Geografic-Information-System/lib/python/script/core.py

"""!@package grass.script.core

@brief GRASS Python scripting module (core functions)

Core functions to be used in Python scripts.

Usage:

@code
from grass.script import core as grass

grass.parser()
...
@endcode

(C) 2008-2014 by the GRASS Development Team
This program is free software under the GNU General Public
License (>=v2). Read the file COPYING that comes with GRASS
for details.

@author Glynn Clements
@author Martin Landa <landa.martin gmail.com>
@author Michael Barton <michael.barton asu.edu>
"""

import os
import sys
import types
import re
import atexit
import subprocess
import shutil
import locale
import codecs

# i18N
import gettext
gettext.install('grasslibs', os.path.join(os.getenv("GISBASE"), 'locale'))

# subprocess wrapper that uses shell on Windows


class Popen(subprocess.Popen):

    def __init__(self, args, bufsize=0, executable=None,
                 stdin=None, stdout=None, stderr=None,
                 preexec_fn=None, close_fds=False, shell=None,
                 cwd=None, env=None, universal_newlines=False,
                 startupinfo=None, creationflags=0):

        if shell == None:
            shell = (sys.platform == "win32")
        if sys.platform == "win32":
            # get full path including file extension for scripts
            fcmd = get_real_command(args[0]) 
            if fcmd.endswith('.py'):
                args[0] = fcmd
                args.insert(0, sys.executable)
        
        subprocess.Popen.__init__(self, args, bufsize, executable,
                                  stdin, stdout, stderr,
                                  preexec_fn, close_fds, shell,
                                  cwd, env, universal_newlines,
                                  startupinfo, creationflags)

PIPE = subprocess.PIPE
STDOUT = subprocess.STDOUT


class ScriptError(Exception):
    def __init__(self, msg):
        self.value = msg

    def __str__(self):
        return self.value

raise_on_error = False  # raise exception instead of calling fatal()


def call(*args, **kwargs):
    return Popen(*args, **kwargs).wait()

# GRASS-oriented interface to subprocess module

_popen_args = ["bufsize", "executable", "stdin", "stdout", "stderr",
               "preexec_fn", "close_fds", "cwd", "env",
               "universal_newlines", "startupinfo", "creationflags"]


def decode(string):
    enc = locale.getdefaultlocale()[1]
    if enc:
        return string.decode(enc)

    return string

def encode(string):
    enc = locale.getdefaultlocale()[1]
    if enc:
        return string.encode(enc)

    return string

def _make_val(val):
    if isinstance(val, types.StringType) or \
            isinstance(val, types.UnicodeType):
        return val
    if isinstance(val, types.ListType):
        return ",".join(map(_make_val, val))
    if isinstance(val, types.TupleType):
        return _make_val(list(val))
    return str(val)


def get_commands():
    """!Create list of available GRASS commands to use when parsing
    string from the command line

    @return list of commands (set) and directory of scripts (collected
    by extension - MS Windows only)

    @code
    >>> cmds = list(get_commands()[0])
    >>> cmds.sort()
    >>> cmds[:5]
    ['d.barscale', 'd.colorlist', 'd.colortable', 'd.correlate', 'd.erase']

    @endcode
    """
    gisbase = os.environ['GISBASE']
    cmd = list()
    scripts = {'.py': list()} if sys.platform == 'win32' else {}

    def scan(gisbase, directory):
        dir_path = os.path.join(gisbase, directory)
        if os.path.exists(dir_path):
            for fname in os.listdir(os.path.join(gisbase, directory)):
                if scripts:  # win32
                    name, ext = os.path.splitext(fname)
                    if ext != '.manifest':
                        cmd.append(name)
                    if ext in scripts.keys():
                        scripts[ext].append(name)
                else:
                    cmd.append(fname)

    for directory in ('bin', 'scripts'):
        scan(gisbase, directory)

    # scan gui/scripts/
    gui_path = os.path.join(gisbase, 'etc', 'gui', 'scripts')
    if os.path.exists(gui_path):
        os.environ["PATH"] = os.getenv("PATH") + os.pathsep + gui_path
        cmd = cmd + os.listdir(gui_path)

    return set(cmd), scripts


# replacement for which function from shutil (not available in all versions)
# from http://hg.python.org/cpython/file/6860263c05b3/Lib/shutil.py#l1068
# added because of Python scripts running Python scripts on MS Windows
# see also ticket #2008 which is unrelated but same function was proposed
def shutil_which(cmd, mode=os.F_OK | os.X_OK, path=None):
    """Given a command, mode, and a PATH string, return the path which
    conforms to the given mode on the PATH, or None if there is no such
    file.

    `mode` defaults to os.F_OK | os.X_OK. `path` defaults to the result
    of os.environ.get("PATH"), or can be overridden with a custom search
    path.

