Skip to content

Add phonetic comparison to search algorithm #33

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
nbeerbower wants to merge 1 commit into
base: main
Choose a base branch
from
Open

Add phonetic comparison to search algorithm #33

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
256 changes: 251 additions & 5 deletions build_docs.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1394,6 +1394,19 @@ def generate_html(docs, logo_data, lang='en'):
font-size: 0.9rem;
}}

.search-phonetic-badge {{
display: inline-block;
margin-left: 0.5rem;
padding: 0.15rem 0.4rem;
background: var(--sage);
color: var(--pine);
border-radius: 4px;
font-size: 0.7rem;
font-style: italic;
text-transform: none;
letter-spacing: normal;
}}

.search-shortcut {{
display: none;
margin-left: 0.5rem;
Expand Down Expand Up @@ -1992,6 +2005,197 @@ def generate_html(docs, logo_data, lang='en'):
let selectedIndex = -1;
let currentResults = [];

// Metaphone phonetic encoding algorithm
// Converts words to phonetic codes for fuzzy matching
function metaphone(word) {{
if (!word) return '';
word = word.toUpperCase();
let result = '';
let i = 0;

// Skip initial silent letters
if (word.length >= 2) {{
const start = word.substring(0, 2);
if (['KN', 'GN', 'PN', 'AE', 'WR'].includes(start)) {{
i = 1;
}}
}}

while (i < word.length && result.length < 6) {{
const c = word[i];
const prev = i > 0 ? word[i - 1] : '';
const next = i < word.length - 1 ? word[i + 1] : '';
const next2 = i < word.length - 2 ? word[i + 2] : '';

// Skip duplicates except CC
if (c === prev && c !== 'C') {{
i++;
continue;
}}

switch (c) {{
case 'A': case 'E': case 'I': case 'O': case 'U':
// Vowels only kept at start
if (i === 0) result += c;
break;

case 'B':
// B silent after M at end
if (!(prev === 'M' && i === word.length - 1)) result += 'B';
break;

case 'C':
if (next === 'I' && next2 === 'A') {{
result += 'X'; // CIA -> X
}} else if (['I', 'E', 'Y'].includes(next)) {{
result += 'S'; // Soft C
}} else if (next === 'H') {{
result += 'X'; // CH -> X
i++;
}} else {{
result += 'K'; // Hard C
}}
break;

case 'D':
if (next === 'G' && ['I', 'E', 'Y'].includes(next2)) {{
result += 'J'; // DGE, DGI, DGY -> J
i++;
}} else {{
result += 'T';
}}
break;

case 'F':
result += 'F';
break;

case 'G':
if (next === 'H') {{
if (!['A', 'E', 'I', 'O', 'U'].includes(next2)) {{
i++; // GH silent
}} else {{
result += 'K';
}}
}} else if (next === 'N' && (i === word.length - 2 || (next2 === 'E' && i === word.length - 3))) {{
// GN at end or GNE at end is silent
}} else if (['I', 'E', 'Y'].includes(next)) {{
result += 'J'; // Soft G
}} else {{
result += 'K'; // Hard G
}}
break;

case 'H':
// H only if followed by vowel and not after vowel
if (['A', 'E', 'I', 'O', 'U'].includes(next) && !['A', 'E', 'I', 'O', 'U'].includes(prev)) {{
result += 'H';
}}
break;

case 'J':
result += 'J';
break;

case 'K':
if (prev !== 'C') result += 'K';
break;

case 'L':
result += 'L';
break;

case 'M':
result += 'M';
break;

case 'N':
result += 'N';
break;

case 'P':
if (next === 'H') {{
result += 'F';
i++;
}} else {{
result += 'P';
}}
break;

case 'Q':
result += 'K';
break;

case 'R':
result += 'R';
break;

case 'S':
if (next === 'H') {{
result += 'X';
i++;
}} else if (next === 'I' && ['O', 'A'].includes(next2)) {{
result += 'X'; // SIO, SIA -> X
}} else {{
result += 'S';
}}
break;

case 'T':
if (next === 'I' && ['O', 'A'].includes(next2)) {{
result += 'X'; // TIO, TIA -> X
}} else if (next === 'H') {{
result += '0'; // TH -> 0 (theta)
i++;
}} else if (next !== 'C' || next2 !== 'H') {{
result += 'T';
}}
break;

case 'V':
result += 'F';
break;

case 'W':
if (['A', 'E', 'I', 'O', 'U'].includes(next)) result += 'W';
break;

case 'X':
result += 'KS';
break;

case 'Y':
if (['A', 'E', 'I', 'O', 'U'].includes(next)) result += 'Y';
break;

case 'Z':
result += 'S';
break;
}}
i++;
}}
return result;
}}

// Get phonetic codes for all words in a string
function getPhoneticCodes(text) {{
if (!text) return [];
const words = text.toLowerCase().replace(/[^a-z\\s]/g, '').split(/\\s+/);
return words.filter(w => w.length > 2).map(w => ({{ word: w, code: metaphone(w) }}));
}}

