I installed Blogg X on my home computer (latest version - am using the windows version.) and added the mambot to my site (latest version). I tried to confirm my site info but no feedback was provided. I then tried to check my site to see if it was OK. Pinged, but timed out. Got message from Blogg X that I was not connected to the internet etc. Not true because I could access other sites.
Removed my site from Blogg X on my computer, still could not connect. Uninstalled the application, still cannot connect to my site...
I looked at the blogg-x mambot for multilanguage support, but I can't do very much with it without access to the user interface in Blogg-x. The metadata has a much better chance of succes, if you can extract the text automatically. If Chris is OK, I'd like to take at shot at it to try to change the mambot to automatically extract metadata, but I need some idea's: - Do I only extract metadata for a new article, and don't update for updated articles, or do I update the metadata each time the article is updated (I'll probably put a switch in the configuration file) - I need some suggestions for what to generate and where, I propose a configuration file on the server in the Blogg-x component directory with (personalized) "standard" keywords and description data that are "always" added, and some rules for adding article-specific metadata. - Presumably, I extract the author and the title, but what else? anybody have a suggestion on how to format all this? - I would imagine something like a section of "text pairs" in the config file, where if the first text is contained in the article, the second text (defaults to the first) is included in the metadata. Is this useful?
Hope you all have some suggestions, keep in mind that I can't touch the user interface, as I don't have the source code. This means that choices will always be site-specific (out of the configuration file), but of course you can "update" the configuration file with FTP if you want to.
I'll post the result somewhere here, just don't expect heroic efforts during the holidays...
some more explanations to make Blogg-x JoomFish-enabled. Here is how JoomFish works: it uses 3 tables in the database (pfx is the table prefix for the Joomla installation): 1)pfx_jf_tableinfo You don't need to concern yourself with this one, it keeps per table the name of the primary key, which for your tables is always the "id" field. 2)pfx_languages This keeps the user-defined languages on the site. Here is the structure:
CREATE TABLE `pfx_languages` ( `id` int(11) NOT NULL auto_increment, `name` varchar(100) NOT NULL default '', `active` tinyint(1) NOT NULL default '0', `iso` varchar(20) default NULL, `code` varchar(20) NOT NULL default '', `shortcode` varchar(20) default NULL, `image` varchar(100) default NULL, `fallback_code` varchar(20) NOT NULL default '', `params` text, `ordering` int(11) NOT NULL default '0', PRIMARY KEY (`id`) )
"id" is the language key The "name" displays the language name, "iso" the ISO-code for the language (en=english, fr=french, de=german, nl=dutch etc) "active" is set to 1 if the language is available in the front-end, code is the internal name (english, french, german, dutch) "shortcode" is user-defined name (on my site, identical to the code), "image" is probably meant for a flag-icon to be displayed in the language-choose module, "fallback_code" is the ordering-value to use if the content is not translated (display the article in english if the german version is not present), "params" is not used to my knowledge, and "ordering" is the order in which to display the languages in the language-choose module.
You will need this one to define the language in blogg-x, and retrieve the language ID. You will probably only need the fields "name" and "id".
3)pfx_jf_content
This is where the translations are kept:
CREATE TABLE `pfx_jf_content` ( `id` int(10) unsigned NOT NULL auto_increment, `language_id` int(11) NOT NULL default '0', `reference_id` int(11) NOT NULL default '0', `reference_table` varchar(100) NOT NULL default '', `reference_field` varchar(100) NOT NULL default '', `value` text NOT NULL, `original_value` varchar(255) default NULL, `original_text` text, `modified` datetime NOT NULL default '0000-00-00 00:00:00', `modified_by` int(11) unsigned NOT NULL default '0', `published` tinyint(1) unsigned NOT NULL default '0', PRIMARY KEY (`id`), KEY `jfCreateContentSQL` (`reference_id`,`reference_field`,`reference_table`,`language_id`,`published`), KEY `jfContent` (`language_id`,`reference_id`,`reference_table`,`reference_field`), KEY `jfReferenceTable` (`reference_table`) ) TYPE=MyISAM AUTO_INCREMENT=547 ;
"id" is the record-ID (auto-increment) "language_id" refers to the id field in pfx_languages, defines the language of this translation "reference_id" refers to the id field of the record of the original element (record in pfx_content for example) "reference_table" is the table where the original element resides (you will probably filter on WHERE 'reference_table' = 'content') "reference_field" is the field that is translated, one content article has a lot of records where the reference_id and reference_table are the same, and different values for reference_field like 'title', 'intro_text', 'full_text' etc. "value" is the translation "original_value" is a hex string, don't know where it is used for "original_text" is either empty or NULL in my tables, not used probably "modified" and "modified by" is the modification time&authorID for the translation "published" is a flag, 1 if the translation is published. ---------------------------------------------------------------------------------- To implement this, you will need to adapt the mambot blogg-x_server.php script.
