python3 export_chats.py /path/to/extracted-export

python3 export_chats.py /path/to/conversations-015.json

python3 export_chats.py /path/to/export.zip

python3 export_chats.py /path/to/export -o another_directory

python3 chats/export_chats.py chats/conversations.json

python3 export_chats.py /path/to/export

python3 chats/export_chats.py /path/to/export -o chats/export_out

^([0-9]{4})-([0-9]{2})-([0-9]{2})_([0-9]{4})-.*\.md$

pandoc "$f" -f gfm-tex_math_dollars -t html

find chats/export_out -maxdepth 1 -type f -name '*.md' -print0 |
while IFS= read -r -d '' f; do
  bn=$(basename "$f")

  [[ $bn =~ ^([0-9]{4})-([0-9]{2})-([0-9]{2})_([0-9]{4})-(.*)-([[:alnum:]]{8})(--.*)?\.md$ ]] || {
    printf 'Skipped unexpected filename: %s\n' "$f" >&2
    continue
  }

  y=${BASH_REMATCH[1]}
  m=${BASH_REMATCH[2]}
  d=${BASH_REMATCH[3]}
  t=${BASH_REMATCH[4]}
  short_id=${BASH_REMATCH[6]}

  outdir="convos/$y/$m/$d"
  mkdir -p "$outdir"

  outfile="$outdir/${y}-${m}-${d}_${t}.php"

  # Preserve the old concise name unless it would overwrite another chat.
  if [[ -e $outfile ]]; then
    outfile="$outdir/${y}-${m}-${d}_${t}-${short_id}.php"
  fi

  pandoc "$f" -f gfm-tex_math_dollars -t html \
    -o "$outfile" \
    2> >(tee -a notes/pandoc_errors.log >&2)
done

cp file-1SFNvyaXUyy2NUYPX4oweB.dat 2016-04-11-neurologist-c.png

file file-1SFNvyaXUyy2NUYPX4oweB.dat

python3 - <<'PY'
import json
import os
import shutil
from pathlib import Path

source_dir = Path(".")
output_dir = Path("recovered_assets")
mapping_file = source_dir / "conversation_asset_file_names.json"

output_dir.mkdir(exist_ok=True)

with mapping_file.open("r", encoding="utf-8") as f:
    mapping = json.load(f)

recovered = 0
missing = 0
collisions = 0

for stored_name, original_name in mapping.items():
    source = source_dir / stored_name

    if not source.is_file():
        print(f"MISSING: {stored_name} -> {original_name}")
        missing += 1
        continue

    # Do not allow paths embedded in an original filename to escape the
    # recovery directory.
    safe_name = Path(original_name).name or stored_name
    destination = output_dir / safe_name

    if destination.exists():
        stem = destination.stem
        suffix = destination.suffix
        counter = 2

        while destination.exists():
            destination = output_dir / f"{stem}--{counter}{suffix}"
            counter += 1

        collisions += 1

    shutil.copy2(source, destination)
    recovered += 1

print(f"Recovered: {recovered}")
print(f"Missing source files: {missing}")
print(f"Renamed filename collisions: {collisions}")
print(f"Output directory: {output_dir.resolve()}")
PY

file file_*.dat file-*.dat

xxd -l 32 unknown-file.dat

python3 recover_assets.py /path/to/export \
    -o recovered_assets

python3 chats/recover_assets.py /path/to/extracted-export \
  -o chats/recovered_assets

cd chats
python3 recover_assets.py . -o recovered_assets

python3 recover_assets.py export.zip -o recovered_assets

<?php
// conversation-index.php v1.0 - Dynamically lists exported conversations by year, month, and day.

declare(strict_types=1);

$page_title = 'Conversation Archive';

$projectRoot = __DIR__;
$convosRoot = $projectRoot . DIRECTORY_SEPARATOR . 'convos';

/*
 * These includes follow the apparent existing page-component order.
 * Remove or rearrange them if your index.php uses a different order.
 */
$metaHeaders = $projectRoot . DIRECTORY_SEPARATOR . 'meta-headers.php';
$bannerFile = $projectRoot . DIRECTORY_SEPARATOR . 'banner.php';
$firstLayerFile = $projectRoot . DIRECTORY_SEPARATOR . 'firstlayer.php';
$footerFile = $projectRoot . DIRECTORY_SEPARATOR . 'footer.php';

/**
 * Escape text for HTML output.
 */
function htmlEscape(string $value): string
{
    return htmlspecialchars($value, ENT_QUOTES | ENT_SUBSTITUTE, 'UTF-8');
}

