Low performance when extracting data

[Jonas Kullenwall]

I have written a macro to extract data from the active document. The macro
should be a part of an algorithm to auto detect headings. The problem is that
I get very long delays when accessing the DOM.

A code-snippet that show this behavior:

Dim nCount as Long
Dim oSents as Sentences
Dim sText as String

Set oSents = ActiveDocument.Sentences
if not oSent is nothing then
nCount = oSents.Count
if nCount > 1 then
sText = oSent.Item(nCount - 1).Text
endif
endif

For certain documents the Sentences::Count will take approx. 10s to execute.
Range::Text takes also approx 10 s to execute. The document is quite simple
with a TOC (~300 items), body text and a lot of heading. The document has 123
pages. I a would try to iterate through all sentences in this document it
would have taken to long time to be acceptable.

However, if I remove the table of content, then the execution time will be
below 1 s. Could someone explain why and is there any settings in Word that I
could modify to get rid of this problem? I have tried to run Word in safe
mode and also protected the document for editing with no performance
differences.

I am running Word 2003 and Win XP SP 2. My computer does not lack in
performance.

Jonas

 

[Cindy M.]

Hi =?Utf-8?B?Sm9uYXMgS3VsbGVud2FsbA==?=,

It sounds to me like the TOC may first be updating (although I'd find that
strange) or the document is repaginating. You may consider a TOC with 300
entries and 123 pages with lots of headings "simple", but calculating the TOC
definitely is not; and depending on the style definitions, calculating the
layout could be very resource-intensive.

You might see what happens if you select the TOC (field) then press Ctrl+F11 to
lock it (so it can't update - Ctrl+Shift+F11 would unlock, and there are object
model equivalents if you need it in your code).

The next thing I'd try is to go into Normal view, then in Tools/Options/General
deactivate Background Pagination (this is only possible in Normal view). See if
this makes any difference when you run your code.

I have written a macro to extract data from the active document. The macro
should be a part of an algorithm to auto detect headings. The problem is that
I get very long delays when accessing the DOM.

A code-snippet that show this behavior:

Dim nCount as Long
Dim oSents as Sentences
Dim sText as String

Set oSents = ActiveDocument.Sentences
if not oSent is nothing then
nCount = oSents.Count
if nCount > 1 then
sText = oSent.Item(nCount - 1).Text
endif
endif

For certain documents the Sentences::Count will take approx. 10s to execute.
Range::Text takes also approx 10 s to execute. The document is quite simple
with a TOC (~300 items), body text and a lot of heading. The document has 123
pages. I a would try to iterate through all sentences in this document it
would have taken to long time to be acceptable.

However, if I remove the table of content, then the execution time will be
below 1 s. Could someone explain why and is there any settings in Word that I
could modify to get rid of this problem? I have tried to run Word in safe
mode and also protected the document for editing with no performance
differences.

I am running Word 2003 and Win XP SP 2. My computer does not lack in
performance.

Cindy Meister
INTER-Solutions, Switzerland
http://homepage.swissonline.ch/cindymeister (last update Jun 17 2005)
http://www.word.mvps.org

This reply is posted in the Newsgroup; please post any follow question or reply
in the newsgroup and not by e-mail

 

[Jonas Kullenwall]

With simple document I ment no tables, images, drawing objects, objects , or
other "special" effects applied to the text. However, I would agree on that
the document is quite large.

I tried your suggestions with no significant result on the performance.
However, when I used the same macro in Word 2000 (XP SP2) I get the same slow
execution time at the first run but the second run was executed under 1 s.
If I was able to achieve the same behavior on Word 2003, that should be
acceptable.
Do you think that it is possible, by tweaking some other settings?


Jonas

 

[Cindy M.]

Hi =?Utf-8?B?Sm9uYXMgS3VsbGVud2FsbA==?=,

With simple document I ment no tables, images, drawing objects, objects , or
other "special" effects applied to the text. However, I would agree on that
the document is quite large.

I tried your suggestions with no significant result on the performance.
However, when I used the same macro in Word 2000 (XP SP2) I get the same slow
execution time at the first run but the second run was executed under 1 s.
If I was able to achieve the same behavior on Word 2003, that should be
acceptable.
Do you think that it is possible, by tweaking some other settings?

Hmmm. There are reports that Word seems to become more responsive if a menu is
first displayed. You might try putting these two lines in before processing
begins and see if that makes any difference

SendKeys ("{Esc}")
CommandBars("Menu bar").Controls("File").Execute

Cindy Meister
INTER-Solutions, Switzerland
http://homepage.swissonline.ch/cindymeister (last update Jun 17 2005)
http://www.word.mvps.org

This reply is posted in the Newsgroup; please post any follow question or reply
in the newsgroup and not by e-mail