Today is: Friday 22 November 2024. |
Level → | 1 | 2 | Psychological Studies ("designated") | 3 | 4 | Conversion MSc PGT | MRes taught postgrad | 9 | ||||
---|---|---|---|---|---|---|---|---|---|---|---|---|
Level | |
---|---|
1 | |
2 | |
Psychological Studies ("designated") | |
3 | |
4 | |
Conversion MSc PGT | |
MRes taught postgrad | |
9 | |
Could redo whole code, using a data structure that put all the courses into one (not 6 levels) of hash array. Would probably make matching easier. Possibly just report multiple hits, not jump straight to one of them. Could do this, but not worth it in this application judging by my trials. If I did then: Go through collecting matches in a queue of codes, and one of descrips Sort -u both queues Go through each queue, expanding a match into a triple Somehow detect dups in result If exactly one match: jump there; if none, re-show query page but say no match. If more than 1 match, query page and list matches under the input box Do hier search e.g. \W boundaries, then slack off. Should search for multiple words sep. and in combo: not just as strings. [I.e. 3 diff. search improvements: report multiple hits; hier. search quality of whole/part words; search for multiple words separately. ]