ecl-data-io 0.0.2

A (lazy) parser and writer for the ecl output format.

ecl-data-io

Parser for the binary ecl format, such as found in output files with extensions .UNRST, .EGRID, .INIT, .UNRST etc. and also the corresponding ascii formats such as .FUNRST, .FEGRID, .FINIT, .FUNRST etc.

The file format comes in two forms: formatted or unformatted, which are ascii and binary respectively (the terminology comes from fortran).

The files consists of named arrays, starting with a 8 character ascii keyword name, followed by a 4 byte signed integer which is the length of the array, then a 4 character keyword describing the data type of elements in the array. Each file consists of a sequence of such arrays.

ecl-data-io does not interpret the output, but simply generate the list of arrays:

import ecl_data_io as eclio

for kw, arr in eclio.read("my_grid.egrid"):
    print(kw)

>>> "FILEHEAD"
>>> "GRIDHEAD"
>>> "COORD"
>>> "ZCORN"
>>> "ACTNUM"
>>> "MAPAXES"

Similarly, the write function will generate such files:

import ecl_data_io as eclio

eclio.write("my_grid.egrid", {"FILEHEAD": [...], "GRIDHEAD": [10,10,10]})

The default format is is binary (unformatted), but it is possible to read and write ascii aswell:

import ecl_data_io as eclio

eclio.write(
    "my_grid.egrid",
    {"FILEHEAD": [...], "GRIDHEAD": [10,10,10]},
    fileformat=eclio.Format.FORMATTED
)

lazy reading

It is also possible to read through the file without loading all arrays into memory, ie. lazily:

import ecl_data_io as eclio

for item in eclio.lazy_read("my_grid.egrid"):
    print(item.read_keyword())

>>> "FILEHEAD"
>>> "GRIDHEAD"
>>> "COORD"
>>> "ZCORN"
>>> "ACTNUM"
>>> "MAPAXES"

Note that eclio.lazy_read in the above example is a generator of array entries and the file will be closed once the generator is finished. Therefore, care will have to be taken in when arrays/keywords are read from the entries. For better control, one can pass the opened file:

import ecl_data_io as eclio

with open("my_grid.egrid", "rb") as f:
    generator = eclio.lazy_read("my_grid.egrid")
    item = next(generator)
    print(item.read_keyword())

>>> "FILEHEAD"

Duplicate keywords

The simple write function works best for keywords with unique names, however, if the file has multiple arays with the same name, list of tuples can be used:

import ecl_data_io as eclio

contents = [ ("COORD", [1.0, ...]), ("COORD", [2.0, ...]), ]

eclio.write("my/file.grdecl", contents)