Update 'CHANGELOG.md' for the latest v2.0.0-beta changes

CHANGELOG.md

@@ -5,15 +5,15 @@
 ## New Features
 
 - Added new overloads for many functions of Adiar's API that use [1] (pure)
-  label predicates, [2] (stateful) label generator functions (coroutines), and
-  [3] iterators. For example, `bdd_exists` is now overloaded with the following
+  label *predicates*, [2] (stateful) label *generator* functions, and [3]
+  *iterators*. For example, `bdd_exists` is now overloaded with the following
   alternatives
 
   - `bdd_exists(f, pred)`: quantify all variables *i* in *f* where *pred(i)*
     evaluates to true.
 
-  - `bdd_exists(f, gen)`: quantify all variables *i* generated by *gen()*. Here
-    it is important that *gen()* provides the variables in descending order.
+  - `bdd_exists(f, gen)`: quantify all variables *i* generated by *gen()*. Here,
+    *gen()* must provides the to-be quantified variables in descending order.
 
   - `bdd_exists(f, begin, end)`: quantify all variables *i* from *begin* to
     *end* (assuming these are sorted in descending order).
@@ -23,36 +23,63 @@
   iterators as an argument, such that they are compatible with any data
   structure you desire.
 
-- `domain_set(...)` Add overload to set domain given a number of variables, a
-  *generator*, or an *iterator*.
+- `domain_set(...)` Add overload to set the global domain given a number of
+  variables, a *generator*, or an *iterator*.
+
+- BDD and ZDD operations now optionally take an `exec_policy` its first
+  argument. This can be used to set parameters for the algorithms to non-default
+  values.
+
+- Added `bdd_top()` and `bdd_bot()` together with `zdd_top(...)` and
+  `zdd_bot(...)` to provide a more uniform interface to quicky switch from BDDs
+  to ZDDs (and vica versa).
+
+- Added `bdd_iscube(f)` and `zdd_ispoint(A)` predicates. Furthermore, added
+  `bdd_cube(...)` and `zdd_point(...)` as aliases of the relevant BDD
+  construction function.
+
+- Added `bdd_isvar(f)`, `bdd_isithvar(f)`, `bdd_isnithvar(f)` predicates.
+
+- The functions `bdd_and(...)` and `bdd_or(...)` now support negated variables.
+  This can either be parsed as a boolean flag or by negating the label.
+
+- Added *<adiar/version.h>* with the library's version number as compile-time
+  known and inlinable integers and strings.
 
 ## Optimisations
 
-- The `access_mode` enum has been added as a new global setting. If set to
-  *AUTO* or to *Random_Access* then `bdd_apply`, `zdd_binop`, and their
-  derivatives will use random access on one of the input (if possible). This
+- The algorithms `bdd_apply`, `zdd_binop`, and their derivatives now use a
+  level-by-level random access on one of the input (if possible). This
   circumvents using an entire priority queue, thereby drastically improving
   performance. This is especially beneficial when applying an operator to a very
   large decision diagram together with a narrow one.
 
-- Added proper algoritms for the quantification of multi-variable
-  quantificiation. To this end, there are now three separate algorithms. The
-  `quantify_mode` enum can be used to pick what algorithm to use (the default is
-  a heuristic that picks between all three algorithms)
+  With `exec_policy::access` one can turn this optimisation off and/or force it
+  to always be used.
 
-  1. The *v1.x* approach of quantifying a single variable at a time still
-     persists (`quantify_mode::Singleton`).
+- The functions `bdd_equal` and `zdd_equal` terminate in constant time, if the
+  width of the two decision diagrams are not the same.
+
+- Added proper support for quantification of multiple variable. To this end, we
+  have implemented three separate algorithms. One can use
+  `exec_policy::quantify` to pick between the three:
+
+  1. The *v1.x* approach of quantifying a single variable at a time is still
+     available (`exec_policy::quantify::Singleton`).
 
