cf7-spam-to-telegram/includes/class-message-formatter.php

<?php
declare(strict_types=1);

namespace Cf7stg;

final class MessageFormatter
{
    private const TG_MAX_LEN = 4096;

    private const LABEL_MAP = [
        'client'  => ['emoji' => '🟢', 'text' => 'похоже клиент'],
        'unclear' => ['emoji' => '🟡', 'text' => 'неясно'],
        'spam'    => ['emoji' => '🔴', 'text' => 'похоже спам'],
        'drop'    => ['emoji' => '⛔', 'text' => 'авто-дроп'],
    ];

    /**
     * @param array{
     *   fields:array<string,mixed>,
     *   classification:array{score:int,label:string,reasons:array},
     *   site_title:string,
     *   form_title:string,
     *   submitted_at:string,
     *   reason:string,
     *   ip:string,
     *   user_agent:string,
     *   is_retro:bool,
     *   dry_run_drop?:bool
     * } $ctx
     *
     * @return array{text:string,reply_markup:array|null}
     */
    public static function build(array $ctx): array
    {
        $fields = $ctx['fields'];
        $class = $ctx['classification'];

        $phone = PhoneParser::parse(self::flatten_fields_for_phone($fields));
        $name = self::pick($fields, ['your-name', 'name', 'имя', 'fio']);
        if ($name === '') $name = FieldDetector::find_name($fields);
        $email = self::pick($fields, ['your-email', 'email', 'e-mail', 'mail']);
        if ($email === '') $email = FieldDetector::find_email($fields);
        $message = self::pick($fields, ['your-message', 'message', 'comment', 'сообщение', 'вопрос']);

        $lines = [];

        $is_dry_run_drop = !empty($ctx['dry_run_drop']);

        if ($is_dry_run_drop) {
            $header = '[БЫЛО БЫ DROPPED]';
        } else {
            $label_info = self::LABEL_MAP[$class['label']] ?? self::LABEL_MAP['unclear'];
            $header = ($ctx['is_retro'] ? '📂 Ретроспектива  ·  ' : '')
                . $label_info['emoji'] . ' ' . $label_info['text'];
        }

        $lines[] = $header;
        $lines[] = '🌐 ' . esc_html($ctx['site_title']) . '  ·  форма «' . esc_html($ctx['form_title']) . '»';
        $lines[] = '🗓 ' . esc_html($ctx['submitted_at']);
        $lines[] = '';
        $lines[] = '📱 ' . ($phone !== null ? esc_html($phone['pretty']) : '(не указан)');
        $lines[] = '👤 ' . ($name !== '' ? esc_html($name) : '(не указано)');
        $lines[] = '📧 ' . ($email !== '' ? esc_html($email) : '(не указан)');
        $lines[] = '';
        $lines[] = '💬 ' . ($message !== '' ? esc_html($message) : '(пусто)');
        $lines[] = '';
        $lines[] = '⚙️ Попало в спам: ' . esc_html($ctx['reason']);
        $lines[] = '📊 Классификация: ' . self::format_reasons($class['reasons']);

        if ($is_dry_run_drop) {
            $rule = '';
            foreach ($class['reasons'] as $r) {
                $text = (string)($r['text'] ?? '');
                if (strpos($text, 'hard-rule:') === 0 || strpos($text, 'score ≤') === 0) {
                    $rule = $text;
                    break;
                }
            }
            if ($rule !== '') $lines[] = '🚫 Сработавшее правило: ' . esc_html($rule);
        }

        $meta_bits = [];
        if ($ctx['ip'] !== '')         $meta_bits[] = 'IP ' . esc_html($ctx['ip']);
        if ($ctx['user_agent'] !== '') $meta_bits[] = 'UA ' . esc_html(self::shorten_ua($ctx['user_agent']));
        if ($meta_bits)                 $lines[] = '🌍 ' . implode(' · ', $meta_bits);

