4 Commits
11 changed files with 97 additions and 44 deletions
+53 -1
View File
@@ -1 +1,53 @@
# RSDO (Razvoj Slovenščine v Digitalnem Okolju) - Concordancer # RSDO (Razvoj Slovenščine v Digitalnem Okolju) - Concordancer
This repository contains the source code of the concordancer which is a part of the [Terminology Portal](https://github.com/clarinsi/rsdo_term_portal).
There are two main parts:
- Web application with backend (API interface) written in ASP.NET Core 6 and frontend written in React
- System Manager which is written in .NET 6 and is used to initially set up the system
## Prerequisites
Web application requires following external dependencies
- OpenSearch v2.3.0 (use docker image: `opensearchproject/opensearch:2.3.0`)
- PostgreSQL (use docker image: `postgres`)
## Building source code
### Frontend (React application)
To build a frontend application successfully, first, [Node.js](https://nodejs.org/en/) must be installed (at least version 16).
Use `npm install` to install all required dependencies, and `npm run build` to build the application.
For more info see the [Readme](src/apps/Rsdo.Concordancer.Api/WebApp/README.md) inside the frontend [root](src/apps/Rsdo.Concordancer.Api/WebApp) folder.
### Backend (ASP.NET Core 6) and System Manager (.NET 6)
Open [Rsdo.Concordancer.sln](src/Rsdo.Concordancer.sln) inside of your IDE and build the solution, or use `dotnet build` command to build the solution. All required dependencies will be installed via NuGet packages.
## Creating docker images
There are two Dockerfile's in the source repository for each part. Both images should be compiled within the context of the project root folder.
### Web application
[Dockerfile](src/apps/Rsdo.Concordancer.Api/Dockerfile) for Web Application accepts additional optional argument (WEBAPP_PUBLIC_URL) which is mapped to React [PUBLIC_URL](https://create-react-app.dev/docs/using-the-public-folder/) variable. If the variable is omited, then application should be deployed on a domain root.
`docker build -t rsdo-concordancer-api -f src/apps/Rsdo.Concordancer.Api/Dockerfile .`
`docker build --build-arg WEBAPP_PUBLIC_URL=/corpus -t rsdo-concordancer-api -f src/apps/Rsdo.Concordancer.Api/Dockerfile .`
### System Manager
Use following [Dockerfile](src/apps/Rsdo.Concordancer.SystemManager/Dockerfile) to build System Manager.
`docker build -t rsdo-concordancer-systemmanager -f src/apps/Rsdo.Concordancer.SystemManager/Dockerfile .`
## Environment variables
### Web application
Web application must be configured with the following Environment variables and values:
- `RSDO:Database:MasterConnectionString` holds connection string to the main database; e.g.: `rsdo_db;Username=rsdo;Password=strongpassword;Database=Concordancer`
- `RSDO:Elastic:ConnectionString` holds connection string to OpenSearch server; e.g.: `http://rsdo_opensearch:9200`
- `RSDO:Tokenizer:ClasslaUrl` holds and url to external service which is used for tokenizing search strings; e.g.: `https://orodja.cjvt.si/oznacevalnik/ajax_api/v1/slv/process`
### System Manager
System manager must be configured with the following Environment variables and values:
- `RSDO:Database:MasterConnectionString` holds connection string to the main database; e.g.: `rsdo_db;Username=rsdo;Password=strongpassword;Database=Concordancer`
- `RSDO:Elastic:ConnectionString` holds connection string to OpenSearch server; e.g.: `http://rsdo_opensearch:9200`
## Using System Manager
To initialy setup the system, you have to use System Manager and run following commands:
1. `dotnet Rsdo.Concordancer.SystemManager.dll createMasterDb --connectionString="Host=rsdo_db;Username=rsdo;Password=strongpassword;Database=postgres"` which will create master database with the name Concordancer. You can select different name with the `--name` option
