mrkkrp
dc63a8f893
Add tests to cover new API for position advancing
...
Close #102 .
2016-05-04 00:53:09 +07:00
mrkkrp
59c01cba67
Improve some existing tests to increase coverage
2016-05-02 21:51:06 +07:00
mrkkrp
ae9ed355aa
Inline functions in ‘Text.Megaparsec.Char’
...
This gives a nice speed up — not too surprising given size of the
functions, although it would be nice if GHC could figure this out by
itself too…
2016-05-02 01:00:59 +07:00
mrkkrp
f4f66222ec
Improve performance of ‘(<*>)’
...
This handwritten version has proved to be more efficient from benchmarks
and profiling.
2016-05-02 00:59:58 +07:00
mrkkrp
df3b3016d5
Make ‘pToken’ lazier (position updating)
...
Evaluation of ‘npos’ is not necessary when we are going to report an
error. Similarly, evaluation of current position is not necessary when
we just need to get incremented position. This seemingly minor change
has profound impact on performance because call to ‘token’ function is
at the base of significant part of parsing process.
2016-05-02 00:53:42 +07:00
mrkkrp
2385cf5867
Add a test for ‘parseMaybe’
2016-04-30 19:02:26 +07:00
mrkkrp
49303f18c4
Add a trivial test for ‘MonadIO’ instance
2016-04-30 18:32:50 +07:00
mrkkrp
2eddd78f0f
Tests for ‘updatePos’ (different types of streams)
2016-04-30 18:22:00 +07:00
mrkkrp
3195a97776
Add new tests for ‘showTokens’ and pretty-printing
2016-04-30 17:31:52 +07:00
mrkkrp
9040284333
Use ‘lts-5.14’ and ‘QuickCheck-2.8.2’
2016-04-29 15:04:26 +07:00
mrkkrp
f651d59ea6
Require at least ‘QuickCheck-2.8.2’ for test suite
...
Because of ‘Arbitrary’ instances for things from ‘contianers’
package (namely ‘Set’).
2016-04-29 14:58:27 +07:00
Mark Karpov
2cad25dd11
Merge pull request #99 from mrkkrp/pos-and-error-reform
...
Position and error reform
2016-04-28 00:10:03 +07:00
mrkkrp
9240896e6b
Update the ‘CHANGELOG.md’ file
2016-04-28 00:07:15 +07:00
mrkkrp
1094408b96
Add ‘pushPosition’ and ‘popPosition’
...
Close #83 .
2016-04-27 00:04:56 +07:00
mrkkrp
8e095bab8c
A correction to ‘getPosition’ and ‘setPosition’
...
Make ‘getPosition’ and ‘setPosition’ work with current file, not with
the whole stack.
2016-04-26 23:01:09 +07:00
mrkkrp
f22cd69ff7
Cosmetic polishing
2016-04-26 00:37:06 +07:00
mrkkrp
13578590da
Adjustments for compatibility
2016-04-25 19:51:13 +07:00
mrkkrp
f576945a4a
Update tests for ‘Text.Megaparsec.Lexer’
2016-04-24 21:56:45 +07:00
mrkkrp
e11066f07a
Update tests for ‘Text.Megaparsec.Expr’
2016-04-24 21:56:45 +07:00
mrkkrp
98b08b752a
Update tests for ‘Text.Megaparsec.Prim’
2016-04-24 21:56:45 +07:00
mrkkrp
643842ac1e
Update tests for ‘Text.Megaparsec.Perm’
2016-04-24 21:56:45 +07:00
mrkkrp
9089ed3325
Update tests for ‘Text.Megaparsec.Combinator’
2016-04-24 21:56:45 +07:00
mrkkrp
6f69306515
Update tests for ‘Text.Megaparsec.Char’
2016-04-24 21:56:45 +07:00
mrkkrp
35dc3a4469
Write tests for the new ‘Text.Megaparsec.Error’
2016-04-24 21:56:40 +07:00
mrkkrp
14b1aeea8a
Write tests for the new ‘Text.Megaparsec.Pos’
2016-04-24 20:21:36 +07:00
mrkkrp
a1df59221b
Update ‘Util’ module in test suite
2016-04-24 20:21:07 +07:00
mrkkrp
be3c1fe987
Derive ‘Read’ instance for ‘Dec’
2016-04-24 20:19:48 +07:00
mrkkrp
495aa1bb87
Add ‘incorrectIndent’, improve indentation errors
...
