Incorrect handling of semicolon after command syntax

[rlivings39]

From thedavidprice/language-matlab-octave#19

Consider he code:

x = 1;
someCommand arg1; x = 1;

should treat x = 1; in the second line as typical code not as a string as part of the command. Notice that the 1 is not highlighted in the second line.

The difference is more pronounced in Atom with the language-matlab-octave extension:

[ghost]

MATLAB-Language-grammar/Matlab.tmbundle/Syntaxes/MATLAB.tmLanguage

Lines 908 to 932 in 69dbf20

	<key>command_dual</key>
	<dict>
	<key>captures</key>
	<dict>
	<key>1</key>
	<dict>
	<key>name</key>
	<string>string.interpolated.matlab</string>
	</dict>
	<key>2</key>
	<dict>
	<key>name</key>
	<string>variable.other.command.matlab</string>
	</dict>
	<key>28</key>
	<dict>
	<key>name</key>
	<string>comment.line.percentage.matlab</string>
	</dict>
	</dict>
	<key>comment</key>
	<string> 1 2 3 4 5 6 7 8 9 10 11 12 13 14 1516 17 18 19 20 21 22 23 24 25 26 27 28</string>
	<key>match</key>
	<string>^\s*((?# A&gt; )([b-df-hk-moq-zA-HJ-MO-Z]\w*|a|an|a([A-Za-mo-z0-9_]\w*|n[A-Za-rt-z0-9_]\w*|ns\w+)|e|ep|e([A-Za-oq-z0-9_]\w*|p[A-Za-rt-z0-9_]\w*|ps\w+)|in|i([A-Za-mo-z0-9_]\w*|n[A-Za-eg-z0-9_]\w*|nf\w+)|I|In|I([A-Za-mo-z0-9_]\w*|n[A-Za-eg-z0-9_]\w*|nf\w+)|j\w+|N|Na|N([A-Zb-z0-9_]\w*|a[A-MO-Za-z0-9_]\w*|aN\w+)|n|na|nar|narg|nargi|nargo|nargou|n([A-Zb-z0-9_]\w*|a([A-Za-mo-qs-z0-9_]\w*|n\w+|r([A-Za-fh-z0-9_]\w*|g([A-Za-hj-nq-z0-9_]\w*|i([A-Za-mo-z0-9_]\w*|n\w+)|o([A-Za-tv-z0-9_]\w*|u([A-Za-su-z]\w*|t\w+))))))|p|p[A-Za-hj-z0-9_]\w*|pi\w+)(?# &lt;A )\s+(((?# B&gt; )([^\s;,%()=.{&amp;|~&lt;&gt;:+\-*/\\@^'"]|(?=')|(?="))(?# &lt;B )|(?# C&gt; )(\.\^|\.\*|\./|\.\\|\.'|\.\(|&amp;&amp;|==|\|\||&amp;(?=[^&amp;])|\|(?=[^\|])|~=|&lt;=|&gt;=|~(?!=)|&lt;(?!=)|&gt;(?!=)|:|\+|-|\*|/|\\|@|\^)(?# &lt;C )(?# D&gt; )([^\s]|\s*(?=%)|\s+$|\s+(,|;|\)|}|\]|&amp;|\||&lt;|&gt;|=|:|\*|/|\\|\^|@|(\.[^\d.]|\.\.[^.])))(?# &lt;D )|(?# E&gt; )(\.[^^*/\\'(\sA-Za-z])(?# &lt;E ))(?# F&gt; )([^%]|'[^']*'|"[^"]*")*(?# &lt;F )|(?# X&gt; )(\.(?=\s)|\.[A-Za-z]|(?={))(?# &lt;X )(?# Y&gt; )([^(=\'"%]|==|'[^']*'|"[^"]*"|\(|\([^)%]*\)|\[|\[[^\]%]*\]|{|{[^}%]*})*(\.\.\.[^%]*)?((?=%)|$)(?# &lt;Y )))(%.*)?$</string>
	</dict>

Looking at the help page, this seems overcomplicated. Dual command syntax is simple.. there's a keyword.other/support.function word and a char vector option at the end (terminated by [\n;%]). A check for missing = or ( after the "word" is the only thing needed.

I'll see what comes of hacking at it

[ghost]

Bugfix @ PR #37.

[ghost]

Could you ping your colleague about the PR on Slack? I've corrected this issue and added enhancements: #37 (comment)

[ghost]

Resolved by #37. This issue needs closing

[dklilley]

Closing - resolved by #37