Experiment with acquisition field in solr

Author: cdrini

conf/solr/conf/managed-schema.xml

@@ -194,6 +194,9 @@
     <field name="time_facet" type="string" stored="false" multiValued="true"/>
     <field name="time_key" type="string" stored="false" multiValued="true"/>
 
+    <!-- only on edition -->
+    <field name="acquisition" type="string" multiValued="true"/>
+
     <!-- Ratings -->
     <field name="ratings_average" type="pfloat"/>
     <field name="ratings_sortable" type="pfloat"/>

openlibrary/book_providers.py

@@ -594,6 +594,33 @@ def get_acquisitions(self, ed_or_solr: Edition | dict) -> list[Acquisition]:
         ]
 
 
+class BetterWorldBooksProvider(AbstractBookProvider):
+    short_name = 'betterworldbooks'
+    long_name = 'Better World Books'
+    identifier_key = 'betterworldbooks'
+
+    def is_own_ocaid(self, ocaid: str) -> bool:
+        return False
+
+    def get_identifiers(self, ed_or_solr: Edition | dict) -> list[str]:
+        # basically just check if it has an isbn?
+        return (ed_or_solr.get('isbn_10') or []) + (ed_or_solr.get('isbn_13') or [])
+
+    def get_acquisitions(
+        self,
+        ed_or_solr: Edition | dict,
+    ) -> list[Acquisition]:
+        return [
+            Acquisition(
+                access='buy',
+                format='web',
+                price=None,
+                url=f'https://www.betterworldbooks.com/product/detail/{self.get_best_identifier(ed_or_solr)}',
+                provider_name=self.short_name,
+            )
+        ]
+
+
 PROVIDER_ORDER: list[AbstractBookProvider] = [
     # These providers act essentially as their own publishers, so link to the first when
     # we're on an edition page
@@ -607,6 +634,8 @@ def get_acquisitions(self, ed_or_solr: Edition | dict) -> list[Acquisition]:
     WikisourceProvider(),
     # Then link to IA
     InternetArchiveProvider(),
+    # Then link to purchase options
+    BetterWorldBooksProvider(),
 ]
 
 

openlibrary/plugins/worksearch/schemes/works.py

@@ -325,8 +325,10 @@ def remove_work_prefix(field: str) -> str:
         # Removes the indicator prefix from queries with the 'work field' before appending them to parameters.
         final_work_query = deepcopy(work_q_tree)
         luqum_replace_field(final_work_query, remove_work_prefix)
+        EDITION_ONLY_FIELDS = {'acquisition'}
         try:
             luqum_remove_field(final_work_query, lambda f: f.startswith('edition.'))
+            luqum_remove_field(final_work_query, lambda f: f in EDITION_ONLY_FIELDS)
         except EmptyTreeError:
             # If the whole tree is removed, we should just search for everything
             final_work_query = luqum_parser('*:*')
@@ -396,6 +398,7 @@ def remove_work_prefix(field: str) -> str:
                 'ia_collection': 'ia_collection',
                 'ia_box_id': 'ia_box_id',
                 'public_scan_b': 'public_scan_b',
+                'acquisition': 'acquisition',
             }
 
             def convert_work_field_to_edition_field(

openlibrary/solr/solr_types.py

@@ -65,6 +65,7 @@ class SolrDocument(TypedDict):
     time: Optional[list[str]]
     time_facet: Optional[list[str]]
     time_key: Optional[list[str]]
+    acquisition: Optional[list[str]]
     ratings_average: Optional[float]
     ratings_sortable: Optional[float]
     ratings_count: Optional[int]

openlibrary/solr/updater/edition.py

@@ -293,6 +293,14 @@ def has_fulltext(self) -> bool:
     def public_scan_b(self) -> bool:
         return self.ebook_access == bp.EbookAccess.PUBLIC
 
+    def acquisition(self) -> list[str]:
+        """Get acquisition methods for the edition."""
+        return [
+            f'{acq.access},{acq.format},{acq.url}'
+            for provider in self._providers
+            for acq in provider.get_acquisitions(self._edition)
+        ]
+
     def build(self) -> SolrDocument:
         """
         Build the solr document for the given edition to store as a nested
@@ -339,6 +347,7 @@ def build(self) -> SolrDocument:
                 'ia_collection': self.ia_collection,
                 'ia_box_id': self.ia_box_id,
                 # Ebook access
+                'acquisition': self.acquisition(),
                 'ebook_access': self.ebook_access.to_solr_str(),
                 'ebook_provider': self.ebook_provider,
                 'has_fulltext': self.has_fulltext,