From 1212abff50b607f82484f1833079baf5690e7637 Mon Sep 17 00:00:00 2001
From: andrewresearch <andrew@andrewresearch.net>
Date: Thu, 2 Nov 2017 16:25:25 +1000
Subject: [PATCH] Minor updates to separate NER and Parse from other pipelines

---
 .travis.yml                                       |  4 ++--
 build.sbt                                         |  2 +-
 .../factorie_nlp_api/AnnotatorPipelines.scala     | 15 +++++++++------
 3 files changed, 12 insertions(+), 9 deletions(-)

diff --git a/.travis.yml b/.travis.yml
index a222ea6..6908934 100644
--- a/.travis.yml
+++ b/.travis.yml
@@ -3,10 +3,10 @@ language: scala
 jdk: oraclejdk8
 
 scala:
-   - 2.12.4
+   - 2.12.3
 
 script:
-   - sbt ++$TRAVIS_SCALA_VERSION -J-Xmx6G test
+   - sbt ++$TRAVIS_SCALA_VERSION -J-Xmx2000m test
 
 # Use container-based infrastructure
 sudo: false
diff --git a/build.sbt b/build.sbt
index 17ec1e1..755566e 100644
--- a/build.sbt
+++ b/build.sbt
@@ -1,6 +1,6 @@
 name := "factorie-nlp-api"
 
-version := "0.5.0"
+version := "0.5.1"
 
 scalaVersion := "2.12.3"
 
diff --git a/src/main/scala/io/nlytx/factorie_nlp_api/AnnotatorPipelines.scala b/src/main/scala/io/nlytx/factorie_nlp_api/AnnotatorPipelines.scala
index e3fe7e4..d0e4c85 100644
--- a/src/main/scala/io/nlytx/factorie_nlp_api/AnnotatorPipelines.scala
+++ b/src/main/scala/io/nlytx/factorie_nlp_api/AnnotatorPipelines.scala
@@ -29,6 +29,7 @@ object AnnotatorPipelines {
   val logger = Logging(system.eventStream, "factorie-nlp-api")
 
   type Pipeline = String => RunnableGraph[Future[Document]]
+  type DocPipeline = Document => RunnableGraph[Future[Document]]
 
   //Make Document
   private lazy val doc = Flow[String].map(new Document(_))
@@ -74,14 +75,14 @@ object AnnotatorPipelines {
 
   val defaultPipeline = lemmaPipeline
 
-  val parsePipeline = (s:String) =>
-    Source.single(s)
-      .via(doc.map(tokeniser).map(segmenter).map(normaliser).map(postagger).map(lemmatiser).map(parser))
+  val parsePipeline = (d:Document) =>
+    Source.single(d)
+      .map(parser)
       .toMat(Sink.head[Document])(Keep.right)
 
-  val nerPipeline = (s:String) =>
-    Source.single(s)
-      .via(doc.map(tokeniser).map(segmenter).map(normaliser).map(postagger).map(lemmatiser).map(parser).mapAsync(2)(nerTagger))
+  val nerPipeline = (d:Document) =>
+    Source.single(d)
+      .mapAsync(2)(nerTagger)
       .toMat(Sink.head[Document])(Keep.right)
 
   val completePipeline = nerPipeline
@@ -89,6 +90,8 @@ object AnnotatorPipelines {
   /* The main method for running a pipeline */
   def process(text:String,pipeline:Pipeline=defaultPipeline):Future[Document] = pipeline(text).run
 
+  def processDoc(doc:Document,pipeline:DocPipeline):Future[Document] = pipeline(doc).run
+
   def profile(text:String,pipeline:Pipeline=defaultPipeline,wait:Int=180):Document = {
     logger.info(s"Profiling pipeline...")
     val start = System.currentTimeMillis()