miq

Thank you for your reply, @Andrew! Although I get your point, to me, this does not sound very consistent, either. First, I have about 200 snippets with keywords that go back almost 15 years now. Back then, I followed a recommendation by TextExpander to use double characters at the start. In Alfreds Tips For Better Snippet Expansion, the same recommendation is included: "Use double characters (e.g. ttime)". All my keywords follow this rule and in my first example, the problem results just from the keywords "Oor" and "ssy". Second, when one uses password-like keywords as you suggest, the mid-string-option seems obsolete at all, as there are very few cases in which non alphanumeric combinations occur in the middle of a word. So to me it seems like the middle-string-option only makes sense for keywords like in your example "ana". Third, I am not sure what the use case for sequential snippets might be. Perhaps for writing code? For natural languages other than English the case I described is problematic: often, words consist of parts that can be 'snippetized' but often do require extra letters between them. Sequential snippets without mid-string expansion do not appear to be designed for natural language text production. Having a "snippet key" like "§" in your case sounds like a very good idea to me. For me though, after getting used to lots of keywords over the years, it is not really an attractive option. Could an extra option "allow sequential snippets" be a solution for both scenarios?

Thank you very much for your reply, @Vero! I understand that this behaviour is intented. I think it might be worth considering to change Alfreds mid-string reset habits? As I said, in German we have many long compound words, but most commonly there are some letters in between, so that pure stringing together snippets is very rarely viable. Do you think it would be disadvantageous to reset the mid-string marker only after a subsequent space character? Or is this actually a feature that I'm missing?

Meanwhile I have made up an English example that is not so realistic but hopefully is easier to follow. I post it here, even if the problem is probably already clear thanks to @deanishe (thanks!). Say we have two snippets: - keyword 'hha' for snippet "hand|" - keyword 'icr' for snippet "craftsman|". 1. We want to type the word "handicap": Typing snippet 'hha' + rest of the word 'icap', we get: "handicap". ✅ But: 2. We want to type "handicraft": Typing snippet 'hha' + rest of the word 'icraft', as soon as we type 'hhaicr', Alfred recognizes 'hha' and 'icr' and expands both to "handicraftsman" even though 'icr' is mid-string. As I said, this example might be a little constructed. In German though, we hava lots of long compound words…

Maybe this is a German-language-only-problem, but snippet expansion does not work as expected when mid-string expansion is unflagged: If the string left of the snippet also is an expanded snippet, it won't be recognized as a string, so the second snippet expands. Example: I have a snippet "Wwi" for German "Organisation" (organization). I also have a snippet "ssy" for German "systemtheoretisch". Now, when I try to type German "Organisationssystem" (organizational system) by typing "Oor"+"system" the "Oor" expands and the mid-string "ssy" is expanded.

Update: It it working now. You were right, it just takes about 10 seconds to display the results (I have 877 .bdskcache files though). You helped me anyway as I had to get the cache files first. Thank you!

OK, thank you very much for your help anyway! Same results with both Mountain Lion and Mavericks.

Now I have the files in ~/Library/Caches/Metadata/edu.ucsd.cs.mmccrack.bibdesk/. I think, in the BibDesk I had to enable "automatic backups". That's the good news. The bad: It still doesn't work. Terminal now gives the output: "<?xml version="1.0"?> <items><item uid="bib/Users/miq/Library/Caches/Metadata/edu.ucsd.cs.mmccrack.bibdesk/Fish:1980.bdskcache" arg="/Users/miq/Library/Caches/Metadata/edu.ucsd.cs.mmccrack.bibdesk/Fish:1980.bdskcache"><title>Is there a Text in this Class? The Authority of Interpretive Communities</title><subtitle>1980: Fish, Stanley</subtitle><icon>icon.png</icon></item></items>" But still, Alfred behaves as described above.

Sebastian, thank you for your reply! 1. The folder does not exist. I only find ~/Library/Caches/edu.ucsd.cs.mmccrack.bibdesk/ containig no .bdskcache files at all. 2. Maybe this is related, but I get no output, but also no error, just a new line ending with $. This might rather be a BibDesk problem? Thanks for your help!

This still doen't work here. When I type "bib", Alfred offers me to "Search in BibDesk library", but as soon as I type the search argument, only the default search engines are displayed. Too bad, I'd really love to use this workflow...