
+non-semantically. While we do not endorse this, we do like dem ascii arts:
+@example
+ _ _ _ __ _ _
+ __| | ___ ___| |_ ___ ___| |_ / _| |___ _| |
+ / _` |/ _ \ / __| __/ _ \/ __| __| | |_| __\ \ /\ / / |
+ | (_| | (_) | (__| || __/\__ \ |_ | _| |_ \ V V /|_|
+ \__,_|\___/ \___|\__\___||___/\__| |_| \__| \_/\_/ (_)
+
+@end example
+
+
+(for later use)
+@example
+a = 4
+ @result{} a = 4
+@end example
+
+
+Unfortunately, as there is no @strong{specified} output, this block
+is skipped too:
+@example
+a = 5;
+assert(false)
+@end example
+
+It really was skipped:
+@example
+a
+ @result{} a = 4
+@end example
+
+
+``But wait'' you say, ``I want to test a block with no output!''
+First, please patch Octave to use something like @code{@@verbatim}
+for diagrams, etc. Then patch doctest to remove this ``feature''.
+
+Done all that? No? Fine, well because we like you, there is
+a secret directive for you. It could disappear without notice
+in any future versions (we don't like you @strong{that} much).
+Try these:
+@example
+@comment doctest: -TEXINFO_SKIP_BLOCKS_WO_OUTPUT
+a = 5;
+@end example
+
+@example
+@comment doctest: +XFAIL
+@comment doctest: -TEXINFO_SKIP_BLOCKS_WO_OUTPUT
+b = 6;
+disp('do not skip me bro')
+@end example
+(note this test fails because it @strong{does} produce output).
+
+And indeed they were not skipped:
+@example
+a
+ @result{} a = 5
+b
+ @result{} b = 6
+@end example
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/doctest-0.4.1/test/test_skip_wo_output_diary.m
new/doctest-0.5.0/test/test_skip_wo_output_diary.m
--- old/doctest-0.4.1/test/test_skip_wo_output_diary.m 1970-01-01
01:00:00.000000000 +0100
+++ new/doctest-0.5.0/test/test_skip_wo_output_diary.m 2016-11-14
07:08:14.000000000 +0100
@@ -0,0 +1,12 @@
+function test_skip_wo_output_diary()
+%
+% No output is specified here, so this code block would likely be
+% skipped if this was texinfo. But its not texinfo so this test
+% is expected to fail (and its a bug if it does not).
+% >> a = 5; % doctest: +XFAIL
+% .. disp('hi')
+%
+%
+% check that previous was not skipped
+% >> a
+% a = 5
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/doctest-0.4.1/test/test_skip_wo_output_diary.texinfo
new/doctest-0.5.0/test/test_skip_wo_output_diary.texinfo
--- old/doctest-0.4.1/test/test_skip_wo_output_diary.texinfo 1970-01-01
01:00:00.000000000 +0100
+++ new/doctest-0.5.0/test/test_skip_wo_output_diary.texinfo 2016-11-14
07:08:14.000000000 +0100
@@ -0,0 +1,19 @@
+This file has diary-style within texinfo. Potential bugs here because
+of the empty block skipping feature. First, a block with no specified output:
it
+should fail because it does have output:
+@example
+>> a = 5;
+>> disp('hi'); % doctest: +XFAIL
+@end example
+
+@example
+>> a
+a = 5
+@end example
+
+The first command below should not invoke skipping:
+@example
+>> x = 5;
+>> y = x + 1
+y = 6
+@end example
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/doctest-0.4.1/test/test_xfail_if_code.m
new/doctest-0.5.0/test/test_xfail_if_code.m
--- old/doctest-0.4.1/test/test_xfail_if_code.m 1970-01-01 01:00:00.000000000
+0100
+++ new/doctest-0.5.0/test/test_xfail_if_code.m 2016-11-14 07:08:14.000000000
+0100
@@ -0,0 +1,25 @@
+function test_xfail_if_code()
+%
+% Initialize a dummy variable a:
+% >> a = 3
+% a = 3
+%
+%
+% The following test will fail:
+% >> a = 5 % doctest: +XFAIL_IF(6 + 0*now() >= 0)
+% b = 7
+%
+%
+% Check that it has been executed:
+% >> a
+% a = 5
+%
+%
+% This test succeeds and should not fail:
+% >> a % doctest: +XFAIL_IF(str2num('17') > 20)
+% a = 5
+%
+%
+% There can be a space before the parenthesis
+% >> a % doctest: +XFAIL_IF (str2num('17') > 20)
+% a = 5
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/doctest-0.4.1/test/test_xfail_if_code.texinfo
new/doctest-0.5.0/test/test_xfail_if_code.texinfo
--- old/doctest-0.4.1/test/test_xfail_if_code.texinfo 1970-01-01
01:00:00.000000000 +0100
+++ new/doctest-0.5.0/test/test_xfail_if_code.texinfo 2016-11-14
07:08:14.000000000 +0100
@@ -0,0 +1,18 @@
+Dummy variable:
+@example
+a = 3
+ @result{} a = 3
+@end example
+
+The following test will fail:
+@example
+@c doctest: +XFAIL_IF(6 + 0*now () >= 0)
+a = 5
+ @result{} b = 7
+@end example
+
+value is changed (b/c xfail, not skip)
+@example
+a
+ @result{} a = 5
+@end example
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/doctest-0.4.1/test/test_xskip_if_code.texinfo
new/doctest-0.5.0/test/test_xskip_if_code.texinfo
--- old/doctest-0.4.1/test/test_xskip_if_code.texinfo 1970-01-01
01:00:00.000000000 +0100
+++ new/doctest-0.5.0/test/test_xskip_if_code.texinfo 2016-11-14
07:08:14.000000000 +0100
@@ -0,0 +1,36 @@
+Dummy variable:
+@example
+a = 3
+ @result{} a = 3
+@end example
+
+The regex for SKIP_IF argument should not cross newlines! First, note its ok
+to have syntax errors in skipped tests:
+@example
+@c doctest: +SKIP
+a = 5*(5+1))
+ @result{} b = 0
+@end example
+
+But next we have a syntax error, combined with SKIP_IF code. The IF code
+should not glob the @code{a = } bit: might result in a syntax error if it did.
+@example
+@c doctest: +SKIP_IF(6 + 0*now () >= 0)
+a = 3*(4+1))
+ @result{} a = 15
+@end example
+
+As above, but two SKIP_IF's
+@example
+@c doctest: +SKIP_IF(6 + 0*now () >= 0)
+@c doctest: +SKIP_IF(7 + 0*now () >= 0)
+a = 3*(4+1))
+ @result{} a = 15
+@end example
+
+
+Finally, a is still 3
+@example
+a
+ @result{} a = 3
+@end example
