Skip to content

Latest commit

 

History

History
183 lines (157 loc) · 9.55 KB

ChangeLog.md

File metadata and controls

183 lines (157 loc) · 9.55 KB
Raw

v0.10.9

  • Made StrictData a default extension
  • Further instrumenting cache operations

v0.10.8

  • Moved Timing to an effect

v0.10.7

  • Added timing utilities

v0.10.6

  • Added category based logging. For now interpreted in terms of previous logger. Adds ability to raise logging priority for a category.

v0.10.4

  • Added splitQ :: Q w m (a, b) -> m (Q w a, Q w b) to Knit.Effect.AtomicCache

v0.10.3.1

  • Removed the ConstraintAbsorber example in RandomExample because it’s not clear what that means in random-1.2

v0.10.3.0

  • Added color in logging output
  • Added ReadStdinStrict function to PandocMonad setup (from Pandoc >= 2.15)

v0.10.2.0

  • Added key to Serialization errors.
  • Bumped Pandoc upper bound to 3, lucid, polysemy and text upper bounds.

v0.10.1.0

  • Changes for streamly-0.8.0: removed deprecated Knit.Utilities.Streamly.streamlyToKnitS
  • Bumped polysemy, doctemplates and pandoc upper bounds
  • Removed Streamly.External.Cereal since we now handle this in Knit.Effect.Serialize

v0.10.0.0

  • Added Foldable and Traversable instances to CacheTime
  • Added explicit RST support
  • Removed separate support for serializing streamly streams. More confusing than helpful.

v0.9.0.0

  • Improved the types underlying the cache system to make the structure clearer and in preparation for adding combinators to make working with caching functions simpler.
  • Added a PrimMonad instance to StreamlyM so that it is easier to use when building mutable structures up from streams.
  • Fixed a performance issue/bug in cached streams. Previously, deserialization of the cached data into the stream would occur even if the stream was not used. Fixed this by pushing error handling for the deserialization into the stream itself.
  • Added fileDependency, updateIf and oldestUnit to Knit.Report.Cache in order to facilitate using the cache machinery to manage dependencies that result in file outputs that are not themselves loaded as inputs. This is useful for a workflow which is intended to keep produced assets data up-to-date.
  • Added mapCachedAction to Knit.Report.Cache for applying stream transformations directly to cached streams.

v0.8.0.0

  • BREAKING: moved config arguments to knit-html and knit-htmls into a data structure KnitConfig. This breaks all existing uses of those functions but is designed to future-proof against additions to the knit configuration. A breaking change was more-or-less necessitated by the addition of a cache directory parameter for the persistent caching effect. See below.

  • New persistent caching effect. Anything which has an instance of Serialize (from the cereal package) may be cached, in memory and backed by a disk-based cache (cache effect will attempt to create directories as required). When knit-html(s) is next run, the cached version can be used in place of running an expensive computation. This comes with some built-in dependency management, via modification times of the files in cache or user-provided time for dependencies on non-cached assets. For more information, see the Readme. For an example, see examples/CacheExample.hs.

  • Added Async into knit-haskell effects. For an example, see examples/AsyncExample.hs.

  • Logging changes:

  1. Added a Debug Int level below Diagnostic with an integer argument to allow various and unlimited debug levels
  2. Changed the previous [LogSeverity] argument given to run functions to (LogSeverity -> Bool) to accomdate the non-Boundedness of the new Debug Int levels.
  3. Made logging thread-aware by making prefixes thread-local and doing IO output of logs via the Say library
  4. Added type synonyms PrefixedLogEffects and PrefixedLogEffectsLE for type-level-lists of logging effects
  • Added functions to handle the failure branch of Maybe and Either Text as errors.
  • Added knitMapError to absorb other types of errors.
  • Updated some dependency bounds (NB: tests now depend on random-fu >= 2.7.3 since previous versions had bad log-domain bounds)
  • Updated for Pandoc >= 2.8. String to Text and template handling changes. All conditional (via CPP) so should work with < 2.8 as well.

v0.7.0.0

  • Updated vega versions in pandoc templates
  • removed (internal) datadir function because we now use the function from the autogenerated Paths_knit-haskell module. Which has its own complications but the other version didn't work at all.
  • Added more constraints to Knit.Report.KnitEffects to simplify things. If you use them to constrain a function to procuce something that can be knitted you will have all the knit effects available.
  • Where possible changed package bounds to >= A.B.C && < D.E. Most were already and some were easy to change. There are some lower bounds that are more specific and I won't generalize those unless I test them.
  • Changed bounds for hvega, polysemy and polysemy-zoo. Downstream changes to account for this.
  • SimpleExample now uses default template so that it also serves as a test that default templates are found.

