PDA

View Full Version : SubHub, a post-OCR correction tool for Srt files, 0.6.0 released


awx
26th May 2008, 04:35
SubHub is a program for dealing with errors in Srt files. At its core, SubHub will read Srt subtitle files into RAM and allow you to run regex queries against the text. Then you can edit the lines to fix any errors you see and resave the file to disk.

The UI is geared for people who want to make corrections to a 100+ files at a time. If you only want to change one file then the UI might seem confusing.

SubHub can be installed and automatically updated directly from the SourceForge website.
Installer with auto-updater:
setup.exe (https://sourceforge.net/projects/subhub/files/subhub/setup.exe/download)

zip file releases:
http://sourceforge.net/project/showfiles.php?group_id=225616

Screenshot:
http://img35.imageshack.us/img35/2640/subhub20090601.png

This software is released under the GPL license. The source code is hosted at SourceForge under the name subhub (http://sourceforge.net/projects/subhub/).

SubHub is written in C# for .NET 4.0. It uses WPF for the GUI and LINQ for much of the searching. Unfortunately, it won't run under the current version of Mono.

Requirements:
.NET 4.0

WARNING:
THIS IS PRE-RELEASE SOFTWARE THAT MAY CONTAIN BUGS. MAKE BACKUPS OF YOUR DATA!

Version history:
Completed in version 0.6.0
* Compiled against .Net 4.0

Completed in version 0.5.3:
* Enabled ClickOnce installer and automatic updater at <http://subhub.sourceforge.net/clickonce/publish.htm>
* Updated application icon with large Vista sizes.

Completed in version 0.5.2:
* Added a non-free Yellow Submarine icon. This icon cannot be included in source files due to license.
* Allow user to specify durations for "too long" and "too short" queries.

Completed in version 0.5.1:
* Bug: DirTreeNode Icon does not update to reflect changes (files added, removed).
* Html Export: use Css classes rather than inline styles for formatting
* Quick searches are no longer case sensitive.
* Dialog added for Revert item function so that user can choose to revert text or timecodes independently.

Completed in version 0.5.0:
* Add AddedItems list to SubtitleClass
* Add Insert Item above/below functions.
* Add an "All Items" query to the Tree Browser.
* Add Delete item function.
* Allow user to open a file with an alternate encoding.
* Bulk language change feature added.
* Subtitle files can now be renamed.
* Added a revert function for subitems
* Made Open Files background thread multithreaded (1 thread per cpu core).
* Item Editor now has native platform line endings, allowing better copy/paste.
* Disable SubtitleEditor when nothing is selected.
* save window size, position and state between launches.
* Search Categories treeview now remembers which major branches were expanded
* Default window size now netbook friendly
* Add an About box.
* Open File and Folder dialogs now remember the last directory opened.

awx
1st June 2008, 08:45
The second release is now available.

Mostly bug fixes and some UI changes.
All queries and file saving are run on a background thread now.
Editing files on a network share should be much faster.

GrofLuigi
1st June 2008, 16:09
Sounds nice, but .Net 35 is a showstopper for me.

My computers(s) are old and weak and not ready to take such a blow. :)

GL

awx
1st June 2008, 17:13
.Net 3.5 is probably the least of your worries. Wpf might make your system crawl.

awx
2nd June 2008, 22:00
Version 0.3.0 adds support for running regex replacement patterns against the text.

Subtitle Text Queries with the lightening bolt icon have a replacement pattern that you can run. After you've selected a text query from the treeview, right-clicking on a file or an individual item will give you the menu item "Apply Regex Replacement Pattern...".

You can fix problems in hundreds of Srt files in a few seconds if you write some good Regex patterns. I've written several for L to I fixes that are useful for English subs.

http://img104.imageshack.us/img104/3126/subhubscreenshot2008060hp0.png (http://imageshack.us)

asarian
5th June 2008, 14:22
Regex-es are cute. But can it globally change/adjust time indices, too? Like, say, I want every subtitle entry to start one second later.

awx
5th June 2008, 17:08
Regex-es are cute. But can it globally change/adjust time indices, too? Like, say, I want every subtitle entry to start one second later.
It's on the ToDo list, in the first post in this thread:
* Write some basic item Timing adjustment functions.
It shouldn't take more than a hour or two to add that feature. Most of the time will be setting up the dialog box. I'll see about getting it into the next version.

But SubHub will always be more of an Post-OCR correction tool and less of a dedicated subtitle editor.

I need a Unicode-enabled subtitle editor that can do 2 or more subs side-by-side. Currently I use GnomeSubtitles in Linux for this. In the very long term, I have been thinking about adding a new window or another exe to SubHub that could do this.

awx
16th June 2008, 02:27
Version 0.4.0 is now up for download. Changes include:

Cleaned up UI.
New quick search feature.
Item Text diff viewer.
Speed improvements for menu/context menu drawing.


Unfortunately, the item timing adjustments won't be included until the next version, in a couple of weeks.

Hopefully the revised UI (picture in 1st post) doesn't bother anyone.

The item text diff viewer:
http://img165.imageshack.us/img165/5751/subhubdiff20080611le7.png (http://imageshack.us)
A white background means no change was made, red is deleted text and blue is added text. Please ignore the odd spacing between the sections- that's the best I could come up with.

There's a noticeable bug that is easy to work around. If all of the menus and toolbar buttons become grayed out, just click somewhere in the treeview and they should appear normal again. I'll try to get a fix for this in the next version.

awx
26th April 2009, 23:27
Version 0.5.0 of SubHub has been released. This is mostly a collection of small enhancements and bugfixes.

Some notable enhancements include:

You can now specify the ANSI code page to be used when opening a file.
There is an option to enable multithreaded file opening (1 thread per CPU core). This gives a significant speedup when opening large numbers of subtitles.


As always, the new version can be downloaded from SubHub's page on sourceforge:
https://sourceforge.net/project/showfiles.php?group_id=225616

awx
1st June 2009, 05:28
This is another small release.

A ClickOnce installer is now available directly from the SourceForge site.

Install the latest version of SubHub by visiting this page:
http://subhub.sourceforge.net/clickonce/publish.htm

ClickOnce requires Firefox or IE for the initial installation. The automatic updater will check for updates whenever you launch the app and ask if you want the latest version to be installed.

krosswindz
2nd June 2009, 00:01
Sweet, regex-es are perfect to fix most of my corrections with srt. It would be great to include a spell check utility. Probably interface it with GNU aspell may be.

awx
26th September 2010, 18:42
v0.6.0 released.

No changes except this version is compiled against .Net 4.0. This will give better on-screen font rendering and the new Windows 7 file dialogs.

The clickonce installer is now available as an .exe file.
setup.exe (https://sourceforge.net/projects/subhub/files/subhub/setup.exe/download)