Files
njump/utils.go

533 lines
16 KiB
Go

package main
import (
"context"
"encoding/json"
"fmt"
"html"
"html/template"
"net/http"
"regexp"
"slices"
"strconv"
"strings"
"sync"
"time"
me "github.com/huantt/plaintext-extractor/markdown"
"github.com/nbd-wtf/go-nostr"
"github.com/nbd-wtf/go-nostr/nip19"
"github.com/puzpuzpuz/xsync/v3"
"mvdan.cc/xurls/v2"
)
const (
XML_HEADER = "<?xml version=\"1.0\" encoding=\"UTF-8\"?>\n"
THIN_SPACE = '\u2009'
INVISIBLE_SPACE = '\U0001D17A'
)
var (
urlSuffixMatcher = regexp.MustCompile(`[\w-_.]+\.[\w-_.]+(\/[\/\w]*)?$`)
nostrEveryMatcher = regexp.MustCompile(`nostr:((npub|note|nevent|nprofile|naddr)1[a-z0-9]+)\b`)
nostrNoteNeventMatcher = regexp.MustCompile(`(?:^|<br/>|\s)nostr:((note|nevent|naddr)1[a-z0-9]+)\b(?:\s|<br/>|$)`)
nostrNpubNprofileMatcher = regexp.MustCompile(`nostr:((npub|nprofile)1[a-z0-9]+)\b`)
urlMatcher = func() *regexp.Regexp {
// hack to only allow these schemes while still using this library
xurls.Schemes = []string{"https"}
xurls.SchemesNoAuthority = []string{"blob"}
xurls.SchemesUnofficial = []string{"http"}
return xurls.Strict()
}()
imageExtensionMatcher = regexp.MustCompile(`.*\.(png|jpg|jpeg|gif|webp|avif)((\?|\#).*)?$`)
videoExtensionMatcher = regexp.MustCompile(`.*\.(mp4|ogg|webm|mov)((\?|\#).*)?$`)
urlRegex = xurls.Strict()
markdownExtractor = me.NewExtractor()
)
var kindNames = map[int]string{
0: "Metadata",
1: "Short Text Note",
2: "Recommend Relay",
3: "Contacts",
4: "Encrypted Direct Messages",
5: "Event Deletion",
6: "Reposts",
7: "Reaction",
8: "Badge Award",
40: "Channel Creation",
41: "Channel Metadata",
42: "Channel Message",
43: "Channel Hide Message",
44: "Channel Mute User",
1063: "File Metadata",
1311: "Live Chat Message",
1984: "Reporting",
9734: "Zap Request",
9735: "Zap",
10000: "Mute List",
10001: "Pin List",
10002: "Relay List Metadata",
13194: "Wallet Info",
22242: "Client Authentication",
23194: "Wallet Request",
23195: "Wallet Response",
24133: "Nostr Connect",
30000: "Categorized People List",
30001: "Categorized Bookmark List",
30008: "Profile Badges",
30009: "Badge Definition",
30017: "Create or update a stall",
30018: "Create or update a product",
30023: "Long-form Content",
30078: "Application-specific Data",
30818: "Wiki article",
30311: "Live Event",
}
var kindNIPs = map[int]string{
0: "01",
1: "01",
2: "01",
3: "02",
4: "04",
5: "09",
6: "18",
7: "25",
8: "58",
40: "28",
41: "28",
42: "28",
43: "28",
44: "28",
1063: "94",
1311: "53",
1984: "56",
9734: "57",
9735: "57",
10000: "51",
10001: "51",
10002: "65",
13194: "47",
22242: "42",
23194: "47",
23195: "47",
24133: "46",
30000: "51",
30001: "51",
30008: "58",
30009: "58",
30017: "15",
30018: "15",
30023: "23",
30078: "78",
30818: "54",
30311: "53",
}
type Style string
const (
StyleTelegram Style = "telegram"
StyleTwitter = "twitter"
StyleFacebook = "facebook" // Both Facebook and Instagram
StyleIOS = "ios"
StyleAndroid = "android"
StyleMattermost = "mattermost"
StyleSlack = "slack"
StyleDiscord = "discord"
StyleWhatsapp = "whatsapp"
StyleIframely = "iframely"
StyleNormal = "normal"
StyleUnknown = "unknown"
)
func getPreviewStyle(r *http.Request) Style {
if style := r.URL.Query().Get("style"); style != "" {
// debug mode
return Style(style)
}
ua := strings.ToLower(r.Header.Get("User-Agent"))
accept := r.Header.Get("Accept")
switch {
case strings.Contains(ua, "telegrambot"):
return StyleTelegram
case strings.Contains(ua, "twitterbot"):
return StyleTwitter
case strings.Contains(ua, "facebookexternalhit"):
return StyleFacebook
case strings.Contains(ua, "iphone"), strings.Contains(ua, "ipad"), strings.Contains(ua, "ipod"):
return StyleIOS
case strings.Contains(ua, "android"):
return StyleAndroid
case strings.Contains(ua, "mattermost"):
return StyleMattermost
case strings.Contains(ua, "slack"):
return StyleSlack
case strings.Contains(ua, "discord"):
return StyleDiscord
case strings.Contains(ua, "whatsapp"):
return StyleWhatsapp
case strings.Contains(ua, "iframely"):
return StyleIframely
case strings.Contains(accept, "text/html"):
return StyleNormal
default:
return StyleUnknown
}
}
func replaceURLsWithTags(input string, imageReplacementTemplate, videoReplacementTemplate string, skipLinks bool) string {
return urlMatcher.ReplaceAllStringFunc(input, func(match string) string {
switch {
case imageExtensionMatcher.MatchString(match):
// Match and replace image URLs with a custom replacement
// Usually is html <img> => ` <img src="%s" alt=""> `
// or markdown !()[...] tags for further processing => `![](%s)`
return fmt.Sprintf(imageReplacementTemplate, match)
case videoExtensionMatcher.MatchString(match):
// Match and replace video URLs with a custom replacement
// Usually is html <video> => ` <video controls width="100%%"><source src="%s"></video> `
// or markdown !()[...] tags for further processing => `![](%s)`
return fmt.Sprintf(videoReplacementTemplate, match)
default:
if skipLinks {
return match
} else {
return "<a href=\"" + match + "\">" + match + "</a>"
}
}
})
}
func replaceNostrURLsWithHTMLTags(matcher *regexp.Regexp, input string) string {
// match and replace npup1, nprofile1, note1, nevent1, etc
names := xsync.NewMapOf[string, string]()
wg := sync.WaitGroup{}
// first we run it without waiting for the results of getNameFromNip19() as they will be async
for _, match := range matcher.FindAllString(input, len(input)+1) {
nip19 := match[len("nostr:"):]
if strings.HasPrefix(nip19, "npub1") || strings.HasPrefix(nip19, "nprofile1") {
ctx, cancel := context.WithTimeout(context.Background(), time.Second*4)
defer cancel()
wg.Add(1)
go func() {
name, _ := getNameFromNip19(ctx, nip19)
names.Store(nip19, name)
wg.Done()
}()
}
}
// in the second time now that we got all the names we actually perform replacement
wg.Wait()
return matcher.ReplaceAllStringFunc(input, func(match string) string {
nip19 := match[len("nostr:"):]
firstChars := nip19[:8]
lastChars := nip19[len(nip19)-4:]
if strings.HasPrefix(nip19, "npub1") || strings.HasPrefix(nip19, "nprofile1") {
name, _ := names.Load(nip19)
return fmt.Sprintf(`<span itemprop="mentions" itemscope itemtype="https://schema.org/Person"><a itemprop="url" href="/%s" class="bg-lavender dark:prose:text-neutral-50 dark:text-neutral-50 dark:bg-garnet px-1"><span>%s</span> (<span class="italic">%s</span>)</a></span>`, nip19, name, firstChars+"…"+lastChars)
} else {
return fmt.Sprintf(`<span itemprop="mentions" itemscope itemtype="https://schema.org/Article"><a itemprop="url" href="/%s" class="bg-lavender dark:prose:text-neutral-50 dark:text-neutral-50 dark:bg-garnet px-1">%s</a></span>`, nip19, firstChars+"…"+lastChars)
}
})
}
func shortenNostrURLs(input string) string {
// match and replace npup1, nprofile1, note1, nevent1, etc
return nostrEveryMatcher.ReplaceAllStringFunc(input, func(match string) string {
nip19 := match[len("nostr:"):]
firstChars := nip19[:8]
lastChars := nip19[len(nip19)-4:]
if strings.HasPrefix(nip19, "npub1") || strings.HasPrefix(nip19, "nprofile1") {
return "@" + firstChars + "…" + lastChars
} else {
return "#" + firstChars + "…" + lastChars
}
})
}
func getNameFromNip19(ctx context.Context, nip19code string) (string, bool) {
metadata, _ := sys.FetchProfileFromInput(ctx, nip19code)
if metadata.Name == "" {
return nip19code, false
}
return metadata.Name, true
}
// replaces an npub/nprofile with the name of the author, if possible.
