Add NBER provider

3 files changed, 64 insertions, 1 deletions

diff --git a/paper2remarkable/providers/__init__.py b/paper2remarkable/providers/__init__.py
index f87a044..c868bc4 100644
--- a/paper2remarkable/providers/__init__.py
+++ b/paper2remarkable/providers/__init__.py
@@ -5,6 +5,7 @@ from .arxiv import Arxiv
 from .citeseerx import CiteSeerX
 from .html import HTML
 from .local import LocalFile
+from .nber import NBER
 from .neurips import NeurIPS
 from .openreview import OpenReview
 from .pdf_url import PdfUrl
@@ -12,11 +13,12 @@ from .pmlr import PMLR
 from .pubmed import PubMed
 from .springer import Springer
 
-# NOTE: Order matters here, PdfUrl should be last
+# NOTE: Order matters here, PdfUrl and HTML should be last
 providers = [
     ACM,
     Arxiv,
     CiteSeerX,
+    NBER,
     NeurIPS,
     OpenReview,
     PMLR,
diff --git a/paper2remarkable/providers/nber.py b/paper2remarkable/providers/nber.py
new file mode 100644
index 0000000..76bc85f
--- /dev/null
+++ b/paper2remarkable/providers/nber.py
@@ -0,0 +1,46 @@
+# -*- coding: utf-8 -*-
+
+"""Provider for NBER
+
+(US) National Bureau of Economic Research
+
+Author: G.J.J. van den Burg
+License: See LICENSE file
+Copyright: 2020, G.J.J. van den Burg
+
+"""
+
+import re
+
+from ._base import Provider
+from ._info import Informer
+from ..exceptions import URLResolutionError
+
+
+class NBERInformer(Informer):
+    def _format_year(self, soup_date):
+        return soup_date.split("-")[0]
+
+
+class NBER(Provider):
+
+    re_abs = "https?://www\.nber\.org/papers/(?P<ref>[a-z0-9]+)$"
+    re_pdf = "https?://www\.nber\.org/papers/(?P<ref>[a-z0-9]+)\.pdf$"
+
+    def __init__(self, *args, **kwargs):
+        super().__init__(*args, **kwargs)
+        self.informer = NBERInformer()
+
+    def get_abs_pdf_urls(self, url):
+        if re.match(self.re_abs, url):
+            abs_url = url
+            pdf_url = url + ".pdf"
+        elif re.match(self.re_pdf, url):
+            pdf_url = url
+            abs_url = url[: -len(".pdf")]
+        else:
+            raise URLResolutionError("NBER", url)
+        return abs_url, pdf_url
+
+    def validate(src):
+        return re.match(NBER.re_abs, src) or re.match(NBER.re_pdf, src)
diff --git a/tests/test_providers.py b/tests/test_providers.py
index d0e3d40..38f88b7 100644
--- a/tests/test_providers.py
+++ b/tests/test_providers.py
@@ -18,6 +18,7 @@ from paper2remarkable.providers import (
     CiteSeerX,
     HTML,
     LocalFile,
+    NBER,
     NeurIPS,
     OpenReview,
     PMLR,
@@ -179,6 +180,20 @@ class TestProviders(unittest.TestCase):
         filename = prov.run(url)
         self.assertEqual(exp, os.path.basename(filename))
 
+    def test_nber_1(self):
+        prov = NBER(upload=False, verbose=VERBOSE)
+        url = "https://www.nber.org/papers/w26752"
+        exp = "Bhattacharya_Packalen_-_Stagnation_and_Scientific_Incentives_2020.pdf"
+        filename = prov.run(url)
+        self.assertEqual(exp, os.path.basename(filename))
+
+    def test_nber_2(self):
+        prov = NBER(upload=False, verbose=VERBOSE)
+        url = "https://www.nber.org/papers/w19152.pdf"
+        exp = "Herbst_Schorfheide_-_Sequential_Monte_Carlo_Sampling_for_DSGE_Models_2013.pdf"
+        filename = prov.run(url)
+        self.assertEqual(exp, os.path.basename(filename))
+
     def test_neurips_1(self):
         prov = NeurIPS(upload=False, verbose=VERBOSE)
         url = "https://papers.nips.cc/paper/325-leaning-by-combining-memorization-and-gradient-descent.pdf"