    """
    # Check that a given file can be accessed with the correct mode.
    # Additionally check that `file` is not a directory, as on Windows
    # directories pass the os.access check.
    def _access_check(fn, mode):
        return (os.path.exists(fn) and os.access(fn, mode)
                and not os.path.isdir(fn))

    # If we're given a path with a directory part, look it up directly rather
    # than referring to PATH directories. This includes checking relative to the
    # current directory, e.g. ./script
    if os.path.dirname(cmd):
        if _access_check(cmd, mode):
            return cmd
        return None

    if path is None:
        path = os.environ.get("PATH", os.defpath)
    if not path:
        return None
    path = path.split(os.pathsep)

    if sys.platform == "win32":
        # The current directory takes precedence on Windows.
        if not os.curdir in path:
            path.insert(0, os.curdir)

        # PATHEXT is necessary to check on Windows.
        pathext = os.environ.get("PATHEXT", "").split(os.pathsep)
        map(lambda x: x.lower(), pathext) # force lowercase
        if '.py' not in pathext:          # we assume that PATHEXT contains always '.py'
            pathext.insert(0, '.py')
        # See if the given file matches any of the expected path extensions.
        # This will allow us to short circuit when given "python.exe".
        # If it does match, only test that one, otherwise we have to try
        # others.
        if any(cmd.lower().endswith(ext.lower()) for ext in pathext):
            files = [cmd]
        else:
            files = [cmd + ext for ext in pathext]
    else:
        # On other platforms you don't have things like PATHEXT to tell you
        # what file suffixes are executable, so just pass on cmd as-is.
        files = [cmd]

    seen = set()
    for dir in path:
        normdir = os.path.normcase(dir)
        if not normdir in seen:
            seen.add(normdir)
            for thefile in files:
                name = os.path.join(dir, thefile)
                if _access_check(name, mode):
                    return name
    return None


# Added because of scripts calling scripts on MS Windows.
# Module name (here cmd) differs from the file name (does not have extension).
# Additionally, we don't run scripts using system executable mechanism,
# so we need the full path name.
# However, scripts are on the PATH and '.PY' in in PATHEXT, so we can use
# shutil.which to get the full file path. Addons are on PATH too.
# An alternative to which function call would be to check the script path and
# addons path. This is proposed improvement for the future.
# Another alternative is to check some global list of scripts but this list
# needs to be created first. The question is what is less expensive.
# Note that getting the full path is only part of the solution,
# the other part is to use the right Python as an executable and pass the full
# script path as a parameter.
# Nevertheless, it is unclear on which places which extensions are added.
# This function also could skip the check for platform but depends
# how will be used, this is most general but not most effective.
def get_real_command(cmd):
    """!Returns the real file commad for a module (cmd)

    For Python scripts on MS Windows it returns full path to the script
    and adds a '.py' extension.
    For other cases it just returns a module (name).
    So, you can just use this function for all without further check.

    >>> get_real_command('g.region')
    'g.region'
    """
    if sys.platform == 'win32':
        # we in fact expect pure module name (without extension)
        # so, lets remove extension
        if os.path.splitext(cmd)[1] == '.py':
            cmd = cmd[:-3]
        full_path = shutil_which(cmd + '.py')
        if full_path:
            return full_path

    return cmd


def make_command(prog, flags="", overwrite=False, quiet=False, verbose=False,
                 **options):
    """!Return a list of strings suitable for use as the args parameter to
    Popen() or call(). Example:

    @code
    >>> make_command("g.message", flags = 'w', message = 'this is a warning')
    ['g.message', '-w', 'message=this is a warning']

    @endcode

    @param prog GRASS module
    @param flags flags to be used (given as a string)
    @param overwrite True to enable overwriting the output (<tt>--o</tt>)
    @param quiet True to run quietly (<tt>--q</tt>)
    @param verbose True to run verbosely (<tt>--v</tt>)
    @param options module's parameters

    @return list of arguments
    """
    args = [prog]
    if overwrite:
        args.append("--o")
    if quiet:
        args.append("--q")
    if verbose:
        args.append("--v")
    if flags:
        if '-' in flags:
            raise ScriptError("'-' is not a valid flag")
        args.append("-%s" % flags)
    for opt, val in options.iteritems():
        if val != None:
            if opt[0] == '_':
                opt = opt[1:]
            args.append("%s=%s" % (opt, _make_val(val)))
    return args


def start_command(prog, flags="", overwrite=False, quiet=False,
                  verbose=False, **kwargs):
    """!Returns a Popen object with the command created by make_command.
    Accepts any of the arguments which Popen() accepts apart from "args"
    and "shell".

