Indexed searching works by searching a pre-built index to locate files, its main advantage is that a search on an index is normally very fast, e.g. sub-second times. So, before you can perform a search you need to create an index. Once the index has been created you can search it by typing information into the search field. Multiple indexes can be created and shared over network drives with other people.
Index based searches can be performed using the Index Search interface, opened by selecting the option Index Search from the criteria drop-down or by using the menu option Search -> Criteria -> Index Search.
This will change the criteria interface to look like this:
The Index Interface has three main sections:
|
Search field. This is a search-as-you-type field and shows previous searches using the drop-down box. The 'Aa' button switches on/off case-sensitive searching, if the specified index supports it.
|
|
Index list. Shows the current index and when dropped-down all the available indexes, to add a new one pick the '<Create New Index...>' option, which will open the Index Manager.
|
|
Index Manager button. Click this button to open the Index Manager. |
Performing a Search
The search terms can be specified using standard Boolean Expressions logic, such as AND, OR, NEAR, NOT, and LIKE. There is, however, one important difference. Whereas a classic Boolean search will match mid-word an index search assumes that the term matches the start of the word. Index searches have an implicit '<' at the start and classic searches have an implicit '*' at the front.
This table illustrates how search terms will be matched:
Search Term |
Index Search |
Classic Search |
fine |
Any word starting with 'fine', e.g.
fine, fined, finest, etc. |
Any word with 'fine' in, e.g.
fine, define, refined etc.
|
*fine |
Any word with 'fine' in, e.g.
fine, define, refined etc.
|
(Same as index search) |
<fine |
Any word starting with 'fine', e.g.
fine, fined, finest, etc.
|
(Same as index search) |
<fine> |
Just the whole word 'fine'.
|
(Same as index search)
|
fin*n |
Any word starting with 'fin' and having 'n' in it, e.g.
finding, finance, etc.
|
Any word with 'fin' somewhere in it, followed by 'n', e.g.
finding, definitions
|
"fine" |
Just the whole word 'fine'. |
Any word with 'fine' in, e.g.
fine, define, refined etc.
|
"fine day" |
Matches the phrase 'fine day' |
(Same as index search) |
Searching Attributes
By default searches will search the file name and the file content (if indexed), however you can specify other attributes of the files using search prefixes.
Prefix |
Purpose |
Example |
name: |
Search on the name of the item. |
To search all files with a name starting 'accounts'
name:accounts
You can also search by extension, so to limit the above search to just .pdf and .txt files:
name:accounts name:*.pdf;*.txt
|
lookin: |
Search on the location of the item.
|
To search all items with Program files in the path:
lookin:"Program Files"
|
ext: |
Search on the file extension.
Note: 'ext:' is implied if you use the pattern *.abc when searching with no prefix or the 'name:' prefix. |
To search all .pdf and .txt files:
ext:pdf;txt
|
moddt: |
Search on the modified date of the item.
|
For all files modified since 1 Nov 2021
moddt:"> 1 Nov 2021"
|
createdt: |
Search on the created date of the item.
|
For all files created between 1 Jan 2020 and 1 Mar 2020:
createdt:"> 1 Jan 2020" createdt:"< 1 Mar 2020"
|
size: |
Search on the size of the item. |
For all files greater than 100 KB
size:"> 100KB"
|
Prefixes can be combined so to search for all .pdf files created since the 1 Jan 2020 for the word "excellent"
Excellent name:*.pdf createdt:"> 1 Jan 2020"
Limitations of Indexed Search
Although indexed searching is very fast there is one very important fact to keep in mind, the term being searched for must be in the index for it to be found. Common issues to be aware of:
File changes |
If files are changed, added or deleted and the index is not updated then searches on the index will be driven by the old information not the new changes. You can use the Index Scheduling functionality to help keep indexes up to date.
|
Search terms |
During the indexing process not all terms are considered 'indexable'. Words that are very common and appear in all documents, such as the or and, are not indexed nor are character sequences that do not seem to be valid language terms, eg. 4*2.
|
Performance Impact of Indexes
The indexing feature is not constantly running in the background and indexes are not loaded unless you try and search them, so there is no performance impact to people who never need to use the indexing functionality. If you don’t use the feature then you won’t notice it because it’s effectively disabled by default.
This method of operation has the advantage that the program is not secretly consuming CPU cycles or eating away at disk space but it has the disadvantage that the index isn’t updated unless you specifically update it. This is fine for large repositories of fairly static data but can cause problems when files are changing frequently. You can schedule index updates using the Index Scheduling feature.