From c862f2ffb61b693752e6a396378098edce9c3e9d Mon Sep 17 00:00:00 2001
From: Erik Schnetter <schnetter@gmail.com>
Date: Tue, 29 Sep 2020 12:19:28 -0400
Subject: [PATCH] Implement Semantic Scholar lookups

---
 README.md                                     |  8 +--
 bin/build.sh                                  |  7 +-
 .../content/scripts/zoterocitationcounts.js   | 67 +++++++++++++++++++
 install.rdf                                   |  2 +-
 update.rdf                                    |  4 +-
 5 files changed, 75 insertions(+), 13 deletions(-)

diff --git a/README.md b/README.md
index f9a1083..ba5bfb1 100644
--- a/README.md
+++ b/README.md
@@ -2,9 +2,9 @@
 
 This is an add-on for Zotero, a research source management tool. The
 add-on can auto-fetch citation counts for journal articles using
-various APIs, including Crossref, Inspire HEP (and in the future
-probably NASA/ADS and Semantic Scholar). (Google Scholar is not
-supported because automated access is against its terms of service.)
+various APIs, including Crossref, Inspire HEP, and and Semantic
+Scholar (and in the future possibly NASA/ADS). Google Scholar is not
+supported because automated access is against its terms of service.
 
 Please report any bugs, questions, or feature requests in the Github
 repository.
@@ -26,7 +26,7 @@ Code for this extension is based on the [Zotero DOI
 - Run Zotero (version 5.x)
 - Go to `Tools -> Add-ons`
 - `Install Add-on From File`
-- Choose the file `zotero-citationcounts-0.1.4.xpi`
+- Choose the file `zotero-citationcounts-0.1.5.xpi`
 - Restart Zotero
 
 ## License
diff --git a/bin/build.sh b/bin/build.sh
index dcc232f..4011f17 100755
--- a/bin/build.sh
+++ b/bin/build.sh
@@ -1,11 +1,6 @@
 #!/bin/sh
 
-version=$1
-
-if [ -z "$version" ]; then
-    echo "Synopsis: $0 <version>"
-    exit 1
-fi
+version='0.1.5'
 
 rm -f zotero-citationcounts-${version}.xpi
 zip -r zotero-citationcounts-${version}.xpi chrome/* defaults/* chrome.manifest install.rdf
diff --git a/chrome/content/scripts/zoterocitationcounts.js b/chrome/content/scripts/zoterocitationcounts.js
index 2ac9053..7c6eecf 100644
--- a/chrome/content/scripts/zoterocitationcounts.js
+++ b/chrome/content/scripts/zoterocitationcounts.js
@@ -148,6 +148,57 @@ async function getInspireCount(item, idtype) {
     return count;
 }
 
+async function getSemanticScholarCount(item, idtype) {
+    let doi = null;
+    if (idtype == 'doi') {
+        doi = item.getField('DOI');
+    } else if (idtype == 'arxiv') {
+        const arxiv = item.getField('url'); // check URL for arXiv id
+        const patt = /(?:arxiv.org[/]abs[/]|arXiv:)([a-z.-]+[/]\d+|\d+[.]\d+)/i;
+        const m = patt.exec(arxiv);
+        if (!m) {
+            // No arxiv id found
+            return -1;
+        }
+        doi = m[1];
+    } else {
+        // Internal error
+        return -1;
+    }
+    if (!doi) {
+        // There is no DOI / arXiv id; skip item
+        return -1;
+    }
+    const edoi = encodeURIComponent(doi);
+
+    const url =
+          "https://api.semanticscholar.org/v1/paper/" +
+          (idtype == 'doi' ? '' : 'arXiv:') + edoi
+    const response = await fetch(url)
+          .then(response => response.json())
+          .catch(err => null);
+
+    if (response === null) {
+        // Something went wrong
+        return -1;
+    }
+
+    let count = null;
+    try {
+        // Semantic Scholar returns the actual citations
+        count = response['citations'].length;
+        // Semantic Scholar imposes a rate limit of 100 requests per 5
+        // minutes. We should keep track of this globally so that we
+        // don't need to rate limit if there are just a few requests.
+        await await new Promise(r => setTimeout(r, 3000));
+    } catch (err) {
+        // There are no citations
+        return -1;
+    }
+
+    return count;
+}
+
 // Preference managers
 
 function getPref(pref) {
@@ -389,6 +440,22 @@ Zotero.CitationCounts.updateItem = async function(item, operation) {
         }
         Zotero.CitationCounts.updateNextItem(operation);
 
+    } else if (operation == "semanticscholar") {
+
+        const count_doi = await getSemanticScholarCount(item, 'doi');
+        const count_arxiv = await getSemanticScholarCount(item, 'arxiv');
+        if (count_doi >= 0 || count_arxiv >= 0) {
+            if (count_doi >= 0) {
+                setCitationCount(item, 'Semantic Scholar/DOI', count_doi);
+            }
+            if (count_arxiv >= 0) {
+                setCitationCount(item, 'Semantic Scholar/arXiv', count_arxiv);
+            }
+            item.saveTx();
+            Zotero.CitationCounts.counter++;
+        }
+        Zotero.CitationCounts.updateNextItem(operation);
+
     } else {
         Zotero.CitationCounts.updateNextItem(operation);
     }
diff --git a/install.rdf b/install.rdf
index c9bdc14..0224e6b 100644
--- a/install.rdf
+++ b/install.rdf
@@ -7,7 +7,7 @@
         RDF:about="urn:mozilla:install-manifest"
         em:id="schnetter@gmail.com"
         em:name="Zotero Citation Counts Manager"
-        em:version="0.1.4"
+        em:version="0.1.5"
         em:type="2"
         em:creator="Erik Schnetter"
         em:description="Automatically fetch and update citation counts"
diff --git a/update.rdf b/update.rdf
index 09424af..90f6dca 100644
--- a/update.rdf
+++ b/update.rdf
@@ -5,13 +5,13 @@
       <rdf:Seq>
         <rdf:li>
           <rdf:Description>
-            <em:version>0.1.4</em:version>
+            <em:version>0.1.5</em:version>
             <em:targetApplication>
               <rdf:Description>
                 <em:id>zotero@chnm.gmu.edu</em:id>
                 <em:minVersion>5.0</em:minVersion>
                 <em:maxVersion>5.*</em:maxVersion>
-                <em:updateLink>https://github.com/eschnett/zotero-citationcounts/releases/download/v0.1.4/zotero-citationcounts-0.1.4.xpi</em:updateLink>
+                <em:updateLink>https://github.com/eschnett/zotero-citationcounts/releases/download/v0.1.5/zotero-citationcounts-0.1.5.xpi</em:updateLink>
               </rdf:Description>
             </em:targetApplication>
           </rdf:Description>