Update changelog & fix version-headings

The entries for the 1.2.3.0 release have been reordered according to
their importance.

The version numbers follow the same heading style as `aeson`'s
changelog (i.e. heading level matches severity of version bump)

Author: hvr

changelog.md

@@ -1,21 +1,24 @@
-1.2.3.0
+### 1.2.3.0
 
-* `unsnoc` implemented
+* Spec compliance: `toCaseFold` now follows the Unicode 9.0 spec
+  (updated from 8.0).
 
 * Bug fix: the lazy `takeWhileEnd` function violated the
-  [lazy text invariant](https://github.com/bos/text/blob/1.2.3.0/Data/Text/Internal/Lazy.hs#L51).
-
-* Spec compliance: `toCaseFold` now follows the Unicode 9.0 spec
-  (updated from 8.0)
+  [lazy text invariant](https://github.com/bos/text/blob/1.2.3.0/Data/Text/Internal/Lazy.hs#L51)
+  (gh-184).
 
 * Bug fix: Fixed usage of size hints causing incorrect behavior (gh-197).
 
-* Improve UTF8 decoder error-recovery (gh-182).
+* New function: `unsnoc` (gh-173).
 
 * Reduce memory overhead in `encodeUTF8` (gh-194).
 
+* Improve UTF-8 decoder error-recovery (gh-182).
+
+* Minor documentation improvements (`@since` annotations, more
+  examples, clarifications).
 
-1.2.2.2
+#### 1.2.2.2
 
 * The `toTitle` function now correctly handles letters that
   immediately follow punctuation. Before, `"there's"` would turn into
@@ -32,7 +35,7 @@
 
 * Bug fix: a logic error in `takeWord16` is fixed.
 
-1.2.2.1
+#### 1.2.2.1
 
 * The switch to `integer-pure` in 1.2.2.0 was apparently mistaken.
   The build flag has been renamed accordingly.  Your army of diligent
@@ -43,7 +46,7 @@
 
 * An STG lint error has been fixed
 
-1.2.2.0
+### 1.2.2.0
 
 * The `integer-simple` package, upon which this package optionally
   depended, has been replaced with `integer-pure`.  The build flag has
@@ -59,37 +62,37 @@
     * if `base` >= 4.7: `PrintfArg`
     * if `base` >= 4.9: `Semigroup`
 
-1.2.1.3
+#### 1.2.1.3
 
 * Bug fix: As it turns out, moving the literal rewrite rules to simplifier
   phase 2 does not prevent competition with the `unpack` rule, which is
   also active in this phase. Unfortunately this was hidden due to a silly
   test environment mistake. Moving literal rules back to phase 1 finally
   fixes GHC Trac #10528 correctly.
 
-1.2.1.2
+#### 1.2.1.2
 
 * Bug fix: Run literal rewrite rules in simplifier phase 2.
   The behavior of the simplifier changed in GHC 7.10.2,
   causing these rules to fail to fire, leading to poor code generation
   and long compilation times. See
   [GHC Trac #10528](https://ghc.haskell.org/trac/ghc/ticket/10528).
 
-1.2.1.1
+#### 1.2.1.1
 
 * Expose unpackCString#, which you should never use.
 
-1.2.1.0
+### 1.2.1.0
 
 * Added Binary instances for both Text types. (If you have previously
   been using the text-binary package to get a Binary instance, it is
   now obsolete.)
 
-1.2.0.6
+#### 1.2.0.6
 
 * Fixed a space leak in UTF-8 decoding
 
-1.2.0.5
+#### 1.2.0.5
 
 * Feature parity: repeat, cycle, iterate are now implemented for lazy
   Text, and the Data instance is more complete
@@ -109,26 +112,26 @@
 * Spec compliance: toCaseFold now follows the Unicode 7.0 spec
   (updated from 6.3)
 
-1.2.0.4
+#### 1.2.0.4
 
 * Fixed an incompatibility with base < 4.5
 
-1.2.0.3
+#### 1.2.0.3
 
 * Update formatRealFloat to correspond to the definition in versions
   of base newer than 4.5 (https://github.com/bos/text/issues/105)
 
-1.2.0.2
+#### 1.2.0.2
 
 * Bumped lower bound on deepseq to 1.4 for compatibility with the
   upcoming GHC 7.10
 
-1.2.0.1
+#### 1.2.0.1
 
 * Fixed a buffer overflow in rendering of large Integers
   (https://github.com/bos/text/issues/99)
 
-1.2.0.0
+## 1.2.0.0
 
 * Fixed an integer overflow in the replace function
   (https://github.com/bos/text/issues/81)
@@ -141,7 +144,7 @@
 
 * Added an instance of IsList for GHC 7.8 and above
 
-1.1.1.0
+### 1.1.1.0
 
 * The Data.Data instance now allows gunfold to work, via a virtual
   pack constructor
@@ -151,12 +154,12 @@
 * Comparing the length of a Text against a number can now
   short-circuit in more cases
 
-1.1.0.1
+#### 1.1.0.1
 
 * streamDecodeUtf8: fixed gh-70, did not return all unconsumed bytes
   in single-byte chunks
 
-1.1.0.0
+## 1.1.0.0
 
 * encodeUtf8: Performance is improved by up to 4x.
 
@@ -175,12 +178,12 @@
   use at your own risk - there are no API stability guarantees for
   internal modules!
 
-1.0.0.1
+#### 1.0.0.1
 
 * decodeUtf8: Fixed a regression that caused us to incorrectly
   identify truncated UTF-8 as valid (gh-61)
 
-1.0.0.0
+# 1.0.0.0
 
 * Added support for Unicode 6.3.0 to case conversion functions