v 0.6.0.1 - (Released 2019-06-21)

  • Moved the Pandoc default templates so they will be installed where Pandoc expects them. This is hard to test!

v 0.6.0.0 - (Released: 2019-06-19)

  • updated to use polysemy-zoo version of constraint absorbers in PandocMonad.
  • Hackage download should now include templates and associated css
  • Added mkPandocWriterConfig and addCss function in Knit.Report.Output to handle loading things in the included directory, which may be installed in a platform independent way.
  • Added two more html templates, pandoc-bootstrap-KH.html and pandoc-adaptive-bootstrap-KH.html with pandoc-bootstrap.css and used them in some of the examples.
  • Added the default html templates into the pandoc-templates directory where Pandoc will expect them if the given template fails.
  • Reorganized data directory with "pandoc-data" "knit-haskell-templates" and "knit-haskell-css"
  • Added output helpers for Pandoc results which are lazy text (html). These create the required parent directories if necessary. See examples for details.
  • Removed Knit.Effect.RandomFu and modified the example which uses it to use the "polysemy-RandomFu" package instead.

v 0.5.0.0

  • Added plots example back since there is a version of plots on hackage with a relaxed upper bound on containers.
  • Changed the Doc effect so that instead of Text name, it carries a polymorphic info type. This is then specified in the Pandoc case to be a Text name and set of template var overrides. This creates several other changes and will break any multi-doc examples since now "newPandoc" takes a first argument of the type PandocInfo (which is just a product of a Text and a Map String String).
  • raised lower bound and relaxed upper bound on polysemy

v 0.4.0.0

  • Added absorbPandocMonad :: PandocEffects r => (forall m. PandocMonad m => m a) -> Sem r a to Knit.Effect.PandocMonad
  • Removed (orphan) instances: PandocMonad (Sem r) and MonadError PandocError (Sem r) in favor of using absorbPandocMonad when required.
  • Deprecated name "Random" in favor of "RandomFu" for clarity and eventual consistency with Polysemy
  • Added absorbMonadRandom :: Member RandomFu r => (forall m. MonadRandom m => m a) -> Sem r a to allow some interoperation with actions constrained by MonadRandom
  • Removed orphan Random.MonadRandom instance from Knit.Effect.RandomFu because orphan instances are bad.
  • Changed return type of Knit.Report.knitError to Sem r a (from Sem r ())
  • Bumped lower bound on polysemy-plugin (because of a buggy version)
  • Bumped lower bound on polysemy
  • Removed plots example in "SimpleExample" and added a diagrams one. Will add plots back once a version issue with containers is resolved.

v 0.3.0.1

  • Examples were "executables" and are now "tests" so that depending on knit-haskell does not pull in the dependencies of the examples.

v 0.3.0.0

  • Added exports of "Colonnade" and "Text.Blaze.Colonnade" to "Knit.Report.Input.Table.Colonnade"
  • Added "knitError" function to Knit.Report to allow user throwing of errors. These will become PandocSomeError and handled as a PandocError.
  • Lowered bound on containers (to 0.5.0) to accomodate use of Plots.
  • Added Knit.Report.Input.Visualization.Diagrams. Adds Diagrams from diagrams-lib via SVG backend.
  • Added some more re-exports (Colonnade, Diagrams.Prelude) from Knit.Report to simplify imports on use.
  • Added an optional figure caption argument to Hvega and Diagrams inputs. (this is a breaking change since it requires another argument).
  • Added a state effect wrapper to facilitate getting unused ids for figures, etc.
  • Changed id argument for Hvega and Diagrams visualizations to "Maybe Text" from Text. Will use built-in facilities for unique ids when set to Nothing.
  • Visualization "addXXX" functions now return the figure id (as "Sem r Text") so it can be referred to elsewhere.
  • Added KnitOne and KnitMany constraint-type-aliases to Knit.Report to simplify constraining doc producing functions.
  • Updated examples. Added a diagrams example to SimpleExample and fixed id and caption arguments throughout.
  • Bumped upper bounds (network)

v 0.2.0.0

  • Documentation Fixes
  • Added export of newPandoc and NamedDoc to Knit.Report to facilitate multi-doc use.
  • Removed redundant imports from Knit.Report.Input.Table.Colonnade
  • Added a multi-doc example
  • (internal) Removed odd, and no longer necessary, "LastMember" constraint from knit functions.
  • Updated effects for polysemy 1.2

v 0.1.0.0

  • Initial version