    @code
    >>> p = start_command("g.gisenv", stdout=subprocess.PIPE)
    >>> print p  # doctest: +ELLIPSIS
    <...Popen object at 0x...>
    >>> print p.communicate()[0]  # doctest: +SKIP
    GISDBASE='/opt/grass-data';
    LOCATION_NAME='spearfish60';
    MAPSET='glynn';
    GRASS_DB_ENCODING='ascii';
    GUI='text';
    MONITOR='x0';

    @endcode

    @param prog GRASS module
    @param flags flags to be used (given as a string)
    @param overwrite True to enable overwriting the output (<tt>--o</tt>)
    @param quiet True to run quietly (<tt>--q</tt>)
    @param verbose True to run verbosely (<tt>--v</tt>)
    @param kwargs module's parameters

    @return Popen object
    """
    options = {}
    popts = {}
    for opt, val in kwargs.iteritems():
        if opt in _popen_args:
            popts[opt] = val
        else:
            if isinstance(val, unicode):
                val = encode(val)
            options[opt] = val

    args = make_command(prog, flags, overwrite, quiet, verbose, **options)

    if debug_level() > 0:
        sys.stderr.write("D1/%d: %s.start_command(): %s\n" % (debug_level(),
                                                              __name__,
                                                              ' '.join(args)))
        sys.stderr.flush()

    return Popen(args, **popts)


def run_command(*args, **kwargs):
    """!Passes all arguments to start_command(), then waits for the process to
    complete, returning its exit code. Similar to subprocess.call(), but
    with the make_command() interface.

    @param args list of unnamed arguments (see start_command() for details)
    @param kwargs list of named arguments (see start_command() for details)

    @return exit code (0 for success)
    """
    ps = start_command(*args, **kwargs)
    return ps.wait()


def pipe_command(*args, **kwargs):
    """!Passes all arguments to start_command(), but also adds
    "stdout = PIPE". Returns the Popen object.

    @code
    >>> p = pipe_command("g.gisenv")
    >>> print p  # doctest: +ELLIPSIS
    <....Popen object at 0x...>
    >>> print p.communicate()[0]  # doctest: +SKIP
    GISDBASE='/opt/grass-data';
    LOCATION_NAME='spearfish60';
    MAPSET='glynn';
    GRASS_DB_ENCODING='ascii';
    GUI='text';
    MONITOR='x0';

    @endcode

    @param args list of unnamed arguments (see start_command() for details)
    @param kwargs list of named arguments (see start_command() for details)

    @return Popen object
    """
    kwargs['stdout'] = PIPE
    return start_command(*args, **kwargs)


def feed_command(*args, **kwargs):
    """!Passes all arguments to start_command(), but also adds
    "stdin = PIPE". Returns the Popen object.

    @param args list of unnamed arguments (see start_command() for details)
    @param kwargs list of named arguments (see start_command() for details)

    @return Popen object
    """
    kwargs['stdin'] = PIPE
    return start_command(*args, **kwargs)


def read_command(*args, **kwargs):
    """!Passes all arguments to pipe_command, then waits for the process to
    complete, returning its stdout (i.e. similar to shell `backticks`).

    @param args list of unnamed arguments (see start_command() for details)
    @param kwargs list of named arguments (see start_command() for details)

    @return stdout
    """
    ps = pipe_command(*args, **kwargs)
    return ps.communicate()[0]


def parse_command(*args, **kwargs):
    """!Passes all arguments to read_command, then parses the output
    by parse_key_val().

    Parsing function can be optionally given by <em>parse</em> parameter
    including its arguments, e.g.

    @code
    parse_command(..., parse = (grass.parse_key_val, { 'sep' : ':' }))
    @endcode

    or you can simply define <em>delimiter</em>

    @code
    parse_command(..., delimiter = ':')
    @endcode

    @param args list of unnamed arguments (see start_command() for details)
    @param kwargs list of named arguments (see start_command() for details)

    @return parsed module output
    """
    parse = None
    parse_args = {}
    if 'parse' in kwargs:
        if type(kwargs['parse']) is types.TupleType:
            parse = kwargs['parse'][0]
            parse_args = kwargs['parse'][1]
        del kwargs['parse']

    if 'delimiter' in kwargs:
        parse_args = {'sep': kwargs['delimiter']}
        del kwargs['delimiter']

    if not parse:
        parse = parse_key_val  # use default fn

    res = read_command(*args, **kwargs)

    return parse(res, **parse_args)


def write_command(*args, **kwargs):
    """!Passes all arguments to feed_command, with the string specified
    by the 'stdin' argument fed to the process' stdin.

    @param args list of unnamed arguments (see start_command() for details)
    @param kwargs list of named arguments (see start_command() for details)

    @return return code
    """
    stdin = kwargs['stdin']
    p = feed_command(*args, **kwargs)
    p.stdin.write(stdin)
    p.stdin.close()
    return p.wait()


def exec_command(prog, flags="", overwrite=False, quiet=False, verbose=False,
                 env=None, **kwargs):
    """!Interface to os.execvpe(), but with the make_command() interface.

