From 49608bfdd473eeae993086c98c572f735073936e Mon Sep 17 00:00:00 2001
From: Bryan Newbold <bnewbold@archive.org>
Date: Tue, 23 Jun 2020 17:24:25 -0700
Subject: ZDB homepage (FIZE) scrape importer

---
 chocula/directories/zdb_fize.py | 34 ++++++++++++++++++++++++++++++++++
 1 file changed, 34 insertions(+)
 create mode 100644 chocula/directories/zdb_fize.py

(limited to 'chocula')

diff --git a/chocula/directories/zdb_fize.py b/chocula/directories/zdb_fize.py
new file mode 100644
index 0000000..a40139a
--- /dev/null
+++ b/chocula/directories/zdb_fize.py
@@ -0,0 +1,34 @@
+import json
+from typing import Iterable, Optional
+
+from chocula.common import DirectoryLoader
+from chocula.database import DirectoryInfo, HomepageUrl
+
+
+class ZdbFizeLoader(DirectoryLoader):
+    """
+    URL metadata scraped from ZDB "FIZE" interface. Consists of just ISSN / URL
+    pair.
+
+    Only interested in the homepage.
+    """
+
+    source_slug = "zdb_fize"
+
+    def open_file(self) -> Iterable:
+        return open(self.config.zdb_fize.filepath, "r")
+
+    def parse_record(self, record) -> Optional[DirectoryInfo]:
+
+        if not record.strip():
+            return None
+        record = json.loads(record)
+
+        info = DirectoryInfo(directory_slug=self.source_slug, issnl=record["issn"])
+
+        homepage = HomepageUrl.from_url(record["homepage"])
+        if homepage:
+            info.homepage_urls.append(homepage)
+        else:
+            return None
+        return info
-- 
cgit v1.2.3