c0a93a4 · streamly · git.iko.soy

CONTRIBUTING.md+1−1

compare with master

unifiedsplit

@@ -191,7 +191,7 @@ folds where there is no incoming state, we start with the initial state

191

1921. When we are building a concurrent stream that needs to share the same `SVar`

193 we pass the incoming state as is.

194−2. In all other cases we must not share the SVar and everytime we pass on the

194+2. In all other cases we must not share the SVar and every time we pass on the

195 state to run a stream we must use `rstState` to reset the `SVar` in the

196 state.

197

Changelog.md+1−1

@@ -355,7 +355,7 @@ concurrently using this.

355* Add `iterate`, `iterateM` stream operations

356

357### Bug Fixes

358−* Fixed a bug that casued unexpected behavior when `pure` was used to inject

358+* Fixed a bug that caused unexpected behavior when `pure` was used to inject

359 values in Applicative composition of `ZipStream` and `ZipAsync` types.

360

361## 0.1.1

bench.sh+2−2

@@ -21,7 +21,7 @@ print_help () {

21 echo "commit is generated, in the 'charts' directory."

22 echo "Use --base and --candidate to select the commits to compare."

23 echo

24− echo "Any arguments after a '--' are passed directly to guage"

24+ echo "Any arguments after a '--' are passed directly to gauge"

25 exit

26}

27

@@ -86,7 +86,7 @@ build_report_progs() {

86

87# We run the benchmarks in isolation in a separate process so that different

88# benchmarks do not interfere with other. To enable that we need to pass the

89−# benchmark exe path to guage as an argument. Unfortunately it cannot find its

89+# benchmark exe path to gauge as an argument. Unfortunately it cannot find its

90# own path currently.

91

92# The path is dependent on the architecture and cabal version.

design/module-organization.md+1−1

@@ -1,6 +1,6 @@

1# Internal vs External

2

3−We keep all modules exposed to faciliate convenient exposure of experimental

3+We keep all modules exposed to facilitate convenient exposure of experimental

4APIs and constructors to users. It allows users of the library to experiment

5much more easily and carry a caveat that these APIs can change in future

6without notice. Since everything is exposed, maintainers do not have to think

design/utf8-decoder.md+3−3

@@ -139,7 +139,7 @@ null-terminated UTF-8 encoded string.

139

140### Error recovery

141

142−It is sometimes desireable to recover from errors when decoding strings

142+It is sometimes desirable to recover from errors when decoding strings

143that are supposed to be UTF-8 encoded. Programmers should be aware that

144this can negatively affect the security properties of their application.

145A common recovery method is to replace malformed sequences with a

@@ -155,8 +155,8 @@ aswell.

155

156The following code implements one such recovery strategy. When an

157unexpected byte is encountered, the sequence up to that point will be

158−replaced and, if the error occured in the middle of a sequence, will

158+replaced and, if the error occurred in the middle of a sequence, will

159−retry the byte as if it occured at the beginning of a string. Note that

159+retry the byte as if it occurred at the beginning of a string. Note that

160the decode function detects errors as early as possible, so the sequence

161`0xED 0xA0 0x80` would result in three replacement characters.

162

examples/ControlFlow.hs+1−1

@@ -69,7 +69,7 @@ mainMaybeBelow = do

69-- of non-determinism below.

70--

71-- Note that this is redundant configuration as the same behavior can be

72−-- acheived with just streamly, using mzero.

72+-- achieved with just streamly, using mzero.

73--

74getSequenceMaybeAbove :: (IsStream t, MonadIO (t m)) => MaybeT (t m) ()

75getSequenceMaybeAbove = do

src/Streamly.hs+1−1

@@ -236,7 +236,7 @@ import qualified Streamly.Prelude as P

236import qualified Streamly.Internal.Prelude as IP

237import qualified Streamly.Streams.StreamK as K

238

239−-- XXX provide good succinct examples of pipelining, merging, splitting ect.

239+-- XXX provide good succinct examples of pipelining, merging, splitting etc.

240-- below.

241--

242-- $streams

src/Streamly/Data/Fold.hs+1−1

@@ -190,7 +190,7 @@ module Streamly.Data.Fold

190 -- ...

191 -- @

192 --

193− -- To compute the average of numbers in a stream without going throught he

193+ -- To compute the average of numbers in a stream without going through the

194 -- stream twice:

195 --

196 -- >>> let avg = (/) <$> FL.sum <*> fmap fromIntegral FL.length

src/Streamly/Internal/Data/Pipe.hs+2−2

@@ -141,7 +141,7 @@ module Streamly.Internal.Data.Pipe

141 -- ...

