insert a .doc file into MSSQL urgent!

[sone_br]

Hello, i´m from BRAZIL...

I need to insert a .doc or .xls file into a database. I´m using MSSQL 2000+PHP3.
- first i upload the .doc file;
- second i read the file and put it in a $string;
- finaly i insert the $string into a database field ('TEXT' field type);

But when i got the data from database, was just retuned a piece of original file data.
(The doc file size is about 70Kb.)What happen?

Please help me with a solution. 🙂
Well thanks for your attention,
[sorry, my english is not good]

Anderson Sone
Brazil-PR-Curitiba

[sone_br]

My php action is like this :

<?php
...
copy("$arquivo","tmp_files/$arquivo_name");

//open file
$data = fopen("tmp_files/$arquivo_name", "r");
//read file
$conteudo = fread($data,filesize("tmp_files/$arquivo_name"));

$conteudo = bin2hex ($conteudo);

$con = mssql_connect("SERVIDOR","sa","");
mssql_select_db("SISAC", $con);

$result =  mssql_query ("INSERT INTO ARQUIVOS (DESCRICAO,DADOS,ARQ_NOME,ARQ_TAM,ARQ_TIPO)
VALUES ('$arquivo','$conteudo','$arquivo_name','$arquivo_size','$arquivo_type')","$con");

...
?>

The real problem is that only part of file was inserted in a database.

[fleymingmasc]

Hmmm...

My first suspicion would be that you are exceeding the maximum length of a PHP string data type. The PHP manual is silent on this issue, so I don't know how long of strings are permitted.

Also, working with blob fields of any kind in MSSQL is necessarily cumbersome, because the blob is actually stored in an internal "file" structure, with a 16-byte pointer to that file actually residing in the table. Our company encountered all kinds of problems with TEXT fields and eventually moved to the solution outlined below.

This will work with MSSQL 7 and up and will kill both these birds with one stone: MSSQL 7+ supports varchars of up to 8K long, so we can create a table where the document is broken up into fragments. Here's how I've done it before:
CREATE TABLE msword_documents
document_id INT,
sequence_id INT,
document_fragment VARCHAR(8000)

Split your .doc file up into 8000-byte sections and insert each fragment into a row, with the document_id identifying the whole document, and the sequence_id incrementing for each fragment of that document. In your original table, store the document_id instead of the TEXT field. When you want to reconstitute the file, do the following SELECT:

"SELECT document_fragment FROM msword_documents WHERE document_id = $doc_id ORDER BY sequence_id"

And then iterate the resultset, concatenating the results.