2. `dotnet Rsdo.Concordancer.SystemManager.dll importSloleks --sourceFile=/data/Sloleks2.0.LMF/sloleks_clarin_2.0.xml` which will import Sloleks (which is used for better searching). XML file is not part of this repository, but it can be downloaded from the following [URL](https://www.clarin.si/repository/xmlui/bitstream/handle/11356/1230/Sloleks2.0.LMF.zip?sequence=3&isAllowed=y)
@@ -4,7 +4,7 @@
<TargetFramework>net6.0</TargetFramework> <TargetFramework>net6.0</TargetFramework>
<GenerateDocumentationFile>true</GenerateDocumentationFile> <GenerateDocumentationFile>true</GenerateDocumentationFile>
<EnableNETAnalyzers>false</EnableNETAnalyzers> <EnableNETAnalyzers>false</EnableNETAnalyzers>
<Version>1.0.1</Version> <Version>1.0.2</Version>
</PropertyGroup> </PropertyGroup>
<ItemGroup> <ItemGroup>
@@ -5,7 +5,7 @@
<TargetFramework>net6.0</TargetFramework> <TargetFramework>net6.0</TargetFramework>
<GenerateDocumentationFile>true</GenerateDocumentationFile> <GenerateDocumentationFile>true</GenerateDocumentationFile>
<EnableNETAnalyzers>false</EnableNETAnalyzers> <EnableNETAnalyzers>false</EnableNETAnalyzers>
<Version>1.0.1</Version> <Version>1.0.2</Version>
</PropertyGroup> </PropertyGroup>
<ItemGroup> <ItemGroup>
@@ -4,7 +4,7 @@
<TargetFramework>net6.0</TargetFramework> <TargetFramework>net6.0</TargetFramework>
<GenerateDocumentationFile>true</GenerateDocumentationFile> <GenerateDocumentationFile>true</GenerateDocumentationFile>
<EnableNETAnalyzers>false</EnableNETAnalyzers> <EnableNETAnalyzers>false</EnableNETAnalyzers>
<Version>1.0.1</Version> <Version>1.0.2</Version>
</PropertyGroup> </PropertyGroup>
<ItemGroup> <ItemGroup>
@@ -4,7 +4,7 @@
<TargetFramework>net6.0</TargetFramework> <TargetFramework>net6.0</TargetFramework>
<GenerateDocumentationFile>true</GenerateDocumentationFile> <GenerateDocumentationFile>true</GenerateDocumentationFile>
<EnableNETAnalyzers>false</EnableNETAnalyzers> <EnableNETAnalyzers>false</EnableNETAnalyzers>
<Version>1.0.1</Version> <Version>1.0.2</Version>
</PropertyGroup> </PropertyGroup>
<ItemGroup> <ItemGroup>
@@ -4,7 +4,7 @@
<TargetFramework>net6.0</TargetFramework> <TargetFramework>net6.0</TargetFramework>
<GenerateDocumentationFile>true</GenerateDocumentationFile> <GenerateDocumentationFile>true</GenerateDocumentationFile>
<EnableNETAnalyzers>false</EnableNETAnalyzers> <EnableNETAnalyzers>false</EnableNETAnalyzers>
<Version>1.0.1</Version> <Version>1.0.2</Version>
</PropertyGroup> </PropertyGroup>
<ItemGroup> <ItemGroup>
@@ -1,5 +1,6 @@
using System; using System;
using System.Collections.Generic; using System.Collections.Generic;
using System.Linq;
using System.Threading.Tasks; using System.Threading.Tasks;
using OpenSearch.Client; using OpenSearch.Client;
using Rsdo.Concordancer.Infrastructure.Search.Extensions; using Rsdo.Concordancer.Infrastructure.Search.Extensions;
@@ -27,23 +28,27 @@ public abstract class BaseAddRecordsHandler<TEntity, TElasticEntity> : IAddRecor
// Get index name // Get index name
var indexName = indexProvider.IndexName; var indexName = indexProvider.IndexName;
// Create request // Batch entities by 1000
var request = new BulkRequest(indexName) foreach (var batch in entities.Chunk(1000))
{ {
Operations = new List<IBulkOperation>(), // Create request
Timeout = TimeSpan.FromMinutes(5), var request = new BulkRequest(indexName)
}; {
Operations = new List<IBulkOperation>(),
Timeout = TimeSpan.FromMinutes(5),
};
// Convert entities to elastic entities // Convert entities to elastic entities
foreach (var entity in entities) foreach (var entity in batch)
{ {
var elasticEntity = ConvertEntity(entity); var elasticEntity = ConvertEntity(entity);