        $utms = self::extract_utms($fields);
        if ($utms !== '')               $lines[] = '🏷 ' . $utms;

        $other = self::extract_other_fields($fields);
        if ($other !== '')              $lines[] = '📋 ' . $other;

        $text = implode("\n", $lines);
        $text = self::truncate($text);

        // Telegram Bot API rejects tel: URLs in inline_keyboard with HTTP 400
        // ("Wrong port number specified"), so no inline button is emitted.
        // The phone number is already shown in the message body (line '📱 ...')
        // and mobile Telegram clients make such numbers tappable automatically.
        return ['text' => $text, 'reply_markup' => null];
    }

    private static function pick(array $fields, array $keys): string
    {
        foreach ($keys as $k) {
            if (isset($fields[$k])) {
                $v = $fields[$k];
                if (is_array($v)) $v = implode(' ', $v);
                return trim((string)$v);
            }
        }
        return '';
    }

    private static function flatten_fields_for_phone(array $fields): string
    {
        $out = [];
        foreach ($fields as $k => $v) {
            if (is_array($v)) $v = implode(' ', array_map('strval', $v));
            $out[] = (string)$v;
        }
        return implode("\n", $out);
    }

    private static function format_reasons(array $reasons): string
    {
        if (!$reasons) return '—';
        $parts = [];
        foreach ($reasons as $r) {
            // Drop reasons use sign='×' weight=0 as a sentinel — render text only
            // to avoid the meaningless "×0 hard-rule: ..." prefix.
            if (($r['sign'] ?? '') === '×' && (int)($r['weight'] ?? 0) === 0) {
                $parts[] = (string)$r['text'];
                continue;
            }
            $parts[] = $r['sign'] . $r['weight'] . ' ' . $r['text'];
        }
        return esc_html(implode(', ', $parts));
    }

    private static function shorten_ua(string $ua): string
    {
        $ua = preg_replace('/\s+/', ' ', $ua) ?? $ua;
        if (mb_strlen($ua) > 80) $ua = mb_substr($ua, 0, 80) . '…';
        return $ua;
    }

    private static function extract_utms(array $fields): string
    {
        $keys = ['utm_source', 'utm_medium', 'utm_campaign', 'utm_term', 'utm_content'];
        $parts = [];
        foreach ($keys as $k) {
            if (!empty($fields[$k])) {
                $v = is_array($fields[$k]) ? implode(' ', $fields[$k]) : (string)$fields[$k];
                $parts[] = esc_html($k) . '=' . esc_html($v);
            }
        }
        return implode(' · ', $parts);
    }

    private static function extract_other_fields(array $fields): string
    {
        $known = [
            'your-name', 'name', 'имя', 'fio',
            'your-email', 'email', 'e-mail', 'mail',
            'your-tel', 'tel', 'phone', 'телефон',
            'your-message', 'message', 'comment', 'сообщение', 'вопрос',
            'utm_source', 'utm_medium', 'utm_campaign', 'utm_term', 'utm_content',
            '_wpcf7', '_wpcf7_version', '_wpcf7_locale', '_wpcf7_unit_tag',
            '_wpcf7_container_post', '_wpcf7_posted_data_hash', 'g-recaptcha-response',
        ];
        $parts = [];
        foreach ($fields as $k => $v) {
            if (in_array($k, $known, true)) continue;
            if (strpos($k, 'wpcf7') === 0) continue;
            if (is_array($v)) $v = implode(', ', array_map('strval', $v));
            $v = trim((string)$v);
            if ($v === '') continue;
            $parts[] = esc_html($k) . ': ' . esc_html($v);
        }
        return implode(' · ', $parts);
    }

    private static function truncate(string $text): string
    {
        if (mb_strlen($text) <= self::TG_MAX_LEN) return $text;
        $suffix = "\n…[обрезано]";
        $cut = self::TG_MAX_LEN - mb_strlen($suffix);
        return mb_substr($text, 0, $cut) . $suffix;
    }
}