Btrieve / Pervasive PSQL

Btrieve (now Pervasive PSQL, Actian Zen) is a key-value database system with optional relational SQL layer. It uses B-tree indexing and is widely deployed in vertical market applications.

What You Can Upload

• .dat or .btr Btrieve data files
• .ddf data dictionary files (FILE.DDF, FIELD.DDF), alongside the data files: a dictionary with no .dat or .btr file has nothing to convert, and the upload is refused
• ZIP archive with complete file sets

What You Get Out

DataMeans converts your data into multiple modern formats:

Output	Description
csv/{TableName}.csv	One CSV file per table with all row data
xlsx/{TableName}.xlsx	Excel workbook per table (omitted for tables over 1,048,575 rows, 16,384 columns or 8,000,000 cells - see note below)
xls/{TableName}.xls	Legacy Excel format per table (omitted for tables over 65,535 rows, 256 columns or 150,000 cells - see note below)
json/{TableName}.json	JSON array of records per table
json/{TableName}.jsonl	Newline-delimited JSON (streaming-friendly)
postgres.sql	PostgreSQL CREATE TABLE + INSERT statements
mysql.sql	The same schema and rows, in MySQL dialect
sqlite.sql	The same schema and rows, in SQLite dialect
ansi.sql	The same schema and rows, in portable ANSI SQL
sqlserver.sql	The same schema and rows, in SQL Server (T-SQL) dialect
report.json	Structured conversion report
report.md	Human-readable conversion summary

How to Export / Obtain Files

1. Locate your application's Btrieve data directory
2. Copy the data files (.dat or .btr)
3. Include the .ddf dictionary files (essential for field names and types)
4. Create a ZIP with all files
5. Upload the ZIP to DataMeans

Supported Features

• Btrieve file formats 3.x to 6.x, checked record for record against an independent Btrieve reader on real files
• The 7.x to 9.x formats written by Pervasive.SQL and Pervasive PSQL, read the same way (see Known Limitations)
• Data dictionary (.ddf) parsing for table and field definitions
• Variable-length records
• Files without a dictionary: key fields are read from the file's own key definitions, and every other byte of each record is laid out around them, so nothing is left out
• Field types including BFLOAT (Microsoft Binary Format), TIMESTAMP and Unicode strings

Known Limitations

• Without the .ddf files, column names cannot be recovered. Key fields are named after their key (KEY0_0), other fields by position (FIELD_3), and the table is marked for review. Fields outside the keys are worked out from the records; anything that cannot be, including the variable-length part of each record, comes out as raw bytes (base64). On the job page you can set the field boundaries and names yourself
• The 7.x to 9.x formats have not yet been checked against a real file of those versions. Each file is checked as it is read, against its own page table and the record count it states: if the count read differs, the report says so, and a file whose pages cannot be read is reported as unreadable rather than returned as an empty table
• Compressed (RDC) files are refused rather than half-read; the upload screen names the file and asks you to export it with Pervasive's BUTIL utility first
• Files protected with an owner name (encrypted) are not detected, and their data comes out unreadable: remove the owner name in your Btrieve tools first
• Actian Zen 13.0-format files are not read
• CURRENCY, BIT and DATETIME fields are delivered as their raw bytes (base64), with a note in the report
• Indexes are not carried over

Note on the Excel formats. An .xlsx sheet holds at most 1,048,575 rows of data and 16,384 columns, and a legacy .xls sheet 65,535 rows and 256 columns. Rather than ship a file that looks complete but is silently cut short, DataMeans leaves a table out of any Excel format it does not fit and notes it in the job report. The .xlsx copy is also left out for tables of more than 8,000,000 cells (1,048,575 rows of 12 columns is 12.6 million), which take too long and too much memory to write, and the .xls copy for tables of more than 150,000 cells, which the legacy format takes far too long to write. The CSV, JSON and SQL outputs always contain every row.

Overview

Btrieve (now Actian Zen) is a transactional database engine using Indexed Sequential Access Method (ISAM) for high-performance data storage and retrieval. Originally developed for embedded applications, it provides record-level access with multi-user concurrency, transactional integrity, and flexible indexing. The engine uses proprietary file formats optimized for speed and reliability in business applications.

History and Background

• 1982: SoftCraft develops Btrieve as a record manager.
• 1987: Novell acquires SoftCraft, integrates Btrieve with NetWare.
• 1994: Novell spins off Btrieve to BTI (Btrieve Technologies Inc.).
• 1996: BTI is renamed Pervasive Software.
• 1998: Pervasive.SQL 7 combines Btrieve 7.0 with Scalable SQL 4.
• 2005: Pervasive PSQL v9 expands maximum file sizes with the 9.0 and 9.5 formats.
• 2007: Pervasive PSQL Summit v10 introduces 64-bit engine support.
• 2010: Pervasive PSQL v11 adds full 64-bit and IPv6 support.
• 2013: Actian acquires Pervasive Software.
• 2019: Zen v14 completes the rebranding from Pervasive PSQL to Actian Zen.
• 2024: Actian Zen v16 introduces the 16.0 file format with long key support.

