Resolver db 2.0 changes

adsmp/solr_updater.py

@@ -446,62 +446,42 @@ def transform_json_record(db_record):
     timestamps = []
     for k, v in DB_COLUMN_DESTINATIONS:
         ts = db_record.get(k + "_updated", None)
-        if ts:
-            ts = time.mktime(ts.timetuple())
-        else:
-            ts = sys.maxsize  # default to use option without timestamp
+        ts = time.mktime(ts.timetuple()) if ts else sys.maxsize # default to use option without timestamp
         timestamps.append((k, v, ts))
     timestamps.sort(key=lambda x: x[2])
 
     # merge data based on timestamps
-    for field, target, _ in timestamps:
+    for field, target, ts in timestamps: # fields = {bib_data, nonbib_data, orcid_claims, metrics ..}
         if db_record.get(field, None):
-            if target:
+            if not target: # bib_data
+                out.update(db_record.get(field))
+            else:
                 if callable(target):
-                    x = target(
+                    enriched_data = target(
                         db_record.get(field), out
                     )  # in the interest of speed, don't create copy of out
-                    if x:
-                        out.update(x)
-                else:
-                    out[target] = db_record.get(field)
-            else:
-                if target is None:
-                    continue
-
-                out.update(db_record.get(field))
-
-        elif field.startswith("#"):
+                    if enriched_data:
+                        out.update(enriched_data)
+                else: # id 
+                    out[target] = db_record.get(field)      
+        elif field.startswith("#"): # timestamps
             if callable(target):
-                x = target(
+                enriched_data = target(
                     db_record, out
                 )  # in the interest of speed, don't create copy of out
-                if x:
-                    out.update(x)
-
-    # override temporal priority for links data
-    if (
-        db_record.get("bib_data", None)
-        and db_record.get("nonbib_data", None)
-        and db_record["bib_data"].get("links_data", None)
-        and db_record["nonbib_data"].get("links_data", None)
+                if enriched_data:
+                    out.update(enriched_data)
+
+    # If both bib and nonbib pipeline provided links data
+    # use nonbib data even if it is older
+    if all(
+        db_record.get(key, {}).get("links_data")
+        for key in ("bib_data", "nonbib_data")
     ):
-        # here if both bib and nonbib pipeline provided links data
-        # use nonbib data even if it is older
+        logger.debug('Both bib and nonbib data provided links data. Using nonbib data: {}'.format(db_record["nonbib_data"]["links_data"]))
         out["links_data"] = db_record["nonbib_data"]["links_data"]
-
-    # override temporal priority for bibgroup and bibgroup_facet, prefer nonbib
-    if db_record.get("nonbib_data", None) and db_record["nonbib_data"].get(
-        "bibgroup", None
-    ):
-        out["bibgroup"] = db_record["nonbib_data"]["bibgroup"]
-    if db_record.get("nonbib_data", None) and db_record["nonbib_data"].get(
-        "bibgroup_facet", None
-    ):
-        out["bibgroup_facet"] = db_record["nonbib_data"]["bibgroup_facet"]
-
-    # if only bib data is available, use it to compute property
-    if db_record.get("nonbib_data", None) is None and db_record.get("bib_data", None):
+    elif db_record.get("bib_data", {}).get("links_data"):
+        logger.debug('Only bib data provided links data. Using bib data: {}'.format(db_record["bib_data"]["links_data"]))
         links_data = db_record["bib_data"].get("links_data", None)
         if links_data:
             try:
@@ -528,6 +508,18 @@ def transform_json_record(db_record):
                         db_record["bibcode"], type(links_data), links_data
                     )
                 )
+
+    # override temporal priority for bibgroup and bibgroup_facet, prefer nonbib
+    if db_record.get("nonbib_data", None) and db_record["nonbib_data"].get(
+        "bibgroup", None
+    ):
+        out["bibgroup"] = db_record["nonbib_data"]["bibgroup"]
+    if db_record.get("nonbib_data", None) and db_record["nonbib_data"].get(
+        "bibgroup_facet", None
+    ):
+        out["bibgroup_facet"] = db_record["nonbib_data"]["bibgroup_facet"]
+
+
     boost_columns = ['doctype_boost', 'recency_boost', 'boost_factor', 'astronomy_final_boost', 'physics_final_boost', \
         'earth_science_final_boost', 'planetary_science_final_boost', 'heliophysics_final_boost', 'general_final_boost']
 
