diff options
Diffstat (limited to 'bucky/search/snippet.js')
| -rw-r--r-- | bucky/search/snippet.js | 47 |
1 files changed, 20 insertions, 27 deletions
diff --git a/bucky/search/snippet.js b/bucky/search/snippet.js index cd0657f..17988d2 100644 --- a/bucky/search/snippet.js +++ b/bucky/search/snippet.js @@ -1,19 +1,11 @@ var util = require('../util/util') var STOPWORDS = require('./stopwords') -function bold_snippet(s, terms) { - return bold_terms(snippet(s, terms), terms) -} -function bold_terms (s, terms) { - s = util.sanitize(s) - terms.forEach( (term) => { - s.replace(new RegExp("\b" + term + "\b", "i"), "<b>" + term + "</b>") - }) -} function snippet(s, terms) { s = util.sanitize(s) - var term_re = new RegExp("\b(" + terms.join("|") + ")\b", "i") - var words = s.split(/\s+/) + var term_set = new Set(terms) + + var words = s.split(/[^a-zA-Z0-9]+/) var snippet = ""; // deduper for matching @words indexes, so we don't add a word twice @@ -26,27 +18,30 @@ function snippet(s, terms) { var aggr = 0; // amount of context to show, in number of words surrounding a match - var $pad = 4; + var pad = 10; // loop over each of the words in the string - words.some((word, i) => { + var word + for (var i = 0, len = words.length; i < len; i++) { + word = words[i] + // if the word matches... - if (term_re.match(word) && ! STOPWORDS.has(word.toLowerCase())) { + if (term_set.has(word.toLowerCase()) && ! STOPWORDS.has(word.toLowerCase())) { // if we aren't already aggregating, add an ellipsis - if (! $aggr) { + if (! aggr) { words_matched.push("...") } - + // look backward $pad words var idx; - for (var j = -pad; j < 1; j++) { + INNER: for (var j = -pad; j < 1; j++) { // create a new index from the offset idx = i + j; // is this a valid index? has it already been encountered? - if (idx < 0) continue; - if (idx > words.length) continue; - if (index_matches[idx]) continue; + if (idx < 0) continue INNER; + if (idx > words.length) continue INNER; + if (index_matches[idx]) continue INNER; // checks out, save this word words_matched.push(words[idx]) @@ -69,20 +64,18 @@ function snippet(s, terms) { // one less word to aggregate aggr--; } - + // keep snippets to a modest length - return words_matched.length > 30; - }) + if (words_matched.length > 30) break + } // add a trailing ellipsis words_matched.push("...") // create the snippet from the saved context words snippet = words_matched.join(" ") - + return snippet } -module.exports = { - bold_snippet, bold_terms, snippet, -}
\ No newline at end of file +module.exports = snippet
\ No newline at end of file |