Add support for OpenReview papers

2 files changed, 55 insertions, 0 deletions

diff --git a/arxiv2remarkable.py b/arxiv2remarkable.py
index 6eb81ff..8ee377b 100755
--- a/arxiv2remarkable.py
+++ b/arxiv2remarkable.py
@@ -511,6 +511,51 @@ class ACMProvider(Provider):
         return dict(title=title, date=date, authors=authors)
 
 
+class OpenReviewProvider(Provider):
+    def __init__(self, *args, **kwargs):
+        super().__init__(*args, **kwargs)
+
+    def get_abs_pdf_urls(self, url):
+        """ Get the pdf and abstract url from a OpenReview url """
+        if re.match(
+                "https?://openreview.net/forum\?id=[A-Za-z0-9]+", url):
+            abs_url = url
+            pdf_url = url.replace('forum', 'pdf')
+        elif re.match(
+                "https?://openreview.net/pdf\?id=[A-Za-z0-9]+", url):
+            abs_url = url.replace('pdf', 'forum')
+            pdf_url = url
+        else:
+            exception("Couldn't figure out OpenReview urls.")
+        return abs_url, pdf_url
+
+    def validate(src):
+        """ Check if the url is a valid OpenReview url. """
+        m = re.match(
+            "https?://openreview.net/(forum|pdf)\?id=[A-Za-z0-9]+", src
+        )
+        return not m is None
+
+    def retrieve_pdf(self, src, filename):
+        """ Download the file and save as filename """
+        _, pdf_url = self.get_abs_pdf_urls(src)
+        self.download_url(pdf_url, filename)
+
+    def get_paper_info(self, src):
+        """ Extract the paper's authors, title, and publication year """
+        abs_url, _ = self.get_abs_pdf_urls(src)
+        self.log("Getting paper info from OpenReview")
+        page = self.get_page_with_retry(abs_url)
+        soup = bs4.BeautifulSoup(page, "html.parser")
+        authors = [
+                x["content"] for x in soup.find_all("meta", {"name": 
+                    "citation_author"})]
+        authors = [x.split(' ')[-1].strip() for x in authors]
+        title = soup.find_all("meta", {"name": "citation_title"})[0]["content"]
+        date = soup.find_all("meta", {"name": 
+            "citation_publication_date"})[0]["content"]
+        return dict(title=title, date=date, authors=authors)
+
 class LocalFileProvider(Provider):
     def __init__(self, *args, **kwargs):
         super().__init__(*args, **kwargs)
diff --git a/test.py b/test.py
index 2ec59d8..201cc9f 100644
--- a/test.py
+++ b/test.py
@@ -15,6 +15,7 @@ from arxiv2remarkable import (
     ArxivProvider,
     PMCProvider,
     ACMProvider,
+    OpenReviewProvider,
     LocalFileProvider,
     PdfUrlProvider,
 )
@@ -73,6 +74,15 @@ class Tests(unittest.TestCase):
         fsize = os.path.getsize(filename)
         self.assertTrue(1691444 < fsize <= 1693444)
 
+    def test_openreview(self):
+        prov = OpenReviewProvider(upload=False)
+        url = "https://openreview.net/forum?id=S1x4ghC9tQ"
+        exp_filename = "Gregor_et_al_-_Temporal_Difference_Variational_Auto-Encoder_2018.pdf"
+        filename = prov.run(url)
+        self.assertEqual(exp_filename, os.path.basename(filename))
+        fsize = os.path.getsize(filename)
+        self.assertTrue(1110316 < fsize <= 1112316)
+
     def test_local(self):
         local_filename = "test.pdf"
         with open(local_filename, "w") as fp: