HXTT PDF
Version
2.0.275
License
Shareware
Trial Limitations
30 days, at most 50 queries per session, SELECT returns the first 100 rows
Operating Systems
Windows, Linux, macOS, Unix, Solaris, AIX, FreeBSD and any OS with a Java VM
System Requirements
Any platform with a Java VM (JDK 1.6 to Java 14), JDBC 4.0 to 4.3 packages
File Size
11.10 MB
Release Date
7/13/2026
Last Updated
11/22/2026
3 views 0 downloads
Pure Java JDBC driver that reads, writes and creates PDF text with SQL

HXTT PDF: Read, Edit and Create PDF Text with SQL from Java

HXTT PDF is a pure Java (type 4) JDBC driver that opens PDF files as databases: each page is a table and each line of text a row, so you can extract text with SELECT, change it with UPDATE and build new PDF pages with INSERT. It's for Java developers who need to pull text out of PDFs or produce simple text PDFs from a database, on any platform. A Single Server License costs $395, one-time.

PDF is built for printing, not for data, so getting text out usually means a PDF library and code that walks content streams and sorts fragments by position. HXTT PDF does that work and hands you rows with the line's text, its page and its X and Y coordinates. SELECT STRING_AGG(textline, '\r\n') FROM invoice.page1 returns the text of page 1, and a WHERE CoordinateY < 100 picks out just the header.

Quick facts

  • Version: 2.0.275, built July 13, 2026
  • Developer: Heng Xing Tian Tai Lab of Xi'an City (HXTT), which also makes JDBC drivers for Access, DBF, Excel and CSV files
  • Platform: any operating system with a Java VM, Java 6 and later; JDBC 4.0, 4.1, 4.2 and 4.3 packages
  • Formats: PDF 1.0 to 1.7, including password-protected files with 128-bit RC4, 128-bit AES or 256-bit AES encryption
  • Price: $395 Single Server, $1,767 100 User Application, $2,536 Enterprise; perpetual, with one year of free upgrades and email support
  • Trial: free 30-day evaluation package; it runs at most 50 queries per session and returns only the first 100 rows of a SELECT
  • Refunds: requested through 2Checkout within 30 days of the order
  • What we checked: the evaluation download pdf.zip (11.1 MB), whose PDF_JDBC40.jar reports version 2.0.275 of July 13, 2026, and its demo SQL script; order page prices of $395, $1,767 and $2,536 for the three licenses, all lifetime, on November 22, 2026. We didn't run it on PDFs of our own.

Download Free TrialBuy Single Server LicenseBuy 100 User Application LicenseBuy Enterprise License

Overview

HXTT PDF is a JDBC driver and a small database engine in one jar. Load com.hxtt.sql.pdf.PDFDriver and connect with a URL such as jdbc:pdf:/c:/data. Each PDF in that folder is a database: testpdf is the whole file, and testpdf.page1, testpdf.page2 and so on are its pages. A page's rows hold one line of text each, with PageNo, CoordinateX, CoordinateY, the textline itself and a Words column, a JSON array with every word's text and position. Pseudo columns such as FormNo and Texts are there too.

It works in both directions. CREATE DATABASE testpdf makes a new PDF, CREATE TABLE testpdf.page2 adds a page, and INSERT writes lines at the coordinates you give. Page size (A0 to A6, Letter, Legal or a custom size), rotation, font and font size are set with SQL variables such as CURRENT_Page and CURRENT_Font.

The engine runs embedded in your application or as a separate server, managed with the DBAdmin tool in the jar, which can run as a Windows service, a Unix or Linux daemon or a normal program.

HXTT DBAdmin Server Information view for a local server with no open connections

DBAdmin, the server manager inside every HXTT driver jar, in an older Windows look, showing the Server Information view of a local server. HXTT's sample connection points at a DBF folder.

Features

Text extraction with positions

HXTT PDF returns every line of text with its page and coordinates, and every word with its own position in the Words JSON column, which you can read with JSON operators such as words#>'{0,text}'. That makes it practical to pull a value from a fixed spot on a form or report, such as an invoice number at the top right, with a WHERE clause on the coordinates.

Edit text in place

HXTT PDF changes existing PDFs with SQL: UPDATE replaces a line or a single word (HXTT's FAQ uses this to fill in a form), DELETE removes lines from a page, and TRUNCATE TABLE leaves a blank page. HXTT says documents are updated incrementally and support multi-user access.

Create PDFs from data

HXTT PDF builds new PDF files and pages with CREATE DATABASE and CREATE TABLE, and fills them with INSERT, or with CREATE TABLE ... SELECT straight from a query. Page size, rotation, standard PDF fonts (Courier, Helvetica, Times) and font size are set in SQL.

Password-protected PDFs

HXTT PDF opens encrypted PDFs at every standard level: 128-bit RC4 (Acrobat 6), 128-bit AES (Acrobat 7) and 256-bit AES (Acrobat X and later). The password is set per connection with the CURRENT_Password variable and isn't visible to other connections.

SQL and access options

HXTT PDF supports nearly full ANSI SQL-92 plus many SQL:2008 additions, more than 250 SQL functions, joins, common table expressions, PIVOT and UNPIVOT, STRING_AGG and ARRAY_AGG, all four transaction levels, and PDFs on local folders, SMB shares, HTTP, HTTPS, FTP and SFTP URLs, or inside ZIP, GZ, BZ2, 7z and TAR archives.

Benefits

  • Text out of PDFs with one query, including where on the page each line sits.
  • No PDF library to learn. If you know SQL and JDBC, you can read and write PDF text.
  • Works in Java tools. DBeaver, JasperSoft Studio and other JDBC clients can browse a folder of PDFs as tables.
  • A perpetual license with a year of upgrades and support included.

