Merge remote-tracking branch 'origin/master' into wip/no-extend-narro…

…w-primop-names

.gitignore

@@ -8,3 +8,7 @@
 /GNUmakefile
 /.ghc.environment.*
 /cabal.project.local
+/cabal.test.project.local
+
+# Test data repo ignored. Please see instruction in tests-and-benchmarks.markdown
+/tests/text-test-data/

benchmarks/LICENSE

@@ -0,0 +1 @@
+../LICENSE

benchmarks/cbits

@@ -0,0 +1 @@
+../cbits

benchmarks/include

@@ -0,0 +1 @@
+../include

benchmarks/python/cut.py

@@ -1,4 +1,4 @@
-#!/usr/bin/env python
+#!/usr/bin/env python3
 
 import utils, sys, codecs
 

benchmarks/python/multilang.py

@@ -1,4 +1,4 @@
-#!/usr/bin/env python
+#!/usr/bin/env python3
 
 import math
 import sys
@@ -12,23 +12,22 @@ def timer(f, count=100):
     a = 1e300
     def g():
         return
-    for i in xrange(3):
+    for i in range(3):
         start = time.time()
-        for j in xrange(count):
+        for j in range(count):
             g()
         a = min(a, (time.time() - start) / count)
 
     b = 1e300
-    for i in xrange(3):
+    for i in range(3):
         start = time.time()
-        for j in xrange(count):
+        for j in range(count):
             f()
         b = min(b, (time.time() - start) / count)
 
     return round(b - a, int(round(math.log(count, 10) - math.log(b - a, 10))))
 
 contents = open('../../tests/text-test-data/yiwiki.xml', 'r').read()
-contents = contents.decode('utf-8')
 
 benchmarks = (
     find_first,
@@ -47,4 +46,4 @@ def g():
 for b in bms:
     sys.stdout.write(b.__name__ + ' ')
     sys.stdout.flush()
-    print b()
+    print(b())

benchmarks/python/sort.py

@@ -1,4 +1,4 @@
-#!/usr/bin/env python
+#!/usr/bin/env python3
 
 import utils, sys, codecs
 

benchmarks/python/strip_tags.py

@@ -1,4 +1,4 @@
-#!/usr/bin/env python
+#!/usr/bin/env python3
 
 import utils, sys
 

benchmarks/python/utils.py

@@ -1,4 +1,4 @@
-#!/usr/bin/env python
+#!/usr/bin/env python3
 
 import sys, time
 

benchmarks/src

@@ -0,0 +1 @@
+../src

benchmarks/text-benchmarks.cabal

@@ -2,17 +2,22 @@ cabal-version:       1.12
 name:                text-benchmarks
 version:             0.0.0.0
 synopsis:            Benchmarks for the text package
-description:         Benchmarks for the text package
+description:         Benchmarks for the text package.
 homepage:            https://bitbucket.org/bos/text
 license:             BSD2
-license-file:        ../LICENSE
+license-file:        LICENSE
 author:              Jasper Van der Jeugt <[email protected]>,
                      Bryan O'Sullivan <[email protected]>,
                      Tom Harper <[email protected]>,
                      Duncan Coutts <[email protected]>
 maintainer:          [email protected]
 category:            Text
 build-type:          Simple
+tested-with:    GHC==8.10.2, GHC==8.8.4, GHC==8.6.5, GHC==8.4.4,
+                GHC==8.2.2, GHC==8.0.2, GHC==7.10.3
+
+extra-source-files:
+    include/*.h
 
 flag bytestring-builder
   description: Depend on the bytestring-builder package for backwards compatibility.
@@ -54,7 +59,7 @@ executable text-benchmarks
     build-depends: bytestring         >= 0.10.4
 
   -- modules for benchmark proper
-  c-sources:      cbits/time_iconv.c
+  c-sources:      cbits-bench/time_iconv.c
   hs-source-dirs: haskell
   main-is:        Benchmarks.hs
   other-modules:
@@ -81,9 +86,9 @@ executable text-benchmarks
 
