I’m using the three pane view in a database holding pdf’s of articles – some of which bear titles that I’d like to continue to use. Others may not have a title or it may not be sufficiently descriptive for my purposes.
When I use the article’s own title, I’ve used it as the first words in the Name and have enclosed it in quotation marks. This means the open quote mark becomes the initial character in the Name. In these cases, DT will sort all those with the initial quote mark at the top of list (alpha sorted). These will be followed by all of the articles not having an opening quote – which will all sort, in alpha order, in the remaining portion of the pane.
I would like DT to ignore the initial open quote mark and sort every entry in the pane alike on an alpha basis using the first words. I’ve seen the “ignore diacritics” in the View menu’s “Sort” selections, but quotation marks aren’t diacritics. This box was already checked – apparently by default upon installation.
I’ve taken a quick trip through 100+ previous posts dealing with sorting and haven’t found my issue discussed. I’m assuming I’m left with the options of a manual sort, or simply leaving the initial quotation marks off the articles’ titles altogether and moving on.
Do I have other options? I’d appreciate learning of any work arounds.
You will not likely find much relief on this. DEVONthink uses standard ASCiI sort order for “by name” sorting. Quotation marks always sort higher than alphabetic characters. I don’t imagine many users would want to see a change from the standard (just my opinion, of course).
Naming files with quotation marks could cause other issues – some file systems do not accept names with quotation marks, which might be a problem if you export your documents and share them.
Seconding korm’s comments, I follow with a STRONG suggestion to NOT include quotes or other punctuation in your filenames. This is an unwise practice and can adversely affect potential workflows, etc. As you can see here it is hindering your progress in DEVONthink already.
Would it be worthwhile to have a script that identifies (and optionally renames) Dt document/group names containing characters that might be problematic when exporting corresponding files/folders for sharing with non-OS X systems? Or maybe it’s sufficient to post-process exported filenames if necessary? Surely there are “sanitize filename” utilities for that purpose.
For some reason Dt automatically changes at least ‘:’ and ‘/’ characters in db item names to ‘-’ in the underlying filenames, which has caused me trouble when comparing original filenames with exported versions that don’t match. Maybe I’ll search for my old post(s) about that issue …
Edit: Found the topic I started on 22 Feb 2010 under the Beta DEVONthink subforum:
• A document titled this.is.a.new.document.txt will be exported as this.is.a.new.txt, dropping the last non-extension component of the original document’s title
It’s actually more like this:
• Creating a document titled with an extension name appended that matches the document type can cause the filename stored in the db to use that extension. Examples:
Creating a Plain Text document titled test.me.txt stores test.me.txt, not test.me.txt.txt
Creating a Rich Text document titled test.me.rtf stores test.me.rtf, not test.me.rtf.rtf
I’ve never intended to explicitly supply an extension in document titles, yet it’ll happen (unexpectedly, without warning) in certain cases which can be problematic when exporting (like I originally reported).
I have and will always despise(d) filename extensions.