/**
 * Convert a relative filesystem path into a safe relative URL.
 */
function relativePathToUrl(string $relativePath): string
{
    $relativePath = str_replace(DIRECTORY_SEPARATOR, '/', $relativePath);
    $parts = explode('/', $relativePath);

    return implode('/', array_map('rawurlencode', $parts));
}

/**
 * Read enough of a generated conversation file to find its first heading.
 *
 * The regex pipeline may comment out the original H1 block, but the literal
 * H1 markup should still be present inside the comment and can be recovered.
 */
function extractConversationTitle(string $filePath): ?string
{
    $handle = @fopen($filePath, 'rb');

    if ($handle === false) {
        return null;
    }

    $sample = fread($handle, 131072);
    fclose($handle);

    if ($sample === false || $sample === '') {
        return null;
    }

    $sample = preg_replace('/^\xEF\xBB\xBF/', '', $sample) ?? $sample;

    if (preg_match('~<h1\b[^>]*>(.*?)</h1>~is', $sample, $match) === 1) {
        $title = strip_tags($match[1]);
        $title = html_entity_decode(
            $title,
            ENT_QUOTES | ENT_HTML5,
            'UTF-8'
        );
        $title = preg_replace('/\s+/u', ' ', $title) ?? $title;
        $title = trim($title);

        if ($title !== '') {
            return $title;
        }
    }

    /*
     * Secondary fallback: inspect the first meaningful source line.
     * This is deliberately conservative so PHP, comments, and raw HTML
     * scaffolding do not become titles.
     */
    $lines = preg_split('/\R/u', $sample, 30);

    if (is_array($lines)) {
        foreach ($lines as $line) {
            $line = trim($line);

            if (
                $line === ''
                || str_starts_with($line, '<?php')
                || str_starts_with($line, '<!--')
                || str_starts_with($line, '<!DOCTYPE')
                || str_starts_with($line, '<html')
                || str_starts_with($line, '<head')
                || str_starts_with($line, '<meta')
            ) {
                continue;
            }

            $candidate = strip_tags($line);
            $candidate = html_entity_decode(
                $candidate,
                ENT_QUOTES | ENT_HTML5,
                'UTF-8'
            );
            $candidate = preg_replace('/\s+/u', ' ', $candidate) ?? $candidate;
            $candidate = trim($candidate);

            if ($candidate !== '') {
                return $candidate;
            }
        }
    }

    return null;
}

/**
 * Extract a displayable date and time from the generated filename.
 *
 * Expected form:
 *     YYYY-MM-DD_HHMM.php
 *
 * Collision-safe variants such as:
 *     YYYY-MM-DD_HHMM-abcdefgh.php
 *
 * are also accepted.
 */
function parseConversationFilename(
    string $filename,
    string $year,
    string $month,
    string $day
): array {
    $result = [
        'sort_key' => $filename,
        'display_time' => null,
        'display_date' => $year . '-' . $month . '-' . $day,
    ];

    if (
        preg_match(
            '/^(\d{4})-(\d{2})-(\d{2})_(\d{2})(\d{2})(?:[-_].*)?\.php$/i',
            $filename,
            $match
        ) === 1
    ) {
        $result['sort_key'] =
            $match[1]
            . $match[2]
            . $match[3]
            . $match[4]
            . $match[5]
            . '-'
            . $filename;

        $result['display_date'] =
            $match[1] . '-' . $match[2] . '-' . $match[3];

        $result['display_time'] =
            $match[4] . ':' . $match[5];
    }

    return $result;
}

/**
 * Month name for headings.
 */
function monthName(string $month): string
{
    $number = (int) $month;

    if ($number < 1 || $number > 12) {
        return $month;
    }

    return date('F', mktime(0, 0, 0, $number, 1, 2000));
}

/*
 * Tree form:
 *
 * $conversations[year][month][day][] = [
 *     filename,
 *     title,
 *     href,
 *     sort_key,
 *     display_time,
 *     display_date
 * ];
 */
$conversations = [];
$unorganized = [];
$readErrors = [];

if (is_dir($convosRoot)) {
    $directory = new RecursiveDirectoryIterator(
        $convosRoot,
        FilesystemIterator::SKIP_DOTS
    );

    $iterator = new RecursiveIteratorIterator(
        $directory,
        RecursiveIteratorIterator::LEAVES_ONLY
    );

    foreach ($iterator as $entry) {
        if (!$entry instanceof SplFileInfo || !$entry->isFile()) {
            continue;
        }

        if (strtolower($entry->getExtension()) !== 'php') {
            continue;
        }