    @param prog GRASS module
    @param flags flags to be used (given as a string)
    @param overwrite True to enable overwriting the output (<tt>--o</tt>)
    @param quiet True to run quietly (<tt>--q</tt>)
    @param verbose True to run verbosely (<tt>--v</tt>)
    @param env directory with environmental variables
    @param kwargs module's parameters

    """
    args = make_command(prog, flags, overwrite, quiet, verbose, **kwargs)

    if env == None:
        env = os.environ
    os.execvpe(prog, args, env)

# interface to g.message


def message(msg, flag=None):
    """!Display a message using `g.message`

    @param msg message to be displayed
    @param flag flags (given as string)
    """
    run_command("g.message", flags=flag, message=msg)


def debug(msg, debug=1):
    """!Display a debugging message using `g.message -d`

    @param msg debugging message to be displayed
    @param debug debug level (0-5)
    """
    if debug_level() >= debug:
        if sys.platform == "win32":
            msg = msg.replace('&', '^&')
            
        run_command("g.message", flags='d', message=msg, debug=debug)

def verbose(msg):
    """!Display a verbose message using `g.message -v`

    @param msg verbose message to be displayed
    """
    message(msg, flag='v')


def info(msg):
    """!Display an informational message using `g.message -i`

    @param msg informational message to be displayed
    """
    message(msg, flag='i')


def percent(i, n, s):
    """!Display a progress info message using `g.message -p`

    @code
    message(_("Percent complete..."))
    n = 100
    for i in range(n):
        percent(i, n, 1)
    percent(1, 1, 1)
    @endcode

    @param i current item
    @param n total number of items
    @param s increment size
    """
    message("%d %d %d" % (i, n, s), flag='p')


def warning(msg):
    """!Display a warning message using `g.message -w`

    @param msg warning message to be displayed
    """
    message(msg, flag='w')


def error(msg):
    """!Display an error message using `g.message -e`

    @param msg error message to be displayed
    """
    message(msg, flag='e')


def fatal(msg):
    """!Display an error message using `g.message -e`, then abort

    Raise exception when raise_on_error is 'True'.

    @param msg error message to be displayed
    """
    global raise_on_error
    if raise_on_error:
        raise ScriptError(msg)

    error(msg)
    sys.exit(1)


def set_raise_on_error(raise_exp=True):
    """!Define behaviour on fatal error (fatal() called)

    @param raise_exp True to raise ScriptError instead of calling
    sys.exit(1) in fatal()

    @return current status
    """
    global raise_on_error
    tmp_raise = raise_on_error
    raise_on_error = raise_exp
    return tmp_raise


def get_raise_on_error():
    """!Return True if a ScriptError exception is raised instead of calling
       sys.exit(1) in case a fatal error was invoked with fatal()
    """
    global raise_on_error
    return raise_on_error

# interface to g.parser


def _parse_opts(lines):
    options = {}
    flags = {}
    for line in lines:
        if not line:
            break
        try:
            [var, val] = line.split('=', 1)
        except:
            raise SyntaxError("invalid output from g.parser: %s" % line)

        if var.startswith('flag_'):
            flags[var[5:]] = bool(int(val))
        elif var.startswith('opt_'):
            options[var[4:]] = val
        elif var in ['GRASS_OVERWRITE', 'GRASS_VERBOSE']:
            os.environ[var] = val
        else:
            raise SyntaxError("invalid output from g.parser: %s" % line)

    return (options, flags)


def parser():
    """!Interface to g.parser, intended to be run from the top-level, e.g.:

    @code
        if __name__ == "__main__":
            options, flags = grass.parser()
            main()
    @endcode

    Thereafter, the global variables "options" and "flags" will be
    dictionaries containing option/flag values, keyed by lower-case
    option/flag names. The values in "options" are strings, those in
    "flags" are Python booleans.
    """
    if not os.getenv("GISBASE"):
        print >> sys.stderr, "You must be in GRASS GIS to run this program."
        sys.exit(1)

    cmdline = [basename(sys.argv[0])]
    cmdline += ['"' + arg + '"' for arg in sys.argv[1:]]
    os.environ['CMDLINE'] = ' '.join(cmdline)

    argv = sys.argv[:]
    name = argv[0]
    if not os.path.isabs(name):
        if os.sep in name or (os.altsep and os.altsep in name):
            argv[0] = os.path.abspath(name)
        else:
            argv[0] = os.path.join(sys.path[0], name)

    p = Popen(['g.parser', '-n'] + argv, stdout=PIPE)
    s = p.communicate()[0]
    lines = s.split('\0')

    if not lines or lines[0] != "@ARGS_PARSED@":
        sys.stdout.write(s)
        sys.exit(p.returncode)

    return _parse_opts(lines[1:])


def separator(sep):
    if sep == "pipe":
        return "|"
    elif sep == "comma":
        return ","
    elif sep == "space":
        return " "
    elif sep == "tab" or sep == "\\t":
        return "\t"
    elif sep == "newline" or sep == "\\n":
        return "\n"
    return sep
    

# interface to g.tempfile


def tempfile(create=True):
    """!Returns the name of a temporary file, created with
    g.tempfile.

