Refactor bookmark import using a generic Netscape parser

Relates to #607
Relates to #608
Relates to #493 (abandoned)

Additions:
- use Composer's autoload to load 3rd-party dependencies under vendor/

Modifications:
- [import] replace the current parser with a generic, stable parser
  - move code to application/NetscapeBookmarkUtils
  - improve status report after parsing
- [router] use the same endpoint for both bookmark upload and import dialog
- [template] update bookmark import options
  - allow adding tags to all imported links
  - allow selecting the visibility (privacy) of imported links
- [tests] ensure bookmarks are properly parsed and imported in the LinkDB
  - reuse reference input from the parser's test data

See:
- https://github.com/shaarli/netscape-bookmark-parser
- https://getcomposer.org/doc/01-basic-usage.md#autoloading

Signed-off-by: VirtualTam <virtualtam@flibidi.net>

1 files changed, 142 insertions, 0 deletions

diff --git a/application/NetscapeBookmarkUtils.php b/application/NetscapeBookmarkUtils.php
index fdbb0ad7..b99a432e 100644
--- a/application/NetscapeBookmarkUtils.php
+++ b/application/NetscapeBookmarkUtils.php
@@ -51,4 +51,146 @@ class NetscapeBookmarkUtils
 
         return $bookmarkLinks;
     }
+
+    /**
+     * Generates an import status summary
+     *
+     * @param string $filename       name of the file to import
+     * @param int    $filesize       size of the file to import
+     * @param int    $importCount    how many links were imported
+     * @param int    $overwriteCount how many links were overwritten
+     * @param int    $skipCount      how many links were skipped
+     *
+     * @return string Summary of the bookmark import status
+     */
+    private static function importStatus(
+        $filename,
+        $filesize,
+        $importCount=0,
+        $overwriteCount=0,
+        $skipCount=0
+    )
+    {
+        $status = 'File '.$filename.' ('.$filesize.' bytes) ';
+        if ($importCount == 0 && $overwriteCount == 0 && $skipCount == 0) {
+            $status .= 'has an unknown file format. Nothing was imported.';
+        } else {
+            $status .= 'was successfully processed: '.$importCount.' links imported, ';
+            $status .= $overwriteCount.' links overwritten, ';
+            $status .= $skipCount.' links skipped.';
+        }
+        return $status;
+    }
+
+    /**
+     * Imports Web bookmarks from an uploaded Netscape bookmark dump
+     *
+     * @param array  $post      Server $_POST parameters
+     * @param array  $file      Server $_FILES parameters
+     * @param LinkDB $linkDb    Loaded LinkDB instance
+     * @param string $pagecache Page cache
+     *
+     * @return string Summary of the bookmark import status
+     */
+    public static function import($post, $files, $linkDb, $pagecache)
+    {
+        $filename = $files['filetoupload']['name'];
+        $filesize = $files['filetoupload']['size'];
+        $data = file_get_contents($files['filetoupload']['tmp_name']);
+
+        // Sniff file type
+        if (! startsWith($data, '<!DOCTYPE NETSCAPE-Bookmark-file-1>')) {
+            return self::importStatus($filename, $filesize);
+        }
+
+        // Overwrite existing links?
+        $overwrite = ! empty($post['overwrite']);
+
+        // Add tags to all imported links?
+        if (empty($post['default_tags'])) {
+            $defaultTags = array();
+        } else {
+            $defaultTags = preg_split(
+                '/[\s,]+/',
+                escape($post['default_tags'])
+            );
+        }
+
+        // links are imported as public by default
+        $defaultPrivacy = 0;
+
+        $parser = new NetscapeBookmarkParser(
+            true,                       // nested tag support
+            $defaultTags,               // additional user-specified tags
+            strval(1 - $defaultPrivacy) // defaultPub = 1 - defaultPrivacy
+        );
+        $bookmarks = $parser->parseString($data);
+
+        $importCount = 0;
+        $overwriteCount = 0;
+        $skipCount = 0;
+
+        foreach ($bookmarks as $bkm) {
+            $private = $defaultPrivacy;
+            if (empty($post['privacy']) || $post['privacy'] == 'default') {
+                // use value from the imported file
+                $private = $bkm['pub'] == '1' ? 0 : 1;
+            } else if ($post['privacy'] == 'private') {
+                // all imported links are private
+                $private = 1;
+            } else if ($post['privacy'] == 'public') {
+                // all imported links are public
+                $private = 0;
+            }                
+
+            $newLink = array(
+                'title' => $bkm['title'],
+                'url' => $bkm['uri'],
+                'description' => $bkm['note'],
+                'private' => $private,
+                'linkdate'=> '',
+                'tags' => $bkm['tags']
+            );
+
+            $existingLink = $linkDb->getLinkFromUrl($bkm['uri']);
+
+            if ($existingLink !== false) {
+                if ($overwrite === false) {
+                    // Do not overwrite an existing link
+                    $skipCount++;
+                    continue;
+                }
+
+                // Overwrite an existing link, keep its date
+                $newLink['linkdate'] = $existingLink['linkdate'];
+                $linkDb[$existingLink['linkdate']] = $newLink;
+                $importCount++;
+                $overwriteCount++;
+                continue;
+            }
+
+            // Add a new link
+            $newLinkDate = new DateTime('@'.strval($bkm['time']));
+            while (!empty($linkDb[$newLinkDate->format(LinkDB::LINK_DATE_FORMAT)])) {
+                // Ensure the date/time is not already used
+                // - this hack is necessary as the date/time acts as a primary key
+                // - apply 1 second increments until an unused index is found
+                // See https://github.com/shaarli/Shaarli/issues/351
+                $newLinkDate->add(new DateInterval('PT1S'));
+            }
+            $linkDbDate = $newLinkDate->format(LinkDB::LINK_DATE_FORMAT);
+            $newLink['linkdate'] = $linkDbDate;
+            $linkDb[$linkDbDate] = $newLink;
+            $importCount++;
+        }
+
+        $linkDb->savedb($pagecache);
+        return self::importStatus(
+            $filename,
+            $filesize,
+            $importCount,
+            $overwriteCount,
+            $skipCount
+        );
+    }
 }