continuation.js - npm Package Compare versions

Comparing version 0.2.0 to 0.2.1

package.json

 {
   "name": "continuation.js",
   "description": "A module for tail call optimization by Continuation Passing Style (CPS) transformation with trampoline technique for Node.js",
-  "version": "0.2.0",
+  "version": "0.2.1",
   "author": "Daishi Kato <[e-mail]>",
@@ -6,0 +6,0 @@ "dependencies": {

README.md

@@ -10,3 +10,4 @@ continuation.js
 Node.js is often used with callback functions,
-which tend to be tail calls (but not necessarily recursions).
+which tend to be tail calls (but not necessarily recursions)
+consuming call stacks.
 
@@ -16,3 +17,3 @@ This module allows to transform native JavaScript code into
 It utilizes so-called trampoline technique to avoid a stack overflow error.
-Transforming all function into CPS is not very easy
+Transforming all functions into CPS is not very easy
 (and sometimes not very efficient),
@@ -30,14 +31,45 @@ hence it has a fallback mechanism, that is, only supported
 
-| NAME                    | continuation.js | [Brushtail][1] | [tailrec.js][2] | [thunk.js][3] | [tail-call][4] |
-|-------------------------|-----------------|----------------|-----------------|---------------|----------------|
-| Tail Call Optimization  | Yes             | Yes            | Yes             | Yes           | Yes            |
-| Mutual TCO              | Yes             | No             | Yes             | No            | No             |
-| Native JavaScript       | Yes             | Yes            | No              | No            | Almost         |
-| `require()` integration | Yes             | No             | No              | No            | No             |
+<table>
+<tr>
+<th>NAME</th>
+<th>continuation.js</th>
+<th><a href="https://github.com/pufuwozu/brushtail">Brushtail</a></th>
+<th><a href="https://github.com/natefaubion/tailrec.js">tailrec.js</a></th>
+<th><a href="https://github.com/jayferd/thunk.js">thunk.js</a></th>
+<th><a href="https://github.com/Gozala/js-tail-call">tail-call</a></th>
+</tr>
+<tr>
+<td>Tail call optimization</td>
+<td>Yes</td>
+<td>Yes</td>
+<td>Yes</td>
+<td>Yes</td>
+<td>Yes</td>
+</tr>
+<tr>
+<td>Mutual recursion</td>
+<td>Yes</td>
+<td>No</td>
+<td>Yes</td>
+<td>No</td>
+<td>No</td>
+</tr>
+<tr>
+<td>Native JavaScript</td>
+<td>Yes</td>
+<td>Yes</td>
+<td>No</td>
+<td>No</td>
+<td>Almost</td>
+</tr>
+<tr>
+<td>`require()` integration</td>
+<td>Yes</td>
+<td>No</td>
+<td>No</td>
+<td>No</td>
+<td>No</td>
+</tr>
+</table>
 
-[1]: https://github.com/pufuwozu/brushtail "pufuwozu/brushtail"
-[2]: https://github.com/natefaubion/tailrec.js "natefaubion/tailrec.js"
-[3]: https://github.com/jayferd/thunk.js "jayferd/thunk.js"
-[4]: https://github.com/Gozala/js-tail-call "Gozala/js-tail-call"
-
 How to use
@@ -139,20 +171,106 @@ ----------
 
