| proximity searches using Fox trot query [message #629] |
Thu, 26 October 2017 02:03  |
don.spady
Messages: 4
Registered: October 2017
|
Junior Member |
|
|
I want to find A near B where A may contain more than 1 word
e.g. I want to find the words "risk assessment" within 30 words of method.
In this instance 'A' is "risk assessment" and 'B' is "method"
I tried various combinations of commands but without any real success.
e.g.
"{30} "risk assessment" methods"
and I get a lot of separated words; i.e. risk "alone" assessment
"alone" plus risk assessment "together"
"{30} ({2} risk assessment) methods"
and I get much the same result, except that even the 2 gets highlighted as
a finding.
Is there a way to nest a query such that you have a close proximity of
words that is then joined to one (or more) more distant words
I have tried various combinations but with no success.
Many thanks.
Don
|
|
|
|
| Re: proximity searches using Fox trot query [message #631 is a reply to message #629] |
Sat, 28 October 2017 10:23  |
FoxTrot Engineering
Messages: 406
Registered: April 2020
|
Senior Member |
|
|
don«|dot|»spady«~at~»gmail«|dot|»com wrote:
> I want to find A near B where A may contain more than 1 word
> e.g. I want to find the words "risk assessment" within 30 words of method.
The easiest way to achieve this is to use the "includes neighboring words" criteria, and quote the consecutive words:
[includes neighboring words] ["risk assessment" method] [around one line]
If you prefer using the FoxTrot query syntax (which allows to specify more precisely the maximum distance):
[matches the FoxTrot query] [{30} "risk assessment" "method"]
With this FoxTrot query syntax, there are two variants for proximity searches:
- when a quoted string begins with a number between braces ( e.g. [matches the FoxTrot query] [x "{10} a b c d" y] ), then the found document must contain at least one occurrence of the first word in the quoted string (a) at the specified maximum distance (10 words) to occurrences of the other words in the quoted string (b, c and d); in addition to the other words outside of the quoted string (x and y) which may be found anywhere in the document.
- when a quoted string is preceded by a number between braces ( e.g. [matches the FoxTrot query] [x {10} "a" "b c" "d" y], then the found document must contain at least one occurrence of the first quoted string (either a single word, or multiple contiguous words) ("a") at the specified maximum distance (10 words) to occurrences of the other immediately following quoted strings ("b c" and "d"); in addition to the other words before the brace, and after the last quoted string (x and y) which may be found anywhere in the document.
> Is there a way to nest a query such that you have a close proximity of
> words that is then joined to one (or more) more distant words
No. However, yo can do something like this:
[matches the FoxTrot query] ["{2} risk assessment" "{30} risk method"]
to find documents that have an occurrence of [risk] very near to [assessment], and an occurrence (but not necessarily the same one) of [risk] relatively near to [method].
Jérôme - CTM Engineering
------------------------------------------------------------ ---------
"Foxtrot is amazingly fast, and the built-in highlighting is a
fantastic idea."
Constantin von Wentzel, Arlington MA
Download a demo version from www.foxtrot.ch
------------------------------------------------------------ ---------
Jérôme - FoxTrot Engineering
|
|
|
|
| Re: proximity searches using Fox trot query [message #633 is a reply to message #629] |
Mon, 30 October 2017 20:00 |
don.spady
Messages: 4
Registered: October 2017
|
Junior Member |
|
|
This helps a great deal.
Many thanks.
Don
On Wednesday, October 25, 2017 at 6:03:29 PM UTC-6, don....@gmail.com wrote:
>
> I want to find A near B where A may contain more than 1 word
> e.g. I want to find the words "risk assessment" within 30 words of method.
> In this instance 'A' is "risk assessment" and 'B' is "method"
> I tried various combinations of commands but without any real success.
> e.g.
> "{30} "risk assessment" methods"
> and I get a lot of separated words; i.e. risk "alone" assessment
> "alone" plus risk assessment "together"
> "{30} ({2} risk assessment) methods"
> and I get much the same result, except that even the 2 gets highlighted as
> a finding.
>
> Is there a way to nest a query such that you have a close proximity of
> words that is then joined to one (or more) more distant words
>
> I have tried various combinations but with no success.
>
> Many thanks.
>
> Don
>
|
|
|
|
] 
FoxTrot Search Forum
Search
Help
Members
Register
Login
Home
