CatgirlIntelligenceAgency/code/process-models/crawling-model/build.gradle
Viktor Lofgren 3113b5a551 (warc) Filter WarcResponses based on X-Robots-Tags
There really is no fantastic place to put this logic, but we need to remove entries with an X-Robots-Tags header where that header indicates it doesn't want to be crawled by Marginalia.
2023-12-16 15:58:27 +01:00

47 lines
1.3 KiB
Groovy

plugins {
id 'java'
id 'jvm-test-suite'
}
java {
toolchain {
languageVersion.set(JavaLanguageVersion.of(21))
}
}
dependencies {
implementation project(':code:common:model')
implementation project(':code:common:db')
implementation project(':code:common:config')
implementation project(':code:common:process')
implementation project(':code:libraries:big-string')
implementation project(':code:api:index-api')
implementation project(':code:common:service-discovery')
implementation project(':code:common:service-client')
implementation project(':code:features-crawl:content-type')
implementation project(':code:libraries:language-processing')
implementation project(':third-party:parquet-floor')
implementation project(':third-party:commons-codec')
implementation libs.bundles.slf4j
implementation libs.notnull
implementation libs.bundles.parquet
implementation libs.jwarc
implementation libs.gson
implementation libs.commons.io
implementation libs.commons.lang3
implementation libs.okhttp3
implementation libs.jsoup
implementation libs.snakeyaml
implementation libs.zstd
testImplementation libs.bundles.slf4j.test
testImplementation libs.bundles.junit
testImplementation libs.mockito
}