1) function get_startup_settings($m) => you retrieve unpublished posts from content table, this should have a "if the language is not the default language" block added, where unpublished translations are retrieved.
To get all the titles of all untranslated articles, here's the SQL query ("Fourth query to get unpublished posts"), assuming you check on whether the language is different from the default language:
$num_unpublished_qry = "SELECT c.title FROM #__jf_content j right outer join #__content c ON j.reference_id=c.id AND j.reference_table='content' AND j.reference_field='title' AND j.language_id=$ACTUAL_language_id WHERE j.reference_id IS NULL")
2) function getpostdata => you should probably add a "function getpostdata_fish" for retrieving the data from the jf_content table. Instead of retrieving one record and storing it in the array, here you will have to retrieve multiple records (with same reference_id and reference_table), and store the value of each record in an array element, depending on the reference_field value.
Here, the SQL query is somewhat more difficult, as you have to retrieve translations from pfx_fj_content, but some other things from pfx_content. Probably better to keep what you have, then retrieve the translations, and replace the fields from the original with the translated "value" wherever they exist. To retrieve all translated fields: $post_data_qry = "SELECT value, reference_field FROM `#__jf_content` WHERE `reference_table`='content' AND `language_id` = $ACTUAL_language_id AND `reference_id` = $contentId";
That's it for the mambot part I think, as get_single function etc. uses things like "$queryOne = $queryOnePartOne . $mosConfig_dbprefix . $queryOnePartTwo;" and therefore the table name except for the prefix is defined in the blogg-x program.
I can't help you on that score, as I don't have the source code.
I think most of us would be very happy if only the text can be modified language-dependant. In JoomFish in the frontend, you can also load different images per translation, you translate all the section and category names etc., but I at least would be more than happy if all those stayed as in the original language.
I hope I didn't make to many mistakes, and I especially hope this helps you.
Blogg-x is great! I would very much appreciate JoomFish functionality added, as I have some people translating articles, but currently doing it in notepad, and copying the text in the Joomla editor screen, because they have a bad internet connection and it keeps dropping (which makes them lose the work they did)
Chris, JoomFish adds a "language chooser" dropdownlist on each page (well, you can configure that, it's just a module), where you can switch languages. When you switch from the default language, you get your article copied from the [prefix]_content table to the [prefix]_jf_content table which has extra things like "language_id" field and "original_text" field. The article gets copied for each language you add. So if you retrieve the available languages from [prefix]_languages (name and id), add a dropdown on the blogg-x interface and fill it with the name values, then if you retrieve and article and it doesn't exist in the [prefix]_jf_content table, get the original article from [prefix]_content, display it in Blogg_x, and when you repost it, save it in the [prefix]_jf_content table with the language id from your combobox. Maybe there is an easier way through the JoomFish API, will look into this, but I have no knowledge of PHP and little of Java (sorry, C# addict), so will be slow going trying to understand.
Anything I can do to help with this? If I can save you some time by sorting out some things, please don't hesitate to ask, I am breathlessly awaiting this functionality.
This has streamlined my content editing in the greatest way!!!! I have many websites that I contribute to and, in most cases, the articles that I submit are the same on each site. It would be fantastic if there was a button for submitting to all sites at once. Perhaps you could assign a default section and category for each site so that way the program will place the articles in all the sites without all the cutting and pasting. Or, if it's easier, perhaps there could be a way for you to just switch between sites without the content in the editor going blank... that way you could just change the site, section and content and then choose publish without having to cut and paste. Just a thought....
Without that, the program is still worth its weight in gold!! And it's pretty heavy!!
Also check and make sure your mambot directory is writable and executable by the web server (chmod 755 - check all the boxes except for WRITE - Group & Other.)
