End of record identification

[mark]

I'm not able to replicate your problem with the following code. Can you? Are you doing something fundamentally different?

<sql "drop table tbldoc"></sql>
<sql "create table tbldoc(id int,ocr varchar(200))"></sql>
<apicp alpostproc 1>
<apicp allinear 1>
<apicp alintersects 1>
<apicp intersects -1>
<apicp alwithin 1>
<apicp minwordlen 5>
<$data="
CLMS
STM What is claimed is:
NUM 1.
PAR 1. sensxr notsensor forming method of forming an annular workpiece which
is pressed between a forming roll and a mandrel moved towards and away
from each other by a moving device, comprising the steps of:
CLMS
STM What is claimed is:
NUM 1.
PAR 1. notsensor forming method of forming an annular workpiece which
is pressed between a forming roll and a mandrel moved towards and away
from each other by a moving device, comprising the steps of:
">
<sql "insert into tbldoc values(1,$data)"></sql>
<sandr "sensxr" "sensor" $data>
<sql "insert into tbldoc values(2,$ret)"></sql>
<$q='sensor'
'sensor w/CLMS'
'sensor w/CLMS w/\P.=>>='
'notsensor w/CLMS w/\P.=>>='
>
<loop $q>
<fmt "Query: %s\n" $q>
<sql row "select id,ocr,mminfo($q,ocr,0,0,1) hit from tbldoc where ocr like $q">
--------------------------------------------------------------
Hit: $id, $hit
</sql>
Hitcount: $loop

</loop>

[kzinda]

I think part of my problem is that the document is broken up into fields, with each filed being searched independently, and then assembled for the final output. For any given field, there is no end of field identifier. Consequently, when trying to search a given section of the overall document, my end delimiter is actually found in a different field. For example:

Field1
Section1......word 1 word 2 ...period

Field 2
Section2 .....word 1 word 2 ...period

To search section 1, I had been using: word1 word2 w/Section1 w/Section2.

In this construct, it seems texis either can't properly identify any of the terms in Section1, or it can only identify terms in the first line of section1. If I use w/Section1 w/\P.=>>=, it seems to start the search at the end of section 2 back up through section1, which seems to explain the problem I was describing earlier.

Given this construct, how can I construct a query to only search section 1?

[mark]

Sorry, I'm confused. Is this directly related to the above or a tangent? If it's related to above, please use similar/same examples and indicate how the previous example code is different than what you're doing?

Are these "fields" texis fields that are searched independantly. Or a blob of text being searched all at once as in the previous thread?

What's your sql query?
What's the data?
What's the answer?
vs. What were you expecting?