Merge pull request haskell-hvr#32 from haskell-hvr/text-2

Support text-2.0

.github/workflows/haskell-ci.yml

@@ -8,9 +8,9 @@
 #
 # For more information, see https://github.com/haskell-CI/haskell-ci
 #
-# version: 0.13.20211029
+# version: 0.13.20211116
 #
-# REGENDATA ("0.13.20211029",["github","cabal.project"])
+# REGENDATA ("0.13.20211116",["github","cabal.project"])
 #
 name: Haskell-CI
 on:
@@ -24,6 +24,8 @@ jobs:
   linux:
     name: Haskell-CI - Linux - ${{ matrix.compiler }}
     runs-on: ubuntu-18.04
+    timeout-minutes:
+      60
     container:
       image: buildpack-deps:bionic
     continue-on-error: ${{ matrix.allow-failure }}
@@ -34,7 +36,7 @@ jobs:
             compilerKind: ghc
             compilerVersion: 9.2.1
             setup-method: ghcup
-            allow-failure: true
+            allow-failure: false
           - compiler: ghc-9.0.1
             compilerKind: ghc
             compilerVersion: 9.0.1
@@ -130,7 +132,7 @@ jobs:
           echo "HCNUMVER=$HCNUMVER" >> "$GITHUB_ENV"
           echo "ARG_TESTS=--enable-tests" >> "$GITHUB_ENV"
           echo "ARG_BENCH=--enable-benchmarks" >> "$GITHUB_ENV"
-          if [ $((HCNUMVER >= 90200)) -ne 0 ] ; then echo "HEADHACKAGE=true" >> "$GITHUB_ENV" ; else echo "HEADHACKAGE=false" >> "$GITHUB_ENV" ; fi
+          echo "HEADHACKAGE=false" >> "$GITHUB_ENV"
           echo "ARG_COMPILER=--$HCKIND --with-compiler=$HC" >> "$GITHUB_ENV"
           echo "GHCJSARITH=0" >> "$GITHUB_ENV"
         env:
@@ -159,17 +161,10 @@ jobs:
           repository hackage.haskell.org
             url: http://hackage.haskell.org/
           EOF
-          if $HEADHACKAGE; then
           cat >> $CABAL_CONFIG <<EOF
-          repository head.hackage.ghc.haskell.org
-             url: https://ghc.gitlab.haskell.org/head.hackage/
-             secure: True
-             root-keys: 7541f32a4ccca4f97aea3b22f5e593ba2c0267546016b992dfadcd2fe944e55d
-                        26021a13b401500c8eb2761ca95c61f2d625bfef951b939a8124ed12ecf07329
-                        f76d08be13e9a61a377a85e2fb63f4c5435d40f8feb3e12eb05905edb8cdea89
-             key-threshold: 3
+          program-default-options
+            ghc-options: $GHCJOBS +RTS -M3G -RTS
           EOF
-          fi
           cat $CABAL_CONFIG
       - name: versions
         run: |
@@ -217,9 +212,6 @@ jobs:
           if [ $((HCNUMVER >= 80200)) -ne 0 ] ; then echo "    ghc-options: -Werror=missing-methods" >> cabal.project ; fi
           cat >> cabal.project <<EOF
           EOF
-          if $HEADHACKAGE; then
-          echo "allow-newer: $($HCPKG list --simple-output | sed -E 's/([a-zA-Z-]+)-[0-9.]+/*:\1,/g')" >> cabal.project
-          fi
           $HCPKG list --simple-output --names-only | perl -ne 'for (split /\s+/) { print "constraints: $_ installed\n" unless /^(text-short)$/; }' >> cabal.project.local
           cat cabal.project
           cat cabal.project.local
@@ -260,6 +252,16 @@ jobs:
       - name: prepare for constraint sets
         run: |
           rm -f cabal.project.local
+      - name: constraint set text-2
+        run: |
+          if [ $((HCNUMVER >= 80000)) -ne 0 ] ; then $CABAL v2-build $ARG_COMPILER --enable-tests --disable-benchmarks --constraint='text ^>=2' --dependencies-only -j2 all ; fi
+          if [ $((HCNUMVER >= 80000)) -ne 0 ] ; then $CABAL v2-build $ARG_COMPILER --enable-tests --disable-benchmarks --constraint='text ^>=2' all ; fi
+          if [ $((HCNUMVER >= 80000)) -ne 0 ] ; then $CABAL v2-test $ARG_COMPILER --enable-tests --disable-benchmarks --constraint='text ^>=2' all ; fi
+      - name: constraint set text-1
+        run: |
+          $CABAL v2-build $ARG_COMPILER --enable-tests --disable-benchmarks --constraint='text ^>=1.2.3.0' --dependencies-only -j2 all
+          $CABAL v2-build $ARG_COMPILER --enable-tests --disable-benchmarks --constraint='text ^>=1.2.3.0' all
+          $CABAL v2-test $ARG_COMPILER --enable-tests --disable-benchmarks --constraint='text ^>=1.2.3.0' all
       - name: constraint set bytestring-0.11
         run: |
           $CABAL v2-build $ARG_COMPILER --enable-tests --disable-benchmarks --constraint='bytestring ==0.11.*' --dependencies-only -j2 all

