<div>
Do the PDFs contain actual text data (as opposed to scanned pages of text, which would need to be OCR'ed first). &nbsp;If they do contain text, there's a command line utility as part of Ghostscript called &quot;pdf2text&quot; that will easily extract out all the textual data from your PDFs.<br />
<br />
If you're running windows, you might be able to easily install it using Cygwin.
</div>


<div>
It is actually probably a bit more complicated question. &nbsp;I've been playing around and it seems there many different formats for PDF documents. &nbsp;I have basically 2 types and a fair bit of software on my computer.<br />
<br />
1. &nbsp;Powerpoint presentations that I want to convert to PDF files. &nbsp;I can do this in several ways:<br />
<br />
a. &nbsp;Use Powerpoint to save them as a PDF.<br />
b. &nbsp;Use the Mac OS print feature to save them as a PDF to disk.<br />
b. &nbsp;Use the Mac OS print feature to save them as an Adobe PDF (I have Adobe Acrobat Pro also).<br />
<br />
2. &nbsp;Various journal articles from medicine that I get from a number of resources.<br />
<br />
<br />
It seems that the journal articles and the documents that I could create using the Adobe PDF creator have more information (i.e. the text) than the ones that I create the other ways, which seem to be bitmaps.<br />
<br />
I do have a program called Wondershare PDF creator that can apparently also convert PDF's to text.<br />
<br />
Just kind of wondering what the best way to go is.<br />
<br />
For some of the journal articles I am able to search the PDF LargeBlob directly and get hits with my search string in MySQL, although it seems to be case sensitive there. &nbsp;Seems like it might be best to create text version of all of the PDF's and store those in the database as well when performing searches since I guess I can do full text searches and get around the case sensitivity as well. &nbsp;Is that true.<br />
<br />
Can you give an example of how to create a query to search for keywords or phrases in the pdf or text doc?
</div>


<div>
I think we need to see your test data. &nbsp;Please post the information you have in your data base blob columns, thanks. ~Ray
</div>


<div>
Well, here are some examples of the PDF created in various ways, as well as the original ppt file, and a converted text file.<br />
<br />
They are at:<br />
<br />
<a href="http://sscotti.com/CaseBreast1adobe.pdf" rel="ugc">http://sscotti.com/CaseBreast1adobe.pdf</a>&nbsp;, saved with save as adobe pdf<br />
<a href="http://sscotti.com/CaseBreast1macos.pdf" rel="ugc">http://sscotti.com/CaseBreast1macos.pdf</a>, saved as pdf with os x print<br />
<a href="http://sscotti.com/CaseBreast1ppt.pdf" rel="ugc">http://sscotti.com/CaseBreast1ppt.pdf</a>, saved as pdf from save as in ppt<br />
<a href="http://sscotti.com/CaseBreast1.ppt" rel="ugc">http://sscotti.com/CaseBreast1.ppt</a>, original powerpoint file<br />
<a href="http://sscotti.com/CaseBreast1.txt" rel="ugc">http://sscotti.com/CaseBreast1.txt</a>, original powerpoint file<br />
<br />
Just examining them with a text/hex editor, using BBEdit, it looks like the one saved as an Adobe PDF file retains the searchable text data, although in binary format since I have it stored as a large blob.<br />
<br />
So question. &nbsp;If I store the raw pdf data as long text instead of a large blob, can I search the raw data without case insensitivity and still use the field to store the pdf.<br />
<br />
I think I have the mysql queries worked out for using the LIKE match.<br />
<br />
DATABASE with table pdf_files has one table with 6 columns:<br />
<br />
pid &nbsp;INT<br />
section VARCHAR &nbsp;(Breast, Cardiac, etc)<br />
name VARCHAR (Title for document)<br />
casenumber INT (case number for document)<br />
pdfdata LONGLBLOB (contains binary data for pdf or other type of doc, i.e. word, ppt, etc., just using pdf for now) &nbsp;can I change to LONGTEXT and do FULLTEXT search and still use for pdfdata, think the pdfdata has to be binary.<br />
pdftext LONGTEXT, trying this to store converted plain text version<br />
<br />
query is like:<br />
<br />
$query = &quot;SELECT * from pdf_files WHERE pdftext LIKE '%&quot;.postUrlStringValue('se<wbr />archText',<wbr />null).&quot;%' ORDER BY section, casenumber, name ASC&quot;;<br />
<br />
or <br />
<br />
$query = &quot;SELECT * from pdf_files WHERE pdfdata LIKE '%&quot;.postUrlStringValue('se<wbr />archText',<wbr />null).&quot;%' ORDER BY section, casenumber, name ASC&quot;;<br />
<br />
I don't know how to do a FULLTEXT search, but you can't do that on a LONGBLOB anyway I think.
</div>


