initial implementation of RSS search feeds

UI integration needs work

3 files changed, 82 insertions, 0 deletions

diff --git a/fatcat_scholar/templates/base.html b/fatcat_scholar/templates/base.html
index 093d4e4..ce74f26 100644
--- a/fatcat_scholar/templates/base.html
+++ b/fatcat_scholar/templates/base.html
@@ -136,6 +136,21 @@
         <div style="display: flex; width: 100%; justify-content: space-between; padding-top: 0.5em;">
           <div></div>
           <div>
+            {% if query %}
+              <form class="" id="rss_feed_form" name="rss_feed_form" action="{{ lang_prefix }}/feed/rss" method="get">
+                <input form="rss_feed_form" type="hidden" name="q" value="{{ query.q }}">
+                {% if query.filter_type %}
+                  <input form="rss_feed_form" type="hidden" name="filter_type" value="{{ query.filter_type }}">
+                {% endif %}
+                {% if query.filter_availability %}
+                  <input form="rss_feed_form" type="hidden" name="filter_availability" value="{{ query.filter_availability }}">
+                {% endif %}
+                <button class="text-button" form="rss_feed_form" type="submit">
+                  {% trans %}RSS Feed{% endtrans %}
+                </button>
+              </form>
+              &nbsp;
+            {% endif %}
             <a rel="help" href="{{ lang_prefix }}/help" style="order: -1;">{{ _("User Guide") }}</a>
           </div>
         </div>
diff --git a/fatcat_scholar/web.py b/fatcat_scholar/web.py
index 83f80ad..ef7a5e1 100644
--- a/fatcat_scholar/web.py
+++ b/fatcat_scholar/web.py
@@ -4,12 +4,14 @@ This contains the FastAPI web application and RESTful API.
 So far there are few endpoints, so we just put them all here!
 """
 
+import datetime
 import logging
 import urllib.parse
 from typing import Any, Dict, List, Optional
 
 import babel.numbers
 import babel.support
+import fastapi_rss
 import fatcat_openapi_client
 import sentry_sdk
 from fastapi import APIRouter, Depends, FastAPI, HTTPException, Query, Request, Response
@@ -176,6 +178,63 @@ def search(query: FulltextQuery = Depends(FulltextQuery)) -> FulltextHits:
     return hits
 
 
+@api.get("/feed/rss", operation_id="get_feed_rss", include_in_schema=False)
+def feed_rss(
+    query: FulltextQuery = Depends(FulltextQuery),
+    lang: LangPrefix = Depends(LangPrefix),
+) -> fastapi_rss.RSSResponse:
+
+    # override some query params for feeds
+    if query.q:
+        query.q += " doc_type:work"
+    query.offset = None
+    query.filter_time = "past_year"
+    query.sort_order = "time_desc"
+
+    hits: FulltextHits = process_query(query)
+
+    rss_items = []
+    for hit in hits.results:
+        scholar_doc = hit["_obj"]
+        abstract: Optional[str] = None
+        if scholar_doc.abstracts:
+            abstract = scholar_doc.abstracts[0].body
+        authors = ", ".join(scholar_doc.biblio.contrib_names) or None
+        pub_date = None
+        if scholar_doc.biblio.release_date:
+            # convert datetime.date to datetime.datetime
+            pub_date = datetime.datetime(
+                *scholar_doc.biblio.release_date.timetuple()[:6]
+            )
+        rss_items.append(
+            fastapi_rss.Item(
+                title=scholar_doc.biblio.title,
+                link=f"https://scholar.archive.org/work/{scholar_doc.work_ident}",
+                description=abstract,
+                author=authors,
+                pub_date=pub_date,
+                guid=fastapi_rss.GUID(content=scholar_doc.key),
+            )
+        )
+
+    last_build_date = None
+    if rss_items:
+        last_build_date = rss_items[0].pub_date
+    feed = fastapi_rss.RSSFeed(
+        title=f"IA Scholar: {query.q}",
+        link="https://scholar.archive.org/",
+        description="Internet Archive Scholar query results feed",
+        language="en",
+        last_build_date=last_build_date,
+        docs="https://scholar.archive.org/help",
+        generator="fatcat-scholar",
+        webmaster="info@archive.org",
+        ttl=60 * 24,  # 24 hours, in minutes
+        item=rss_items,
+    )
+    return fastapi_rss.RSSResponse(feed)
+
+
 @api.get("/work/{work_ident}", operation_id="get_work")
 def get_work(work_ident: str = Query(..., min_length=20, max_length=20)) -> dict:
     doc = get_es_scholar_doc(f"work_{work_ident}")
diff --git a/proposals/2022-03-25_search_rss_feed.md b/proposals/2022-03-25_search_rss_feed.md
new file mode 100644
index 0000000..88eb85b
--- /dev/null
+++ b/proposals/2022-03-25_search_rss_feed.md
@@ -0,0 +1,8 @@
+
+status: brainstorm
+
+Idea is to have a way to generate a "feed" for any search query, as RSS, which
+can be used to keep up with recent publications.
+
+This idea was originally discussed in 2021 here: <https://github.com/internetarchive/fatcat-scholar/issues/37>
+