ChangeLog.md

@@ -1,3 +1,7 @@
+## 0.1.5
+
+  * text-2.0 support
+
 ## 0.1.4
 
   * Fix `fromString` for single character strings.

cabal.haskell-ci

@@ -5,3 +5,14 @@ constraint-set bytestring-0.11
   constraints: bytestring ==0.11.*
   tests: True
   run-tests: True
+
+constraint-set text-1
+  constraints: text ^>=1.2.3.0
+  tests: True
+  run-tests: True
+
+constraint-set text-2
+  ghc: >= 8.0
+  constraints: text ^>=2
+  tests: True
+  run-tests: True

src/Data/Text/Short/Internal.hs

@@ -126,7 +126,6 @@ import           Data.Maybe                     (fromMaybe, isNothing)
 import           Data.Semigroup
 import qualified Data.String                    as S
 import qualified Data.Text                      as T
-import qualified Data.Text.Encoding             as T
 import           Foreign.C
 import           GHC.Base                       (assert, unsafeChr)
 import qualified GHC.CString                    as GHC
@@ -149,6 +148,13 @@ import           Text.Printf                    (PrintfArg, formatArg,
 
 import qualified Language.Haskell.TH.Syntax     as TH
 
+#if MIN_VERSION_text(2,0,0)
+import qualified Data.Text.Internal             as TI
+import qualified Data.Text.Array                as TA
+#else
+import qualified Data.Text.Encoding             as T
+#endif
+
 import qualified PrimOps
 
 -- | A compact representation of Unicode strings.
@@ -653,12 +659,16 @@ foldr1 f st
 --
 -- prop> (toText . fromText) t == t
 --
--- This is currently not \(\mathcal{O}(1)\) because currently 'T.Text' uses UTF-16 as its internal representation.
--- In the event that 'T.Text' will change its internal representation to UTF-8 this operation will become \(\mathcal{O}(1)\).
+-- This is \(\mathcal{O}(1)\) with @text-2@.
+-- Previously it wasn't because 'T.Text' used UTF-16 as its internal representation.
 --
 -- @since 0.1
 toText :: ShortText -> T.Text
+#if MIN_VERSION_text(2,0,0)
+toText (ShortText (BSSI.SBS ba)) = TI.Text (TA.ByteArray ba) 0 (I# (GHC.Exts.sizeofByteArray# ba))
+#else
 toText = T.decodeUtf8 . toByteString
+#endif
 
 ----
 
@@ -692,7 +702,22 @@ fromString s = case s of
 --
 -- @since 0.1
 fromText :: T.Text -> ShortText
+#if MIN_VERSION_text(2,0,0)
+fromText (TI.Text (TA.ByteArray ba) off len) =
+    ShortText (BSSI.SBS (case sliceByteArray (TA.ByteArray ba) off len of TA.ByteArray ba' -> ba'))
+
+sliceByteArray :: TA.Array -> Int -> Int -> TA.Array
+sliceByteArray ta@(TA.ByteArray ba) 0 len
+    | len == I# (GHC.Exts.sizeofByteArray# ba)
+    = ta
+sliceByteArray ta off len = TA.run $ do
+    ma <- TA.new len
+    TA.copyI len ma 0 ta off
+    return ma
+
+#else
 fromText = fromByteStringUnsafe . T.encodeUtf8
+#endif
 
 -- | \(\mathcal{O}(n)\) Construct 'ShortText' from UTF-8 encoded 'ShortByteString'
 --

text-short.cabal

@@ -1,7 +1,6 @@
 cabal-version:       1.18
 name:                text-short
-version:             0.1.4
-x-revision:          2
+version:             0.1.5
 synopsis:            Memory-efficient representation of Unicode text strings
 license:             BSD3
 license-file:        LICENSE
@@ -37,7 +36,7 @@ library
                      , bytestring  >= 0.10.4 && < 0.12
                      , hashable    >= 1.2.6  && < 1.5
                      , deepseq     >= 1.3    && < 1.5
-                     , text        >= 1.0    && < 1.3
+                     , text        >= 1.0    && < 1.3 || >=2.0 && <2.1
                      , binary      >= 0.7.1  && < 0.9
                      , ghc-prim    >= 0.3.1  && < 0.9
                      , template-haskell >= 2.9.0.0 && <2.19