    @param create True to create a file

    @return path to a tmp file
    """
    flags = ''
    if not create:
        flags += 'd'

    return read_command("g.tempfile", flags=flags, pid=os.getpid()).strip()


def tempdir():
    """!Returns the name of a temporary dir, created with g.tempfile."""
    tmp = tempfile(create=False)
    os.mkdir(tmp)

    return tmp


class KeyValue(dict):
    """A general-purpose key-value store.

    KeyValue is a subclass of dict, but also allows entries to be read and
    written using attribute syntax. Example:

    \code
    >>> reg = KeyValue()
    >>> reg['north'] = 489
    >>> reg.north
    489
    >>> reg.south = 205
    >>> reg['south']
    205

    \endcode
    """

    def __getattr__(self, key):
        return self[key]

    def __setattr__(self, key, value):
        self[key] = value

# key-value parsers


def parse_key_val(s, sep='=', dflt=None, val_type=None, vsep=None):
    """!Parse a string into a dictionary, where entries are separated
    by newlines and the key and value are separated by `sep' (default: `=')

    @param s string to be parsed
    @param sep key/value separator
    @param dflt default value to be used
    @param val_type value type (None for no cast)
    @param vsep vertical separator (default os.linesep)

    @return parsed input (dictionary of keys/values)
    """
    result = KeyValue()

    if not s:
        return result

    if vsep:
        lines = s.split(vsep)
        try:
            lines.remove('\n')
        except ValueError:
            pass
    else:
        lines = s.splitlines()

    for line in lines:
        kv = line.split(sep, 1)
        k = kv[0].strip()
        if len(kv) > 1:
            v = kv[1].strip()
        else:
            v = dflt

        if val_type:
            result[k] = val_type(v)
        else:
            result[k] = v

    return result


def _compare_projection(dic):
    """
        !Check if projection has some possibility of duplicate names like
        Universal Transverse Mercator and Universe Transverse Mercator and
        unify them

        @param dic The dictionary containing information about projection

        @return The dictionary with the new values if needed

    """
    # the lookup variable is a list of list, each list contains all the
    # possible name for a projection system
    lookup = [['Universal Transverse Mercator', 'Universe Transverse Mercator']]
    for lo in lookup:
        for n in range(len(dic['name'])):
            if dic['name'][n] in lo:
                dic['name'][n] = lo[0]
    return dic


def _compare_units(dic):
    """
        !Check if units has some possibility of duplicate names like
        meter and metre and unify them

        @param dic The dictionary containing information about units

        @return The dictionary with the new values if needed

    """
    # the lookup variable is a list of list, each list contains all the
    # possible name for a units
    lookup = [['meter', 'metre'], ['meters', 'metres'], ['kilometer',
              'kilometre'], ['kilometers', 'kilometres']]
    for l in lookup:
        for n in range(len(dic['unit'])):
            if dic['unit'][n] in l:
                dic['unit'][n] = l[0]
        for n in range(len(dic['units'])):
            if dic['units'][n] in l:
                dic['units'][n] = l[0]
    return dic


def _text_to_key_value_dict(filename, sep=":", val_sep=",", checkproj=False,
                            checkunits=False):
    """
        !Convert a key-value text file, where entries are separated
        by newlines and the key and value are separated by `sep',
        into a key-value dictionary and discover/use the correct
        data types (float, int or string) for values.

        @param filename The name or name and path of the text file to convert
        @param sep The character that separates the keys and values, default is ":"
        @param val_sep The character that separates the values of a single key, default is ","
        @param checkproj True if it has to check some information about projection system
        @param checkproj True if it has to check some information about units
        
        @return The dictionary

        A text file with this content:
        \code
        a: Hello
        b: 1.0
        c: 1,2,3,4,5
        d : hello,8,0.1
        \endcode

        Will be represented as this dictionary:
        \code
        {'a': ['Hello'], 'c': [1, 2, 3, 4, 5], 'b': [1.0], 'd': ['hello', 8, 0.1]}
        \endcode

