# Scraping off a Word document? **URL:** \ **Category:** Ruby\ **Created:** [January 16, 2009, 4:58pm UTC](https://www.ruby-forum.com/t/scraping-off-a-word-document/157927 "2009-01-16T16:58:40Z")\ **Posts on this page:** 9\ **Page:** 1
**Author:** ![gregarican](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/g/7feea3/32.png) [@gregarican](https://www.ruby-forum.com/u/gregarican)\ **Post date:** [January 16, 2009, 4:58pm UTC](https://www.ruby-forum.com/t/scraping-off-a-word-document/157927/1 "2009-01-16T16:58:40Z")
Here’s a conceptual question. I have a Word mail merge, with a few dozen documents. There’s a certain field, let’s call in the Employee field, on each page. These documents are sorted in order of this field. What I’d like to do is save off each group of pages into its own Word document under that field. So if it’s Employee: Joe Schmoe on the first five pages I’d want to save off just those pages and name the file “Joe Schmoe.Doc” and so on. The mail merge itself is pretty much hard-coded into a big group of documents, so that’s my basis. Any suggestions about what Ruby modules and methods I’d start out delving into? I’m thinking win32ole of course, but have a good-sized task ahead of me I have to deliver in relatively short order :-/ ---
**Author:** ![gregarican](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/g/7feea3/32.png) [@gregarican](https://www.ruby-forum.com/u/gregarican)\ **Post date:** [January 16, 2009, 6:06pm UTC](https://www.ruby-forum.com/t/scraping-off-a-word-document/157927/2 "2009-01-16T18:06:16Z")
“gregarican” [removed\_email\_address@domain.invalid](mailto:removed_email_address@domain.invalid) wrote in message news:removed\_email\_address@domain.invalid… > and methods I’d start out delving into? I’m thinking win32ole of > course, but have a good-sized task ahead of me I have to deliver in > relatively short order :-/ I’m not sure you can do what you want to do. You open a Word doc, then want to save a portion based on the Employee Name to it’s own file? Then, after that Save is finished, keep the file open, advance the database to the next Employee Name and repeat the save, then repeat the entire process until you get through all of the Employee Names? It occurs to me that Word can’t do that task because the Employee Name field in the document is an unknown until the actual time of the merge. There is only one DOC file for any given letter, and when I do these kinds of merge all I get to see is where the variables are that get filled in during the merge. You have to fill the variable from the database then save the result, advance the database to the next record and fill the variable again to save that result. You are going to create a file for each employee for each letter, and this seems to me to defeat the whole reason to merge data into a document. The reason I merge is because I want one file for everybody, I specifically do not want a separate file for each person. ---
**Author:** ![gregarican](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/g/7feea3/32.png) [@gregarican](https://www.ruby-forum.com/u/gregarican)\ **Post date:** [January 17, 2009, 3:35am UTC](https://www.ruby-forum.com/t/scraping-off-a-word-document/157927/3 "2009-01-17T03:35:49Z")
gregarican wrote: > and methods I’d start out delving into? I’m thinking win32ole of > course, but have a good-sized task ahead of me I have to deliver in > relatively short order :-/ Write what you need using the VBA built into Word. Intellisense will make that rather easy. Then either replicate your VBA calls using Ruby’s win32ole… …or just shell directly from Ruby to your VBA! ---
**Author:** ![gregarican](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/g/7feea3/32.png) [@gregarican](https://www.ruby-forum.com/u/gregarican)\ **Post date:** [January 17, 2009, 11:06am UTC](https://www.ruby-forum.com/t/scraping-off-a-word-document/157927/4 "2009-01-17T11:06:45Z")
On 16.01.2009 16:51, gregarican wrote: > and methods I’d start out delving into? I’m thinking win32ole of > course, but have a good-sized task ahead of me I have to deliver in > relatively short order :-/ I’d do it with VB from inside Word. An alternative might be to use OpenOffice, read the word, write OO’s format (XML in ZIP) and the manipulate the XML. But this sounds pretty awkward. Can’t you force the mail merge to produce multiple documents? Cheers robert ---
**Author:** ![gregarican](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/g/7feea3/32.png) [@gregarican](https://www.ruby-forum.com/u/gregarican)\ **Post date:** [January 17, 2009, 5:21pm UTC](https://www.ruby-forum.com/t/scraping-off-a-word-document/157927/5 "2009-01-17T17:21:14Z")
Robert K. wrote: > I’d do it with VB from inside Word. An alternative might be to use > OpenOffice, read the word, write OO’s format (XML in ZIP) and the > manipulate the XML. But this sounds pretty awkward. I suspect Word can also barf out an XML representation. It may be awkward (get ready for the horror when you open that file!), but it’s probably the best way. All word processing is heading towards XML for its interoperability. ---
