Personal Video Database

Hi,

I have some problems using the Advanced Search. I did create a test database with three movies

- one movie with German and English language
- one movie with Italian language
- one movie with German and Italian language

All movies:


If I try to define some condition in the advanced search they didn't work as I expected.

First screenshot is with the conditions to select only movies with German AND Italian language but PVD doesn't show any movie at all. I expected it would show the one movie which has Italian and German language.



Second screenshot is a select to NOT show movies with German language. PVD shows all movies instead of the one movie which has only Italian as language as I expected.



Any guess where are my faults defining the conditions? 

Thanks in advance



PS: System is XPSP3 with PVD 1.0.2.3 beta

Hi,

I tried the advanced search on a custom field (checkbox).

The Advanced Search box offers the following values (only -1 - take a look at the first screenshot).



If I press Ok PVD presents the following message box with the error described in the subject. I tested it with a new created database.



Looking in the firebird FAQ I found the following advice. Maybe it can help to resolve the mystery.

http://www.firebirdfaq.org/faq184/

This is 1.0.2.2 Beta

I didn't find an already reported bug for this error in the forum. If I shall deliver further information tell me.

Thanks

Hello,

I have some problems with a probable bug in the skin templates (xml). The bug occurs using comments in xml.
Putting comments before markup (not inside!) changes the behaviour of the tags in the visualization.
Until now I found one example with the tag videostreamlabel and another with the tag videodatalabel. Putting a comment before the former tag the comment changes its alignment in the visualization. Putting a comment before the latter one makes vanish completely the data which should be shown.

Included you find a working example and a not working example to compare the effect.
(Tip: delete first comment and see the effect then delete the second comment and see the second effect)

It would be nice if someone could verify this.

Thanks

Example (not working - comments at wrong place)
<?xml version="1.0" encoding="utf-8"?>
<xml>
    <pvd_skin_movie>
        <skinname>PVD Midnight bug</skinname>
        <version>0.9.0.0</version>
        <author>goddert</author>
        <description>Color skin</description>
        <preview>pvd_movie_midnight.png</preview>
        <!-- section with filepath, (quality, duration, size, mediatype, medialabel, medialocation), videostreamdata, audiostreamdata, subtitle and feature rows -->
<section>
<!-- row with videostreamdata column -->
<row>
<!-- column with videostreamdata expandable section -->
<column>
<!-- expandable section with videostreamdata rows -->
<expandsection>
<!-- row with videostreamlabel and videodatalabel columns -->
<row>
<!-- don't comment here; corrupts alignment of label -->
<column>
<videostreamlabel></videostreamlabel>
</column>
<!-- don't comment here; makes vanish data -->
<column>
<videodatalabel></videodatalabel>
</column>
</row>
<row>
<column>
<fieldlabel field="videocodec"></fieldlabel>
<videocodec/>
</column>
</row>
</expandsection>
</column>
</row>
</section>
    </pvd_skin_movie>
</xml>
Example (working - no comments at wrong place)
<?xml version="1.0" encoding="utf-8"?>
<xml>
    <pvd_skin_movie>
        <skinname>PVD Midnight nobug</skinname>
        <version>0.9.0.0</version>
        <author>goddert</author>
        <description>Color skin</description>
        <preview>pvd_movie_midnight.png</preview>
        <!-- section with filepath, (quality, duration, size, mediatype, medialabel, medialocation), videostreamdata, audiostreamdata, subtitle and feature rows -->
<section>
<!-- row with videostreamdata column -->
<row>
<!-- column with videostreamdata expandable section -->
<column>
<!-- expandable section with videostreamdata rows -->
<expandsection>
<!-- row with videostreamlabel and videodatalabel columns -->
<row>
<column>
<videostreamlabel></videostreamlabel>
</column>
<column>
<videodatalabel></videodatalabel>
</column>
</row>
<row>
<column>
<fieldlabel field="videocodec"></fieldlabel>
<videocodec/>
</column>
</row>
</expandsection>
</column>
</row>
</section>
    </pvd_skin_movie>
</xml>

Hello,

attached you'll find a link to a package of scripts and SQL-procedures which I developed to read video, audio and subtitle information from movie files and import these information into PVD with help of the database interface.

Motivation of this package was the current limitations of the mediainfo.dll and ffmpeg bundled with PVD.
Unfortunately ffmpeg doesn't read always the correct informations from a movie file (it isn't neither its scope). Also mediainfo.dll (being an older version) doesn't get always the correct info. For various reasons in the end I approached the problem using the database interface.

The package mainly uses SQL-procedures, mediainfo (command line version), awk and some script.
It is possible to process a great number of movie files in one run (I tested it with more than thousand files) and update the database.

I don't know whether someone is interested but I wanted to share it anyway. If you are, I invite you to read the included Readme.txt. Additionally all procedures and scripts are well documented.

Obviously if anyone has questions I'm here to answer them.

Thanks

Link to Mediainfo Import  package: http://www.mediafire.com/download.php?urkbmezi2z1wuk5

Hello,

I decided to put this request under feature suggestions as I think it could be useful for others and because I didn't find a satisfying workaround.

In the preferences for the file scanner exists the table "Find and replace". Currently, as I understood, this table replaces strings/characters in the file name before PDV it uses in the regular expressions and compares it to the various fields like origtitle, title, ecc.

A common problem is that the original title may contain special characters like '?', ':' etc., characters which are not allowed in filenames. Currently using the "Find and Replace" one must find a particular replace character for every not allowed special character. E.g. '?'<-'@' and ':'<-'%' to replace these in filenames with the corrisponding special character in the original title. E.g. "Wo is Fred@.mkv" becomes "Who is Fred?" ecc. This is a bit tedious and creates particular strange filenames. Unfortunately the presence of those special characters inside the various fields like origtitle, eptitle ecc. doesn't allow circumventing them in the usual regular expressions in the file scanner dialog.
If the table mentioned above allowed both directions of replacing we could replace special characters, e.g., in original title (obviously only in memory and not in the field itself) before comparing it to the filename. This way one could define a set of special characters which would be replaced with ONE other character or better with nothing. It would avoid curious filenames and a n->1 mapping of a replace string in a useful direction. E.g. (\?|:) would be replaced with "" (nothing). In the example above "Who is Fred?" would become "Who is Fred.mkv". I think a more elegant solution for a cumbersome problem.

If I missed something or one has a nice solution for the special character problem in original titles please tell me.

 

Thanks