hextra_mirror/layouts/_partials/utils/fragments.html

{{- /*
  fragments.html - Split page content into searchable fragments
  
  This partial processes a Hugo page and splits its content into fragments based on headings,
  creating a data structure suitable for search indexing. It supports different fragment types
  and handles hierarchical heading structures (h1, h2).
  
  Parameters:
  - .context (Page): The Hugo page to process
  - .type (string): Fragment type - "content" (default), "heading", "title", or "summary"
  
  Returns:
  - dict: Map of heading keys to content fragments
  
  Example:
  Input page with content:
    # Introduction
    This is the intro text.
    ## Setup
    Setup instructions here.
    # Configuration  
    Config details here.
  
  Output (type "content"):
  {
    "": "This is the intro text.",
    "intro#Introduction": "This is the intro text. Setup instructions here.",
    "setup#Setup": "Setup instructions here.", 
    "config#Configuration": "Config details here."
  }
  
  Fragment types:
  - "content": Splits page content by headings (default)
  - "heading": Returns heading keys with empty content
  - "title": Returns empty content (title handled elsewhere)
  - "summary": Returns page summary only
*/ -}}

{{- /* Extract page context and fragment type */ -}}
{{- $page := .context -}}
{{- $type := .type | default "content" -}}

{{- /* Process all headings */ -}}
{{- $s := newScratch -}}
{{- $s.Set "keys" slice -}}
{{- $s.Set "titles" slice -}}

{{- partial "utils/extract-headings.html" (dict "target" $page.Fragments "scratch" $s) -}}

{{- $headingKeys := $s.Get "keys" -}}
{{- $headingTitles := $s.Get "titles" -}}

{{- $content := $page.Content | htmlUnescape -}}
{{- $len := len $headingKeys -}}
{{- $data := dict -}}

{{ if eq $type "content" }}
  {{/* Include full content of the page */}}
  {{ if eq $len 0 }}
    {{ $data = $data | merge (dict "" ($page.Plain | htmlUnescape | strings.TrimSpace)) }}
  {{ else }}
    {{/* Split the raw content from bottom to top */}}
    {{ range seq $len }}
      {{ $i := sub $len . }}
      {{ $headingKey := index $headingKeys $i }}
      {{ $headingTitle := index $headingTitles $i }}

      {{ if eq $i 0 }}
        {{ $data = $data | merge (dict $headingKey ($content | plainify | htmlUnescape | strings.TrimSpace)) }}
      {{ else }}
        {{ $parts := split $content (printf "%s" $headingTitle) }}
        {{ $lastPart := index $parts (sub (len $parts) 1) }}

        {{ $data = $data | merge (dict $headingKey ($lastPart | plainify | htmlUnescape | strings.TrimSpace)) }}
        {{ $content = strings.TrimSuffix $lastPart $content }}
        {{ $content = strings.TrimSuffix (printf "%s" $headingTitle) $content }}
      {{ end }}
    {{ end }}
  {{ end }}
{{ else if (eq $type "heading" ) }}
  {{/* Put heading keys with empty content to the data object */}}
  {{ $data = dict "" "" }}
  {{ range $headingKeys }}
    {{ $data = $data | merge (dict . "") }}
  {{ end }}
{{ else if (eq $type "title") }}
  {{/* Use empty data object since title is included in search-data.json */}}
  {{ $data = $data | merge (dict "" "") }}
{{ else if (eq $type "summary" ) }}
  {{ $data = $data | merge (dict "" ($page.Summary | plainify | htmlUnescape | strings.TrimSpace)) }}
{{ end }}

{{ return $data }}