// Check if query matches any word phonetically
function phoneticMatch(queryCodes, targetCodes) {{
for (const qc of queryCodes) {{
for (const tc of targetCodes) {{
if (qc.code && tc.code && qc.code === tc.code && qc.word !== tc.word) {{
return tc.word; // Return the matched word for highlighting
}}
}}
}}
return null;
}}

// Build search index from PAGES
function buildSearchIndex() {{
const index = [];
Expand Down Expand Up @@ -2023,13 +2227,19 @@ def generate_html(docs, logo_data, lang='en'):
.trim()
.substring(0, 200);

// Compute phonetic codes for title and headings
const titleCodes = getPhoneticCodes(title);
const headingCodes = headings.flatMap(h => getPhoneticCodes(h));

index.push({{
title: title,
pageId: page.id,
section: section,
headings: headings,
content: page.content.toLowerCase(),
preview: preview
preview: preview,
titleCodes: titleCodes,
headingCodes: headingCodes
}});
}}
return index;
Expand All @@ -2042,12 +2252,14 @@ def generate_html(docs, logo_data, lang='en'):
if (!query || query.length < 2) return [];

const q = query.toLowerCase().trim();
const queryCodes = getPhoneticCodes(q);
const results = [];

for (const item of searchIndex) {{
let score = 0;
let matchedHeading = null;
let matchContext = '';
let phoneticMatchWord = null;

// Check title (highest priority)
const titleLower = item.title.toLowerCase();
Expand Down Expand Up @@ -2084,6 +2296,32 @@ def generate_html(docs, logo_data, lang='en'):
if (end < item.content.length) matchContext = matchContext + '...';
}}

// Phonetic matching (finds similar-sounding words even with typos)
if (score === 0 && queryCodes.length > 0) {{
// Check title phonetically
const titleMatch = phoneticMatch(queryCodes, item.titleCodes);
if (titleMatch) {{
score = 35; // Phonetic title match
phoneticMatchWord = titleMatch;
}}

// Check headings phonetically
if (!titleMatch) {{
const headingMatch = phoneticMatch(queryCodes, item.headingCodes);
if (headingMatch) {{
score = Math.max(score, 25); // Phonetic heading match
phoneticMatchWord = headingMatch;
// Find the heading containing this match
for (const heading of item.headings) {{
if (heading.toLowerCase().includes(headingMatch)) {{
matchedHeading = heading;
break;
}}
}}
}}
}}
}}

if (score > 0) {{
results.push({{
title: item.title,
Expand All @@ -2092,7 +2330,8 @@ def generate_html(docs, logo_data, lang='en'):
score: score,
matchedHeading: matchedHeading,
preview: matchContext || item.preview,
query: q
query: q,
phoneticMatch: phoneticMatchWord
}});
}}
}}
Expand All @@ -2118,13 +2357,20 @@ def generate_html(docs, logo_data, lang='en'):
}}

const html = results.map((result, index) => {{
const titleHtml = highlightText(result.title, query);
const previewHtml = highlightText(result.preview, query);
// Highlight both query and phonetic match word
let titleHtml = highlightText(result.title, query);
let previewHtml = highlightText(result.preview, query);
if (result.phoneticMatch) {{
titleHtml = highlightText(titleHtml, result.phoneticMatch);
previewHtml = highlightText(previewHtml, result.phoneticMatch);
}}
const selectedClass = index === selectedIndex ? ' selected' : '';
const phoneticIndicator = result.phoneticMatch ?
`<span class="search-phonetic-badge">sounds like "${{result.phoneticMatch}}"</span>` : '';

return `
<div class="search-result${{selectedClass}}" data-index="${{index}}" data-page="${{result.pageId}}">
${{result.section ? `<div class="search-result-section">${{result.section}}</div>` : ''}}
${{result.section ? `<div class="search-result-section">${{result.section}}${{phoneticIndicator}}</div>` : phoneticIndicator}}
<div class="search-result-title">${{titleHtml}}</div>
${{result.matchedHeading ? `<div class="search-result-preview">${{highlightText(result.matchedHeading, query)}}</div>` : ''}}
<div class="search-result-preview">${{previewHtml}}</div>
Expand Down
Loading