diff --git a/.github/workflows/maven.yml b/.github/workflows/maven.yml index 3a4d7e32..15241faf 100644 --- a/.github/workflows/maven.yml +++ b/.github/workflows/maven.yml @@ -10,7 +10,7 @@ jobs: build: strategy: matrix: - jdk: [8, 11, 17, 19] + jdk: [8, 11, 17, 21] runs-on: ubuntu-latest timeout-minutes: 30 diff --git a/contrib/src/main/java/org/archive/modules/extractor/ExtractorYoutubeDL.java b/contrib/src/main/java/org/archive/modules/extractor/ExtractorYoutubeDL.java index 467c9ac9..aa9f949d 100644 --- a/contrib/src/main/java/org/archive/modules/extractor/ExtractorYoutubeDL.java +++ b/contrib/src/main/java/org/archive/modules/extractor/ExtractorYoutubeDL.java @@ -63,8 +63,7 @@ import com.google.gson.stream.JsonReader; import com.google.gson.stream.JsonToken; /** - * Extracts links to media by running youtube-dl in a subprocess. Runs only on - * html. + * Extracts links to media by running yt-dlp in a subprocess. Runs only on html. * *
* Also implements {@link WARCRecordBuilder} to write youtube-dl json to the
@@ -436,11 +435,11 @@ public class ExtractorYoutubeDL extends Extractor
*/
protected YoutubeDLResults runYoutubeDL(CrawlURI uri) {
/*
- * --format=best
- *
- * best: Select the best quality format represented by a single file
- * with video and audio.
- * https://github.com/ytdl-org/youtube-dl/blob/master/README.md#format-selection
+ * https://github.com/yt-dlp/yt-dlp#format-selection-examples
+ * updated for yt-dlp v.2023.07.06 and higher
+ * Download the best video with best vcodec no better than h264 and
+ * the best audio with best acodec no better than aac and
+ * with neither height nor width larger than 576.
*/
ProcessBuilder pb = new ProcessBuilder("yt-dlp", "--ignore-config",
"--simulate", "--dump-single-json", "-S vcodec:h264,res:576,acodec:aac",
diff --git a/modules/pom.xml b/modules/pom.xml
index 32613751..a6755776 100644
--- a/modules/pom.xml
+++ b/modules/pom.xml
@@ -30,12 +30,12 @@