Command-line tools¶

pycdfpp comes with two command-line tools. They show how a CDF file is laid out on disk, record by record. That’s useful to debug a file writer, to understand why a file is big, or to compare with what NASA’s tools report.

To just look at a file’s variables and attributes, use print(pycdfpp.load(path)) or the CDFpp Explorer instead.

Install them with the cli extra:

$ python -m pip install "pycdfpp[cli]"

cdfdump¶

cdfdump prints every internal record of the file, in the order they are stored, with all their fields:

$ cdfdump ac_h0_mfi_20200101_v07.cdf
ac_h0_mfi_20200101_v07.cdf
├── @8 CDR
│   ├── GDRoffset: 320
│   ├── Version: 3
│   ├── Release: 7
│   ├── Encoding: "network"
│   ...
├── @320 GDR
│   ├── rVDRhead: 33674
...

Each node is one record: @8 is its byte offset in the file, and CDR its type. The CDF specification describes every record type.

Add --irsdump to get the text format of NASA’s cdfirsdump tool instead.

cdfirsdump¶

cdfirsdump reproduces the output of NASA’s cdfirsdump tool, byte for byte. So you can diff its output against the real tool’s, or feed it to scripts that expect that format.

By default, it prints a summary of what takes space in the file:

$ cdfirsdump ac_h0_mfi_20200101_v07.cdf
...
Summary...

  Total bytes: 253980 (+16 if with checksum)
   Used bytes: 253980, 100.000%
 Unused bytes:      0,   0.000%

     IR count:    349
    ...
    VVR count:      9,  43550 bytes, 17.147%
   CVVR count:      7, 168326 bytes, 66.275%

Here, compressed variable records (CVVR) take two thirds of the file.

The main options:

Option	Effect
`--level full`	Dump every record, not only the summary.
`--data`	With `--level full`, also hex-dump the variable data.
`--offset N`	Start at byte `N`.
`--radix 16`	Print offsets in hexadecimal.
`--output FILE`	Write to a file instead of the terminal.
`--no-summary`	Leave out the summary table.

Unlike NASA’s tool, options use two dashes: --level full, not -full. Run cdfirsdump --help for the complete list.

From Python¶

Both tools are built on the pycdfpp.debug module. Use it directly to inspect records from a script. Here, we list the compression records of the ACE file from the Quickstart:

import urllib.request
import pycdfpp

url = ("https://spdf.gsfc.nasa.gov/pub/data/ace/mag/level_2_cdaweb/"
       "mfi_h0/2020/ac_h0_mfi_20200101_v07.cdf")
urllib.request.urlretrieve(url, "ac_h0_mfi_20200101_v07.cdf")

from pycdfpp.debug import for_each_record

for offset, record_type, fields in for_each_record("ac_h0_mfi_20200101_v07.cdf"):
    if record_type == "CPR":
        print(offset, fields)