request.Operations.Add(new BulkIndexOperation<TElasticEntity>(elasticEntity)); request.Operations.Add(new BulkIndexOperation<TElasticEntity>(elasticEntity));
}
// Get response
var response = await client.BulkAsync(request);
response.ThrowIfInvalid();
} }
// Get response
var response = await client.BulkAsync(request);
response.ThrowIfInvalid();
} }
protected abstract TElasticEntity ConvertEntity(TEntity entity); protected abstract TElasticEntity ConvertEntity(TEntity entity);
@@ -4,7 +4,7 @@
<TargetFramework>net6.0</TargetFramework> <TargetFramework>net6.0</TargetFramework>
<GenerateDocumentationFile>true</GenerateDocumentationFile> <GenerateDocumentationFile>true</GenerateDocumentationFile>
<EnableNETAnalyzers>false</EnableNETAnalyzers> <EnableNETAnalyzers>false</EnableNETAnalyzers>
<Version>1.0.1</Version> <Version>1.0.2</Version>
</PropertyGroup> </PropertyGroup>
<ItemGroup> <ItemGroup>
@@ -109,7 +109,7 @@ public class ImportTermListHandler : IRequestHandler<ImportTermList, ExecutionRe
Lemma = termObj["kandidat"].Value<string>(), Lemma = termObj["kandidat"].Value<string>(),
Msd = termObj["POSoznake"].Value<string>(), Msd = termObj["POSoznake"].Value<string>(),
Weight = termObj["ranking"].Value<decimal>(), Weight = termObj["ranking"].Value<decimal>(),
Frequency = termObj["pogostostpojavljanja"].Value<int>(), Frequency = termObj["pogostostpojavljanja"].Values<int>().First(),
TermListAutoId = termList.AutoId, TermListAutoId = termList.AutoId,
}.ApplyCreateValues(); }.ApplyCreateValues();
@@ -90,39 +90,35 @@ public class IndexTextHandler : IRequestHandler<IndexText, ExecutionResult>
private static List<Concordance> GetConcordances(Text text, Paragraph paragraph, List<Token> tokens) private static List<Concordance> GetConcordances(Text text, Paragraph paragraph, List<Token> tokens)
{ {
// Get window size (max 10)
var windowSize = Math.Min(tokens.Count - 1, 10);
// Create list of tokens with positions
var tokensIdx = tokens.Select((t, i) => new KeyValuePair<int, Token>(i, t)).ToList();
// Get tokens which will appear in current concordance
List<KeyValuePair<int, Token>> concordanceTokens;
var concordances = new List<Concordance>(); var concordances = new List<Concordance>();
while ((concordanceTokens = tokensIdx.Where(x => x.Key >= -windowSize && x.Key <= windowSize).ToList().OrderBy(x => x.Key).ToList()).Any()) for (var i = 0; i < tokens.Count; i++)
{ {
if (concordanceTokens.Last().Key < 0)
{
break;
}
var concordance = new Concordance() var concordance = new Concordance()
{ {
ParagraphId = paragraph.Id, ParagraphId = paragraph.Id,
TextId = text.Id, TextId = text.Id,
}; };
// Loop through tokens and set it in the position // set center token
foreach (var concordanceToken in concordanceTokens) concordance.SetToken(tokens[i], 0);
// set left and right context tokens
for (var c = 1; c <= 10; c++)
{ {
concordance.SetToken(concordanceToken.Value, concordanceToken.Key); // left context
if (i - c >= 0)
{
concordance.SetToken(tokens[i - c], -c);
}
// right context
if (i + c < tokens.Count)
{
concordance.SetToken(tokens[i + c], c);
}
} }
concordances.Add(concordance); concordances.Add(concordance);
// Decrease indexes of tokens (shift tokens to left, relative to window size)
tokensIdx = tokensIdx.Select(x => new KeyValuePair<int, Token>(x.Key - 1, x.Value)).ToList();
} }
return concordances; return concordances;
@@ -4,7 +4,7 @@
<TargetFramework>net6.0</TargetFramework> <TargetFramework>net6.0</TargetFramework>
<GenerateDocumentationFile>true</GenerateDocumentationFile> <GenerateDocumentationFile>true</GenerateDocumentationFile>
<EnableNETAnalyzers>false</EnableNETAnalyzers> <EnableNETAnalyzers>false</EnableNETAnalyzers>
<Version>1.0.1</Version> <Version>1.0.2</Version>
</PropertyGroup> </PropertyGroup>
<ItemGroup> <ItemGroup>