@geoharkat/dlis-parser

Parse RP66 V1 (DLIS) well-log binary files — in the browser and Node.js with zero runtime dependencies.

---

Features

• Full RP66 V1 / DLIS binary parsing (Storage Unit Label, Visible Records, Logical Record Segments)
• Correct EFLR parsing: CHANNEL, FRAME, ORIGIN, PARAMETER, ZONE, EQUIPMENT, TOOL, CALIBRATION families
• FDATA IFLR decoding with proper per-frame UVARI frame-number handling
• All 25 RP66 V1 representation codes (including IBM float, VAX float, FSHORT, complex)
• Multi-dimensional and waveform array channels
• Export to CSV and LAS 2.0
• Vendor quirks: Baker Hughes ATLAS ABSATR extension, Schlumberger VR version byte variants
• Works in browser (fromBuffer) and Node.js (fromFile)
• Full TypeScript type declarations included

---

Install

npm install @geoharkat/dlis-parser

---

Quick start

Node.js

import { DLISFile } from '@geoharkat/dlis-parser';

const file = await DLISFile.fromFile('well.dlis');

// Well metadata
const origin = file.origin;
console.log('Well:', origin.wellName, '|', 'Field:', origin.fieldName);
console.log('Company:', origin.company);

// List frames and channels
for (const frame of file.frames.values()) {
  console.log(`Frame "${frame.name}"  index: ${frame.indexType}  channels: ${frame.channelNames.join(', ')}`);
}

// Decode a frame
const frame = file.logicalFiles[0].getFrame('MAIN');
const { frameCount, data, channels } = frame.decode();

console.log(`${frameCount} depth samples`);
console.log('DEPTH[0]:', data.DEPTH[0], channels.find(c => c.name === 'DEPTH').units);
console.log('GR[0]:',    data.GR[0]);

// Export
const csv = frame.toCSV();
const las = frame.toLAS();

Browser

<input type="file" id="fileInput" accept=".dlis">
<script type="module">
import { DLISFile } from 'https://cdn.jsdelivr.net/npm/@geoharkat/dlis-parser/src/index.js';

document.getElementById('fileInput').addEventListener('change', async e => {
  const buf = await e.target.files[0].arrayBuffer();
  const file = DLISFile.fromBuffer(buf);

  for (const frame of file.frames.values()) {
    const result = frame.decode();
    if (result) console.log(frame.name, result.frameCount, 'frames');
  }
});
</script>

---

API

DLISFile

Member	Description
DLISFile.fromBuffer(buffer: ArrayBuffer)	Parse from an in-memory buffer (sync)
DLISFile.fromFile(path: string)	Parse from a file path (async, Node.js only)
.sul	Storage Unit Label ({ version, structure, maxRecordLength, storageSetId, … })
.logicalFiles	Array of LogicalFile
.origin	Shortcut → logicalFiles[0].origin
.frames	Shortcut → logicalFiles[0].frames
.channels	Shortcut → logicalFiles[0].channels
.warnings	Non-fatal parse warnings (string[])

LogicalFile

Member	Description
.id	File ID string (from FILE-HEADER)
.origin	Well/acquisition metadata (OriginInfo)
.channels	Map<key, ChannelInfo> keyed by 'origin/copy/name'
.frames	Map<key, Frame> keyed by 'origin/copy/name'
.parameters	ParameterInfo[] (KB, TD, BHT, mud weight, …)
.getFrame(name)	Find Frame by name (case-insensitive)
.getChannel(name)	Find ChannelInfo by name (case-insensitive)

Frame

Member	Description
.name	Frame name string
.indexType	e.g. 'BOREHOLE-DEPTH', 'TIME'
.direction	'INCREASING' or 'DECREASING'
.spacing	Nominal depth/time step (number or null)
.channelNames	string[] in recording order
.channels	ChannelInfo[] fully resolved
.decode()	Returns DecodeResult
.toCSV(opts?)	Export scalar channels to CSV string
.toLAS(opts?)	Export scalar channels to LAS 2.0 string

DecodeResult

{
  frameCount:   number;
  frameNumbers: Int32Array;          // RP66 frame numbers (1-based)
  channels:     { name, longName, units, repcode, dimension }[];
  data:         { [channelName]: Float64Array };  // length = frameCount * stride
  strides:      { [channelName]: number };        // 1 for scalars, N for array channels
}

Reading array channels:

// DTCO_WAVE has dimension [512] → stride = 512
const wave0 = result.data.DTCO_WAVE.slice(0, 512); // waveform at first depth
const wave1 = result.data.DTCO_WAVE.slice(512, 1024);

---

Representation codes

import { RC } from '@geoharkat/dlis-parser';

// RC.FSINGL = 2  (IEEE 754 single)
// RC.FDOUBL = 7  (IEEE 754 double)
// RC.IDENT  = 17 (short ASCII string)
// etc.

---

Null values

DLIS has no universal null. Common vendor nulls are detected by isNullVal() inside the library but all values are returned as-is so you can apply your own masking:

import { NULL_VALUES } from '@geoharkat/dlis-parser';

const masked = Array.from(result.data.GR).map(v =>
  !Number.isFinite(v) || NULL_VALUES.has(v) ? NaN : v
);

---

Contact

Ismail Harkat — geoharkat@gmail.com

Issues: https://github.com/geoharkat/dlis-parser/issues

---

License

MIT © 2025 Ismail Harkat