Reservoir Fortran Output

5.0.1 verified Fri May 01 auth: no python

A lazy parser and writer for reservoir simulator Fortran output format (unformatted and formatted). Current version 5.0.1 with strict typing and type stubs. Maintenance via Equinor, release cadence irregular.

pip install resfo

Common errors

error TypeError: write() got an unexpected keyword argument 'fileformat' ↓

cause In older versions, fileformat was a positional argument; in v5 it is a keyword-only argument.

fix

Use write(filename, data, fileformat=Format.FORMATTED) instead of positional.

error ModuleNotFoundError: No module named 'resfo.lazy_read' ↓

cause lazy_read is a function, not a submodule; incorrect import path.

fix

Use 'from resfo import lazy_read'.

error ValueError: data parameter must be an iterable of Entry objects, not dict ↓

cause Version 5 no longer accepts dict for write(). Data must be a list/tuple of Entry.

fix

Construct a list of Entry objects with keyword and data (numpy array).

Warnings

breaking In version 5.0.0, write() no longer accepts a dict for the data argument. Use a list of Entry objects instead. ↓

fix Replace dict with a list of Entry objects: write(filename, [Entry('KEYWORD', numpy_array)], fileformat=Format.FORMATTED)

gotcha lazy_read returns an iterator, not a list. It reads data lazily; iterating multiple times will reset the file pointer or raise a StopIteration. ↓

fix If you need random access, use read() instead, which reads all entries into memory.

deprecated Passing a dict to write() is removed in v5. The old pattern write('file', {'KEY': [1,2]}) no longer works. ↓

fix Use a list of Entry objects: from resfo import Entry; write('file', [Entry('KEY', np.array([1,2]))])

gotcha Memory usage: lazy_read keeps only one entry in memory at a time, but the iterator must be consumed completely. If you stop early, the file handle may not be closed immediately. ↓

fix Always iterate fully or use a context manager if available (not yet, but wrap in try/finally).

Imports

lazy_read
wrong
```
from resfo.lazy_read import ...
```
correct
```
from resfo import lazy_read
```
lazy_read is a top-level function, not a submodule.
read
wrong
```
import resfo; resfo.read()
```
correct
```
from resfo import read
```
read is also top-level, but lazy_read is preferred for large files.
write
```
from resfo import write
```
write is top-level; arguments changed in v5: no longer accepts dict for data.
Format
```
from resfo import Format
```
Enum for FORMATTED/UNFORMATTED file format.
Entry
```
from resfo import Entry
```
TypedDict representing a single keyword entry.

Quickstart

Lazily read an unformatted reservoir output file and print each keyword name and data size.

from resfo import lazy_read, Format

# Path to a small .FEGRID file
filepath = 'sample.FEGRID'  # replace with actual path
for entry in lazy_read(filepath, fileformat=Format.UNFORMATTED):
    print(f"Keyword: {entry.keyword}, Size: {len(entry.data)}")