Segyio

Latest version: v1.9.12

Safety actively analyzes 683530 Python packages for vulnerabilities to keep your Python projects secure.

Scan your dependencies

Page 2 of 8

2016.11

* Fixed some condtions where a failed write would corrupt trace data
* Fixed a memory leak bug
* VERSION string added to python
* Experimental memory-mapped file support
* Line-oriented C functions are offset aware
* Python offset property exposes offset numbers, not just count
* Support for pre-stack files, with new subindexing syntax, line[n, offset]
* Improved python repl (shell, read-eval-print-loop) support
* The widgets have color- and layout selectors

2016.10

* Matlab tests can optionally be turned off
* The application Segyviewer is embeddable and provided by the segyview sub
library
* libcwrap has been replaced fully by the python C api
* OS X and experimental Windows support
* A new sub mode for traces, raw, for eager reading of trace data


Planned breaking changes for segyio 2

This documents describe the planned breaking changes for segyio >= 2.

libsegyio
removal of linkage of ebcdic2ascii and friends, and ibm2ieee

The conversion functions are linkable today, but the symbols are not exposed in
any headers. Giving these functions public linkage (for testing purposes) was a
mistake, but hopefully no users will be affected as the symbols never appears in
the header file. Affected functions:

- `ebcdic2ascii`
- `ascii2ebcdic`
- `ibm2ieee`
- `ieee2ibm`
- `segy_seek`
- `segy_ftell`

python
accessing closed files raises ValueError

Calling methods on closed files should not raise `IOError`, but `ValueError`,
in order to be uniform with python's own file object.

Most users shouldn't (or wouldn't) use this error for control flow or recovery,
but in order to account for it, the change is postponed.

f.text no longer implicitly resolves to f.text[0]

The implicit accessing of the mandatory text header is inconsistent and error
prone, and won't work in segyio2. In segyio2, the only way to grab the
mandatory text header is `f.text[0]`

str(f.text[n]) removed

The implicit string conversion should be either tied to fmt or other explicit
methods, instead of implicitly resolving into a newline-interleaved
80-linewidth string

f.text[n] always returns 3200 bytes buffer object
Currently, the returned buffer assumes the contents are somewhat string-like,
and has some wonky behaviour when there are 0-bytes in the text headers, and
not proper whitespace paddings. In segyio2, this behaviour will be reviewed for
consistency, so that 0-bytes-in-buffers are handled gracefully, and textheaders
can be indexed into reliably and safely.

1.9.12

* Distribute wheels for Python 3.12
* Distribute wheels for Apple ARM

1.9.11

* Fixed a bug where segyio used the Delay Recording Time without it's scaling
factor when setting up the sample offsets on open.

1.9.10

* Distribute wheels for Python 3.11

1.9.9

* Packaging updates

Page 2 of 8

© 2024 Safety CLI Cybersecurity Inc. All Rights Reserved.