        $absolutePath = $entry->getPathname();

        $relativePath = substr(
            $absolutePath,
            strlen($convosRoot) + 1
        );

        if ($relativePath === false || $relativePath === '') {
            continue;
        }

        $relativePath = str_replace('\\', '/', $relativePath);
        $pathParts = explode('/', $relativePath);

        $title = extractConversationTitle($absolutePath);

        if ($title === null || $title === '') {
            $title = pathinfo($entry->getFilename(), PATHINFO_FILENAME);
            $readErrors[] = $relativePath;
        }

        $record = [
            'filename' => $entry->getFilename(),
            'title' => $title,
            'href' => 'convos/' . relativePathToUrl($relativePath),
            'sort_key' => $relativePath,
            'display_time' => null,
            'display_date' => null,
        ];

        /*
         * The existing export loop creates:
         *
         * convos/YYYY/MM/DD/filename.php
         */
        if (
            count($pathParts) >= 4
            && preg_match('/^\d{4}$/', $pathParts[0]) === 1
            && preg_match('/^\d{2}$/', $pathParts[1]) === 1
            && preg_match('/^\d{2}$/', $pathParts[2]) === 1
        ) {
            $year = $pathParts[0];
            $month = $pathParts[1];
            $day = $pathParts[2];

            $fileInfo = parseConversationFilename(
                $entry->getFilename(),
                $year,
                $month,
                $day
            );

            $record = array_merge($record, $fileInfo);
            $conversations[$year][$month][$day][] = $record;
        } else {
            /*
             * Do not discard unexpected PHP files. Completeness has priority,
             * so anything outside the expected structure appears separately.
             */
            $unorganized[] = $record;
        }
    }
}

/*
 * Sort all levels chronologically in ascending order.
 */
ksort($conversations, SORT_STRING);

foreach ($conversations as &$months) {
    ksort($months, SORT_STRING);

    foreach ($months as &$days) {
        ksort($days, SORT_STRING);

        foreach ($days as &$files) {
            usort(
                $files,
                static function (array $a, array $b): int {
                    return strnatcasecmp($a['sort_key'], $b['sort_key']);
                }
            );
        }

        unset($files);
    }

    unset($days);
}

unset($months);

usort(
    $unorganized,
    static function (array $a, array $b): int {
        return strnatcasecmp($a['sort_key'], $b['sort_key']);
    }
);

$totalConversations = 0;

foreach ($conversations as $months) {
    foreach ($months as $days) {
        foreach ($days as $files) {
            $totalConversations += count($files);
        }
    }
}

$totalConversations += count($unorganized);

/*
 * Existing site header components.
 */
if (is_file($metaHeaders)) {
    include $metaHeaders;
} else {
    ?>
<!DOCTYPE html>
<html lang="en">
<head>
    <meta charset="UTF-8">
    <meta
        name="viewport"
        content="width=device-width, initial-scale=1"
    >
    <title><?= htmlEscape($page_title) ?></title>
    <link rel="stylesheet" href="prime-style.css">
</head>
<body>
    <?php
}

if (is_file($bannerFile)) {
    include $bannerFile;
}

if (is_file($firstLayerFile)) {
    include $firstLayerFile;
}
?>

<main>
    <h1><?= htmlEscape($page_title) ?></h1>

    <p>
        <?= number_format($totalConversations) ?>
        stored conversation<?= $totalConversations === 1 ? '' : 's' ?> found.
    </p>

    <?php if (!is_dir($convosRoot)): ?>
        <p>
            The conversation directory was not found:
            <code><?= htmlEscape($convosRoot) ?></code>
        </p>

    <?php elseif ($totalConversations === 0): ?>
        <p>No conversation PHP files were found.</p>

    <?php else: ?>

        <?php foreach ($conversations as $year => $months): ?>
            <section class="conversation-year">
                <h2><?= htmlEscape($year) ?></h2>

                <?php foreach ($months as $month => $days): ?>
                    <section class="conversation-month">
                        <h3>
                            <?= htmlEscape(monthName($month)) ?>
                            <?= htmlEscape($year) ?>
                        </h3>

                        <?php foreach ($days as $day => $files): ?>
                            <section class="conversation-day">
                                <h4>
                                    <?= htmlEscape(
                                        $year . '-' . $month . '-' . $day
                                    ) ?>
                                </h4>