    """
    text = open(filename, "r").readlines()
    kvdict = KeyValue()

    for line in text:
        if line.find(sep) >= 0:
            key, value = line.split(sep)
            key = key.strip()
            value = value.strip()
        else:
            # Jump over empty values
            continue
        values = value.split(val_sep)
        value_list = []

        for value in values:
            not_float = False
            not_int = False

            # Convert values into correct types
            # We first try integer then float
            try:
                value_converted = int(value)
            except:
                not_int = True
            if not_int:
                try:
                    value_converted = float(value)
                except:
                    not_float = True

            if not_int and not_float:
                value_converted = value.strip()

            value_list.append(value_converted)

        kvdict[key] = value_list
    if checkproj:
        kvdict = _compare_projection(kvdict)
    if checkunits:
        kvdict = _compare_units(kvdict)
    return kvdict


def compare_key_value_text_files(filename_a, filename_b, sep=":",
                                 val_sep=",", precision=0.000001,
                                 proj=False, units=False):
    """
    !Compare two key-value text files

    This method will print a warning in case keys that are present in the first
    file are not present in the second one.
    The comparison method tries to convert the values into their native format
    (float, int or string) to allow correct comparison.

    An example key-value text file may have this content:
    \code
    a: Hello
    b: 1.0
    c: 1,2,3,4,5
    d : hello,8,0.1
    \endcode

    @param filename_a name of the first key-value text file
    @param filenmae_b name of the second key-value text file
    @param sep character that separates the keys and values, default is ":"
    @param val_sep character that separates the values of a single key, default is ","
    @param precision precision with which the floating point values are compared
    @param proj True if it has to check some information about projection system
    @param units True if it has to check some information about units

    @return True if full or almost identical, False if different
    """
    dict_a = _text_to_key_value_dict(filename_a, sep, checkproj=proj,
                                     checkunits=units)
    dict_b = _text_to_key_value_dict(filename_b, sep, checkproj=proj,
                                     checkunits=units)

    if sorted(dict_a.keys()) != sorted(dict_b.keys()):
        return False

    # We compare matching keys
    for key in dict_a.keys():
        # Floating point values must be handled separately
        if isinstance(dict_a[key], float) and isinstance(dict_b[key], float):
            if abs(dict_a[key] - dict_b[key]) > precision:
                return False
        elif isinstance(dict_a[key], float) or isinstance(dict_b[key], float):
            warning(_("Mixing value types. Will try to compare after "
                      "integer conversion"))
            return int(dict_a[key]) == int(dict_b[key])
        elif key == "+towgs84":
            # We compare the sum of the entries
            if abs(sum(dict_a[key]) - sum(dict_b[key])) > precision:
                return False
        else:
            if dict_a[key] != dict_b[key]:
                return False
    return True

def diff_files(filename_a, filename_b):
    """!Diffs two text files and returns difference.

    @param filename_a first file path    
    @param filename_b second file path
    
    @return list of strings
    """
    import difflib
    differ = difflib.Differ()
    fh_a = open(filename_a, 'r')
    fh_b = open(filename_b, 'r')
    result = list(differ.compare(fh_a.readlines(),
                                 fh_b.readlines()))
    return result


# interface to g.gisenv


def gisenv():
    """!Returns the output from running g.gisenv (with no arguments), as a
    dictionary. Example:

    @code
    >>> env = gisenv()
    >>> print env['GISDBASE']  # doctest: +SKIP
    /opt/grass-data

    @endcode

    @return list of GRASS variables
    """
    s = read_command("g.gisenv", flags='n')
    return parse_key_val(s)

# interface to g.region


def locn_is_latlong():
    """!Tests if location is lat/long. Value is obtained
    by checking the "g.region -pu" projection code.

    @return True for a lat/long region, False otherwise
    """
    s = read_command("g.region", flags='pu')
    kv = parse_key_val(s, ':')
    if kv['projection'].split(' ')[0] == '3':
        return True
    else:
        return False


def region(region3d=False, complete=False):
    """!Returns the output from running "g.region -gu", as a
    dictionary. Example:

    @param region3d True to get 3D region

    @code
    >>> curent_region = region()
    >>> # obtain n, s, e and w values
    >>> [curent_region[key] for key in "nsew"]  # doctest: +ELLIPSIS
    [..., ..., ..., ...]
    >>> # obtain ns and ew resulutions
    >>> (curent_region['nsres'], curent_region['ewres'])  # doctest: +ELLIPSIS
    (..., ...)

    @endcode

    @return dictionary of region values
    """
    flgs = 'gu'
    if region3d:
        flgs += '3'
    if complete:
        flgs += 'cep'

    s = read_command("g.region", flags=flgs)
    reg = parse_key_val(s, val_type=float)
    for k in ['rows',  'cols',  'cells',
              'rows3', 'cols3', 'cells3', 'depths']:
        if k not in reg:
            continue
        reg[k] = int(reg[k])

    return reg


def region_env(region3d=False, **kwargs):
    """!Returns region settings as a string which can used as
    GRASS_REGION environmental variable.

    If no 'kwargs' are given then the current region is used. Note
    that this function doesn't modify the current region!

