I had messed up, I'm OK now, the list to dbs was not in the proper shape.

Anyway no 1 is still growing big, what I've done here is to try and
emulate the er file in the app demo. In that er file no non-ref/key
relation is mapped anywhere in the call to dbs, why?

I've just started stepping through file no 1 and I see that the
+WordCount references are there, this is my dbs call:

(dbs
   (2 +User +Feed)                                 # 1
   (3 +Article +ArticleLinks)                      # 2
   (3 +Aword +Awords)                              # 3
   (2 +Tag +FeedTag)                               # 4
   (3 +Read +Likes +ArticleTag)                    # 5
   (4 +WordCount +Similarity)                      # 6 <-------
   (2 (+User uname uid feeds))                     # 7
   (3 (+Feed fid xmlUrl htmlUrl))                  # 8
   (3 (+Article aid htmlUrl picoDate feed words))  # 9
   (3 (+ArticleLinks from to))                     # 10
   (3 (+Aword word))                               # 11
   (3 (+WordCount word article))                   # 12 <--------
   (3 (+Similarity a1 a2))                         # 13
   (3 (+Tag tag))                                  # 14
   (3 (+FeedTag user tag feed))                    # 15
   (4 (+ArticleTag user tag article))              # 16
   (3 (+Likes user article))                       # 17
   (3 (+Read user article)))

And the +WordCount definition looks like this:

(class +WordCount +Entity) #
(rel word      (+Aux +Ref +Link) (article) NIL (+Aword)) #
(rel article   (+Ref +Link) NIL (+Article)) #
(rel count     (+Number))

Note how the count +Number is missing from the dbs mapping, could that
fact have anything to do with it?

Anyway, this is a couple of examples of what it looks like when I'm
stepping no 1:

{;2} (+WordCount)
   count 1
   article {2-1}
   word {3-34}
(seq Obj)
!
{;3} (+WordCount)
   count 1
   article {2-1}
   word {3-35}
(seq Obj)
!
{;4} (+WordCount)
   count 1
   article {2-1}
   word {3-36}
(seq Obj)

Any ideas of why the counts end up in no 1?

/Henrik


On Tue, Aug 18, 2009 at 9:08 AM, Henrik Sarvell<hsarv...@gmail.com> wrote:
> Thanks for that, will try it out tonight!
>
> /Henrik
>
>
> On Tue, Aug 18, 2009 at 7:21 AM, Alexander Burger<a...@software-lab.de> wr=
ote:
>> Hi Henrik,
>>
>>> I've just used dbs to split my database into several files which seems
>>> to be working but if I recall correctly the catch all will be file no
>>> 1 and it's growing out of proportion very rapidly when I import my
>>> data. Out of proportion with regards to what I believe should be in it
>>> that is.
>>
>> The first thing I would do is load "lib/too.l" and then call (dbfCheck).
>> This will list all objects in the DB that are not explicitly assigned to
>> a file.
>>
>>
>>> let me simply walk through an arbitrary database file and list
>>> whatever is in there. If I could I would be able to "debug" this
>>> situation.
>>
>> This can be done with 'seq'. (seq 7) gives the first object of file 7,
>> and calling 'seq' with that object will give the next sequential object.
>> For example, to show all articles in 'app'
>>
>> =A0 : (for (Obj (seq 3) Obj (seq Obj)) (show Obj))
>>
>> or, if you don't know the file number
>>
>> =A0 : (for (Obj (seq (db: +Item)) Obj (seq Obj)) (show Obj))
>>
>>
>> Usually, I put a breakpoint before 'seq', to be able to look at each
>> object
>>
>> =A0 : (for (Obj (seq (db: +Item)) Obj (! seq Obj)) (show Obj))
>>
>> then just hitting ENTER will step through the file. You can terminate
>> the process with Ctrl-X.
>>
>> Cheers,
>> - Alex
>> --
>> UNSUBSCRIBE: mailto:picol...@software-lab.de?subject=3dunsubscribe
>>
>
-- 
UNSUBSCRIBE: mailto:picol...@software-lab.de?subject=unsubscribe

Reply via email to