hugolib: Consider summary in front matter for .Summary

Add the ability to have a `summary` page variable that overrides
the auto-generated summary.  Logic for obtaining summary becomes:

  * if summary divider is present in content, use the text above it
  * if summary variables is present in page metadata, use that
  * auto-generate summary from first _x_ words of the content

Fixes #5800
This commit is contained in:
Jim McDonald 2019-04-05 18:11:04 +01:00 committed by Bjørn Erik Pedersen
parent ebab291c0e
commit 3a62d54745
8 changed files with 115 additions and 15 deletions

View file

@ -42,6 +42,12 @@ import (
// SummaryDivider denotes where content summarization should end. The default is "<!--more-->".
var SummaryDivider = []byte("<!--more-->")
var (
openingPTag = []byte("<p>")
closingPTag = []byte("</p>")
paragraphIndicator = []byte("<p")
)
// ContentSpec provides functionality to render markdown content.
type ContentSpec struct {
BlackFriday *BlackFriday
@ -580,6 +586,21 @@ func (c *ContentSpec) TruncateWordsToWholeSentence(s string) (string, bool) {
return strings.TrimSpace(s[:endIndex]), endIndex < len(s)
}
// TrimShortHTML removes the <p>/</p> tags from HTML input in the situation
// where said tags are the only <p> tags in the input and enclose the content
// of the input (whitespace excluded).
func (c *ContentSpec) TrimShortHTML(input []byte) []byte {
first := bytes.Index(input, paragraphIndicator)
last := bytes.LastIndex(input, paragraphIndicator)
if first == last {
input = bytes.TrimSpace(input)
input = bytes.TrimPrefix(input, openingPTag)
input = bytes.TrimSuffix(input, closingPTag)
input = bytes.TrimSpace(input)
}
return input
}
func isEndOfSentence(r rune) bool {
return r == '.' || r == '?' || r == '!' || r == '"' || r == '\n'
}

View file

@ -29,6 +29,28 @@ import (
const tstHTMLContent = "<!DOCTYPE html><html><head><script src=\"http://two/foobar.js\"></script></head><body><nav><ul><li hugo-nav=\"section_0\"></li><li hugo-nav=\"section_1\"></li></ul></nav><article>content <a href=\"http://two/foobar\">foobar</a>. Follow up</article><p>This is some text.<br>And some more.</p></body></html>"
func TestTrimShortHTML(t *testing.T) {
tests := []struct {
input, output []byte
}{
{[]byte(""), []byte("")},
{[]byte("Plain text"), []byte("Plain text")},
{[]byte(" \t\n Whitespace text\n\n"), []byte("Whitespace text")},
{[]byte("<p>Simple paragraph</p>"), []byte("Simple paragraph")},
{[]byte("\n \n \t <p> \t Whitespace\nHTML \n\t </p>\n\t"), []byte("Whitespace\nHTML")},
{[]byte("<p>Multiple</p><p>paragraphs</p>"), []byte("<p>Multiple</p><p>paragraphs</p>")},
{[]byte("<p>Nested<p>paragraphs</p></p>"), []byte("<p>Nested<p>paragraphs</p></p>")},
}
c := newTestContentSpec()
for i, test := range tests {
output := c.TrimShortHTML(test.input)
if bytes.Compare(test.output, output) != 0 {
t.Errorf("Test %d failed. Expected %q got %q", i, test.output, output)
}
}
}
func TestStripHTML(t *testing.T) {
type test struct {
input, expected string

View file

@ -64,6 +64,8 @@ type pageMeta struct {
title string
linkTitle string
summary string
resourcePath string
weight int
@ -361,6 +363,9 @@ func (pm *pageMeta) setMetadata(p *pageState, frontmatter map[string]interface{}
case "linktitle":
pm.linkTitle = cast.ToString(v)
pm.params[loki] = pm.linkTitle
case "summary":
pm.summary = cast.ToString(v)
pm.params[loki] = pm.summary
case "description":
pm.description = cast.ToString(v)
pm.params[loki] = pm.description

View file

@ -128,6 +128,14 @@ func newPageContentOutput(p *pageState) func(f output.Format) (*pageContentOutpu
cp.summary = helpers.BytesToHTML(summary)
}
}
} else if cp.p.m.summary != "" {
html := cp.p.s.ContentSpec.RenderBytes(&helpers.RenderingContext{
Content: []byte(cp.p.m.summary), RenderTOC: false, PageFmt: cp.p.m.markup,
Cfg: p.Language(),
DocumentID: p.File().UniqueID(), DocumentName: p.File().Path(),
Config: cp.p.getRenderingConfig()})
html = cp.p.s.ContentSpec.TrimShortHTML(html)
cp.summary = helpers.BytesToHTML(html)
}
}
@ -271,7 +279,7 @@ func (p *pageContentOutput) WordCount() int {
}
func (p *pageContentOutput) setAutoSummary() error {
if p.p.source.hasSummaryDivider {
if p.p.source.hasSummaryDivider || p.p.m.summary != "" {
return nil
}