@@ -562,5 +554,5 @@ def transform_json_record(db_record):
                 if any([char.isalnum() for char in out_field]):
                     has.append(field)
         out["has"] = has
-
-    return out
+    logger.debug('Out: {}'.format(out))
+    return out

adsmp/tasks.py

@@ -279,21 +279,21 @@ def reindex_records(bibcodes, force=False, update_solr=True, update_metrics=True
 
     # check if we have complete record
     for bibcode in bibcodes:
-        r = app.get_record(bibcode, load_only=fields)
+        record = app.get_record(bibcode, load_only=fields)
 
-        if r is None:
+        if record is None:
             logger.error('The bibcode %s doesn\'t exist!', bibcode)
             continue
 
-        augments_updated = r.get('augments_updated', None)
-        bib_data_updated = r.get('bib_data_updated', None)
-        fulltext_updated = r.get('fulltext_updated', None)
-        metrics_updated = r.get('metrics_updated', None)
-        nonbib_data_updated = r.get('nonbib_data_updated', None)
-        orcid_claims_updated = r.get('orcid_claims_updated', None)
+        augments_updated = record.get('augments_updated', None)
+        bib_data_updated = record.get('bib_data_updated', None)
+        fulltext_updated = record.get('fulltext_updated', None)
+        metrics_updated = record.get('metrics_updated', None)
+        nonbib_data_updated = record.get('nonbib_data_updated', None)
+        orcid_claims_updated = record.get('orcid_claims_updated', None)
 
         year_zero = '1972'
-        processed = r.get('processed', adsputils.get_date(year_zero))
+        processed = record.get('processed', adsputils.get_date(year_zero))
         if processed is None:
             processed = adsputils.get_date(year_zero)
 
@@ -314,26 +314,25 @@ def reindex_records(bibcodes, force=False, update_solr=True, update_metrics=True
                               metrics_updated, augments_updated))
             # build the solr record
             if update_solr:
-                solr_payload = solr_updater.transform_json_record(r)
-
+                solr_payload = solr_updater.transform_json_record(record)
                 # ADS microservices assume the identifier field exists and contains the canonical bibcode:
                 if 'identifier' not in solr_payload:
                     solr_payload['identifier'] = []
                 if 'bibcode' in solr_payload and solr_payload['bibcode'] not in solr_payload['identifier']:
                     solr_payload['identifier'].append(solr_payload['bibcode'])
                 logger.debug('Built SOLR record for %s', solr_payload['bibcode'])
                 solr_checksum = app.checksum(solr_payload)
-                if ignore_checksums or r.get('solr_checksum', None) != solr_checksum:
+                if ignore_checksums or record.get('solr_checksum', None) != solr_checksum:
                     solr_records.append(solr_payload)
                     solr_records_checksum.append(solr_checksum)
                 else:
                     logger.debug('Checksum identical, skipping solr update for: %s', bibcode)
 
             # get data for metrics
             if update_metrics:
-                metrics_payload = r.get('metrics', None)
+                metrics_payload = record.get('metrics', None)
                 metrics_checksum = app.checksum(metrics_payload or '')
-                if (metrics_payload and ignore_checksums) or (metrics_payload and r.get('metrics_checksum', None) != metrics_checksum):
+                if (metrics_payload and ignore_checksums) or (metrics_payload and record.get('metrics_checksum', None) != metrics_checksum):
                     metrics_payload['bibcode'] = bibcode
                     logger.debug('Got metrics: %s', metrics_payload)
                     metrics_records.append(metrics_payload)
@@ -342,10 +341,10 @@ def reindex_records(bibcodes, force=False, update_solr=True, update_metrics=True
                     logger.debug('Checksum identical or no metrics data available, skipping metrics update for: %s', bibcode)
 
             if update_links and links_url:
-                datalinks_payload = app.generate_links_for_resolver(r)
+                datalinks_payload = app.generate_links_for_resolver(record)
                 if datalinks_payload:
                     datalinks_checksum = app.checksum(datalinks_payload)
-                    if ignore_checksums or r.get('datalinks_checksum', None) != datalinks_checksum:
+                    if ignore_checksums or record.get('datalinks_checksum', None) != datalinks_checksum:
                         links_data_records.append(datalinks_payload)
                         links_data_records_checksum.append(datalinks_checksum)
         else: