Category: Tech

A maintenance release, 2.0.1, was released today for Sphider-PDO, Sphider PostgreSQL, and Sphider SQLite.

There were no functionality changes, but a number of corrections were made to fix coding errors in database error handling procedures.

New versions can be found on the Sphider Search Home download page, or accessed from the Downloads tab of this website.

The Sphider Help Forum has moved to https://forum.worldspaceflight.com.

All posts and users have been ported, so there is nothing you need to do except maybe change your bookmarks.

All links from this blog (hopefully) have been updated.

Work is in progress to automatically redirect to the new url.

Sphider – a PHP spider and search engine

While this blog will continue to provide news and information about Sphider, and links to downloads will continue to be provided from the blog, the principal home for Sphider is now:

http://www.sphidersearch.com or https://www.sphider.worldspaceflight.com. Either url will bring you to the same page. The Spider Forum has not moved and is accessible from the new domain.

Besides the main page, there is a downloads page, an About page, a document page from which the Sphider User’s Guide may be downloaded, and a changelog page. Other pages will be added as the need arises.

All the current releases of Sphider had a minor bug when doing an image search by url. The corrected code is available on the downloads page. The main Sphider 2.0 is designated by an “a” suffix. All of the PDO versions have a “c” suffix.

 

The ONLY file changed is search.php. And in search.php, there is only one line altered. A passed parameter “type” was having uppercase characters stripped. A column in the database image table “images” is “imgUrl”. The uppercase “U” was stripped and the query failed when it couldn’t find the column “imgrl”!

If you don’t use the “Image Search” your version will work fine. If you DO use the “Image Search”, the ONLY file you need to replace is search.php. There is no need to do a reinstall.

The embarrassing part of it all is that this problem WAS caught and corrected during testing prior to the 2.0 release. HOWEVER, that corrected piece of code wasn’t placed into the zip files, which shipped with the uncorrected version of search.php. 😳

A minor problem was found affecting the PDO versions (PDO, SQLite PDO, and PostgreSQL PDO) of Sphider.

During indexing, if the “Use site map” switch was set, but the site map was not found or not usable, the code to update the database to turn the switch off was failing to execute.

The code has been corrected to enable the database to update. The updated downloads are reflected as a “b” version.

The non-PDO version was unaffected. This was strictly a PDO issue.

Thanks go out to Webbo for the catch.

It has come to our attention there are typos in the code for all PDO versions. For the normal PDO (MySQL/MariaDB), spider.php and spiderfuncs.php have been slightly modified. Spider.php had a single typo. Sphiderfuncs.php was missing 5 lines of code. While the version number is unchanged, the new download is designated as 2.0.0a.

For the PostgreSQL and SQLite versions, only sphider.php contained a single typo each. No other files are affected. As with the regular PDO, the version is unchanged by the download designations are shown as 2.0.0a.

Our apologies for the inconvenience. During testing of all these versions. these anomalies were uncaught and thus it seems that, for the most part, crawling functionality was not adversely impacted, although it COULD be under certain circumstances.

Our deepest thanks go out to Ed Parrish for having caught these issues.

Sphider 2.0.0 has been released and may be obtained from the Downloads tab.

Sphider 2.0.0 is under going final testing and will be released probably by mid-October.

Virtually every file has gone at least some alteration. The features of Sphider 2.0.0 are:
– Better page charset handling to ensure that the database receives only UTF-8 input. UTF-8 encodeing of web pages already in UTF-8 format is avoided to eliminate garbled entries.
– Phrase searches have been improved.
– This version is PHP 7.1 ready.
– Integrated indexing of images, with the option to NOT index images. An image search page is also provided.
– RSS content may also be indexed and searched.
– Jquery has been updated to a more recent version.
– While not fully PSR-2 compliant when it comes to PHP coding standards, the code is a LOT closer than it ever has been. This involved the renaming of many functions, the elimination of a few functions which were found to be obsolete (and thus, unused). Coding style had to be changed virtually every module. This is why so much code has been altered, affecting nearly every Sphider PHP code segment.
– The search page is integrated for legacy, RSS, and image searches. Knowing that RSS and images are something not every user will be interested in, an updated (as in 2.0.x compliant) version of the 1.6.x search page is provided. The revised 1.6.x search form, it will work fine with 2.0.x. It will need to be renamed to replace the provided search.php.

Also, finding that porting PDO to databases other than MySQL was messier than anticipated (too many DB specific requirements for each), Sphider 2.0.0 will actually have 4 flavors. The “kits” for PostgreSQL and SQLite were too cumbersome and confusing.
1) The legacy Sphider, using the MySQL database (or MariaDB) and using MySQLi and MySQLnd.
2) PDO Sphider, also using the MySQL database (or MariaDB), but using a PDO implementation (for installations lacking MySQLnd support).
3) PostgreSQL version using a PostgresSQL database and accessed via PDO,
4) SQLite version, using a SQLite database accessed via PDO.

All flavors are testing well and it seems no more coding changes will be needed, after working out some “peculiarities” for each. Now each version must have a final full set of operations performed to ensure everything works. This includes new installation via PHP script, installation using SQL queries, upgrade installation, adding sites, indexing sites, deleting sites, adding, editing, and deleting categories. Also the same is done for RSS indexing. The search functions need to be tested for various situations. We have found a few websites which have, uh…., what you might call “unusual” methods resulting in unusual problems. (Ever seen an image “alt” tag with text running in excess of 1000 characters? We have!)

This is what an integrated search page may look like for a Sphider capable of not just indexing sites, but also images on those sites, plus the ability to index RSS feeds.

A bug was just discovered in PDO versions of Sphider that allowed duplicate keywords to be indexed. This has been corrected in a new release of the PDO version, Sphider 1.6.1-PDO. The issue does not exist in the “vanilla” Sphider 1.6.0.