thread-master/backend/internal/library/placement/parse_threads.go

package placement

import (
	"regexp"
	"strings"
)

var threadsPostURLRe = regexp.MustCompile(`(?i)threads\.(?:com|net)/@([^/]+)/post/([^/?#]+)`)

type ParsedThreadsPost struct {
	Permalink  string
	ExternalID string
	Author     string
	Text       string
}

func ParseThreadsPostFromWebResult(title, snippet, url string) (ParsedThreadsPost, bool) {
	permalink := normalizeThreadsPermalink(url)
	if permalink == "" {
		return ParsedThreadsPost{}, false
	}
	match := threadsPostURLRe.FindStringSubmatch(permalink)
	if len(match) < 3 {
		return ParsedThreadsPost{}, false
	}
	text := strings.TrimSpace(strings.Join(filterNonEmpty([]string{strings.TrimSpace(title), strings.TrimSpace(snippet)}), " — "))
	if text == "" {
		text = strings.TrimSpace(title)
	}
	if text == "" {
		text = strings.TrimSpace(snippet)
	}
	if len([]rune(text)) < 6 {
		return ParsedThreadsPost{}, false
	}
	return ParsedThreadsPost{
		Permalink:  permalink,
		ExternalID: match[2],
		Author:     match[1],
		Text:       text,
	}, true
}

func filterNonEmpty(items []string) []string {
	out := make([]string, 0, len(items))
	for _, item := range items {
		if strings.TrimSpace(item) != "" {
			out = append(out, item)
		}
	}
	return out
}

func normalizeThreadsPermalink(raw string) string {
	raw = strings.TrimSpace(raw)
	if raw == "" {
		return ""
	}
	raw = strings.Split(raw, "?")[0]
	raw = strings.Split(raw, "#")[0]
	if threadsPostURLRe.MatchString(raw) {
		return raw
	}
	return ""
}