Add citeseerx provider

3 files changed, 90 insertions, 0 deletions

diff --git a/paper2remarkable/providers/__init__.py b/paper2remarkable/providers/__init__.py
index 45148fd..fabdcfe 100644
--- a/paper2remarkable/providers/__init__.py
+++ b/paper2remarkable/providers/__init__.py
@@ -2,6 +2,7 @@
 
 from .acm import ACM
 from .arxiv import Arxiv
+from .citeseerx import CiteSeerX
 from .local import LocalFile
 from .neurips import NeurIPS
 from .openreview import OpenReview
@@ -14,6 +15,7 @@ from .springer import Springer
 providers = [
     ACM,
     Arxiv,
+    CiteSeerX,
     NeurIPS,
     OpenReview,
     PMLR,
diff --git a/paper2remarkable/providers/citeseerx.py b/paper2remarkable/providers/citeseerx.py
new file mode 100644
index 0000000..fdc0e8a
--- /dev/null
+++ b/paper2remarkable/providers/citeseerx.py
@@ -0,0 +1,64 @@
+# -*- coding: utf-8 -*-
+
+"""Provider for CiteSeerX
+
+Author: G.J.J. van den Burg
+License: See LICENSE file
+Copyright: 2019, G.J.J. van den Burg
+
+"""
+
+import re
+
+from ._base import Provider
+from ._info import Informer
+from ..utils import exception
+
+
+class CiteSeerXInformer(Informer):
+
+    meta_author_key = "citation_authors"
+    meta_date_key = "citation_year"
+
+    def _format_authors(self, soup_authors):
+        op = lambda x: x[0].split(",")
+        return super()._format_authors(soup_authors, sep=" ", idx=-1, op=op)
+
+
+class CiteSeerX(Provider):
+
+    re_abs = "^https?:\/\/citeseerx.ist.psu.edu\/viewdoc\/summary\?doi=(?P<doi>[0-9\.]+)"
+    re_pdf = "^https?:\/\/citeseerx.ist.psu.edu\/viewdoc\/download(\;jsessionid=[A-Z0-9]+)?\?doi=(?P<doi>[0-9\.]+)&rep=rep1&type=pdf"
+
+    def __init__(self, *args, **kwargs):
+        super().__init__(*args, **kwargs)
+        self.informer = CiteSeerXInformer()
+
+    def _get_doi(self, url):
+        m = re.match(self.re_abs, url) or re.match(self.re_pdf, url)
+        if m:
+            return m["doi"]
+        exception("Couldn't retrieve CiteSeerX publication doi.")
+
+    def get_abs_pdf_urls(self, url):
+        """ Get the pdf and abstract url from a OpenReview url """
+        if re.match(self.re_abs, url):
+            abs_url = url
+            doi = self._get_doi(abs_url)
+            pdf_url = "http://citeseerx.ist.psu.edu/viewdoc/download?doi={doi}&rep=rep1&type=pdf".format(
+                doi=doi
+            )
+        elif re.match(self.re_pdf, url):
+            pdf_url = url
+            doi = self._get_doi(pdf_url)
+            abs_url = "http://citeseerx.ist.psu.edu/viewdoc/summary?doi={doi}".format(
+                doi=doi
+            )
+        else:
+            exception("Couldn't figure out CiteSeerX urls.")
+        return abs_url, pdf_url
+
+    def validate(src):
+        return re.match(CiteSeerX.re_abs, src) or re.match(
+            CiteSeerX.re_pdf, src
+        )
diff --git a/tests/test_providers.py b/tests/test_providers.py
index e0c98a2..75703ff 100644
--- a/tests/test_providers.py
+++ b/tests/test_providers.py
@@ -15,6 +15,7 @@ import unittest
 from paper2remarkable.providers import (
     ACM,
     Arxiv,
+    CiteSeerX,
     LocalFile,
     NeurIPS,
     OpenReview,
@@ -166,6 +167,29 @@ class TestProviders(unittest.TestCase):
         filename = prov.run(url)
         self.assertEqual(exp, os.path.basename(filename))
 
+    def test_citeseerx_1(self):
+        prov = CiteSeerX(upload=False, verbose=VERBOSE)
+        url = "http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.89.6548"
+        exp = "Aaronson_-_Is_P_Versus_NP_Formally_Independent_2003.pdf"
+        filename = prov.run(url)
+        self.assertEqual(exp, os.path.basename(filename))
+
+    def test_citeseerx_2(self):
+        prov = CiteSeerX(upload=False, verbose=VERBOSE)
+        url = "http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.101.6521&rep=rep1&type=pdf"
+        exp = "Everingham_et_al_-_The_2005_Pascal_Visual_Object_Classes_Challenge_2006.pdf"
+        filename = prov.run(url)
+        self.assertEqual(exp, os.path.basename(filename))
+
+    def test_citeseerx_3(self):
+        prov = CiteSeerX(upload=False, verbose=VERBOSE)
+        url = (
+            "http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.109.4049"
+        )
+        exp = "Brin_Page_-_The_Anatomy_of_a_Large-Scale_Hypertextual_Web_Search_Engine_1998.pdf"
+        filename = prov.run(url)
+        self.assertEqual(exp, os.path.basename(filename))
+
 
 if __name__ == "__main__":
     unittest.main()