Merge pull request #17 from eggzilla/jsonAlien

Json alien

Author: eggzilla

.travis.yml

@@ -1,110 +1,28 @@
-#Copied as is from https://www.fpcomplete.com/blog/2016/02/updated-haskell-travis-config 
-#Copy these contents into the root directory of your Github project in a file
-# named .travis.yml
+sudo: required
 
-# Use new container infrastructure to enable caching
-sudo: false
-
-# Choose a lightweight base image; we provide our own build tools.
 language: c
 
-# Caching so the next build will be fast too.
-cache:
-  directories:
-  - $HOME/.ghc
-  - $HOME/.cabal
-  - $HOME/.stack
-
-# The different configurations we want to test. We have BUILD=cabal which uses
-# cabal-install, and BUILD=stack which uses Stack. More documentation on each
-# of those below.
-#
-# We set the compiler values here to tell Travis to use a different
-# cache file per set of arguments.
-#
-# If you need to have different apt packages for each combination in the
-# matrix, you can use a line such as:
-#     addons: {apt: {packages: [libfcgi-dev,libgmp-dev]}}
-matrix:
-  include:
-  # We grab the appropriate GHC and cabal-install versions from hvr's PPA. See:
-  # https://github.com/hvr/multi-ghc-travis
-  - env: BUILD=cabal GHCVER=8.0.2 CABALVER=1.24
-    compiler: ": #GHC 8.0.2"
-    addons: {apt: {packages: [cabal-install-1.24,ghc-8.0.2], sources: [hvr-ghc]}}
-
-  # The Stack builds. We can pass in arbitrary Stack arguments via the ARGS
-  # variable, such as using --stack-yaml to point to a different file.
-  #- env: BUILD=stack ARGS="--resolver lts-6.22"
-  #  compiler: ": #stack 7.10.3"
-  #  addons: {apt: {packages: [ghc-7.10.3], sources: [hvr-ghc]}}
-
-  # Nightly builds are allowed to fail
-  #- env: BUILD=stack ARGS="--resolver nightly"
-  #  compiler: ": #stack nightly"
-  #  addons: {apt: {packages: [libgmp-dev]}}
-
-  # Build on OS X in addition to Linux
-  #- env: BUILD=stack ARGS="--resolver lts-6.22"
-  #  compiler: ": #stack 7.10.3 osx"
-  #  os: osx
-
-  #- env: BUILD=stack ARGS="--resolver nightly"
-  #  compiler: ": #stack nightly osx"
-  #  os: osx
-
-  allow_failures:
-  - env: BUILD=cabal GHCVER=8.0.1 CABALVER=1.24
-  - env: BUILD=stack ARGS="--resolver nightly"
-
-before_install:
-# Using compiler above sets CC to an invalid value, so unset it
-- unset CC
-
-# We want to always allow newer versions of packages when building on GHC HEAD
-- CABALARGS=""
-- if [ "x$GHCVER" = "xhead" ]; then CABALARGS=--allow-newer; fi
-
-# Download and unpack the stack executable
-- export PATH=/opt/ghc/$GHCVER/bin:/opt/cabal/$CABALVER/bin:$HOME/.local/bin:$PATH
-- mkdir -p ~/.local/bin
-- |
-  if [ `uname` = "Darwin" ]
-  then
-    curl --insecure -L https://www.stackage.org/stack/osx-x86_64 | tar xz --strip-components=1 --include '*/stack' -C ~/.local/bin
-  else
-    curl -L https://www.stackage.org/stack/linux-x86_64 | tar xz --wildcards --strip-components=1 -C ~/.local/bin '*/stack'
-  fi
+services:
+  - docker
 
-install:
-- echo "$(ghc --version) [$(ghc --print-project-git-commit-id 2> /dev/null || echo '?')]"
-- if [ -f configure.ac ]; then autoreconf -i; fi
-- |
-  case "$BUILD" in
-    stack)
-      stack --no-terminal --install-ghc $ARGS test --only-dependencies
-      ;;
-    cabal)
-      cabal --version
-      travis_retry cabal update
-      cabal install --only-dependencies --enable-tests --enable-benchmarks --force-reinstalls --ghc-options=-O0 --reorder-goals --max-backjumps=-1 $CABALARGS
-      ;;
-  esac
+before_script:
+  - docker build --tag devel -f Dockerfile.dev .
 
 script:
-- |
-  case "$BUILD" in
-    stack)
-      stack --no-terminal $ARGS test --haddock --no-haddock-deps
-      ;;
-    cabal)
-      cabal configure --enable-tests --enable-benchmarks -v2 --ghc-options="-O0 -Werror"
-      cabal build
-      cabal check || [ "$CABALVER" == "1.16" ]
-      cabal test
-      cabal sdist
-      cabal copy
-      SRC_TGZ=$(cabal info . | awk '{print $2;exit}').tar.gz && \
-        (cd dist && cabal install --force-reinstalls "$SRC_TGZ")
-      ;;
-  esac
+  - docker create --name develcontainer devel
+  - mkdir RNAlien
+  - docker cp develcontainer:/source/dist-newstyle/build/x86_64-linux/ghc-8.4.3/RNAlien-1.6.0/x/RNAlien/build/RNAlien/RNAlien RNAlien
+  - docker cp develcontainer:/source/dist-newstyle/build/x86_64-linux/ghc-8.4.3/RNAlien-1.6.0/x/RNAlienStatistics/build/RNAlienStatistics/RNAlienStatistics RNAlien
+  - docker cp develcontainer:/source/dist-newstyle/build/x86_64-linux/ghc-8.4.3/RNAlien-1.6.0/x/cmsearchToBed/build/cmsearchToBed/cmsearchToBed RNAlien
+  - docker cp develcontainer:/source/dist-newstyle/build/x86_64-linux/ghc-8.4.3/RNAlien-1.6.0/x/RNAcentralHTTPRequest/build/RNAcentralHTTPRequest/RNAcentralHTTPRequest RNAlien
+  - cp LICENSE RNAlien
+  - tar -cvzf RNAlien.tar.gz RNAlien
+
+
+deploy:
+  provider: releases
+  skip_cleanup: true
+  api_key: $GITHUB_TOKEN
+  file: "RNAlien.tar.gz" 
+  on:
+    tags: true

src/Bio/RNAcentralHTTPRequest.hs renamed to Biobase/RNAcentralHTTPRequest.hs

@@ -6,7 +6,7 @@
 module Main where
 
 import System.Console.CmdArgs
-import Bio.RNAcentralHTTP
+import Biobase.RNAlien.RNAcentralHTTP
 
 data Options = Options
   { inputSequence :: String

src/Bio/RNAlien.hs renamed to Biobase/RNAlien.hs

@@ -3,20 +3,21 @@
 
 -- | Unsupervized construction of RNA family models
 --   For more information on RNA family models consult <http://>
---   Testcommand: dist/build/RNAlien/RNAlien -i ~egg/initialfasta/RybB.fa -c 3 -o /scr/kronos/egg/temp/ > ~egg/Desktop/alieninitialtest
+--   Usage example: RNAlien -i /path/input.fa -c 5 -o /outdir/
+--   Usage example offline mode: RNAlien -i /path/input.fa -b /backup/blast/nt_v5 -o /outdir/ -c 5 -t 1396 -j 
 module Main where
 
 import System.Console.CmdArgs
 import System.Directory
-import Bio.Sequence.Fasta
-import Bio.RNAlienData
-import Bio.RNAlienLibrary
+import Biobase.RNAlien.Types
+import Biobase.RNAlien.Library
 import Data.Maybe
 import Data.Either.Unwrap
 import Data.Time
 import qualified System.FilePath as FP
 import Paths_RNAlien (version)
 import Data.Version (showVersion)
+--import Biobase.Fasta.Streaming
 
 data Options = Options
   { inputFastaFilePath :: String,
@@ -35,7 +36,8 @@ data Options = Options
     taxonomyRestriction :: Maybe String,
     sessionIdentificator :: Maybe String,
     performEvaluation :: Bool,
-    checkSetup :: Bool
+    checkSetup :: Bool,
+    offlineMode :: Bool
   } deriving (Show,Data,Typeable)
 
 options :: Options
@@ -45,7 +47,7 @@ options = Options
     inputTaxId = Nothing &= name "t" &= help "NCBI taxonomy ID number of input RNA organism",
     inputnSCICutoff = Just (1 :: Double) &= name "z" &= help "Only candidate sequences with a normalized structure conservation index (nSCI) higher than this value are accepted. Default: 1",
     inputEvalueCutoff = Just (0.001 :: Double) &= name "e" &= help "Evalue cutoff for cmsearch filtering. Default: 0.001",
-    inputBlastDatabase = Just "nt" &= name "b" &= help "Specify name of blast database to use. Default: nt",
+    inputBlastDatabase = Just "nt" &= name "b" &= help "Specify name of blast database to use, in offline mode the filepath to the blast database (/home/user/nt_v5). Default: nt",
     lengthFilter = True &= name "l" &= help "Filter blast hits per genomic length. Default: True",
     coverageFilter = True &= name "a" &= help "Filter blast hits by coverage of at least 80%. Default: True",
     singleHitperTax = False &= name "s" &= help "Only the best blast hit per taxonomic entry is considered. Default: False",
@@ -56,8 +58,9 @@ options = Options
     taxonomyRestriction = Nothing &= name "r" &= help "Restrict search space to taxonomic kingdom (bacteria,archea,eukaryia). Default: not set",
     sessionIdentificator = Nothing &= name "d" &= help "Optional session id that is used instead of automatically generated one.",
     performEvaluation = True &= name "x" &= help "Perform evaluation step. Default: True",
-    checkSetup = False &= name "g" &= help "Just prints installed tool versions and performs connection check. Default: False"
-  } &= summary ("RNAlien " ++ alienVersion) &= help "Florian Eggenhofer, Ivo L. Hofacker, Christian Hoener zu Siederdissen - 2013 - 2017" &= verbosity
+    checkSetup = False &= name "g" &= help "Just prints installed tool versions and performs connection check. Default: False",
+    offlineMode = False &= name "j" &= help "Uses locally installed blast and databases. Default: False"
+  } &= summary ("RNAlien " ++ alienVersion) &= help "Florian Eggenhofer, Ivo L. Hofacker, Christian Hoener zu Siederdissen - 2013 - 2019" &= verbosity
 
