by The standard FP search form starts with the following statements - - -
"Comment: Text Search - This page lets you search through the default
text index that is created whenever web pages are saved or web links are
recalculated. No customization is required.
Use the form below to search for documents in this web containing
specific words or combinations of words. The text search engine will
display a weighted list of matching documents, with better matches shown
first. Each list item is a link to a matching document; if the document
has a title it will be shown, otherwise only the document's file name is
displayed."
It goes on to say that it allows the use of AND, OR and NOT () and * as
a wildcard after one or more letters.
So far so good!
(In case it's relevant, I'm using FP97 and FP2000 Premium)
1. My ISP reports that my searches using this search form method are NOT
confined to the directories/folders containing my web pages. In other
words, it's searching for documents *outside* of "this web containing
specific words or combinations of words" (as quoted above).
The ISP cannot find a way of restricting the search to the area intended
so that I've had to remove the search facility until a solution is
found.
The search form properties mentions "Word list to be search" which seems
to default to "all". Is this the clue? What should be entered in this
box?
2. How should the weighting indicated on the weighted list be
interpreted? Anything more than "with better matches shown first"?
Maybe I've answered my own question because I've just conducted an
experiment. I formed two files, both containing *only* the characters
xyz
(and these being the only files in my web containing xyz) and saved them
in my web.
If I search for xyz the two files are identified with scores of 707
If I search for xyz* the two files are identified but with a score 1414
Exactly the same scores respectively appear if the two files contain xy
and xy* respectively.
These scores are highly reminiscent of the reciprocal of the square root
of 2, and the square root of two, respectively, but three orders of
magnitude larger.
Any ideas anyone? It would have been a lot simpler if the score had been
a count of the occurrences of the text being sought!
I'm now more curious than ever to understand the strange results I've
obtained.
3. Is there any way of making it search for a string of words?
A common enough search when looking for someone, such as Mr Paul
Smith
Searching for (Mr AND Paul AND Smith) does not produce the required
result.
4. How can I format the items that appear as results? eg. Size, bold,
italic, colour, make the table have zero borders, etc . .
5. Rather more generally, where can I find the definitions of all/many
of the webbots - and IMPORTANTLY, a list of their variable parameters
and their functions? I've asked this question before but even an MVP
didn't seem to understand what I thought was a reasonable question.
--
Best regards
Peter Child