    See also use_temp_region() for alternative method how to define
    temporary region used for raster-based computation.

    \param region3d True to get 3D region
    \param kwargs g.region's parameters like 'rast', 'vect' or 'region'
    \code
    os.environ['GRASS_REGION'] = grass.region_env(region='detail')
    grass.mapcalc('map=1', overwrite=True)
    os.environ.pop('GRASS_REGION')
    \endcode

    @return string with region values
    @return empty string on error
    """
    # read proj/zone from WIND file
    env = gisenv()
    windfile = os.path.join(env['GISDBASE'], env['LOCATION_NAME'],
                            env['MAPSET'], "WIND")
    fd = open(windfile, "r")
    grass_region = ''
    for line in fd.readlines():
        key, value = map(lambda x: x.strip(), line.split(":", 1))
        if kwargs and key not in ('proj', 'zone'):
            continue
        if not kwargs and not region3d and \
                key in ('top', 'bottom', 'cols3', 'rows3',
                        'depths', 'e-w resol3', 'n-s resol3', 't-b resol'):
            continue

        grass_region += '%s: %s;' % (key, value)

    if not kwargs:  # return current region
        return grass_region

    # read other values from `g.region -gu`
    flgs = 'ug'
    if region3d:
        flgs += '3'

    s = read_command('g.region', flags=flgs, **kwargs)
    if not s:
        return ''
    reg = parse_key_val(s)

    kwdata = [('north',     'n'),
              ('south',     's'),
              ('east',      'e'),
              ('west',      'w'),
              ('cols',      'cols'),
              ('rows',      'rows'),
              ('e-w resol', 'ewres'),
              ('n-s resol', 'nsres')]
    if region3d:
        kwdata += [('top',        't'),
                   ('bottom',     'b'),
                   ('cols3',      'cols3'),
                   ('rows3',      'rows3'),
                   ('depths',     'depths'),
                   ('e-w resol3', 'ewres3'),
                   ('n-s resol3', 'nsres3'),
                   ('t-b resol',  'tbres')]

    for wkey, rkey in kwdata:
        grass_region += '%s: %s;' % (wkey, reg[rkey])

    return grass_region


def use_temp_region():
    """!Copies the current region to a temporary region with "g.region save=",
    then sets WIND_OVERRIDE to refer to that region. Installs an atexit
    handler to delete the temporary region upon termination.
    """
    name = "tmp.%s.%d" % (os.path.basename(sys.argv[0]), os.getpid())
    run_command("g.region", save=name, overwrite=True)
    os.environ['WIND_OVERRIDE'] = name
    atexit.register(del_temp_region)


def del_temp_region():
    """!Unsets WIND_OVERRIDE and removes any region named by it."""
    try:
        name = os.environ.pop('WIND_OVERRIDE')
        run_command("g.remove", quiet=True, region=name)
    except:
        pass

# interface to g.findfile


def find_file(name, element='cell', mapset=None):
    """!Returns the output from running g.findfile as a
    dictionary. Example:

    @code
    >>> result = find_file('elevation', element='cell')
    >>> print result['fullname']
    elevation@PERMANENT
    >>> print result['file']  # doctest: +ELLIPSIS
    /.../PERMANENT/cell/elevation

    @endcode

    @param name file name
    @param element element type (default 'cell')
    @param mapset mapset name (default all mapsets in search path)

    @return parsed output of g.findfile
    """
    if element == 'raster' or element == 'rast':
        verbose(_('Element type should be "cell" and not "%s"') % element)
        element = 'cell'
    s = read_command("g.findfile", flags='n', element=element, file=name,
                     mapset=mapset)
    return parse_key_val(s)

# interface to g.list


def list_grouped(type, check_search_path=True):
    """!List elements grouped by mapsets.

    Returns the output from running g.list, as a dictionary where the
    keys are mapset names and the values are lists of maps in that
    mapset. Example:

    @code
    >>> list_grouped('rast')['PERMANENT']  # doctest: +ELLIPSIS
    [..., 'lakes', ..., 'slope', ...

    @endcode

    @param type element type (rast, vect, rast3d, region, ...)
    @param check_search_path True to add mapsets for the search path with no
                             found elements

    @return directory of mapsets/elements
    """
    if type == 'raster' or type == 'cell':
        verbose(_('Element type should be "rast" and not "%s"') % type)
        type = 'rast'
    dashes_re = re.compile("^----+$")
    mapset_re = re.compile("<(.*)>")
    result = {}
    if check_search_path:
        for mapset in mapsets(search_path=True):
            result[mapset] = []

    mapset = None
    for line in read_command("g.list", type=type).splitlines():
        if line == "":
            continue
        if dashes_re.match(line):
            continue
        m = mapset_re.search(line)
        if m:
            mapset = m.group(1)
            if mapset not in result.keys():
                result[mapset] = []
            continue
        if mapset:
            result[mapset].extend(line.split())

    return result


def _concat(xs):
    result = []
    for x in xs:
        result.extend(x)
    return result


def list_pairs(type):
    """!List of elements as tuples.

