diff options
| author | Bjørn Erik Pedersen <[email protected]> | 2024-09-10 08:54:03 +0200 |
|---|---|---|
| committer | Bjørn Erik Pedersen <[email protected]> | 2024-09-10 11:03:47 +0200 |
| commit | 3d6baedaec306300f2c6f7ed471e774dca0f112a (patch) | |
| tree | 4a7b7f62c337aceb9983f8a0490b7e153a7b3d23 /resources/page/page_markup.go | |
| parent | 84ee00bbc24328295237695a39e6e876ed186312 (diff) | |
| download | hugo-3d6baedaec306300f2c6f7ed471e774dca0f112a.tar.gz hugo-3d6baedaec306300f2c6f7ed471e774dca0f112a.zip | |
Don't count HTML markup in auto summaries
This commit also fixes a bug where a `</picture>` end tag was wrongly used to detect a end paragraph. This should be very rare, though.
Closes #12837
Diffstat (limited to 'resources/page/page_markup.go')
| -rw-r--r-- | resources/page/page_markup.go | 20 |
1 files changed, 19 insertions, 1 deletions
diff --git a/resources/page/page_markup.go b/resources/page/page_markup.go index ef4a56e3a..44980e8b0 100644 --- a/resources/page/page_markup.go +++ b/resources/page/page_markup.go @@ -161,6 +161,16 @@ func (s *HtmlSummary) resolveParagraphTagAndSetWrapper(mt media.Type) tagReStart return ptag } +// Avoid counting words that are most likely HTML tokens. +var ( + isProbablyHTMLTag = regexp.MustCompile(`^<\/?[A-Za-z]+>?$`) + isProablyHTMLAttribute = regexp.MustCompile(`^[A-Za-z]+=["']`) +) + +func isProbablyHTMLToken(s string) bool { + return s == ">" || isProbablyHTMLTag.MatchString(s) || isProablyHTMLAttribute.MatchString(s) +} + // ExtractSummaryFromHTML extracts a summary from the given HTML content. func ExtractSummaryFromHTML(mt media.Type, input string, numWords int, isCJK bool) (result HtmlSummary) { result.source = input @@ -173,6 +183,14 @@ func ExtractSummaryFromHTML(mt media.Type, input string, numWords int, isCJK boo var count int countWord := func(word string) int { + word = strings.TrimSpace(word) + if len(word) == 0 { + return 0 + } + if isProbablyHTMLToken(word) { + return 0 + } + if isCJK { word = tpl.StripHTML(word) runeCount := utf8.RuneCountInString(word) @@ -193,7 +211,7 @@ func ExtractSummaryFromHTML(mt media.Type, input string, numWords int, isCJK boo for j := result.WrapperStart.High; j < high; { s := input[j:] - closingIndex := strings.Index(s, "</"+ptag.tagName) + closingIndex := strings.Index(s, "</"+ptag.tagName+">") if closingIndex == -1 { break |