From 60e2ca08beb9a9d5795f0a44df096fc6c0bc2583 Mon Sep 17 00:00:00 2001
From: Arun Isaac
Date: Wed, 26 Oct 2022 14:23:34 +0530
Subject: Add search.

* gn3/api/search.py: New file.
* gn3/app.py: Register the search blueprint.
---
 gn3/api/search.py | 62 +++++++++++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 62 insertions(+)
 create mode 100644 gn3/api/search.py

(limited to 'gn3/api/search.py')

diff --git a/gn3/api/search.py b/gn3/api/search.py
new file mode 100644
index 0000000..78a3245
--- /dev/null
+++ b/gn3/api/search.py
@@ -0,0 +1,62 @@
+"""Search using Xapian index."""
+
+import json
+import urllib.parse
+
+from flask import abort, Blueprint, jsonify, request
+import xapian
+
+from gn3.monads import MonadicDict
+from gn3.db_utils import xapian_database
+
+search = Blueprint("search", __name__)
+
+@search.route("/")
+def search_results():
+    """Search Xapian index and return a list of results."""
+    args = request.args
+    search_type = args.get("type", default="gene")
+    querystring = args.get("query", default="")
+    page = args.get("page", default=1, type=int)
+    if page < 1:
+        abort(404, description="Requested page does not exist")
+    results_per_page = args.get("per_page", default=100, type=int)
+    maximum_results_per_page = 10000
+    if results_per_page > maximum_results_per_page:
+        abort(400, description="Requested too many search results")
+
+    queryparser = xapian.QueryParser()
+    queryparser.set_stemmer(xapian.Stem("en"))
+    queryparser.set_stemming_strategy(queryparser.STEM_SOME)
+    queryparser.add_boolean_prefix("author", "A")
+    queryparser.add_boolean_prefix("species", "XS")
+    queryparser.add_boolean_prefix("group", "XG")
+    queryparser.add_boolean_prefix("tissue", "XI")
+    queryparser.add_boolean_prefix("dataset", "XDS")
+    queryparser.add_boolean_prefix("symbol", "XY")
+    queryparser.add_boolean_prefix("chr", "XC")
+    queryparser.add_boolean_prefix("peakchr", "XPC")
+    queryparser.add_prefix("description", "XD")
+    for i, prefix in enumerate(["mean:", "peak:", "mb:", "peakmb:", "additive:", "year:"]):
+        queryparser.add_rangeprocessor(xapian.NumberRangeProcessor(i, prefix))
+    query = queryparser.parse_query(querystring)
+    traits = []
+    # pylint: disable=invalid-name
+    with xapian_database() as db:
+        enquire = xapian.Enquire(db)
+        # Filter documents by type.
+        enquire.set_query(xapian.Query(xapian.Query.OP_FILTER,
+                                       query,
+                                       xapian.Query(f"XT{search_type}")))
+        for xapian_match in enquire.get_mset((page-1)*results_per_page, results_per_page):
+            trait = MonadicDict(json.loads(xapian_match.document.get_data()))
+            # Add PubMed link to phenotype search results.
+            if search_type == "phenotype":
+                trait["pubmed_link"] = trait["pubmed_id"].map(
+                    lambda pubmed_id: "http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?"
+                    + urllib.parse.urlencode({"cmd": "Retrieve",
+                                              "db": "PubMed",
+                                              "list_uids": pubmed_id,
+                                              "dopt": "Abstract"}))
+            traits.append(trait.data)
+    return jsonify(traits)
-- 
cgit v1.2.3