View file

@ -45,6 +45,16 @@ const (
simplePageRFC3339Date = "---\ntitle: RFC3339 Date\ndate: \"2013-05-17T16:59:30Z\"\n---\nrfc3339 content"
simplePageWithoutSummaryDelimiter = `---
title: SimpleWithoutSummaryDelimiter
---
[Lorem ipsum](https://lipsum.com/) dolor sit amet, consectetur adipiscing elit, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua. Ut enim ad minim veniam, quis nostrud exercitation ullamco laboris nisi ut aliquip ex ea commodo consequat. Duis aute irure dolor in reprehenderit in voluptate velit esse cillum dolore eu fugiat nulla pariatur. Excepteur sint occaecat cupidatat non proident, sunt in culpa qui officia deserunt mollit anim id est laborum.
Additional text.
Further text.
`
simplePageWithSummaryDelimiter = `---
title: Simple
---
@ -52,6 +62,16 @@ Summary Next Line
<!--more-->
Some more text
`
simplePageWithSummaryParameter = `---
title: SimpleWithSummaryParameter
summary: "Page with summary parameter and [a link](http://www.example.com/)"
---
Some text.
Some more text.
`
simplePageWithSummaryDelimiterAndMarkdownThatCrossesBorder = `---
@ -519,6 +539,22 @@ func TestCreateNewPage(t *testing.T) {
testAllMarkdownEnginesForPages(t, assertFunc, settings, simplePage)
}
func TestPageSummary(t *testing.T) {
t.Parallel()
assertFunc := func(t *testing.T, ext string, pages page.Pages) {
p := pages[0]
checkPageTitle(t, p, "SimpleWithoutSummaryDelimiter")
// Source is not Asciidoctor- or RST-compatibile so don't test them
if ext != "ad" && ext != "rst" {
checkPageContent(t, p, normalizeExpected(ext, "<p><a href=\"https://lipsum.com/\">Lorem ipsum</a> dolor sit amet, consectetur adipiscing elit, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua. Ut enim ad minim veniam, quis nostrud exercitation ullamco laboris nisi ut aliquip ex ea commodo consequat. Duis aute irure dolor in reprehenderit in voluptate velit esse cillum dolore eu fugiat nulla pariatur. Excepteur sint occaecat cupidatat non proident, sunt in culpa qui officia deserunt mollit anim id est laborum.</p>\n\n<p>Additional text.</p>\n\n<p>Further text.</p>\n"), ext)
checkPageSummary(t, p, normalizeExpected(ext, "Lorem ipsum dolor sit amet, consectetur adipiscing elit, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua. Ut enim ad minim veniam, quis nostrud exercitation ullamco laboris nisi ut aliquip ex ea commodo consequat. Duis aute irure dolor in reprehenderit in voluptate velit esse cillum dolore eu fugiat nulla pariatur. Excepteur sint occaecat cupidatat non proident, sunt in culpa qui officia deserunt mollit anim id est laborum. Additional text."), ext)
}
checkPageType(t, p, "page")
}
testAllMarkdownEnginesForPages(t, assertFunc, nil, simplePageWithoutSummaryDelimiter)
}
func TestPageWithDelimiter(t *testing.T) {
t.Parallel()
assertFunc := func(t *testing.T, ext string, pages page.Pages) {
@ -532,6 +568,22 @@ func TestPageWithDelimiter(t *testing.T) {
testAllMarkdownEnginesForPages(t, assertFunc, nil, simplePageWithSummaryDelimiter)
}
func TestPageWithSummaryParameter(t *testing.T) {
t.Parallel()
assertFunc := func(t *testing.T, ext string, pages page.Pages) {
p := pages[0]
checkPageTitle(t, p, "SimpleWithSummaryParameter")
checkPageContent(t, p, normalizeExpected(ext, "<p>Some text.</p>\n\n<p>Some more text.</p>\n"), ext)
// Summary is not Asciidoctor- or RST-compatibile so don't test them
if ext != "ad" && ext != "rst" {
checkPageSummary(t, p, normalizeExpected(ext, "Page with summary parameter and <a href=\"http://www.example.com/\">a link</a>"), ext)
}
checkPageType(t, p, "page")
}
testAllMarkdownEnginesForPages(t, assertFunc, nil, simplePageWithSummaryParameter)
}
// Issue #3854
// Also see https://github.com/gohugoio/hugo/issues/3977
func TestPageWithDateFields(t *testing.T) {

View file

@ -55,6 +55,9 @@ func TestRSSOutput(t *testing.T) {
if c != rssLimit {
t.Errorf("incorrect RSS item count: expected %d, got %d", rssLimit, c)
}
// Encoded summary
th.assertFileContent(filepath.Join("public", rssURI), "<?xml", "description", "A &lt;em&gt;custom&lt;/em&gt; summary")
}
// Before Hugo 0.49 we set the pseudo page kind RSS on the page when output to RSS.

View file

@ -586,6 +586,7 @@ date = "2012-01-01"
publishdate = "2012-01-01"
my_param = "baz"
my_date = 2010-05-27T07:32:00Z
summary = "A _custom_ summary"
categories = [ "hugo" ]
+++
Front Matter with Ordered Pages 4. This is longer content`

View file

@ -15,7 +15,6 @@
package transform
import (
"bytes"
"html"
"html/template"
@ -91,12 +90,6 @@ func (ns *Namespace) HTMLUnescape(s interface{}) (string, error) {
return html.UnescapeString(ss), nil
}
var (
markdownTrimPrefix = []byte("<p>")
markdownTrimSuffix = []byte("</p>\n")
markdownParagraphIndicator = []byte("<p")
)
// Markdownify renders a given input from Markdown to HTML.
func (ns *Namespace) Markdownify(s interface{}) (template.HTML, error) {
ss, err := cast.ToStringE(s)
@ -114,14 +107,9 @@ func (ns *Namespace) Markdownify(s interface{}) (template.HTML, error) {
)
// Strip if this is a short inline type of text.
first := bytes.Index(m, markdownParagraphIndicator)
last := bytes.LastIndex(m, markdownParagraphIndicator)
if first == last {
m = bytes.TrimPrefix(m, markdownTrimPrefix)
m = bytes.TrimSuffix(m, markdownTrimSuffix)
}
m = ns.deps.ContentSpec.TrimShortHTML(m)
return template.HTML(m), nil
return helpers.BytesToHTML(m), nil
}
// Plainify returns a copy of s with all HTML tags removed.