I wanted to make a smart group that contains all files which are named with a certain format. That is, I have a whole bunch of files that have names which begin with dates, e.g. 1932-2-22 or 1932-5 (some don’t have days). Is there a way for a smart group to filter files so that I can get those that have the date format YYYY-M?
Thanks for any suggestions!
I don’t think it’s possible to match a pattern that specific because (AFAICT) the ‘-’ character is ignored in searches. For example, a name search for “wi-fi” matched documents containing “wi-fi” and “wifi” in their names.
Surprisingly, I couldn’t find any documentation (e.g. in the built-in help appendix) specifically mentioning which characters are ignored in searches. Hopefully DTech will document that information or make it easier to find if it already exists.
Yes.
Create a new smart group and give it the criteria, name matches 1932-*
the asterisk will match any string.
I think waiheke123 might want a general pattern match, like:
[0-9][0-9][0-9][0-9]-[0-9]
For some reason my earlier testing with ‘-’ in a search string ignored it, but now it’s being honored. Can’t reproduce what I’d done before.
That worked! You’re an f-ing genius.
It IS important to note that DT does not “count” the dashes “-”. So that type of smart group solution will turn up “1931-9” and 19319.
Thank you so much for your help!!!
hmmm. Try escaping the dash with a backslash
the release note say that absolute not is working in searches and not is a dash
try this line:
1932-*
see if that will knock the 193219 result out. Otherwise, we’ll have to find out if the escape character is something else.
No luck,
Writing [0-9][0-9][0-9][0-9]-[0-9]*
As my “Name Is” entry doesn’t work
It still comes up with files whose names include the following sequences, and others like them–
8162/5
1954 .5
000387
Any other guesses?
Try “[0-9][0-9][0-9][0-9] [0-9]*” (with the quotes). That’s looking for a phrase (multiple words separated by non alphanumeric characters).
On the other hand, a search like word1-word2 is basically identical to word1word2 OR “word1 word2”. Very useful to search for certain words with different spellings, e.g. e-mail, DEVON-think, DEVON-technologies etc.
Christian,
how can you search for a literal dash in the record name?
For instance how cam I make a smart group that will find
1997-12-7
1997-3
but not
1997563421
and not
1997wasagoodyear
The above phrase will do this as your examples are also multiple words separated by nonalphanumeric characters (in this case ‘-’).
So if he wanted a specific year, from his example, he could do
“[1932] [1-9]*”
that’s pretty slick!!!
Almost. The right query is “1932 [1-9]*” as [1932] means either 1, 2, 3 or 9.
I see. So this is basically a perl regex?
I’m new to regular expressions.
It’s similar but not as complex as regular expressions.
Does that mean that a Smart Group with this criteria:
Name matches wi-fi
… is matching “wi” and “fi” as “word1 word2” in a document titled “this is a wi-fi example”? And as “word1word2” in one titled “this is a wifi document”?
But it’s confusing if you’re trying to only match for words that contain hyphens, e.g. “wi-fi” only matches “this is a wi-fi example” above and not also “this is a wifi document”.
…that’s pretty slick. It does the a pretty good job of returning the right files.
the only extras i get now have something like 1945. 6 (yes with a space there too) or 8902,2 or 2980/ 9 .
In the future, is it better to code using searchable signs, like slashes or backslashes or commas? Or, perhaps it’s possible to sift the other number sequences which are non linked by dashes, such as a string of “NOT” commands like, perhaps?
e.g.
NOT [0-9][0-9][0-9][0-9]/[0-9]
Thanks again for your help!!!
I’m sure having trouble wrapping my mind around DT search patterns, e.g. why this Smart Group:
Name matches wi*fi
… matches “this is a wifi document” but not “this is a wi-fi document”.
Wildcards are applied to words but wi-fi are two words.
Is there a single pattern that matches “wifi” and “wi-fi”?
I’m puzzled by differences in results using Tools -> Search… (Shift-Command-F) vs. a Name matches pattern Smart Group …
A Shift-Command-F search for “wifi” (and “Search for: Name” set) will find a Web Archive document named:
Agile Web Solutions Support Forums - View Single Post - WiFi sync broken with B25
… but a Name matches wifi Smart Group doesn’t, even if Kind is Web Archive is explicitly set.
A Shift-Command-F search for “wi-fi” finds that same WA document and a Bookmark document with “Wi-Fi” in its name, but a Name matches wi-fi SG only finds the latter. Also, when this SG is created it doesn’t display the number of documents matched until it’s expanded (and still displays that number after it’s collapsed). Also, Right-Arrow will expand the SG but Left-Arrow doesn’t collapse it; I reported that problem several months ago. Also, the expanded SG will be collapsed after closing and reopening the db. More issues there than just the non-matching WA document. 
See above, just search for wi-fi. This will match and highlight both of them.
I’ve just created such a document, it’s found both by the Search window and the smart group. Could you send me a small demo database? Thanks!
Only smart groups not using the “matches” options display the number immediately whereas smart groups using the “matches” options are only evaluated if necessary.