Provide path to tree_0_0_0.dat file and it will print out units for all fields.

ctree_units.py

import functools
import re
import sys

from unyt import \
    unyt_array, \
    unyt_quantity
from unyt.dimensions import \
    dimensionless, \
    length
from unyt.unit_registry import \
    UnitRegistry
from unyt.exceptions import \
    UnitParseError

_unit_registry = UnitRegistry()
_unit_registry.add('h', 0.7, dimensionless)
for my_unit in ["m", "pc", "AU"]:
    new_unit = "%scm" % my_unit
    _unit_registry.add(
        new_unit, _unit_registry.lut[my_unit][0],
        length, _unit_registry.lut[my_unit][3])
quan = functools.partial(unyt_quantity,
                         registry=_unit_registry)

def get_ctrees_units(filename):
    fields = []
    fi = {}
    fdb = {}
    rems = ["%s%s%s" % (s[0], t, s[1])
            for s in [("(", ")"), ("", "")]
            for t in ["physical, peculiar",
                      "comoving", "physical"]]

    f = open(filename, "r")
    # Read the first line as a list of all fields.
    # Do some footwork to remove awkard characters.
    rfl = f.readline()[1:].strip().split()
    reg = re.compile(r"\(\d+\)$")
    for pf in rfl:
        match = reg.search(pf)
        if match is None:
            fields.append(pf)
        else:
            fields.append(pf[:match.start()])

    # Now grab a bunch of things from the header.
    while True:
        line = f.readline()
        if line is None:
            raise IOError(
                "Encountered enexpected EOF reading %s." %
                filename)
        elif not line.startswith("#"):
            _ntrees = int(line.strip())
            _hoffset = f.tell()
            break

        # cosmological parameters
        if "Omega_M" in line:
            pars = line[1:].split(";")
            for j, par in enumerate(["omega_matter",
                                     "omega_lambda",
                                     "hubble_constant"]):
                v = float(pars[j].split(" = ")[1])
                print ("%s: %s." % (par, v))

        # box size
        elif "Full box size" in line:
            pars = line.split("=")[1].strip().split()
            box = pars
            print ("Box size: %s." % box)

        # These are lines describing the various fields.
        # Pull them apart and look for units.
        elif ":" in line:
            tfields, desc = line[1:].strip().split(":", 1)

            # Units are enclosed in parentheses.
            # Pull out what's enclosed and remove things like
            # "comoving" and "physical".
            if "(" in line and ")" in line:
                punits = desc[desc.find("(")+1:desc.rfind(")")]
                for rem in rems:
                    while rem in punits:
                        pre, mid, pos = punits.partition(rem)
                        punits = pre + pos
                try:
                    quan(1, punits)
                except UnitParseError:
                    punits = ""
                punits = punits.strip()
            else:
                punits = ""

            # Multiple fields together on the same line.
            for sep in ["/", ","]:
                if sep in tfields:
                    tfields = tfields.split(sep)
                    break
            if not isinstance(tfields, list):
                tfields = [tfields]

            # Assign units and description.
            for tfield in tfields:
                fdb[tfield.lower()] = {"description": desc.strip(),
                                       "units": punits}

    f.close()

    # Fill the field info with the units found above.
    for i, field in enumerate(fields):
        if "(" in field and ")" in field:
            cfield = field[:field.find("(")]
        else:
            cfield = field
        fi[field] = fdb.get(cfield.lower(),
                            {"description": "",
                             "units": ""})
        fi[field]["column"] = i
    
    return fi

if __name__ == "__main__":
    fn = sys.argv[1]
    fi = get_ctrees_units(fn)
    ### can also get "description" from each entry
    for field, entry in fi.items():
        print ("%s: \"%s\"" % (field, entry.get("units", "")))