Nesting

[Ursus]

It can be done, I know I have seen it, but for the heck I can't remember.
Nesting the substitute function within a script.

Here's what I have. Fmp 6.0v4, win98SE.

A global, containing text.
A plugin, to export this text to a file.

The contents has to be pure html before exporting, so I have scripted a
replace-all-special-charachters. The script contains 45 set field steps,
using the replace function. like this:

Set Field [gExport, Substitute(gExport;"à";"&agrave")]

I think if I can nest these it would all go a lot faster. Other ideas also
welcome.

--
Vintage Ursus

[Bridget Eley]

You can only nest them if each subsequent Substitute function is working on
the result of the previous one, eg

Substitute(Substitute(Substitute(text, search*string, replace*string),
search*string, replace*string), search*string, replace*string)

Bridget Eley



in article 3f85aba3$0$429$1b62eedf@news.wanadoo.nl, Ursus at [email]pooh@bear.com[/email]
wrote on 10/10/03 4:38 AM:

> It can be done, I know I have seen it, but for the heck I can't remember.
> Nesting the substitute function within a script.
>
> Here's what I have. Fmp 6.0v4, win98SE.
>
> A global, containing text.
> A plugin, to export this text to a file.
>
> The contents has to be pure html before exporting, so I have scripted a
> replace-all-special-charachters. The script contains 45 set field steps,
> using the replace function. like this:
>
> Set Field [gExport, Substitute(gExport;"à";"&agrave")]
>
> I think if I can nest these it would all go a lot faster. Other ideas also
> welcome.

--

(to email direct, replace "DOT" with "." and remove ".invalid")

[NeoLuddite]

In article <BBAC02C5.3EDA%bridgeteley@ihugDOTcomDOTau.invalid >,
Bridget Eley <bridgeteley@ihugDOTcomDOTau.invalid> wrote:

> You can only nest them if each subsequent Substitute function is working on
> the result of the previous one, eg
>
> Substitute(Substitute(Substitute(text, search*string, replace*string),
> search*string, replace*string), search*string, replace*string)

It's a little less confusing if you write it this way

Substitute(
Substitute(
Substitute(
text,
search*string, replace*string),
search*string, replace*string),
search*string, replace*string)

--
Real Email-> neoluddite0comcast.net <- replace 0 with @
Blue & White G3/400MHz MacOS 9.1, Starmax 3000/200MHz MacOS 8.6

[Helpful Harry]

In article
<see_signature_for_address-D4B825.16525609102003@news.comcast.giganews.c
om>, NeoLuddite <see_signature_for_address@invalid.com> wrote:

> In article <BBAC02C5.3EDA%bridgeteley@ihugDOTcomDOTau.invalid >,
> Bridget Eley <bridgeteley@ihugDOTcomDOTau.invalid> wrote:
>

> > You can only nest them if each subsequent Substitute function is working on
> > the result of the previous one, eg
> >
> > Substitute(Substitute(Substitute(text, search*string, replace*string),
> > search*string, replace*string), search*string, replace*string)

>
> It's a little less confusing if you write it this way
>
> Substitute(
> Substitute(
> Substitute(
> text,
> search*string, replace*string),
> search*string, replace*string),
> search*string, replace*string)

It'd be intersting to know if there is actually any speed increase over
using the separate Substitute functions - I'd guess there's no
noticeable difference, unless you're doing MANY, MANY, MANY
substitutions and/or records in which case it may or may not be faster.

Helpful Harry
"Just trying to help whenever I can." :o)

[Bridget Eley]

There would be a big difference if you had forgotten to use the Go to Field
step at the beginning of a series of Set Field steps (which would not be
necessary in the nested Substitute version).

To tell if there was a speed difference between a series of Set Fields
preceded by a Go to Field step, and a single Set Field with nested
Substitute functions, you would have to do a speed test using large numbers
of records.

Bridget Eley

in article 101020031249303351%helpful_harry@nom.de.plume.com, Helpful Harry
at [email]helpful_harry@nom.de.plume.com[/email] wrote on 10/10/03 9:49 AM:

> In article
> <see_signature_for_address-D4B825.16525609102003@news.comcast.giganews.c
> om>, NeoLuddite <see_signature_for_address@invalid.com> wrote:
>

>> In article <BBAC02C5.3EDA%bridgeteley@ihugDOTcomDOTau.invalid >,
>> Bridget Eley <bridgeteley@ihugDOTcomDOTau.invalid> wrote:
>>

>>> You can only nest them if each subsequent Substitute function is working on
>>> the result of the previous one, eg
>>>
>>> Substitute(Substitute(Substitute(text, search*string, replace*string),
>>> search*string, replace*string), search*string, replace*string)

>>
>> It's a little less confusing if you write it this way
>>
>> Substitute(
>> Substitute(
>> Substitute(
>> text,
>> search*string, replace*string),
>> search*string, replace*string),
>> search*string, replace*string)

>
> It'd be intersting to know if there is actually any speed increase over
> using the separate Substitute functions - I'd guess there's no
> noticeable difference, unless you're doing MANY, MANY, MANY
> substitutions and/or records in which case it may or may not be faster.
>
> Helpful Harry
> "Just trying to help whenever I can." :o)

--

(to email direct, replace "DOT" with "." and remove ".invalid")

[Ursus]

Bridget and others,

I already thought the same last night in bed. And tried it now.

File with 2409 records
WITH nested substitutes timed at 6:46 minutes for the whole process (script
of about 20 steps. Using plug-ins)
WITH separated substitutes timed at 7:36 min.

So the nesting seems to be plenty faster.

Now I'm still hoping to speed things up a bit more. So I hope to find out if
this is possible.

Thanks Ursus

"Bridget Eley" <bridgeteley@ihugDOTcomDOTau.invalid> schreef in bericht
news:BBAC3273.3EE7%bridgeteley@ihugDOTcomDOTau.inv alid...

> There would be a big difference if you had forgotten to use the Go to

Field

> step at the beginning of a series of Set Field steps (which would not be
> necessary in the nested Substitute version).
>
> To tell if there was a speed difference between a series of Set Fields
> preceded by a Go to Field step, and a single Set Field with nested
> Substitute functions, you would have to do a speed test using large

numbers

> of records.
>
> Bridget Eley
>