                                <ul>
                                    <?php foreach ($files as $conversation): ?>
                                        <li>
                                            <a
                                                href="<?= htmlEscape(
                                                    $conversation['href']
                                                ) ?>"
                                            >
                                                <?php if (
                                                    $conversation['display_time']
                                                    !== null
                                                ): ?>
                                                    <?= htmlEscape(
                                                        $conversation[
                                                            'display_time'
                                                        ]
                                                    ) ?>
                                                    &mdash;
                                                <?php endif; ?>

                                                <?= htmlEscape(
                                                    $conversation['title']
                                                ) ?>
                                            </a>
                                        </li>
                                    <?php endforeach; ?>
                                </ul>
                            </section>
                        <?php endforeach; ?>
                    </section>
                <?php endforeach; ?>
            </section>
        <?php endforeach; ?>

        <?php if ($unorganized !== []): ?>
            <section class="conversation-unorganized">
                <h2>Files outside the expected date structure</h2>

                <p>
                    These files were included for completeness but were not
                    located under <code>convos/YYYY/MM/DD/</code>.
                </p>

                <ul>
                    <?php foreach ($unorganized as $conversation): ?>
                        <li>
                            <a
                                href="<?= htmlEscape(
                                    $conversation['href']
                                ) ?>"
                            >
                                <?= htmlEscape($conversation['title']) ?>
                            </a>

                            <small>
                                <?= htmlEscape($conversation['sort_key']) ?>
                            </small>
                        </li>
                    <?php endforeach; ?>
                </ul>
            </section>
        <?php endif; ?>

    <?php endif; ?>

    <?php if ($readErrors !== []): ?>
        <details>
            <summary>
                <?= count($readErrors) ?>
                file<?= count($readErrors) === 1 ? '' : 's' ?>
                had no readable H1 title
            </summary>

            <p>
                Their filenames were used as link text instead.
            </p>

            <ul>
                <?php foreach ($readErrors as $relativePath): ?>
                    <li><code><?= htmlEscape($relativePath) ?></code></li>
                <?php endforeach; ?>
            </ul>
        </details>
    <?php endif; ?>
</main>

<?php
if (is_file($footerFile)) {
    include $footerFile;
} elseif (!is_file($metaHeaders)) {
    ?>
</body>
</html>
    <?php
}

<a href="conversation-index.php">Browse stored conversations</a>

<h1>Morosyne Secretary-Coach</h1>

<body>
  <?php include 'banner.php'; ?>
  <?php include 'firstlayer.php'; ?>
  <?php include 'conversation-list.php'; ?>
</body>

<!-- v1.0.2 - added conversation archive listing -->
<!DOCTYPE html>
<html lang="en">
<?php include 'meta-headers.php'; ?>
<body>
  <?php include 'banner.php'; ?>
  <?php include 'firstlayer.php'; ?>
  <?php include 'conversation-list.php'; ?>
</body>
<?php include 'footer.php'; ?>
</html>

<!DOCTYPE html>
<html>
<head>
<body>

<?php
$convoList = __DIR__ . '/conversation-list.php';

if (is_file($convoList)) {
    include $convoList;
} else {
    echo '<p>(conversation list module not found)</p>';
}
?>

<?php include 'conversation-list.php'; ?>

<?php include 'conversation-list.php'; ?>

php -l conversation-list.php

php conversation-list.php > /tmp/conversation-list-test.html

ls -lh /tmp/conversation-list-test.html
head -n 20 /tmp/conversation-list-test.html

ini_set('display_errors', '1');
error_reporting(E_ALL);

sudo tail -n 50 /var/log/apache2/error.log

sudo journalctl -u php-fpm -n 50

declare(strict_types=1);
str_starts_with(...)

php --version

find convos -type f -name '*.php' | wc -l

function convoListEscape(string $value): string

function convoListEscape($value): string
{
    return htmlspecialchars(
        (string) $value,
        ENT_QUOTES | ENT_SUBSTITUTE,
        'UTF-8'
    );
}

function convoListMonthName($month): string
{
    $monthText = (string) $month;
    $number = (int) $month;

    if ($number < 1 || $number > 12) {
        return $monthText;
    }

    return date('F', mktime(0, 0, 0, $number, 1, 2000));
}

ksort($conversations, SORT_STRING);

krsort($conversations, SORT_STRING);

ksort($months, SORT_STRING);

ksort($days, SORT_STRING);

krsort($months, SORT_STRING);

krsort($days, SORT_STRING);

return strnatcasecmp($a['sort_key'], $b['sort_key']);

return strnatcasecmp($b['sort_key'], $a['sort_key']);