-| Suite name      | Original      | CPS transformed | Improved? |
-|-----------------|---------------|-----------------|-----------|
-| Richards        | 324 ops/sec   | 38.28 ops/sec   | No        |
-| DeltaBlue       | 188 ops/sec   | 12.90 ops/sec   | No        |
-| Encrypt         | 162 ops/sec   | 122 ops/sec     | No        |
-| Decrypt         | 8.62 ops/sec  | 6.86 ops/sec    | No        |
-| RayTrace        | 19.71 ops/sec | 4.10 ops/sec    | No        |
-| Earley          | 278 ops/sec   | 33.91 ops/sec   | No        |
-| Boyer           | 18.19 ops/sec | 1.60 ops/sec    | No        |
-| RegExp          | 7.12 ops/sec  | 5.14 ops/sec    | No        |
-| Splay           | 123 ops/sec   | 75.21 ops/sec   | No        |
-| NavierStokes    | 2.40 ops/sec  | 3.13 ops/sec    | Yes       |
-| PdfJS           | 2.85 ops/sec  | 2.68 ops/sec    | No        |
-| Gameboy         | 0.98 ops/sec  | 0.35 ops/sec    | No        |
-| CodeLoadClosure | 370 ops/sec   | 372 ops/sec     | Yes       |
-| CodeLoadJQuery  | 8.83 ops/sec  | 10.61 ops/sec   | Yes       |
-| Box2D           | 2.33 ops/sec  | 2.54 ops/sec    | Yes       |
+<table>
+<tr>
+<th>Suite name</th>
+<th>Original</th>
+<th>CPS transformed</th>
+<th>Improved?</th>
+</tr>
+<tr>
+<td>Richards</td>
+<td>324 ops/sec</td>
+<td>38.28 ops/sec</td>
+<td>No</td>
+</tr>
+<tr>
+<td>DeltaBlue</td>
+<td>188 ops/sec</td>
+<td>12.90 ops/sec</td>
+<td>No</td>
+</tr>
+<tr>
+<td>Encrypt</td>
+<td>162 ops/sec</td>
+<td>122 ops/sec</td>
+<td>No</td>
+</tr>
+<tr>
+<td>Decrypt</td>
+<td>8.62 ops/sec</td>
+<td>6.86 ops/sec</td>
+<td>No</td>
+</tr>
+<tr>
+<td>RayTrace</td>
+<td>19.71 ops/sec</td>
+<td>4.10 ops/sec</td>
+<td>No</td>
+</tr>
+<tr>
+<td>Earley</td>
+<td>278 ops/sec</td>
+<td>33.91 ops/sec</td>
+<td>No</td>
+</tr>
+<tr>
+<td>Boyer</td>
+<td>18.19 ops/sec</td>
+<td>1.60 ops/sec</td>
+<td>No</td>
+</tr>
+<tr>
+<td>RegExp</td>
+<td>7.12 ops/sec</td>
+<td>5.14 ops/sec</td>
+<td>No</td>
+</tr>
+<tr>
+<td>Splay</td>
+<td>123 ops/sec</td>
+<td>75.21 ops/sec</td>
+<td>No</td>
+</tr>
+<tr>
+<td>NavierStokes</td>
+<td>2.40 ops/sec</td>
+<td>3.13 ops/sec</td>
+<td>Yes</td>
+</tr>
+<tr>
+<td>PdfJS</td>
+<td>2.85 ops/sec</td>
+<td>2.68 ops/sec</td>
+<td>No</td>
+</tr>
+<tr>
+<td>Gameboy</td>
+<td>0.98 ops/sec</td>
+<td>0.35 ops/sec</td>
+<td>No</td>
+</tr>
+<tr>
+<td>CodeLoadClosure</td>
+<td>370 ops/sec</td>
+<td>372 ops/sec</td>
+<td>Yes</td>
+</tr>
+<tr>
+<td>CodeLoadJQuery</td>
+<td>8.83 ops/sec</td>
+<td>10.61 ops/sec</td>
+<td>Yes</td>
+</tr>
+<tr>
+<td>Box2D</td>
+<td>2.33 ops/sec</td>
+<td>2.54 ops/sec</td>
+<td>Yes</td>
+</tr>
+</table>
 
+Since trampoline is costly,
+performance drops in most suites especially basic ones.
+Whereas in relatively complex suites, there are some cases
+when performance is improved. (Anybody interested why this is happening?)
+
 Limitations
@@ -168,4 +286,3 @@ -----------
 
-* Transform all tail recursive calls into CPS.
 * Work with try...catch and throw.
 * Transform non-tail recursive calls into CPS.

No alert changes

Improved metrics

Total package byte prevSize

162516

increased by0.17%

Number of lines in readme file

284

increased by70.06%

No dependency changes