Skip to content

Commit 4a30253

Browse files
committed
Added words to statistics count. Also ordered metrics from smallest quantity (lines) up through largest quantity (characters).
1 parent 2c49401 commit 4a30253

File tree

1 file changed

+6
-3
lines changed

1 file changed

+6
-3
lines changed

single-page-apps/text_stats_analyzer.html

Lines changed: 6 additions & 3 deletions
Original file line numberDiff line numberDiff line change
@@ -94,6 +94,9 @@ <h5 class="card-title">Analysis Results</h5>
9494
// Basic stats
9595
const numCharacters = text.length;
9696
const numLines = text.split(/\r?\n/).length;
97+
98+
// Word count using regex to split on whitespace
99+
const numWords = text.trim().split(/\s+/).filter(word => word.length > 0).length;
97100

98101
// Check if GPTTokenizer_o200k_base is loaded and has the encode method
99102
if (
@@ -111,10 +114,10 @@ <h5 class="card-title">Analysis Results</h5>
111114
// Display the results
112115
const output = `
113116
<ul>
114-
<li><strong>Characters:</strong> ${numCharacters}</li>
115117
<li><strong>Lines:</strong> ${numLines}</li>
116-
<li><strong>Tokens*:</strong> ${numTokens}</li>
117-
<li><i>*Per GPT Tokenizer (4o, o1, etc)</i></li>
118+
<li><strong>Words:</strong> ${numWords}</li>
119+
<li><strong>Tokens*:</strong> ${numTokens}, <i>*Per GPT Tokenizer (4o, o1, etc)</i></li>
120+
<li><strong>Characters:</strong> ${numCharacters}</li>
118121
</ul>
119122
`;
120123
document.getElementById('output').innerHTML = output;

0 commit comments

Comments
 (0)