 main :: IO ()
 main = do
@@ -91,7 +94,7 @@ main = do
            writeFile (temporaryDirectoryPath ++ "log/warnings") ("")
            logMessage ("Timestamp: " ++ (show timestamp) ++ "\n") temporaryDirectoryPath
            logMessage ("Temporary Directory: " ++ temporaryDirectoryPath ++ "\n") temporaryDirectoryPath
-           inputFasta <- readFasta inputFastaFilePath
+           inputFasta <- readFastaFile inputFastaFilePath
            if null inputFasta
              then do
                putStrLn "Error: Input fasta file is empty."
@@ -106,10 +109,10 @@ main = do
                  else do
                    logToolVersions inputQuerySelectionMethod temporaryDirectoryPath
                    let inputSequence = reformatFasta (head inputFasta)
-                   initialTaxId <- setInitialTaxId inputBlastDatabase temporaryDirectoryPath inputTaxId inputSequence
+                   initialTaxId <- setInitialTaxId offlineMode threads inputBlastDatabase temporaryDirectoryPath inputTaxId inputSequence
                    let checkedTaxonomyRestriction = checkTaxonomyRestriction taxonomyRestriction
-                   let staticOptions = StaticOptions temporaryDirectoryPath sessionId (fromJust inputnSCICutoff) inputTaxId singleHitperTax inputQuerySelectionMethod inputQueryNumber lengthFilter coverageFilter blastSoftmasking threads inputBlastDatabase checkedTaxonomyRestriction (setVerbose verboseLevel)
-                   let initialization = ModelConstruction iterationNumber inputSequence [] initialTaxId Nothing (fromJust inputEvalueCutoff) False [] []
+                   let staticOptions = StaticOptions temporaryDirectoryPath sessionId (fromJust inputnSCICutoff) inputTaxId singleHitperTax inputQuerySelectionMethod inputQueryNumber lengthFilter coverageFilter blastSoftmasking threads inputBlastDatabase checkedTaxonomyRestriction (setVerbose verboseLevel) offlineMode
+                   let initialization = ModelConstruction iterationNumber inputFasta [] initialTaxId Nothing (fromJust inputEvalueCutoff) False [] []
                    logMessage (show initialization) temporaryDirectoryPath
                    modelConstructionResults <- modelConstructer staticOptions initialization
                    let resultTaxonomyRecordsCSVTable = constructTaxonomyRecordsCSVTable modelConstructionResults

src/Bio/InfernalParser.hs renamed to Biobase/RNAlien/InfernalParser.hs

@@ -1,7 +1,7 @@
 -- | This module contains parsing functions for Infernal programs
 
-module Bio.InfernalParser (
-                           module Bio.RNAlienData,
+module Biobase.RNAlien.InfernalParser (
+                           module Biobase.RNAlien.Types,
                            readCMSearch,
                            readCMSearches,
                            parseCMSearch,
@@ -12,8 +12,8 @@ module Bio.InfernalParser (
 where
 
 import Text.ParserCombinators.Parsec
-import Bio.RNAlienData
-import qualified Data.ByteString.Lazy.Char8 as L
+import Biobase.RNAlien.Types
+import qualified Data.ByteString.Char8 as B
 import qualified Control.Exception.Base as CE
 
 -- | parse from input filePath              
@@ -240,7 +240,7 @@ genParserCMsearchHit = do
   newline
   optional (try (string " ------ inclusion threshold ------"))
   optional (try newline)
-  return $ CMsearchHit (readInt hitRank') hitSignificant' (readDouble hitEValue') (readDouble hitScore') (readDouble hitBias') (L.pack hitSequenceHeader') (readInt hitStart') (readInt hitEnd') hitStrand' (L.pack hitModel') (L.pack hitTruncation') (readDouble hitGCcontent') (L.pack hitDescription')
+  return $ CMsearchHit (readInt hitRank') hitSignificant' (readDouble hitEValue') (readDouble hitScore') (readDouble hitBias') (B.pack hitSequenceHeader') (readInt hitStart') (readInt hitEnd') hitStrand' (B.pack hitModel') (B.pack hitTruncation') (readDouble hitGCcontent') (B.pack hitDescription')
 
 -- | parse from input filePath              
 parseCMstat :: String -> Either ParseError CMstat