text-annotator - npm Package Compare versions

Comparing version 0.7.3 to 0.7.4

build/text-annotator.js

@@ -13,3 +13,5 @@ "use strict";
 // used to distinguish between browser and Node.js environments
-const isBrowser = typeof window !== 'undefined' && typeof window.document !== 'undefined';
+// is it possible to relax so as to allow jsdom
+const isBrowser = typeof window !== 'undefined' && typeof window.document !== 'undefined'; // div inside span is a bad idea
+
 const blockElements = ['address', 'article', 'aside', 'blockquote', 'canvas', 'dd', 'div', 'dl', 'dt', 'fieldset', 'figcaption', 'figure', 'footer', 'form', 'h1', 'h2', 'h3', 'h4', 'h5', 'h6', 'header', 'hgroup', 'hr', 'li', 'main', 'nav', 'noscript', 'ol', 'output', 'p', 'pre', 'section', 'table', 'tfoot', 'ul', 'video'];
@@ -31,3 +33,3 @@
 
-    this.sentences = []; // one highlight can have more than one location because of the potential issue in tag insertion***
+    this.sentences = []; // future work: one highlight can have more than one location because of the potential issue in tag insertion
 
@@ -64,4 +66,5 @@ this.highlights = [];
       return highlightIndex;
-    }
+    } // experimental feature
 
+
     if (fuzzySearchOptions) {
@@ -73,3 +76,4 @@ highlightIndex = this.fuzzySearch(prefix, str, postfix, fuzzySearchOptions);
       }
-    } // eager search only works in (particular) browsers
+    } // experimental feature
+    // eager search only works in (particular) browsers
 
@@ -86,3 +90,4 @@
     return highlightIndex;
-  } // only support direct search for now
+  } // experimental feature
+  // only support direct search for now
 
@@ -131,4 +136,5 @@
     }
-  }
+  } // experimental feature
 
+
   highlightAll(highlightIndexes, options = {}) {
@@ -162,2 +168,3 @@ // either containerId or content is required
     if (highlightIndex !== -1) {
+      // content is undefined if containerId and returnContent falsy
       return {
@@ -188,3 +195,4 @@ highlightIndex,
 
-      let newContent = content;
+      let newContent = content; // need to change when one annotation => more than one highlight
+
       const loc = this.adjustLoc(highlightIdPattern, highlightIndex, highlightClass);
@@ -494,4 +502,4 @@ const openTagLength = TextAnnotator.getOpenTagLength(highlightIdPattern, highlightIndex, highlightClass);
     return highlightIndex;
-  } // further improvement when one annotation binds with more than one highlight***
-  // block elements are only used to check in the = condition for now
+  } // future work: further improvement when one annotation binds with more than one highlight
+  // includeRequiredTag used in = condition only
 
@@ -501,6 +509,7 @@
     const isCloseTag = tag.startsWith('</');
-    const tagType = isCloseTag ? tag.split('</')[1].split('>')[0] : tag.split(' ')[0].split('<')[1].split('>')[0];
+    const tagName = isCloseTag ? tag.split('</')[1].split('>')[0] : tag.split(' ')[0].split('<')[1].split('>')[0];
     let included = false;
     let requiredTagNumber = 1;
-    let requiredTagCount = 0; // outer
+    let requiredTagCount = 0; // if both the start tag and the end tag are at the borders, place the tags outside the borders
+    // if the close tag is at the border, check backwards until the start of the highlight
 
@@ -516,5 +525,5 @@ if (isCloseTag) {
 
-          if (tag2.startsWith('</' + tagType)) {
+          if (tag2.startsWith('</' + tagName)) {
             requiredTagNumber++;
-          } else if (tag2.startsWith('<' + tagType)) {
+          } else if (tag2.startsWith('<' + tagName)) {
             requiredTagCount++;
@@ -529,24 +538,25 @@ }
       }
-    } else {
-      for (let i2 = i + 1; i2 < this.tagLocations.length; i2++) {
-        const tagLoc2 = this.tagLocations[i2];
+    } // if the start tag is at the border, check forwards until the end of the highlight
+    else {
+        for (let i2 = i + 1; i2 < this.tagLocations.length; i2++) {
+          const tagLoc2 = this.tagLocations[i2];
 
-        if (highlightLoc[1] < tagLoc2[0]) {
-          break;
-        } else {
-          const tag2 = this.originalContent.substring(tagLoc2[0] + tagLoc2[2], tagLoc2[0] + tagLoc2[2] + tagLoc2[1]);
+          if (highlightLoc[1] < tagLoc2[0]) {
+            break;
+          } else {
+            const tag2 = this.originalContent.substring(tagLoc2[0] + tagLoc2[2], tagLoc2[0] + tagLoc2[2] + tagLoc2[1]);
 
