Looking for cool hints and tricks to make code smaller/better

<p>I am trying to learn Go right now and I am looking for some hints from the pros on how to optimize code in terms of LOC as well as performance/efficiency. An example of code I was working on here: <a href="http://play.golang.org/p/ehC53xA4OX">http://play.golang.org/p/ehC53xA4OX</a> Which I think is generally already okay, though I would like to use something like &#34;type bigram [2]rune&#34; instead of the []string. For comparison I pulled up an old implementation of the algorithm I did a few years back in ruby: <a href="https://gist.github.com/anonymous/7fc2780832b5396d038d">https://gist.github.com/anonymous/7fc2780832b5396d038d</a> readability isn&#39;t great though</p> <hr/>**评论：**<br/><br/>mkaz: <pre><p>LOC should not be what you optimize for; readability and comprehension are much better goals. Your future self with thank you.</p></pre>lwe: <pre><p>It&#39;s more of a challenge really. Same as writing write only Perl code </p></pre>dchapes: <pre><p>Exactly, my first thought when I saw the title was to say &#34;smaller != better&#34;.</p></pre>lwe: <pre><p>On the other hand more verbose code is not automatically better <a href="https://medium.com/@webseanhickey/the-evolution-of-a-software-engineer-db854689243" rel="nofollow">https://medium.com/@webseanhickey/the-evolution-of-a-software-engineer-db854689243</a> </p></pre>dchapes: <pre><p>&#34;smaller != better&#34; is not at all the same thing as saying or implying that &#34;verbose == better&#34;. There is very little relationship between terse/verbose and quality.</p></pre>a9y: <pre><p>A similar implementation of the same algorithm can be found <a href="https://groups.google.com/d/msg/golang-nuts/TTKPzAjwOV4/5315kJgCfWgJ">here</a>.</p> <p>It doesn&#39;t perform as well as yours, but it does handle unicode and duplicates as one would expect.</p> <p>The author chose to store the runes in a struct, however using a rune array for the <em>Bigram</em> type as you suggested would work just as well.</p> <p>One quick suggestion about your implementation is that you should call <em>strings.ToLower</em> on the entire string, not on each slice. You get a noticeable performance hit calling it that many times.</p></pre>lwe: <pre><p>I found that one earlier. It&#39;s where I got the idea to use [2]rune. But I am still working out how to write that efficiently.</p> <p>You&#39;re right about ToLower of course, I fixed that. Also found a bug in the first implementation. p2 = &#34;&#34; works on a copy not the actual array element. Had to change that.</p></pre>