Opened 2 years ago
Closed 19 months ago
#1487907 closed Feature Patches (fixed)
Enhancement for LDAP fuzzy search
| Reported by: | sbutler | Owned by: | |
|---|---|---|---|
| Priority: | 5 | Milestone: | 0.7-beta |
| Component: | LDAP connection | Version: | git-master |
| Severity: | normal | Keywords: | fuzzy_search ldap |
| Cc: |
Description
My organization's directory pre-composes some permutations for the common name field. For example, for:
displayName: Stephen Joseph Butler
We have for CN the values:
- cn: Stephen Joseph Butler
- cn: Stephen Butler
- cn: Butler, Stephen
The RC default fuzzy match of '*<term>*' make inefficient use of the indexes in our case. When I'd really like is for RC to fuzzy match '<term>*'. The attached patch expands the 'fuzzy_search' property to have these values:
- true or 'both': maintain the old behavior of '*<term>*'
- 'begin': match '*<term>'
- 'end': match '<term>*'
Attachments (1)
Change History (6)
Changed 2 years ago by sbutler
comment:1 Changed 2 years ago by alec
- Milestone changed from later to 0.6-beta
comment:2 Changed 2 years ago by thomasb
I'd rather name it 'prefix' instead of 'end' because this denotes a prefix-matching. This term is maybe more clear to tech-savvy people who will configure Roundcube.
I've also never used a suffix-match (i.e. 'begin') in the wild.
comment:3 Changed 2 years ago by sbutler
No objections. I added the 'begin' for completeness sake, not because I had a use for it.
I can submit a new patch with 'prefix' as the setting. Any thoughts about how to name the old wildcard matching? infix?
comment:4 Changed 20 months ago by alec
I'd like to see a patch that merges this solution with #1486564 and a user-configurable preference for this behaviour.
comment:5 Changed 19 months ago by alec
- Resolution set to fixed
- Status changed from new to closed
Implemented in [f21a04c0].
"begin" looks useless.