142 -- @

143 --

144− -- To compute the average of numbers in a stream without going throught he

144+ -- To compute the average of numbers in a stream without going through the

145 -- stream twice:

146 --

147 -- >>> let avg = (/) <$> FL.sum <*> fmap fromIntegral FL.length

@@ -1805,7 +1805,7 @@ classifyKeepAliveChunks spanout = classifyChunksBy spanout True

1805-- All the input elements belonging to a session are collected using the fold

1806-- @f@. The session key and the fold result are emitted in the output stream

1807-- when the session is purged either via the session close event or via the

1808−-- session liftime timeout.

1808+-- session lifetime timeout.

1809--

1810-- @since 0.7.0

1811{-# INLINABLE classifySessionsBy #-}

src/Streamly/Internal/Data/SVar.hs+2−2

@@ -1674,7 +1674,7 @@ estimateWorkers workerLimit svarYields gainLossYields

1674 -- maxWorkerLatency.

1675 --

1676 let

1677− -- How many workers do we need to acheive the required rate?

1677+ -- How many workers do we need to achieve the required rate?

1678 --

1679 -- When the workers are IO bound we can increase the throughput by

1680 -- increasing the number of workers as long as the IO device has enough

@@ -1691,7 +1691,7 @@ estimateWorkers workerLimit svarYields gainLossYields

1691 -- use that to determine the max rate of workers, and also take the CPU

1692 -- bandwidth into account. We can also discover the IO bandwidth if we

1693 -- know that we are not CPU bound, then how much steady state rate are

1694− -- we able to acheive. Design tests for CPU bound and IO bound cases.

1694+ -- we able to achieve. Design tests for CPU bound and IO bound cases.

1695

1696 -- Calculate how many yields are we ahead or behind to match the exact

1697 -- required rate. Based on that we increase or decrease the effective

src/Streamly/Internal/FileSystem/File.hs+1−1

@@ -22,7 +22,7 @@

22-- session consisting of multiple reads and writes to the handle, these APIs

23-- are one shot read or write APIs. These APIs open the file handle, perform

24-- the requested operation and close the handle. Thease are safer compared to

25−-- the handle based APIs as there is no possiblity of a file descriptor

25+-- the handle based APIs as there is no possibility of a file descriptor

26-- leakage.

27--

28-- > import qualified Streamly.Internal.FileSystem.File as File

src/Streamly/Internal/Prelude.hs+2−2

@@ -2077,7 +2077,7 @@ insertBy cmp x m = fromStreamS $ S.insertBy cmp x (toStreamS m)

2077-- Deleting

2078------------------------------------------------------------------------------

2079

2080−-- | Deletes the first occurence of the element in the stream that satisfies

2080+-- | Deletes the first occurrence of the element in the stream that satisfies

2081-- the given equality predicate.

2082--

2083-- @

@@ -3644,7 +3644,7 @@ classifyKeepAliveChunks spanout = classifyChunksBy spanout True

3644-- All the input elements belonging to a session are collected using the fold

3645-- @f@. The session key and the fold result are emitted in the output stream

3646-- when the session is purged either via the session close event or via the

3647−-- session liftime timeout.

3647+-- session lifetime timeout.

3648--

3649-- @since 0.7.0

3650{-# INLINABLE classifySessionsBy #-}

src/Streamly/Prelude.hs+1−1

@@ -781,7 +781,7 @@ import Streamly.Internal.Prelude

781-- left fold reconstructs in a LIFO style, thereby reversing the order of

782-- elements..

783-- 3. A right fold has termination control and therefore can terminate early

784−-- without going throught the entire input, a left fold cannot terminate

784+-- without going through the entire input, a left fold cannot terminate

785-- without consuming all of its input. For example, a right fold

786-- implementation of 'or' can terminate as soon as it finds the first 'True'

787-- element, whereas a left fold would necessarily go through the entire input

src/Streamly/Streams/Serial.hs+1−1

@@ -308,7 +308,7 @@ instance IsStream WSerialT where

308------------------------------------------------------------------------------

309

310-- Additionally we can have m elements yield from the first stream and n

311−-- elements yeilding from the second stream. We can also have time slicing

311+-- elements yielding from the second stream. We can also have time slicing

312-- variants of positional interleaving, e.g. run first stream for m seconds and

313-- run the second stream for n seconds.

314--

src/Streamly/Streams/StreamD.hs+1−1

@@ -579,7 +579,7 @@ enumerateFromThenIntegral from next =

579-- 9007199254740992 + 2 :: Double => 9.007199254740994e15

580

581-- Instead we accumulate the increment counter and compute the increment

582−-- everytime before adding it to the starting number.

582+-- every time before adding it to the starting number.

583--

584-- This works for Integrals as well as floating point numbers, but

585-- enumerateFromStepIntegral is faster for integrals.