// meant to be used when plaintext is expected, not formatted HTML.
func replaceUserReferencesWithNames(ctx context.Context, input []string, prefix string) []string {
// match and replace npup1 or nprofile1
ctx, cancel := context.WithTimeout(ctx, time.Second*3)
defer cancel()
for i, line := range input {
input[i] = strings.TrimSpace(
nostrNpubNprofileMatcher.ReplaceAllStringFunc(line, func(match string) string {
submatch := nostrNpubNprofileMatcher.FindStringSubmatch(match)
nip19code := submatch[1]
name, ok := getNameFromNip19(ctx, nip19code)
if ok {
return prefix + strings.ReplaceAll(name, " ", string(THIN_SPACE))
}
return nip19code[0:10] + "…" + nip19code[len(nip19code)-5:]
}),
)
}
return input
}
// replace nevent and note with their text, HTML-formatted
func renderQuotesAsHTML(ctx context.Context, input string, usingTelegramInstantView bool) string {
ctx, cancel := context.WithTimeout(ctx, time.Second*3)
defer cancel()
quotes := xsync.NewMapOf[string, string]()
wg := sync.WaitGroup{}
// first we run it without waiting for the results of getEvent() as they will be async
for _, submatches := range nostrNoteNeventMatcher.FindAllStringSubmatch(input, len(input)+1) {
nip19 := submatches[1]
ctx, cancel := context.WithTimeout(context.Background(), time.Second*4)
defer cancel()
wg.Add(1)
go func() {
event, _, err := getEvent(ctx, nip19, false)
if err == nil {
quotedEvent := basicFormatting(submatches[0], false, usingTelegramInstantView, false)
var content string
if event.Kind == 30023 {
content = mdToHTML(event.Content, usingTelegramInstantView)
} else {
content = basicFormatting(event.Content, false, usingTelegramInstantView, false)
}
content = fmt.Sprintf(
`<blockquote class="border-l-05rem border-l-strongpink border-solid"><div class="-ml-4 bg-gradient-to-r from-gray-100 dark:from-zinc-800 to-transparent mr-0 mt-0 mb-4 pl-4 pr-2 py-2">quoting %s </div> %s </blockquote>`, quotedEvent, content)
quotes.Store(submatches[0], content)
}
wg.Done()
}()
}
// in the second time now that we got all the quoted events we actually perform replacement
wg.Wait()
return nostrNoteNeventMatcher.ReplaceAllStringFunc(input, func(match string) string {
quote, ok := quotes.Load(match)
if !ok {
return match
}
return quote
})
}
func linkQuotes(input string) string {
return nostrNoteNeventMatcher.ReplaceAllStringFunc(input, func(match string) string {
nip19 := match[len("nostr:"):]
firstChars := nip19[:8]
lastChars := nip19[len(nip19)-4:]
return fmt.Sprintf(`<a href="/%s">%s</a>`, nip19, firstChars+"…"+lastChars)
})
}
func basicFormatting(input string, skipNostrEventLinks bool, usingTelegramInstantView bool, skipLinks bool) string {
nostrMatcher := nostrEveryMatcher
if skipNostrEventLinks {
nostrMatcher = nostrNpubNprofileMatcher
}
imageReplacementTemplate := ` <img src="%s"> `
videoReplacementTemplate := `<video controls width="100%%" class="max-h-[90vh] bg-neutral-300 dark:bg-zinc-700"><source src="%s"></video>`
if usingTelegramInstantView {
// telegram instant view doesn't like when there is an image inside a blockquote (like <p><img></p>)
// so we use this custom thing to stop all blockquotes before the images, print the images then
// start a new blockquote afterwards -- we do the same with the markdown renderer for <p> tags on mdToHtml
imageReplacementTemplate = "</blockquote>" + imageReplacementTemplate + "<blockquote>"
videoReplacementTemplate = "</blockquote>" + videoReplacementTemplate + "<blockquote>"
}
lines := strings.Split(input, "\n")
for i, line := range lines {
line = replaceURLsWithTags(line, imageReplacementTemplate, videoReplacementTemplate, skipLinks)
line = replaceNostrURLsWithHTMLTags(nostrMatcher, line)
lines[i] = line
}
return strings.Join(lines, "<br/>")
}
func previewNotesFormatting(input string) string {
lines := strings.Split(input, "\n")
var processedLines []string
for _, line := range lines {
if strings.TrimSpace(line) == "" {
continue
}
processedLine := shortenNostrURLs(line)
processedLines = append(processedLines, processedLine)
}
return strings.Join(processedLines, "<br/>")
}
func trimProtocolAndEndingSlash(relay string) string {
