From be28fd5d4a600bc29507418fddfdc95802abc98f Mon Sep 17 00:00:00 2001
From: Bryan Newbold <bnewbold@robocracy.org>
Date: Fri, 21 May 2021 17:46:37 -0700
Subject: ingest: add per-container ingest type overrides

---
 python/fatcat_tools/transforms/ingest.py | 18 +++++++++++++++++-
 python/tests/transform_ingest.py         |  6 ++++++
 2 files changed, 23 insertions(+), 1 deletion(-)

(limited to 'python')

diff --git a/python/fatcat_tools/transforms/ingest.py b/python/fatcat_tools/transforms/ingest.py
index 59831017..9aaeaa84 100644
--- a/python/fatcat_tools/transforms/ingest.py
+++ b/python/fatcat_tools/transforms/ingest.py
@@ -1,4 +1,19 @@
 
+INGEST_TYPE_CONTAINER_MAP = {
+    # Optica
+    "twtpsm6ytje3nhuqfu3pa7ca7u": "html",
+    # Optics Express
+    "cg4vcsfty5dfvgmat5wm62wgie": "html",
+    # First Monday
+    "svz5ul6qozdjhjhk7d627avuja": "html",
+    # D-Lib Magazine
+    "ugbiirfvufgcjkx33r3cmemcuu": "html",
+    # Distill (distill.pub)
+    "lx7svdzmc5dl3ay4zncjjrql7i": "html",
+    # NLM technical bulletin
+    "lovwr7ladjagzkhmoaszg7efqu": "html",
+}
+
 def release_ingest_request(release, ingest_request_source='fatcat', ingest_type=None):
     """
     Takes a full release entity object and returns an ingest request (as dict),
@@ -15,7 +30,8 @@ def release_ingest_request(release, ingest_request_source='fatcat', ingest_type=
     if release.state != 'active':
         return None
 
-    # TODO: infer ingest type based on release_type or container metadata?
+    if (not ingest_type) and release.container_id:
+        ingest_type = INGEST_TYPE_CONTAINER_MAP.get(release.container_id)
     if not ingest_type:
         ingest_type = 'pdf'
 
diff --git a/python/tests/transform_ingest.py b/python/tests/transform_ingest.py
index c7044bc0..a61aa4a7 100644
--- a/python/tests/transform_ingest.py
+++ b/python/tests/transform_ingest.py
@@ -54,3 +54,9 @@ def test_rich_ingest_release():
     assert ir['base_url'] == 'https://doi.org/10.123/456'
     assert ir['ext_ids']['doi'] == '10.123/456'
     assert ir['ext_ids'].get('pmcid') is None
+    assert ir['ingest_type'] == 'pdf'
+
+    # check ingest type ("d-lib")
+    r.container_id = "ugbiirfvufgcjkx33r3cmemcuu"
+    ir = release_ingest_request(r)
+    assert ir['ingest_type'] == 'html'
-- 
cgit v1.2.3