From dccbaa5c1b0ba556449de6024540ba05d67ef6a0 Mon Sep 17 00:00:00 2001
From: Martin Czygan <martin.czygan@gmail.com>
Date: Tue, 21 Sep 2021 15:55:52 +0200
Subject: matching: run an additional es query for fuzzy matching

---
 fuzzycat/matching.py   | 74 +++++++++++++++++++++++++++++++++++++++++++++++++-
 tests/test_matching.py | 22 +++++++++++++--
 2 files changed, 93 insertions(+), 3 deletions(-)

diff --git a/fuzzycat/matching.py b/fuzzycat/matching.py
index 33e130e..310dfc2 100644
--- a/fuzzycat/matching.py
+++ b/fuzzycat/matching.py
@@ -73,12 +73,83 @@ def match_release_fuzzy(
         if r:
             return [r]
 
+
+    if release.title is not None and release.contribs is not None:
+        names = " ".join([c.raw_name for c in release.contribs])
+        body = {
+            "track_total_hits": True,
+            "query": {
+                "bool": {
+                    "must": [
+                        {
+                            "match": {
+                                "title": {
+                                    "query": release.title,
+                                    "operator": "AND",
+                                    "fuzziness": "AUTO",
+                                },
+                            }
+                        },
+                        {
+                            "match": {
+                                "contrib_names": {
+                                    "query": names,
+                                    "operator": "AND",
+                                    "fuzziness": "AUTO",
+                                }
+                            }
+                        },
+                    ],
+                },
+            },
+            "size": size,
+        }
+        resp = es.search(body=body, index="fatcat_release")
+        if es_compat_hits_total(resp) > 0:
+            return response_to_entity_list(resp, entity_type=ReleaseEntity, size=size, api=api)
+
+        body = {
+            "track_total_hits": True,
+            "query": {
+                "bool": {
+                    "should": [
+                        {
+                            "match": {
+                                "title": {
+                                    "query": release.title,
+                                    "operator": "AND",
+                                    "fuzziness": "AUTO",
+                                },
+                            }
+                        },
+                        {
+                            "match": {
+                                "contrib_names": {
+                                    "query": names,
+                                    "operator": "AND",
+                                    "fuzziness": "AUTO",
+                                }
+                            }
+                        },
+                    ],
+                },
+            },
+            "size": size,
+        }
+        resp = es.search(body=body, index="fatcat_release")
+        if es_compat_hits_total(resp) > 0:
+            return response_to_entity_list(resp, entity_type=ReleaseEntity, size=size, api=api)
+
+    # Note: If the title is short, we will get lots of results here; do we need
+    # to check for title length or result set length length or result set
+    # length here?
     body = {
+        "track_total_hits": True,
         "query": {
             "match": {
                 "title": {
                     "query": release.title,
-                    "operator": "AND"
+                    "operator": "AND",
                 }
             }
         },
@@ -91,6 +162,7 @@ def match_release_fuzzy(
     # Get fuzzy.
     # https://www.elastic.co/guide/en/elasticsearch/reference/current/common-options.html#fuzziness
     body = {
+        "track_total_hits": True,
         "query": {
             "match": {
                 "title": {
diff --git a/tests/test_matching.py b/tests/test_matching.py
index 2122144..c2e26f3 100644
--- a/tests/test_matching.py
+++ b/tests/test_matching.py
@@ -71,6 +71,24 @@ def test_match_release_fuzzy(es_client, caplog):
             "title": "digital libraries",
             "ext_ids": {}
         }, 5),
+        ({
+            "title": "unlikelytitle",
+            "ext_ids": {}
+        }, 0),
+        ({
+            "title": "Imminent dystopia",
+            "ext_ids": {}
+        }, 2),
+        ({
+            "title": "",
+            "contribs": [{"raw_name": "Aristoteles"}],
+            "ext_ids": {}
+        }, 5),
+        ({
+            "title": "Letter",
+            "contribs": [{"raw_name": "Claudel"}],
+            "ext_ids": {}
+        }, 1),
         ({
             "title": "The Future of Digital Scholarship",
             "contribs": [{
@@ -83,6 +101,6 @@ def test_match_release_fuzzy(es_client, caplog):
         entity = entity_from_dict(doc, ReleaseEntity)
         result = match_release_fuzzy(entity, es=es_client)
         with caplog.at_level(logging.INFO):
-            logging.info("[{}] given {}, found {}, {}".format(i, entity.title, len(result),
+            logging.info("[{}] given title '{}', found {}, {}".format(i, entity.title, len(result),
                                                               [v.title for v in result]))
-        assert len(result) == count
+        assert len(result) == count, doc
-- 
cgit v1.2.3