<div>
Hey, thanks for that. &nbsp;I have gone back and forth with storing the data in a database vs. on the server. &nbsp;I have another app for radiology that works in a similar fashion, although the BLOB data are mostly pictures there and generally are not any larger than 500K at most each, with a sprinkling of other documents. &nbsp;I like having them in a database as it is a single file, a little more portable and in some ways easier to manage. &nbsp;With the arrangement I have now I probably would be better of storing them on the server and putting other parameters in the database.<br />
<br />
The actual app is here: &nbsp;sscotti.com<br />
<br />
Haven't implemented your script, but will, thanks.<br />
<br />
Stephen
</div>


<div>
Thanks for the points. &nbsp;This is a really great question, ~Ray
</div>


<div>
Here is another question. &nbsp;In the PDF documents that I've created it looks like the pdf creator is putting text on the slides from the ppt presentation encloded between parentheses: &nbsp;e.g.<br />
<br />
&quot;(52 yo male with palpable )Tj 2.042 -1.182 Td [(lump left breast)&quot;<br />
<br />
surrounded by formatting data. &nbsp;The text that I actually want to search in the pdf is the text between the parentheses. &nbsp;Is there a way to use a regex in the preg_match function to look for an occurrence of my search text in between parentheses? &nbsp;If I can do that things may actually work OK. &nbsp;I don't need preg_match_all. &nbsp;I am just looking for at least one match, so preg_match works. &nbsp;I tried it in my code, and the performance seems similar to searching the database with SQL using LIKE &quot;%match%&quot;.<br />
<br />
Thanks.<br />
<br />

</div>


<div>
Yes, you can search for the text between parentheses. &nbsp;Your REGEX will need to escape the parens (the are metacharacters to REGEX).
</div>


<div>
<div class="content wysiwyg-content">
I have a bunch of pdf files that I have stored as large blobs in a database, which seems to be working well for managing them and displaying them. &nbsp;I also want to be able to search the text in the documents using PHP and MySQL to look for keywords or phrases using either the LIKE mysql function or FULLTEXT search. &nbsp;It seems that I would best off converting the PDF blob data into a simple text version with the extracted data in free text form. &nbsp;Is there a PHP class or function that can do this, and can you suggest how I would then go about searching the free text data for a keyword or phrase. &nbsp;<br />
<br />
I already have the PDF's in a MySQL database table called pdf_files, with the large blob field simply called pdf.
</div>
</div>


I have a bunch of pdf files that I have stored as large blobs in a database, which seems to be working well for managing them and displaying them.  I also want to be able to search the text in the documents using PHP and MySQL to look for keywords or phrases using either the LIKE mysql function or FULLTEXT search.  It seems that I would best off converting the PDF blob data into a simple text version with the extracted data in free text form.  Is there a PHP class or function that can do this, and can you suggest how I would then go about searching the free text data for a keyword or phrase.  

I already have the PDF's in a MySQL database table called pdf_files, with the large blob field simply called pdf.

Convert pdf to free text form for fulltext search

MySQL is an open source, relational database management system that runs as a server providing multi-user access to a number of databases. Acquired by Oracle in 2009, it is frequently used in combination with PHP installations, powering most of the WordPress installations.

MySQL Server

PHP is a widely-used server-side scripting language especially suited for web development, powering tens of millions of sites from Facebook to personal WordPress blogs. PHP is often paired with the MySQL relational database, but includes support for most other mainstream databases. By utilizing different Server APIs, PHP can work on many different web servers as a server-side scripting language.