Close #90 .
2016-04-24 15:09:52 +07:00
mrkkrp
1ba128521f
Make ‘tokens’ use “canonicalized” token streams
...
And introduce the notion in docs of ‘MessageItem’.
2016-04-23 22:31:09 +07:00
mrkkrp
91955013ef
Adjust ‘old-tests’ test suite to new code base
2016-04-22 19:53:36 +07:00
mrkkrp
0879af8103
Add more dependencies to ‘tests’ component
2016-04-22 19:52:55 +07:00
mrkkrp
dd0cf0fe4a
Add one more argument to ‘token’
...
The new argument allows to specify what to report as expected token when
input stream happens to be empty.
2016-04-22 19:51:14 +07:00
mrkkrp
c0c9b836dd
Fix a bug (hints assymetry with ‘<|>’)
2016-04-22 19:48:55 +07:00
mrkkrp
17a6bb2baa
Remove ‘MonadThrow’ instance, add ‘Dec’
...
‘Dec’ stands for “default error component”. We need this because
‘String’ looses information and thus we cannot write good enough tests
with it.
2016-04-22 19:48:34 +07:00
mrkkrp
eedf69761e
Remove redundant constraints from lexer module
2016-04-18 21:52:45 +07:00
mrkkrp
84038a4fab
Export ‘parseErrorPretty’ from ‘Text.Megaparsec’
2016-04-18 20:48:31 +07:00
mrkkrp
f3f0aef44e
Improve docs of ‘Text.Megaparsec.Prim’
2016-04-18 19:49:52 +07:00
mrkkrp
006900e257
Add missing ‘@since’ labels to new functions
2016-04-18 19:49:20 +07:00
mrkkrp
fee19134e8
Improve new version of ‘Text.Megaparsec.Error’
...
Compatibility and documentation improvements.
2016-04-18 19:05:34 +07:00
mrkkrp
641114a442
Improve new version of ‘Text.Megaparsec.Pos’
...
Documentation improvements. The commit also makes it compile with all
supported GHC versions.
2016-04-18 17:59:32 +07:00
mrkkrp
acbae63a21
First attempt to outline Megaparsec 5.0.0
...
This is rather a sketch, we need to work on documentation, tests, and
perhaps on performance, but it should show the direction Megaparsec
5.0.0 is taking.
2016-04-17 20:08:13 +07:00
mrkkrp
a3254f5371
Move language extensions to files
...
This turns out that it's easier for readers to lookup the list of
extensions in current file than go to Cabal file.
2016-04-10 19:26:54 +07:00
mrkkrp
747993e0bb
Add ‘skipBlockCommentNested’ function
...
Close #96 .
2016-03-30 14:50:35 +06:00
mrkkrp
18d192ba70
Use ‘Scientific’ as target type for floats
...
Close #95 .
Here we introduce ‘scientific’ parser that can parse arbitrary big
numbers without error or memory overflow. ‘float’ still returns
‘Double’, but it's defined in terms of ‘scientific’ now. Since
‘Scientific’ type can reliably represent integer values as well as
floating point values, ‘number’ now returns ‘Scientific’ instead of
‘Either Integer Double’ (‘Integer’ or ‘Double’ can be extracted from
‘Scientific’ value anyway). This in turn makes ‘signed’ parser more
natural and general, because we do not need ad-hoc ‘Signed’ type class
anymore.
2016-03-29 16:42:51 +06:00
Mark Karpov
cd7f12963c
Merge pull request #94 from mrkkrp/position-advancing
...
More flexible position-advancing function
2016-03-29 10:39:01 +07:00
mrkkrp
aa805864af
Add simple tests for ‘Text’ and ‘ByteString’
...
These are primitive tests, but nevertheless they check ‘uncons’ and
‘updatePos’ — everything we need.
2016-03-29 09:19:50 +06:00
mrkkrp
00ae88a8e7
Add a test to check that ‘tokens’ never consumes
2016-03-28 22:51:07 +06:00
mrkkrp
0b4abd81ae
Update the ‘CHANGELOG.md’ file
2016-03-28 21:23:23 +06:00
mrkkrp
39b9ffb083
Add “since 5.0.0” to haddocks of new functions
2016-03-28 21:10:00 +06:00
mrkkrp
0f7a2c180c
Improve efficiency of the library
2016-03-28 15:43:03 +06:00