**Author:** ![gregarican](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/g/7feea3/32.png) [@gregarican](https://www.ruby-forum.com/u/gregarican)\ **Post date:** [January 18, 2009, 3:20am UTC](https://www.ruby-forum.com/t/scraping-off-a-word-document/157927/6 "2009-01-18T03:20:56Z")
On Jan 17, 11:18 am, Phlip [removed\_email\_address@domain.invalid](mailto:removed_email_address@domain.invalid) wrote: > Robert K. wrote: > > > I’d do it with VB from inside Word. An alternative might be to use > > OpenOffice, read the word, write OO’s format (XML in ZIP) and the > > manipulate the XML. But this sounds pretty awkward. > > I suspect Word can also barf out an XML representation. > > It may be awkward (get ready for the horror when you open that file!), but it’s > probably the best way. All word processing is heading towards XML for its > interoperability. I wound up writing a C# console program to do the work. I just referred to ugly underbelly of all of the Word COM stuff and was able to grab what I needed. It took awhile though, since my text was contained within text frames. So I had to work with the Document.Shapes property and whatnot. In searching for a solution I did run across a VBA code snippet that would save off each document separately after the mail merge completed. At least now I have a totally automated solution, although it’s cobbled together from various sources. First I pull my data from a SQL DB using Ruby, dumping that to an Excel data source. Then I have a C# program that takes that data source, uses a Word mail merge template and delivers the final document set. Finally, I have a Ruby program that looks in that save directory and e-mails the documents to the individual employees. Eventually it’d be a lot cleaner and easier to maintain if I had all of the work done in a single program written in a single language. But that’s another fight for another day 🙂 ---
**Author:** ![gregarican](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/g/7feea3/32.png) [@gregarican](https://www.ruby-forum.com/u/gregarican)\ **Post date:** [January 18, 2009, 1:30pm UTC](https://www.ruby-forum.com/t/scraping-off-a-word-document/157927/7 "2009-01-18T13:30:50Z")
On 18.01.2009 03:18, gregarican wrote: > I wound up writing a C# console program to do the work. I just > referred to ugly underbelly of all of the Word COM stuff and was able > to grab what I needed. It took awhile though, since my text was > contained within text frames. So I had to work with the > Document.Shapes property and whatnot. I’d say that’s pretty fast. Good job! > in a single language. But that’s another fight for another day 🙂 > May I suggest a different approach? Since your primary step is pulling > data from a relational DB using Ruby, you could as well do this: open > the mail merge Word template, replace mail merge fields with text with > special formatting (for example “\<\<\>\>” or whatever doesn’t > collide with RTF meta sequences). Then you save this as RTF file (ASCII > readable). Now you only need to read in the mail template file from > Ruby, do all the replacements and then write it out in Ruby again once > for each record. Sounds pretty simple IMHO. Kind regards robert ---
**Author:** ![gregarican](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/g/7feea3/32.png) [@gregarican](https://www.ruby-forum.com/u/gregarican)\ **Post date:** [January 18, 2009, 6:15am UTC](https://www.ruby-forum.com/t/scraping-off-a-word-document/157927/8 "2009-01-18T06:15:48Z")
“gregarican” [removed\_email\_address@domain.invalid](mailto:removed_email_address@domain.invalid) wrote in message news:removed\_email\_address@domain.invalid… On Jan 17, 11:18 am, Phlip [removed\_email\_address@domain.invalid](mailto:removed_email_address@domain.invalid) wrote: > interoperability. > I wound up writing a C# console program to do the work. I just > referred to ugly underbelly of all of the Word COM stuff and was able > to grab what I needed. It took awhile though, since my text was > contained within text frames. So I had to work with the > Document.Shapes property and whatnot. In searching for a solution I did run across a VBA code snippet that would save off each document separately after the mail merge completed. At least now I have a totally automated solution, although it’s cobbled together from various sources. First I pull my data from a SQL DB using Ruby, dumping that to an Excel data source. Then I have a C# program that takes that data source, uses a Word mail merge template and delivers the final document set. Finally, I have a Ruby program that looks in that save directory and e-mails the documents to the individual employees. Eventually it’d be a lot cleaner and easier to maintain if I had all of the work done in a single program written in a single language. But that’s another fight for another day 🙂 Well, if anybody can figure this out, it's you. ---
**Author:** ![gregarican](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/g/7feea3/32.png) [@gregarican](https://www.ruby-forum.com/u/gregarican)\ **Post date:** [January 18, 2009, 5:27pm UTC](https://www.ruby-forum.com/t/scraping-off-a-word-document/157927/9 "2009-01-18T17:27:20Z")
Robert K. wrote: > On 18.01.2009 03:18, gregarican wrote: > > > I wound up writing a C# console program to do the work. I just > > referred to ugly underbelly of all of the Word COM stuff and was able > > to grab what I needed. It took awhile though, since my text was > > contained within text frames. So I had to work with the > > Document.Shapes property and whatnot. > > I’d say that’s pretty fast. Good job! > > > in a single language. But that’s another fight for another day 🙂 > > May I suggest a different approach? Since your primary step is pulling > > data from a relational DB using Ruby, you could as well do this: open > > the mail merge Word template, replace mail merge fields with text with > > special formatting (for example “\<\<\>\>” or whatever doesn’t > > collide with RTF meta sequences). Then you save this as RTF file (ASCII > > readable). Now you only need to read in the mail template file from > > Ruby, do all the replacements and then write it out in Ruby again once > > for each record. Sounds pretty simple IMHO. > > Kind regards > > robert FYI, a similar (though not necessarily better) solution using Find & Replace in Word is demonstrated here: [Ruby on Windows: Find & Replace with MS Word](http://rubyonwindows.blogspot.com/2007/11/find-replace-with-ms-word.html) Greg: If you’re willing to share your C# code for automating Word, I, for one, would like to see it. Feel free to email me, if you like. David