Updated changelog.

CHANGES.html

@@ -348,6 +348,12 @@ ul.auto-toc {
 <p>As many people have been using D:YAML from git master since the 0.4 release, each change
 is prefixed by the year the change was introduced.</p>
 <ul>
+<li><dl class="first docutils">
+<dt><tt class="docutils literal">2014</tt> The <tt class="docutils literal">cdc.d</tt> build script has been removed; dub is now the only 'official'</dt>
+<dd><p class="first last">way to build D:YAML.</p>
+</dd>
+</dl>
+</li>
 <li><p class="first"><tt class="docutils literal">2014</tt> Broke compatibility with all DMD versions before 2.066</p>
 </li>
 <li><dl class="first docutils">
@@ -402,18 +408,31 @@ and to minimize internal exception handling during parsing.</li>
 input, and UTF-32 inputs can be encoded into UTF-8 in-place without allocating.
 UTF-16 inputs still need an allocation. This also gets rid of all dchar[]-&gt;char[]
 conversions which were a significant source of GC allocations.</li>
+<li>Various optimizations in <tt class="docutils literal">Reader</tt>/<tt class="docutils literal">Scanner</tt>, especially for mostly-ASCII files
+containing plain scalars (most common YAML files). Measured speedup of ~80% when
+parsing mostly-ASCII
+files, slowdown of ~12% for mostly non-ASCII files (all tested files were UTF-8).</li>
 <li>Less GC usage during YAML node construction.</li>
-<li>The <tt class="docutils literal">Scanner</tt> subsystems is now mostly <tt class="docutils literal">&#64;nogc</tt>; it never allocates memory for
-token values, using slices into the input buffer instead.</li>
+<li><tt class="docutils literal">Scanner</tt> is now mostly <tt class="docutils literal">&#64;nogc</tt>; it never allocates memory for token values, using
+slices into the input buffer instead.</li>
 <li>Custom, <tt class="docutils literal">&#64;nogc</tt> UTF decoding/encoding code based on <tt class="docutils literal">std.utf</tt> to enable more use
 of <tt class="docutils literal">&#64;nogc</tt> in D:YAML internals and to improve performance.</li>
-<li>Replaced <tt class="docutils literal">std.stream.EndianStream</tt> with custom endian-fixing code.</li>
+<li>Less memory allocations during scanning in general, including manual allocations.</li>
+<li>Default <tt class="docutils literal">Constructor</tt>/<tt class="docutils literal">Resolver</tt> are now only constructed if the user doesn't
+specify their own.</li>
+<li>Replaced <tt class="docutils literal">std.stream.EndianStream</tt> with
+<a class="reference external" href="https://github.com/kiith-sa/tinyendian">tinyendian</a>.</li>
 <li>D:YAML is now a DUB package.</li>
 <li>Removed static data structures such as default Constructor and Resolver.</li>
 <li>Compile-time checks for size of data structures that should be small.</li>
 <li>Better error messages.</li>
 <li>Various refactoring changes, using more 'modern' D features, better tests.</li>
-<li>Updated documentation, examples.</li>
+<li>Updated documentation, examples to reflect changes in 0.5.</li>
+<li>Updated the <tt class="docutils literal">yaml_bench</tt> example/tool with options to keep the input file in memory
+instead of reloading it for repeated parses, and to only benchmark scanning time
+instead of the entire parser.</li>
+<li>The <tt class="docutils literal">yaml_gen</tt> example/tool can now generate strings from user-specified alphabets
+which may contain non-ASCII characters.</li>
 </ul>
 </div>
 <div class="section" id="bugfixes">

CHANGES.rst

@@ -9,6 +9,8 @@ Breaking changes
 As many people have been using D:YAML from git master since the 0.4 release, each change
 is prefixed by the year the change was introduced.
 
+- ``2014`` The ``cdc.d`` build script has been removed; dub is now the only 'official'
+           way to build D:YAML.
 - ``2014`` Broke compatibility with all DMD versions before 2.066
 - ``2014`` ``Loader`` API depending on ``std.stream`` is now deprecated and will be
            removed in the next release.
@@ -50,18 +52,31 @@ Other improvements
   input, and UTF-32 inputs can be encoded into UTF-8 in-place without allocating.
   UTF-16 inputs still need an allocation. This also gets rid of all dchar[]->char[]
   conversions which were a significant source of GC allocations.
+- Various optimizations in ``Reader``/``Scanner``, especially for mostly-ASCII files 
+  containing plain scalars (most common YAML files). Measured speedup of ~80% when 
+  parsing mostly-ASCII 
+  files, slowdown of ~12% for mostly non-ASCII files (all tested files were UTF-8).
 - Less GC usage during YAML node construction.
-- The ``Scanner`` subsystems is now mostly ``@nogc``; it never allocates memory for
-  token values, using slices into the input buffer instead.
+- ``Scanner`` is now mostly ``@nogc``; it never allocates memory for token values, using
+  slices into the input buffer instead.
 - Custom, ``@nogc`` UTF decoding/encoding code based on ``std.utf`` to enable more use
   of ``@nogc`` in D:YAML internals and to improve performance.
-- Replaced ``std.stream.EndianStream`` with custom endian-fixing code.
+- Less memory allocations during scanning in general, including manual allocations.
+- Default ``Constructor``/``Resolver`` are now only constructed if the user doesn't 
+  specify their own.
+- Replaced ``std.stream.EndianStream`` with 
+  `tinyendian <https://github.com/kiith-sa/tinyendian>`_.
 - D:YAML is now a DUB package.
 - Removed static data structures such as default Constructor and Resolver.
 - Compile-time checks for size of data structures that should be small.
 - Better error messages.
 - Various refactoring changes, using more 'modern' D features, better tests.
-- Updated documentation, examples.
+- Updated documentation, examples to reflect changes in 0.5.
+- Updated the ``yaml_bench`` example/tool with options to keep the input file in memory
+  instead of reloading it for repeated parses, and to only benchmark scanning time
+  instead of the entire parser.
+- The ``yaml_gen`` example/tool can now generate strings from user-specified alphabets
+  which may contain non-ASCII characters.
 
 
 ^^^^^^^^