Ideas for converting .doc files to HTML to a Form then to a Database?

Cherphas

Hi,

Is there a way to read a Word document with php and detecting styles (headings and such, I think this is used to convert a .doc file to xml?) of the document?

I want to parse a word document, seperate the Titles, Subtitles and Body and store them in each its own column in an MySQL database.

The person writing my Word documents would use a certain format. He would use a certain heading(does this have anything to do with macros?) for the Titles and a certain heading for the Paragraphs, therefore when php reads it and incounters a certain heading, it would know in which column to store it.

Better yet! I have an html form that I use to populate the database. This way I don't have to work with id indexing to relate multiple tables, everything is done automatically. Is there a way that I can read the document, populate the text fields and textareas in the same manner as I explained above, then I could submit it to the database?

I also need to preserve the html tags for tables, lists, lines breaks andimages. The rest can be striped out.

I was thinking I could create a php page with a button, when pressed, it would open the document, converting it to HTML, cleaning it somehow(Dreamweaver doesn't do a good job with that!) and preserve only the tags I need. While reading the document, php would detect titles and columns and would put a custom tag around it, something like <title></title>. Then it would load the form I created and populate it, putting everything in <title> in the Title field and everything in <body> in the definition field. Then I could submit it to the database! Whew!!

Sorry for writing so much text, I was getting these ideas while I was writing.

Before I start coding myself to death, I just wan to know if this seems possible?

Cherphas

After re-reading my post, I guess what I really want to know is if it is possible to detect properties or styles in a word document. I think all the rest can be done fairly easy.

drawmack

You could do a couple of things, one attack it as if word were encrypting the plain text (which it is kind of) and you are making a known plain text attack. Here is the process:

1) Create serveral documents in the format of the document you wish to process (20 should be sufficient) with each document having different information of different lengths for the different sections.

2) Analyze the documents in a text or hex editor a find a pattern that exists before and after each of the seperate elements.

3) Write a php script that uses the unique patterns from step 2 to parse the document into the fields

This shouldn't be alll that hard since it's uses a markup language.

The other method is to use an html file generated by word. It will still involve the same process as above but it will be much simplier as you know the markup language it is using.

The third method is to trash word and have him create textfiles in the format of

[title]This is the title[/title]
[subject]This is the subject[/subject]
[body]Look ma an article.[/body]

and then to just write a processor for that.

Cherphas

Anymore ideas?