-          if (tag2.startsWith('<' + tagType)) {
-            requiredTagNumber++;
-          } else if (tag2.startsWith('</' + tagType)) {
-            requiredTagCount++;
-          }
+            if (tag2.startsWith('<' + tagName)) {
+              requiredTagNumber++;
+            } else if (tag2.startsWith('</' + tagName)) {
+              requiredTagCount++;
+            }
 
-          if (requiredTagNumber === requiredTagCount) {
-            included = true;
-            break;
+            if (requiredTagNumber === requiredTagCount) {
+              included = true;
+              break;
+            }
           }
         }
       }
-    }
 
@@ -569,3 +579,3 @@ return included;
       else if (highlightLoc[1] === tagLoc[0]) {
-          const tag = this.originalContent.substring(tagLoc[0] + tagLoc[2], tagLoc[0] + tagLoc[2] + tagLoc[1]);
+          const tag = this.originalContent.substring(tagLoc[0] + tagLoc[2], tagLoc[0] + tagLoc[2] + tagLoc[1]); // if end tag, not block element and include the required close tag, add right to the tag
 
@@ -580,3 +590,3 @@ if (!tag.endsWith('/>') && tag.startsWith('</') && !blockElements.includes(tag.split('</')[1].split('>')[0]) && this.includeRequiredTag(i, highlightLoc, tag)) {
             if (highlightLoc[0] === tagLoc[0]) {
-              const tag = this.originalContent.substring(tagLoc[0] + tagLoc[2], tagLoc[0] + tagLoc[2] + tagLoc[1]);
+              const tag = this.originalContent.substring(tagLoc[0] + tagLoc[2], tagLoc[0] + tagLoc[2] + tagLoc[1]); // if self close tag or end tag or block element or not include the required close tag, add right to the tag
 
@@ -592,6 +602,9 @@ if (tag.startsWith('</') || tag.endsWith('/>') || blockElements.includes(tag.split(' ')[0].split('<')[1].split('>')[0]) || !this.includeRequiredTag(i, highlightLoc, tag)) {
     } // step 2: check locations of other highlights
+    // all span (no blocks)
+    // stored in a different array than tags
+    // can intersect
 
 
     for (let i = 0; i < this.highlights.length; i++) {
-      const highlight = this.highlights[i];
+      const highlight = this.highlights[i]; // only check the highlighted
 
@@ -606,13 +619,15 @@ if (highlight.highlighted) {
           locInc[1] += openTagLength + closeTagLength;
-        } else if (highlightLoc[0] < loc[1] && highlightLoc[0] > loc[0] && highlightLoc[1] > loc[1]) {
-          locInc[0] += openTagLength;
-          locInc[1] += openTagLength + closeTagLength;
-        } else if (highlightLoc[0] <= loc[0] && highlightLoc[1] >= loc[1]) {
-          locInc[1] += openTagLength + closeTagLength;
-        } else if (highlightLoc[0] < loc[0] && highlightLoc[1] > loc[0] && highlightLoc[1] < loc[1]) {
-          locInc[1] += openTagLength;
-        } else if (highlightLoc[0] >= loc[0] && highlightLoc[1] <= loc[1]) {
-          locInc[0] += openTagLength;
-          locInc[1] += openTagLength;
-        }
+        } // syntactical correct but semantical incorrect
+        else if (highlightLoc[0] < loc[1] && highlightLoc[0] > loc[0] && highlightLoc[1] > loc[1]) {
+            locInc[0] += openTagLength;
+            locInc[1] += openTagLength + closeTagLength;
+          } else if (highlightLoc[0] <= loc[0] && highlightLoc[1] >= loc[1]) {
+            locInc[1] += openTagLength + closeTagLength;
+          } // syntactical correct but semantical incorrect
+          else if (highlightLoc[0] < loc[0] && highlightLoc[1] > loc[0] && highlightLoc[1] < loc[1]) {
+              locInc[1] += openTagLength;
+            } else if (highlightLoc[0] >= loc[0] && highlightLoc[1] <= loc[1]) {
+              locInc[0] += openTagLength;
+              locInc[1] += openTagLength;
+            }
       }
@@ -673,3 +688,3 @@ }
     return (0, _sbd.default)(text, options).map(raw => {
-      // can tokenizer return location directly***
+      // future work: can tokenizer return location directly
       const index = text.indexOf(raw);
@@ -689,3 +704,3 @@ return {
       // step 1: derive best substr
-      // /s may be better***
+      // future work: /s may be better
       const words = str.split(' ');
@@ -692,0 +707,0 @@

package.json

 {
   "name": "text-annotator",
-  "version": "0.7.3",
+  "version": "0.7.4",
   "description": "A JavaScript library for locating and annotating plain text in HTML",
@@ -5,0 +5,0 @@ "main": "build/text-annotator.js",

src/text-annotator.js

 import getSentences from './ext/sbd'
 
 // used to distinguish between browser and Node.js environments
+// is it possible to relax so as to allow jsdom
 const isBrowser =
   typeof window !== 'undefined' && typeof window.document !== 'undefined'
 