Real-World Uses

  • Invoice and statement capture: a nightly job reads supplier PDFs from an SFTP folder and pulls the invoice number, date and total by their position on page 1.
  • Searching a document archive: a service loads the text of every PDF in a ZIP into a search index with STRING_AGG.
  • Simple generated documents: an application writes a plain text PDF, such as a packing list or a log summary, straight from a query.
  • Correcting a typo across files: an UPDATE replaces a wrong word on the same line of a batch of generated PDFs.

Who It's For, and Who Should Skip It

HXTT PDF is for Java developers and integrators who need the text of PDFs, with positions, inside a Java application or JDBC tool, or who want to produce simple text PDFs from data, on any operating system.

Skip it if you need to design rich PDFs with tables, images and layout: a reporting tool or a PDF generation library fits better. It reads text, not scanned images, so scanned paper documents need OCR first. If you don't write Java, a desktop PDF editor or converter is the right tool. And for heavy PDF manipulation (merging, signing, redaction), use a full PDF library.

Usage Guide

  1. Download the evaluation package (pdf.zip, about 11 MB) and unzip it. It holds the JDBC 4.0 jar, documentation and a demo SQL script; other JDBC packages are available after registering on HXTT's site.
  2. Add PDF_JDBC40.jar to your classpath (Java 6 or later).
  3. Load the driver com.hxtt.sql.pdf.PDFDriver and connect with jdbc:pdf:/your/folder, or jdbc:pdf://host:port/folder for a remote server.
  4. Read a page: SELECT PageNo, CoordinateY, textline FROM file.page1. For an encrypted file, set CURRENT_Password first.
  5. Try writing with the statements in the demo's sql.txt, which creates a test PDF, adds pages and lines and edits them.
  6. Buy the license that fits how you'll deploy it before the evaluation period ends.

Where it falls short

  • The trial is tight. The evaluation jar runs at most 50 queries per session, returns only the first 100 rows of a SELECT, and stops after 30 days, so test a few short documents.
  • Text only. Scanned PDFs without a text layer give nothing back. Tables and layout aren't exposed as such, only lines of text with positions.
  • Listed versions stop at PDF 1.7 and Java 14, and the most recent entries in the version log are from 2024.
  • Updates are included for one year. After that, upgrades and support are a yearly renewal. Redistributing it in your product needs the $1,767 license.

Comparisons

HXTT PDF vs a PDF library such as Apache PDFBox

A free PDF library can extract text, render pages, merge files and much more, but you write Java against its object model, and positions take extra code. HXTT PDF covers a narrower job, text in and out, through SQL and JDBC, with coordinates and word positions ready to filter on. Choose the library for full PDF manipulation; choose the driver when you want PDF text to behave like database rows.

Editions and Pricing

LicensePriceWhat it coversBuy
Single Server License$395Unlimited connections on a single serverBuy Single Server
100 User Application License$1,767Unlimited connections; redistribute 100 times with one commercial productBuy 100 User
Enterprise License$2,536Unlimited connections and servers, within one company in one countryBuy Enterprise

Every license includes one year of free online upgrades and technical support by email and forum. HXTT offers bulk discounts on the order page.

FAQs

How do I get all the text of a PDF with HXTT PDF?

HXTT PDF returns the whole document's text with SELECT STRING_AGG(textline, '\r\n') AS content FROM testpdf, or one page's text by querying testpdf.page1 instead.

Can HXTT PDF read password-protected PDF files?

Yes. HXTT PDF opens PDFs encrypted with 128-bit RC4, 128-bit AES and 256-bit AES once you set the password with the CURRENT_Password variable for that connection.

Can HXTT PDF read scanned documents?

No. HXTT PDF reads the text stored in a PDF; a scanned page is an image with no text layer, so it needs OCR before the driver can return anything.

Can HXTT PDF create a new PDF file?

Yes. HXTT PDF creates a PDF with CREATE DATABASE, adds pages with CREATE TABLE and writes text lines with INSERT, at coordinates you choose, with page size and font set in SQL.

Which Java versions does HXTT PDF need?

HXTT PDF needs Java 6 or later; it comes in JDBC 4.0, 4.1, 4.2 and 4.3 packages, and HXTT lists versions up to Java 14.

Which license do I need to ship HXTT PDF inside my own software?

To redistribute HXTT PDF with a commercial product, the 100 User Application License covers 100 copies of one product. The Single Server and Enterprise licenses are for running it on your own servers.

How do I get a refund?

Orders for HXTT PDF placed online are handled by 2Checkout, which accepts refund requests up to 30 days from the order date.

System Requirements

  • Any operating system with a Java VM: Windows, Linux, macOS, Unix, Solaris, AIX, FreeBSD and others
  • Java 6 (JDK 1.6) through Java 14, with the matching JDBC 4.0 to 4.3 package

Conclusion

HXTT PDF is worth its $395 Single Server price if your Java code has to read text out of many PDFs, with positions, or write simple text PDFs, and you'd rather use SQL than a PDF object model. The reason to buy it is that page text arrives as rows with coordinates and word positions, from encrypted files and remote folders too. The reason to skip it is that it does one job: for rich layouts, scanned documents or full PDF editing you need other tools. Check it against your own documents with the 30-day evaluation first.

Download Free TrialBuy Single Server LicenseBuy 100 User Application LicenseBuy Enterprise License

Other Software Products by HXTT:
1) HXTT Json
2) HXTT Word
3) HXTT XML
4) HXTT Cobol
5) HXTT Excel
6) HXTT Paradox
7) HXTT Text (CSV)
8) HXTT DBF
9) HXTT Access

We may earn a commission when you buy through links on this page.
Copyright © 2026, Ivertech. All rights reserved.