relay = strings.TrimPrefix(relay, "wss://")
relay = strings.TrimPrefix(relay, "ws://")
relay = strings.TrimPrefix(relay, "wss:/") // some browsers replace upfront '//' with '/'
relay = strings.TrimPrefix(relay, "ws:/") // some browsers replace upfront '//' with '/'
relay = strings.TrimSuffix(relay, "/")
return relay
}
func normalizeWebsiteURL(u string) string {
if strings.HasPrefix(u, "http") {
return u
}
return "https://" + u
}
func limitAt[V any](list []V, n int) []V {
if len(list) < n {
return list
}
return list[0:n]
}
func maxIndex(slice []int) int {
maxIndex := -1
maxVal := 0
for i, val := range slice {
if val > maxVal {
maxVal = val
maxIndex = i
}
}
return maxIndex
}
func getUTCOffset(loc *time.Location) string {
// Get the offset from UTC
_, offset := time.Now().In(loc).Zone()
// Calculate the offset in hours
offsetHours := offset / 3600
// Format the UTC offset string
sign := "+"
if offsetHours < 0 {
sign = "-"
offsetHours = -offsetHours
}
return fmt.Sprintf("UTC%s%d", sign, offsetHours)
}
func toJSONHTML(evt *nostr.Event) template.HTML {
if evt == nil {
return ""
}
tagsHTML := "["
for t, tag := range evt.Tags {
tagsHTML += "\n ["
for i, item := range tag {
cls := `"text-zinc-500 dark:text-zinc-50"`
if i == 0 {
cls = `"text-amber-500 dark:text-amber-200"`
}
tagsHTML += "\n <span class=" + cls + ">"
// if it's tagging another event, pubkey or address, make it a clickable link
linkCls := "underline underline-offset-4 text-amber-700 dark:text-amber-100 hover:text-amber-600 dark:hover:text-amber-200"
if i == 1 && tag[0] == "e" && nostr.IsValid32ByteHex(item) {
var relayHints []string
var authorHint string
if len(tag) > 2 {
relayHints = []string{tag[2]}
if len(tag) > 4 {
authorHint = tag[4]
}
}
nevent, _ := nip19.EncodeEvent(item, relayHints, authorHint)
tagsHTML += `<a class="` + linkCls + `" href="/` + nevent + `">"` + item + `"</a>`
} else if spl := strings.Split(item, ":"); i == 1 && tag[0] == "a" && len(spl) == 3 && nostr.IsValidPublicKey(spl[1]) {
var relayHints []string
if len(tag) > 2 {
relayHints = []string{tag[2]}
}
kind, _ := strconv.Atoi(spl[0])
naddr, _ := nip19.EncodeEntity(spl[1], kind, spl[2], relayHints)
tagsHTML += `<a class="` + linkCls + `" href="/` + naddr + `">"` + item + `"</a>`
} else if i == 1 && strings.ToLower(tag[0]) == "p" && nostr.IsValidPublicKey(item) {
var relayHints []string
if len(tag) > 2 {
relayHints = []string{tag[2]}
}
nprofile, _ := nip19.EncodeProfile(item, relayHints)
tagsHTML += `<a class="` + linkCls + `" href="/` + nprofile + `">"` + item + `"</a>`
} else {
// otherwise just print normally
itemJSON, _ := json.Marshal(item)
tagsHTML += html.EscapeString(string(itemJSON))
}
if i < len(tag)-1 {
tagsHTML += ","
} else {
tagsHTML += "\n "
}
}
tagsHTML += "]"
if t < len(evt.Tags)-1 {
tagsHTML += ","
} else {
tagsHTML += "\n "
}
}
tagsHTML += "]"
contentJSON, _ := json.Marshal(evt.Content)
keyCls := "text-purple-700 dark:text-purple-300"
return template.HTML(fmt.Sprintf(
`{
<span class="`+keyCls+`">"id":</span> <span class="text-zinc-500 dark:text-zinc-50">"%s"</span>,
<span class="`+keyCls+`">"pubkey":</span> <span class="text-zinc-500 dark:text-zinc-50">"%s"</span>,
<span class="`+keyCls+`">"created_at":</span> <span class="text-green-600">%d</span>,
<span class="`+keyCls+`">"kind":</span> <span class="text-amber-500 dark:text-amber-200">%d</span>,
<span class="`+keyCls+`">"tags":</span> %s,
<span class="`+keyCls+`">"content":</span> <span class="text-zinc-500 dark:text-zinc-50">%s</span>,
<span class="`+keyCls+`">"sig":</span> <span class="text-zinc-500 dark:text-zinc-50 content">"%s"</span>
}`, evt.ID, evt.PubKey, evt.CreatedAt, evt.Kind, tagsHTML, html.EscapeString(string(contentJSON)), evt.Sig),
)
}
func isValidShortcode(s string) bool {
for _, r := range s {
if !('a' <= r && r <= 'z' || 'A' <= r && r <= 'Z' || '0' <= r && r <= '9' || r == '_') {
return false
}
}
return true
}
func appendUnique[I comparable](arr []I, item ...I) []I {
for _, item := range item {
if slices.Contains(arr, item) {
return arr
}
arr = append(arr, item)
}
return arr
}