   2. The *v1.x* algorithm has been generalized to (partially) handle multiple
-     variables at once in a single sweep (`quantify_mode::Partial`).
+     variables at once in a single sweep (`exec_policy::quantify::Partial`).
      Furthermore, this generalized sweep can be rerun on its own unreduced
      result (skipping the possibly costly and unecessary Reduce sweep).
 
   3. Implemented the *nested sweeping* framework to add support for an
      I/O-efficient simulation of BDD operations that (may) recurse on
      intermediate results. This provides a completely new multi-variable
      quantification operations that is similar to the one in other BDD packages
-     (`quantify_mode::Nested`).
+     (`exec_policy::quantify::Nested`).
+
+  By default, one or more of the above algorithms are used in conjunction
+  (`exec_policy::quantify::Auto`).
 
 ## Bug Fixes
 
@@ -66,10 +93,10 @@
 ## Breaking Changes
 
 There has been a major rewrite of the internal logic of Adiar to pay off
-technical debt and to get Adiar ready for implementing the I/O-efficient version
-of the much more complex BDD operations. At the same time, this also brings
-Adiar much closer to support other types of decision diagrams (e.g. QMDDs) and
-to store BDDs on disk and loading them again later.
+technical debt and to get Adiar ready for an I/O-efficient implementation of the
+much more complex BDD operations, e.g. multi-variable quantification. At the
+same time, this also brings Adiar much closer to support other types of decision
+diagrams (e.g. QMDDs) and to store BDDs on disk and loading them again later.
 
 - All internal logic in *<adiar/internal/...>* has been moved into its nested
   namespace `adiar::internal`. If you use anything from this namespace (e.g. the
@@ -80,16 +107,9 @@ to store BDDs on disk and loading them again later.
   `{prefix}_{nocase}` naming scheme. For example, `adiar::adiar_set_domain(...)`
   has been renamed into `adiar::domain_set(...)`.
 
-- Fundamental data types in *<adiar/data.h>* have been moved to each their own
-  file (most of which are placed in the *<adiar/internal/data_types/...>*).
-  These have been converted from C-like structs with manipulating functions to
-  C++ (immutable) classes. That is, the entire interface for *node*, *arc* etc.
-  have completely changed.
-
 - Files, streams and their writers have been moved, rewritten, and renamed.
-  Since the v2.0 API uses *predicates* and *generator* functions, Adiar's files
-  do not need to be used directly. Hence, these have been delegated to the
-  *adiar::internal* namespace.
+  Since newly added *predicates* and *generator* functions superseed any use of
+  `adiar::file<...>`, all file-based overloads have been removed.
 
 That is, this rewrite only results in breaking changes if you have been
 interacting with Adiar's files directly, e.g. you have used the `bdd_restrict`,
@@ -107,16 +127,21 @@ Other breaking changes are:
     `zdd_singleton` function.
 
 - `bdd_satmin(f)` and `bdd_satmax(f)`, resp. `zdd_minelem(A)` and `zdd_maxelem(A)`:
-  - The return type when called without a *callback function* has been turned
-    into a `bdd`, resp. `zdd`. This ought to make their return type more usable
-    in a symbolic context.
+  - The return type when called without a *consumer* or iterators has been
+    turned into a `bdd`, resp. `zdd`. This ought to make their return type more
+    usable in a symbolic context.
+
+  - Semantics has changed such that it only reports the variables on the path
+    traversed in the decision diagram. That is, if a level is skipped on some
+    path, then that variable is also skipped in the output.
+
+- Renamed `bdd_varprofile(...)` to `bdd_support(...)` (and the ZDD variant
+  similarly). This improves its discoverability and better reflects its content.
 
-  - Semantics has been changed if the global domain has been set. If so, then it
-    will return the assignment to the variables in the domain rather than
-    specifically the levels present in the BDD.
+- The `memory_mode` enum introduced in *v1.2* on/off has been moved to
+  `exec_policy::memory`.
 
-- Every deprecated function from *v1.x* that was moved to *<adiar/deprecated.h>*
-  has been removed.
+- All deprecated function from *v1.x* have been removed.
 
 - The two-level granularity of statistics has been removed. If you want to
   compile Adiar with statistics you just have to set the CMake variable