function ostr = mlstripcommentsstr(istr)
%MLSTRIPCOMMENTSSTR Strip comments from a string with MATLAB code.
%
% OSTR = MLSTRIPCOMMENTSSTR(ISTR) takes the input string ISTR, assuming it
% contains MATLAB code, strips all MATLAB comments, and returns the result as
% the output string OSTR. The input string must be a row vector or a cell
% array of row vectors. Any input string may be a multi-line string, that is
% a string containing newline characters, for example a string containing the
% entire contents of an m-file.
%
% Example: Read all MATLAB code from an m-file opened for reading on file
% identifier "fid" and strip all comments:
%
% str = fread(fid); % data as numerical column vector
% str = char(str)'; % data as character row vector
% str = mlstripcommentsstr(str); % strip all MATLAB comments
%
% See also MLSTRIPCOMMENTSFILE, MLSTRIPCOMMENTSFID.
% Author: Peter J. Acklam
% Time-stamp: 2004-03-18 08:23:19 +0100
% E-mail: pjacklam@online.no
% URL: http://home.online.no/~pjacklam
% Regex for removing a comment on a line possibly containing MATLAB code.
%
% Use a persistent variable so that the variable is set up only once.
%
% Find the part of the line which is not a comment and then replace the
% whole line with the part that was not a comment.
%
% A comment is started by a percent sign, but not all percent signs start a
% comment. A percent sign might be embedded in a string, in which case the
% percent sign obviously does not start a comment. So we must look for
% percent signs which are not embedded in strings. To do this we must find
% all strings. Strings are delimited by single quotes, but not all single
% quotes start strings. A single quote might also be a tranpose operator.
% So we must look for single quotes which are not tranpose operators.
%
% The MATLAB grammar (syntax and semantics) is not available to the public,
% but from my knowledge, a single quote is a transpose operator if and only
% if one of the following criteria is satisfied
%
% 1) it follows a right closing delimiter
% 2) it follows right after a variable name or function name
% 3) it follows right after a period (as part of the ".'" operator)
% 4) it follows right after another tranpose operator
%
% This can be immediately after a right closing bracket ("]"), right closing
% parenthesis (")"), right closing brace ("}"), an English letter ("a" to
% "z" and "A" to "Z", a digit ("0" to "9"), an underline ("_"), a period
% ("."), or another single quote ("'").
persistent mainregex
if isempty(mainregex)
mainregex = [ ...
' ( ' ... % Grouping parenthesis (content goes to $1).
' ( ^ | \n ) ' ... % Beginning of string or beginning of line.
' ( ' ... % Non-capturing grouping parenthesis.
' ' ...
'' ... % Match anything that is neither a comment nor a string...
' ( ' ... % Non-capturing grouping parenthesis.
' [\]\)}\w.]' ... % Either a character followed by
' ''+ ' ... % one or more transpose operators
' | ' ... % or else
' [^''%] ' ... % any character except single quote (which
' ' ... % starts a string) or a percent sign (which
' ' ... % starts a comment).
' )+ ' ... % Match one or more times.
' ' ...
'' ... % ...or...
' | ' ...
' ' ...
'' ... % ...match a string.
' '' ' ... % Opening single quote that starts the string.
' [^''\n]* ' ... % Zero or more chars that are neither single
' ' ... % quotes (special) nor newlines (illegal).
' ( ' ... % Non-capturing grouping parenthesis.
' '''' ' ... % An embedded (literal) single quote character.
' [^''\n]* ' ... % Again, zero or more chars that are neither
' ' ... % single quotes nor newlines.
' )* ' ... % Match zero or more times.
' '' ' ... % Closing single quote that ends the string.
' ' ...
' )* ' ... % Match zero or more times.
' ) ' ...
' [^\n]* ' ... % What remains must be a comment.
];
% Remove all the blanks from the regex.
mainregex = mainregex(~isspace(mainregex));
end
% Initialize output.
ostr = istr;
% If input is a character array, put it into a cell array. We'll later make
% sure output is a character array if input is a character array.
if ischar(ostr)
ostr = {ostr};
end
% Iterate over each element in the cell array.
for i = 1 : numel(ostr)
% Get the ith input string.
str = ostr{i};
% All character arrays must be row vectors.
if (ndims(str) > 2) | (any(size(str) > 0) & (size(str, 1) ~= 1))
error('All character arrays must be row vectors.');
end
% Note: With MATLAB 6.5.1 it seems that the '^' and '$' anchors in the
% regexes '(^|\n)' and '($|\n)' are not working (bug?), so use a
% workaround: prepend and append a newline and remove them later.
use_workaround = 1;
if use_workaround
lf = sprintf('\n'); % LF = Line feed character
str = [lf, str, lf]; % prepend and append an LF
end
% Remove comment lines.
str = regexprep(str, '(^|\n)[^\S\n]*%[^\n]*', '$1', 'tokenize');
% Remove a comment on a line possibly containing MATLAB code.
str = regexprep(str, mainregex, '$1', 'tokenize');
% Remove trailing whitespace from each line.
str = regexprep(str, '[^\S\n]+($|\n)', '$1', 'tokenize');
% Compress multiple blank lines.
%str = regexprep(str, '\n\n\n+', sprintf('\n\n'));
if use_workaround
str = str(2:end-1); % remove leading and trailing LF
end
% Insert ith string into output cell array.
ostr{i} = str;
end
% If the input was a character array, make sure output is so too.
if ischar(istr)
ostr = ostr{1};
end