+// div inside span is a bad idea
 const blockElements = [
@@ -66,3 +68,3 @@ 'address',
     this.sentences = []
-    // one highlight can have more than one location because of the potential issue in tag insertion***
+    // future work: one highlight can have more than one location because of the potential issue in tag insertion
     this.highlights = []
@@ -106,2 +108,3 @@
 
+    // experimental feature
     if (fuzzySearchOptions) {
@@ -119,2 +122,3 @@ highlightIndex = this.fuzzySearch(
 
+    // experimental feature
     // eager search only works in (particular) browsers
@@ -136,2 +140,3 @@ if (isBrowser && eagerSearchOptions) {
 
+  // experimental feature
   // only support direct search for now
@@ -193,2 +198,3 @@ searchAll(str, options = {}) {
 
+  // experimental feature
   highlightAll(highlightIndexes, options = {}) {
@@ -220,2 +226,3 @@ // either containerId or content is required
     if (highlightIndex !== -1) {
+      // content is undefined if containerId and returnContent falsy
       return {
@@ -248,2 +255,3 @@ highlightIndex,
       let newContent = content
+      // need to change when one annotation => more than one highlight
       const loc = this.adjustLoc(
@@ -599,7 +607,7 @@ highlightIdPattern,
 
-  // further improvement when one annotation binds with more than one highlight***
-  // block elements are only used to check in the = condition for now
+  // future work: further improvement when one annotation binds with more than one highlight
+  // includeRequiredTag used in = condition only
   includeRequiredTag(i, highlightLoc, tag) {
     const isCloseTag = tag.startsWith('</')
-    const tagType = isCloseTag
+    const tagName = isCloseTag
       ? tag.split('</')[1].split('>')[0]
@@ -615,3 +623,4 @@ : tag
     let requiredTagCount = 0
-    // outer
+    // if both the start tag and the end tag are at the borders, place the tags outside the borders
+    // if the close tag is at the border, check backwards until the start of the highlight
     if (isCloseTag) {
@@ -627,5 +636,5 @@ for (let i2 = i - 1; i2 >= 0; i2--) {
           )
-          if (tag2.startsWith('</' + tagType)) {
+          if (tag2.startsWith('</' + tagName)) {
             requiredTagNumber++
-          } else if (tag2.startsWith('<' + tagType)) {
+          } else if (tag2.startsWith('<' + tagName)) {
             requiredTagCount++
@@ -639,3 +648,5 @@ }
       }
-    } else {
+    }
+    // if the start tag is at the border, check forwards until the end of the highlight
+    else {
       for (let i2 = i + 1; i2 < this.tagLocations.length; i2++) {
@@ -650,5 +661,5 @@ const tagLoc2 = this.tagLocations[i2]
           )
-          if (tag2.startsWith('<' + tagType)) {
+          if (tag2.startsWith('<' + tagName)) {
             requiredTagNumber++
-          } else if (tag2.startsWith('</' + tagType)) {
+          } else if (tag2.startsWith('</' + tagName)) {
             requiredTagCount++
@@ -685,2 +696,3 @@ }
         )
+        // if end tag, not block element and include the required close tag, add right to the tag
         if (
@@ -704,2 +716,3 @@ !tag.endsWith('/>') &&
           )
+          // if self close tag or end tag or block element or not include the required close tag, add right to the tag
           if (
@@ -727,4 +740,8 @@ tag.startsWith('</') ||
     // step 2: check locations of other highlights
+    // all span (no blocks)
+    // stored in a different array than tags
+    // can intersect
     for (let i = 0; i < this.highlights.length; i++) {
       const highlight = this.highlights[i]
+      // only check the highlighted
       if (highlight.highlighted) {
@@ -741,3 +758,5 @@ const openTagLength = TextAnnotator.getOpenTagLength(
           locInc[1] += openTagLength + closeTagLength
-        } else if (
+        }
+        // syntactical correct but semantical incorrect
+        else if (
           highlightLoc[0] < loc[1] &&
@@ -751,3 +770,5 @@ highlightLoc[0] > loc[0] &&
           locInc[1] += openTagLength + closeTagLength
-        } else if (
+        }
+        // syntactical correct but semantical incorrect
+        else if (
           highlightLoc[0] < loc[0] &&
@@ -816,3 +837,3 @@ highlightLoc[1] > loc[0] &&
     return getSentences(text, options).map(raw => {
-      // can tokenizer return location directly***
+      // future work: can tokenizer return location directly
       const index = text.indexOf(raw)
@@ -838,3 +859,3 @@ return { raw, index }
       // step 1: derive best substr
-      // /s may be better***
+      // future work: /s may be better
       const words = str.split(' ')
@@ -841,0 +862,0 @@ while (words.length) {

No alert changes

Improved metrics

Total package byte prevSize

2649184

increased by0.08%

Lines of code

2367

increased by1.37%

No dependency changes