A Version represents the version of a software entity.

Instances of Eq and Ord are provided, which gives exact equality and lexicographic ordering of the version number components (i.e. 2.1 > 2.0, 1.2.3 > 1.2.2, etc.).

This type is opaque and distinct from the Version type in Data.Version since Cabal-2.0. The difference extends to the Binary instance using a different (and more compact) encoding.

Since: Cabal-syntax-2.0.0.2

Version 0. A lower bound of Version.

Since: Cabal-syntax-2.2

Construct Version from list of version number components.

For instance, mkVersion [3,2,1] constructs a Version representing the version 3.2.1.

All version components must be non-negative. mkVersion [] currently represents the special null version; see also nullVersion.

Since: Cabal-syntax-2.0.0.2

Variant of mkVersion which converts a Data.Version Version into Cabal's Version type.

Since: Cabal-syntax-2.0.0.2

Unpack Version into list of version number components.

This is the inverse to mkVersion, so the following holds:

(versionNumbers . mkVersion) vs == vs

Since: Cabal-syntax-2.0.0.2

Constant representing the special null Version

The nullVersion compares (via Ord) as less than every proper Version value.

Since: Cabal-syntax-2.0.0.2

Apply function to list of version number components

alterVersion f == mkVersion . f . versionNumbers

Since: Cabal-syntax-2.0.0.2

The version range -any. That is, a version range containing all versions.

withinRange v anyVersion = True

The empty version range -none, that is a version range containing no versions.

This can be constructed using any unsatisfiable version range expression, for example < 0.

withinRange v noVersion = False

The version range == v.

withinRange v' (thisVersion v) = v' == v

The version range /= v.

withinRange v' (notThisVersion v) = v' /= v

The version range > v.

withinRange v' (laterVersion v) = v' > v

The version range < v.

withinRange v' (earlierVersion v) = v' < v

The version range >= v.

withinRange v' (orLaterVersion v) = v' >= v

The version range <= v.

withinRange v' (orEarlierVersion v) = v' <= v

The version range vr1 || vr2.

  withinRange v' (unionVersionRanges vr1 vr2)
= withinRange v' vr1 || withinRange v' vr2

The version range vr1 && vr2.

  withinRange v' (intersectVersionRanges vr1 vr2)
= withinRange v' vr1 && withinRange v' vr2

The version range == v.*.

For example, for version 1.2, the version range == 1.2.* is the same as >= 1.2 && < 1.3.

withinRange v' (withinVersion v) = v' >= v && v' < upper v
  where
    upper (Version lower t) = Version (init lower ++ [last lower + 1]) t

The version range ^>= v.

For example, for version 1.2.3.4, the version range ^>= 1.2.3.4 is the same as >= 1.2.3.4 && < 1.3.

Note that ^>= 1 is equivalent to >= 1 && < 1.1.

Since: Cabal-syntax-2.0.0.2

Does this version fall within the given range?

This is the evaluation function for the VersionRange type.

Does this VersionRange place any restriction on the Version or is it in fact equivalent to AnyVersion.

Note this is a semantic check, not simply a syntactic check. So for example the following is True (for all v).

isAnyVersion (EarlierVersion v `UnionVersionRanges` orLaterVersion v)

This is the converse of isAnyVersion. It check if the version range is empty, if there is no possible version that satisfies the version range.

For example this is True (for all v):

isNoVersion (EarlierVersion v `IntersectVersionRanges` LaterVersion v)

Is this version range in fact just a specific version?

For example the version range ">= 3 && <= 3" contains only the version 3.

Simplify a VersionRange expression. For non-empty version ranges this produces a canonical form. Empty or inconsistent version ranges are left as-is because that provides more information.

If you need a canonical form use fromVersionIntervals . toVersionIntervals

It satisfies the following properties:

withinRange v (simplifyVersionRange r) = withinRange v r

    withinRange v r = withinRange v r'
==> simplifyVersionRange r = simplifyVersionRange r'
 || isNoVersion r
 || isNoVersion r'

Fold over the basic syntactic structure of a VersionRange.

This provides a syntactic view of the expression defining the version range. The syntactic sugar ">= v", "<= v" and "== v.*" is presented in terms of the other basic syntax.

For a semantic view use asVersionIntervals.

Normalise VersionRange.

In particular collapse (== v || > v) into >= v, and so on.

Remove VersionRangeParens constructors.

Since version 3.4 this function is id, there aren't VersionRangeParens constructor in VersionRange anymore.

Since: Cabal-syntax-2.2

Does the version range have an upper bound?

Since: Cabal-syntax-1.24.0.0

Does the version range have an explicit lower bound?

Note: this function only considers the user-specified lower bounds, but not the implicit >=0 lower bound.

Since: Cabal-syntax-1.24.0.0

F-Algebra of VersionRange. See cataVersionRange.

Since: Cabal-syntax-2.2

Fold VersionRange.

Since: Cabal-syntax-2.2

Unfold VersionRange.

Since: Cabal-syntax-2.2

Refold VersionRange.

Since: Cabal-syntax-2.2

Generic destructor for VersionRange.

Since: Cabal-syntax-2.2

Generic constructor for VersionRange.

Since: Cabal-syntax-2.2

Increment the last version number.

Example: For 1.2 this returns 1.3 so that it can be used as upper bound when resolving == 1.2.*. For 0.4.1 it returns 0.4.2.

Since: Cabal-syntax-2.2

Compute next greater major version to be used as upper bound.

Example: 0.4.1 produces the version 0.5 which then can be used to construct a range >= 0.4.1 && < 0.5

Since: Cabal-syntax-2.2

Given a version range, remove the highest upper bound. Example: (>= 1 && < 3) || (>= 4 && < 5) is converted to (>= 1 && || (= 4).

Given a version range, remove the lowest lower bound. Example: (>= 1 && || (= 4 && < 5) is converted to (>= 0 && || (= 4 && < 5).

Rewrite ^>= x.y.z into >= x.y.z && < x.(y+1)

Since: Cabal-syntax-3.6.0.0

Rewrite ^>= x.y.z into >= x.y.z

Since: Cabal-syntax-3.6.0.0

Rewrite ^>= x.y.z into <x.(y+1)

Since: Cabal-syntax-3.6.0.0

View a VersionRange as a union of intervals.

This provides a canonical view of the semantics of a VersionRange as opposed to the syntax of the expression used to define it. For the syntactic view use foldVersionRange.

Each interval is non-empty. The sequence is in increasing order and no intervals overlap or touch. Therefore only the first and last can be unbounded. The sequence can be empty if the range is empty (e.g. a range expression like && 2).

Other checks are trivial to implement using this view. For example:

isNoVersion vr | [] <- asVersionIntervals vr = True
               | otherwise                   = False

isSpecificVersion vr
   | [(LowerBound v  InclusiveBound
      ,UpperBound v' InclusiveBound)] <- asVersionIntervals vr
   , v == v'   = Just v
   | otherwise = Nothing

A complementary representation of a VersionRange. Instead of a boolean version predicate it uses an increasing sequence of non-overlapping, non-empty intervals.

The key point is that this representation gives a canonical representation for the semantics of VersionRanges. This makes it easier to check things like whether a version range is empty, covers all versions, or requires a certain minimum or maximum version. It also makes it easy to check equality or containment. It also makes it easier to identify 'simple' version predicates for translation into foreign packaging systems that do not support complex version range expressions.

Convert a VersionRange to a sequence of version intervals.

Convert a VersionIntervals value back into a VersionRange expression representing the version intervals.

Inspect the list of version intervals.