   -- Source code for IUT (implementation under test)
   -- "borrowed" from parent folder
-  include-dirs:   ../include
-  c-sources:      ../cbits/cbits.c
-  hs-source-dirs: ..
+  include-dirs:   include
+  c-sources:      cbits/cbits.c
+  hs-source-dirs: src
   other-modules:
     Data.Text
     Data.Text.Array
@@ -96,6 +101,7 @@ executable text-benchmarks
     Data.Text.Internal.Builder.Functions
     Data.Text.Internal.Builder.Int.Digits
     Data.Text.Internal.Builder.RealFloat.Functions
+    Data.Text.Internal.ByteStringCompat
     Data.Text.Internal.Encoding.Fusion
     Data.Text.Internal.Encoding.Fusion.Common
     Data.Text.Internal.Encoding.Utf16
@@ -132,6 +138,7 @@ executable text-benchmarks
 
   default-language: Haskell2010
   default-extensions: NondecreasingIndentation
+  other-extensions: DeriveAnyClass
 
 
 executable text-multilang

cabal.haskell-ci

@@ -0,0 +1,8 @@
+benchmarks: >=7.4
+jobs-selection: any
+doctest: True
+doctest-options: -fobject-code -DINTEGER_GMP
+branches: master
+
+installed: +all -text -parsec
+install-dependencies: False

cabal.project

@@ -1,3 +1,6 @@
 -- See http://cabal.readthedocs.io/en/latest/nix-local-build-overview.html
-packages: ., benchmarks, th-tests
+packages: .
+packages: tests
+packages: th-tests
+packages: benchmarks
 tests: True

cabal.tests.project

@@ -0,0 +1,5 @@
+-- this project doesn't have local 'text' package,
+-- so tests build faster.
+
+packages: tests
+tests: True

changelog.md

@@ -1,3 +1,9 @@
+### 1.2.4.1
+
+* Support `template-haskell-2.17.0.0`
+* Support `bytestring-0.11`
+* Add `take . drop` related RULE
+
 ### 1.2.4.0
 
 * Add TH `Lift` instances for `Data.Text.Text` and `Data.Text.Lazy.Text` (gh-232)

scripts/tests.sh

@@ -0,0 +1,30 @@
+#!/bin/sh
+
+set -ex
+
+runtest() {
+  HC=$1
+  shift
+
+  # EDIT last line to pass arguments
+
+  cabal run text-tests:test:tests \
+      --project-file=cabal.tests.project \
+      --builddir="dist-newstyle/$HC" \
+      --with-compiler="$HC" \
+      -- "$@"
+}
+
+runtest ghc-8.10.2 "$@"
+runtest ghc-8.8.4  "$@"
+runtest ghc-8.6.5  "$@"
+runtest ghc-8.4.4  "$@"
+runtest ghc-8.2.2  "$@"
+runtest ghc-8.0.2  "$@"
+
+runtest ghc-7.10.3 "$@"
+runtest ghc-7.8.4  "$@"
+runtest ghc-7.6.3  "$@"
+runtest ghc-7.4.2  "$@"
+runtest ghc-7.2.2  "$@"
+runtest ghc-7.0.4  "$@"

Data/Text.hs → src/Data/Text.hs

@@ -259,6 +259,11 @@ import qualified Language.Haskell.TH.Syntax as TH
 import Text.Printf (PrintfArg, formatArg, formatString)
 #endif
 
+-- $setup
+-- >>> import Data.Text
+-- >>> import qualified Data.Text as T
+-- >>> :seti -XOverloadedStrings
+
 -- $character_definition
 --
 -- This package uses the term /character/ to denote Unicode /code points/.
@@ -1224,6 +1229,8 @@ drop n t@(Text arr off len)
     drop n t = unstream (S.drop n (stream t))
 "TEXT drop -> unfused" [1] forall n t.
     unstream (S.drop n (stream t)) = drop n t
+"TEXT take . drop -> unfused" [1] forall len off t.
+    unstream (S.take len (S.drop off (stream t))) = take len (drop off t)
   #-}
 
 -- | /O(n)/ 'dropEnd' @n@ @t@ returns the prefix remaining after
@@ -1352,13 +1359,19 @@ splitAt n t@(Text arr off len)
 -- a pair whose first element is the longest prefix (possibly empty)
 -- of @t@ of elements that satisfy @p@, and whose second is the
 -- remainder of the list.
+--
+-- >>> T.span (=='0') "000AB"
+-- ("000","AB")
 span :: (Char -> Bool) -> Text -> (Text, Text)
 span p t = case span_ p t of
              (# hd,tl #) -> (hd,tl)
 {-# INLINE span #-}
 
 -- | /O(n)/ 'break' is like 'span', but the prefix returned is
 -- over elements that fail the predicate @p@.
+--
+-- >>> T.break (=='c') "180cm"
+-- ("180","cm")
 break :: (Char -> Bool) -> Text -> (Text, Text)
 break p = span (not . p)
 {-# INLINE break #-}