@ckirby/sbd - npm Package Compare versions

Comparing version 2.1.2 to 2.2.0

lib/Match.js

@@ -0,2 +1,5 @@
+// @ts-check
+/** @type {string[]} */
 var abbreviations;
+/** @type {string[]} */
 var englishAbbreviations = [
@@ -64,2 +67,3 @@ "al",
 
+/** @type {(abbr: string[]) => void} */
 exports.setAbbreviations = function(abbr) {
@@ -73,2 +77,3 @@ if(abbr){
 
+/** @type {(str: string) => boolean} */
 exports.isCapitalized = function(str) {
@@ -78,3 +83,6 @@ return /^[A-Z][a-z].*/.test(str) || this.isNumber(str);
 
-// Start with opening quotes or capitalized letter
+/**
+ * Start with opening quotes or capitalized letter
+ * @type {(str: string) => boolean}
+ */
 exports.isSentenceStarter = function(str) {
@@ -84,20 +92,23 @@ return this.isCapitalized(str) || /``|"|'/.test(str.substring(0,2));
 
+/** @type {(str: string) => boolean} */
 exports.isCommonAbbreviation = function(str) {
-    str = str.replace(/\W+/g, '')
-    if (abbreviations.includes(str)) {
-      return true;
-    }
-    // if str is all-caps, try lower-case and sentence-case
-    if (!/[a-z]/.test(str)) {
-      str = str.toLowerCase()
-      if (abbreviations.includes(str)) {
+    // if the word has a / in it, consider whether the last part is an
+    // abbreviation to allow for things like "Mexico/U.S. border"
+    var last = str.replace(/[^\w/]+/g, '').split('/').pop() || '';
+    if (abbreviations.includes(last)) {
         return true;
-      }
-      str = str.replace(/(\w)(.+)/, function (_, first, rest) {
-        return first.toUpperCase() + rest
-      })
-      if (abbreviations.includes(str)) {
-        return true;
-      }
     }
+    // if last is all-caps, try lower-case and sentence-case
+    if (!/[a-z]/.test(last)) {
+        last = last.toLowerCase()
+        if (abbreviations.includes(last)) {
+            return true;
+        }
+        last = last.replace(/(\w)(.+)/, function (_, first, rest) {
+            return first.toUpperCase() + rest
+        })
+        if (abbreviations.includes(last)) {
+            return true;
+        }
+    }
     return false;
@@ -104,0 +115,0 @@ }

lib/tokenizer.js

@@ -6,3 +6,6 @@ /*jshint node:true, laxcomma:true */
 
-// Split the entry into sentences.
+
+/**
+ * Split the entry into sentences.
+ */
 exports.sentences = function(text, user_options) {
@@ -9,0 +12,0 @@ if (!text || typeof text !== "string" || !text.length) {

package.json

 {
   "name": "@ckirby/sbd",
-  "version": "2.1.2",
+  "version": "2.2.0",
   "description": "Split text into sentences with Sentence Boundary Detection (SBD).",
   "main": "lib/tokenizer.js",
+  "types": "lib/tokenizer.d.ts",
   "files": [
@@ -15,3 +16,3 @@ "lib"
     "type": "git",
-    "url": "https://github.com/chadkirby/sbd.git"
+    "url": "git+https://github.com/chadkirby/sbd.git"
   },
@@ -18,0 +19,0 @@ "keywords": [

New alerts

License Policy Violation

License

This package is not allowed per your license policy. Review the package's license to ensure compliance.

Found 1 instance in 1 package

Fixed alerts

License Policy Violation

License

This package is not allowed per your license policy. Review the package's license to ensure compliance.

Found 1 instance in 1 package

Improved metrics

Total package byte prevSize

14942

increased by6.52%

Number of package files

6

increased by20%

Lines of code

364

increased by8.66%

No dependency changes