File Format Specifications

Btrieve uses proprietary ISAM-based file formats with efficient indexing and storage.

File Extensions:

• .btr: Conventional extension for Btrieve data files (binary ISAM format)
• .mkd: Conventional extension for MicroKernel Database Engine data files
• .ddf: Data dictionary files describing table schema for the relational engine
• No separate index files - indexes embedded in the data file

File Structure:

• Fixed-length or variable-length records
• B-tree indexes for fast key-based access
• Page-based storage with page sizes from 512 to 16,384 bytes
• Optional record compression and page compression (9.5 format onward)
• Transaction log segments for recovery (.log files)

Key Components:

• Data pages: Store actual record data
• Index pages: B-tree nodes for key navigation
• File control record: Metadata including format version and page size
• Page allocation tables: Map logical pages to physical pages (6.0 format onward)
• Variable-tail allocation tables: Locate overflow portions of variable-length records (6.15 onward)

Data Types and Structures

Btrieve supports the following documented key types:

Type Code	Type	Size	Description
0	String	Variable	Character data ordered left to right, blank-padded
1	Integer	1, 2, 4, or 8 bytes	Signed binary integer
2	Float	4 or 8 bytes	IEEE single or double precision
3	Date	4 bytes	Day, month, and year (0001 to 9999)
4	Time	4 bytes	Hundredths, seconds, minutes, hours
5	Decimal	Variable	Packed decimal, two digits per byte
6	Money	Variable	Decimal representation with two implied decimal places
7	Logical	1 or 2 bytes	Boolean collated as a string
8	Numeric	Variable	Zoned decimal stored as ASCII digits
9	BFloat	4 or 8 bytes	Microsoft Basic floating point (legacy)
10	LString	Variable	Length-prefixed (Pascal-style) string
11	ZString	Variable	Null-terminated (C-style) byte string
12	Note	Variable	Legacy text type, superseded
13	LVar	Variable	Legacy variable-length type, superseded
14	Unsigned Binary	Variable	Unsigned integer compared byte by byte
15	Autoincrement	2, 4, or 8 bytes	Automatically assigned ascending integer
25	WString	Variable	UCS-2 Unicode string, not null-terminated
26	WZString	Variable	Double-null-terminated Unicode string

Database Structure:

• Records: Fixed-length or variable-length data units
• Keys: Indexed fields for fast access (up to 119 keys per file)
• Segments: Multi-field composite keys (up to 420 segments per file in the 9.5 format with 8 KB or larger pages)
• Duplicates: Keys may be unique or allow duplicate values
• Referential integrity: Enforced by the relational engine using data dictionary definitions

Version Differences

Version	Year	Key Features	File Format
Btrieve 5.x	1988	NetWare-native engine; transaction logging in 5.1	5.x, pre-image paging
Btrieve 6.0	1992	Shadow paging	6.0, page allocation tables
Btrieve 6.15	1994	MicroKernel Database Engine	6.x, variable-tail allocation tables
Pervasive.SQL 7	1998	Btrieve 7.0 bundled with Scalable SQL 4	7.x
Pervasive PSQL v9	2005	Larger databases, improved Linux support	9.0 (128 GB); 9.5 (256 GB)
Pervasive PSQL v10	2007	First 64-bit engine support	No format change
Pervasive PSQL v11	2010	Full 64-bit and IPv6 support	No format change
PSQL v13	2017	8-byte autoincrement, Android and iOS support	13.0 (terabyte file sizes)
Actian Zen v14	2019	Zen rebranding, AES-256 encryption, JSON filtering	No format change
Actian Zen v15	2021	SQL windowing functions	No format change
Actian Zen v16	2024	Long keys, Kafka integration	16.0

Compatibility Notes:

• Later engines open files created in earlier formats; 5.x and older files are read-only until rebuilt
• Btrieve 6.x and earlier engines cannot open files in the 7.0 or later formats
• Automatic in-place format upgrade applies only from 8.x to 9.0
• The Rebuild utility converts files between format versions
• Page compression requires the 9.5 file format or later

Technical References

• Wikipedia: Btrieve
• Wikipedia: Actian Zen
• Actian Zen Data Types
• Actian Zen Converting Data Files
• BTR File Format