Fix parser sorting

webis-de · Nov 24, 2023 · 219684b · 219684b
1 parent 318b692
commit 219684b
Show file tree

Hide file tree

Showing 5 changed files with 5 additions and 5 deletions.
diff --git a/archive_query_log/parsers/url_offset.py b/archive_query_log/parsers/url_offset.py
@@ -119,7 +119,7 @@ def _url_offset_parsers(
     parsers: Iterable[UrlOffsetParser] = (
         UrlOffsetParser.search(using=config.es.client)
         .filter(Term(provider__id=provider_id))
-        .sort("-priority")
+        .query(RankFeature(field="priority", saturation={}))
         .scan()
     )
     parsers = safe_iter_scan(parsers)

diff --git a/archive_query_log/parsers/url_page.py b/archive_query_log/parsers/url_page.py
@@ -118,7 +118,7 @@ def _url_page_parsers(
     parsers: Iterable[UrlPageParser] = (
         UrlPageParser.search(using=config.es.client)
         .filter(Term(provider__id=provider_id))
-        .sort("-priority")
+        .query(RankFeature(field="priority", saturation={}))
         .scan()
     )
     parsers = safe_iter_scan(parsers)

diff --git a/archive_query_log/parsers/url_query.py b/archive_query_log/parsers/url_query.py
@@ -121,7 +121,7 @@ def _url_query_parsers(
     parsers: Iterable[UrlQueryParser] = (
         UrlQueryParser.search(using=config.es.client)
         .filter(Term(provider__id=provider_id))
-        .sort("-priority")
+        .query(RankFeature(field="priority", saturation={}))
         .scan()
     )
     parsers = safe_iter_scan(parsers)

diff --git a/archive_query_log/parsers/warc_query.py b/archive_query_log/parsers/warc_query.py
@@ -103,7 +103,7 @@ def _warc_query_parsers(
     parsers: Iterable[WarcQueryParser] = (
         WarcQueryParser.search(using=config.es.client)
         .filter(Term(provider__id=provider_id))
-        .sort("-priority")
+        .query(RankFeature(field="priority", saturation={}))
         .scan()
     )
     parsers = safe_iter_scan(parsers)

diff --git a/archive_query_log/parsers/warc_snippets.py b/archive_query_log/parsers/warc_snippets.py
@@ -153,7 +153,7 @@ def _warc_snippets_parsers(
     parsers: Iterable[WarcSnippetsParser] = (
         WarcSnippetsParser.search(using=config.es.client)
         .filter(Term(provider__id=provider_id))
-        .sort("-priority")
+        .query(RankFeature(field="priority", saturation={}))
         .scan()
     )
     parsers = safe_iter_scan(parsers)