You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
* make clientlib module
* gradle fixes for compilation
* gradle fixes for compilation
* gradle changes
* rename org.apache.platypus to com.yelp.platypus
* fix groupId in clientlib build.gradle
* Revert "rename org.apache.platypus to com.yelp.platypus"
This reverts commit a1640e8.
* rename org.apache.platypus to com.yelp.nrtsearch
* fix publication groupID
* Change all traces of "platypus" (except for repo names) to "nrtSearch"
* fix unfound java src in clientlib error
* fixes for building grpc-gateway
* fixes for building grpc-gateway
Copy file name to clipboardexpand all lines: README.md
+13-13
Original file line number
Diff line number
Diff line change
@@ -1,4 +1,4 @@
1
-
# Platypus
1
+
# nrtSearch
2
2
A high performance gRPC server on top of [Apache Lucene](http://lucene.apache.org/) version 8.x source, exposing lucene's
3
3
core functionality over a simple gRPC based API.
4
4
@@ -17,8 +17,8 @@ Fields must first be registered with the *registerFields* command, where you exp
17
17
There is no transaction log, so you must call *commit* yourself periodically to make recent changes durable on disk. This means that if a node crashes, all indexed documents since the last commit are lost.
18
18
19
19
# Indexing a stream of documents
20
-
platypus supports client side gRPC streaming for its *addDocuments* endpoint. This means that the server API accepts a stream of documents . The client can choose to stream the documents however it wishes.
21
-
The example platypus client implemented here reads a CSV file and streams documents from it over to the server. The server can index chunks of documents the size of which is configurable as the client
20
+
nrtSearch supports client side gRPC streaming for its *addDocuments* endpoint. This means that the server API accepts a stream of documents . The client can choose to stream the documents however it wishes.
21
+
The example nrtSearch client implemented here reads a CSV file and streams documents from it over to the server. The server can index chunks of documents the size of which is configurable as the client
22
22
continues to send more documents over its stream. gRPC enables this with minimal application code and yields higher performance compared to JSON. TODO[citation needed]: Add performance numbers of stream based indexing for some datasets.
23
23
24
24
# Near-real-time-replication
@@ -48,20 +48,20 @@ Note: This code has been tested on *Java13*
@@ -138,11 +138,11 @@ This should create a src/main/docs/index.html file that can be seen in your loca
138
138
This tool indexes yelp reviews available at [Yelp dataset challenge](https://www.yelp.com/dataset/challenge). It runs a default version with only 1k reviews of the `reviews.json` or you could download the yelp dataset and place the review.json in the user.home dir and the tool will use that instead. The complete review.json should have close to 7Million reviews. The tool runs multi-threaded indexing and a search thread in parallel reporting the `totalHits`. Command to run this specific test:
This test indexes businesses, creates an Infix Suggester and fetches suggestions. It requires a host, a port and a writeable directory in a standalone Platypus server.
146
+
This test indexes businesses, creates an Infix Suggester and fetches suggestions. It requires a host, a port and a writeable directory in a standalone nrtSearch server.
147
147
148
-
```./gradlew test -DsuggestTmp=remoteServerDir -DsuggestHost=yourStandaloneServerHost -DsuggestPort=yourStandaloneServerHost --tests "org.apache.platypus.server.YelpSuggestTest"```
148
+
```./gradlew test -DsuggestTmp=remoteServerDir -DsuggestHost=yourStandaloneServerHost -DsuggestPort=yourStandaloneServerHost --tests "com.yelp.nrtsearch.server.YelpSuggestTest"```
RUN for GOOS in darwin linux windows; do for GOARCH in 386 amd64; do echo "Building $GOOS-$GOARCH"; export GOOS=$GOOS; export GOARCH=$GOARCH; go build -o bin/http_wrapper-$GOOS-$GOARCH http_wrapper.go; done; done
77
+
RUN for GOOS in darwin linux windows; do \
78
+
for GOARCH in 386 amd64; do \
79
+
echo "Building $GOOS-$GOARCH"; \
80
+
export GOOS=$GOOS; \
81
+
export GOARCH=$GOARCH; \
82
+
go build -o bin/http_wrapper-$GOOS-$GOARCH http_wrapper.go; \
0 commit comments