    Returns the output from running g.list, as a list of (map, mapset)
    pairs. Example:

    @code
    >>> list_pairs('rast')  # doctest: +ELLIPSIS
    [..., ('lakes', 'PERMANENT'), ..., ('slope', 'PERMANENT'), ...

    @endcode

    @param type element type (rast, vect, rast3d, region, ...)

    @return list of tuples (map, mapset)
    """
    return _concat([[(map, mapset) for map in maps]
                    for mapset, maps in list_grouped(type).iteritems()])


def list_strings(type):
    """!List of elements as strings.

    Returns the output from running g.list, as a list of qualified
    names. Example:

    @code
    >>> list_strings('rast')  # doctest: +ELLIPSIS
    [..., 'lakes@PERMANENT', ..., 'slope@PERMANENT', ...

    @endcode

    @param type element type

    @return list of strings ('map@@mapset')
    """
    return ["%s@%s" % pair for pair in list_pairs(type)]

# interface to g.mlist


def mlist_strings(type, pattern=None, mapset=None, flag=''):
    """!List of elements as strings.

    Returns the output from running g.mlist, as a list of qualified
    names.

    @param type element type (rast, vect, rast3d, region, ...)
    @param pattern pattern string
    @param mapset mapset name (if not given use search path)
    @param flag pattern type: 'r' (basic regexp), 'e' (extended regexp), or ''
                (glob pattern)

    @return list of elements
    """
    if type == 'raster' or type == 'cell':
        verbose(_('Element type should be "rast" and not "%s"') % type)
        type = 'rast'
    result = list()
    for line in read_command("g.mlist",
                             quiet=True,
                             flags='m' + flag,
                             type=type,
                             pattern=pattern,
                             mapset=mapset).splitlines():
        result.append(line.strip())

    return result


def mlist_pairs(type, pattern=None, mapset=None, flag=''):
    """!List of elements as pairs

    Returns the output from running g.mlist, as a list of
    (name, mapset) pairs

    @param type element type (rast, vect, rast3d, region, ...)
    @param pattern pattern string
    @param mapset mapset name (if not given use search path)
    @param flag pattern type: 'r' (basic regexp), 'e' (extended regexp), or ''
                (glob pattern)

    @return list of elements
    """
    return [tuple(map.split('@', 1)) for map in mlist_strings(type, pattern,
                                                              mapset, flag)]


def mlist_grouped(type, pattern=None, check_search_path=True, flag=''):
    """!List of elements grouped by mapsets.

    Returns the output from running g.mlist, as a dictionary where the
    keys are mapset names and the values are lists of maps in that
    mapset. Example:

    @code
    >>> mlist_grouped('vect', pattern='*roads*')['PERMANENT']
    ['railroads', 'roadsmajor']

    @endcode

    @param type element type (rast, vect, rast3d, region, ...)
    @param pattern pattern string
    @param check_search_path True to add mapsets for the search path with no
                             found elements
    @param flag pattern type: 'r' (basic regexp), 'e' (extended regexp), or ''
                (glob pattern)

    @return directory of mapsets/elements
    """
    if type == 'raster' or type == 'cell':
        verbose(_('Element type should be "rast" and not "%s"') % type)
        type = 'rast'
    result = {}
    if check_search_path:
        for mapset in mapsets(search_path=True):
            result[mapset] = []

    mapset = None
    for line in read_command("g.mlist", quiet=True, flags="m" + flag,
                             type=type, pattern=pattern).splitlines():
        try:
            name, mapset = line.split('@')
        except ValueError:
            warning(_("Invalid element '%s'") % line)
            continue

        if mapset in result:
            result[mapset].append(name)
        else:
            result[mapset] = [name, ]

    return result

# color parsing

named_colors = {
    "white":   (1.00, 1.00, 1.00),
    "black":   (0.00, 0.00, 0.00),
    "red":     (1.00, 0.00, 0.00),
    "green":   (0.00, 1.00, 0.00),
    "blue":    (0.00, 0.00, 1.00),
    "yellow":  (1.00, 1.00, 0.00),
    "magenta": (1.00, 0.00, 1.00),
    "cyan":    (0.00, 1.00, 1.00),
    "aqua":    (0.00, 0.75, 0.75),
    "grey":    (0.75, 0.75, 0.75),
    "gray":    (0.75, 0.75, 0.75),
    "orange":  (1.00, 0.50, 0.00),
    "brown":   (0.75, 0.50, 0.25),
    "purple":  (0.50, 0.00, 1.00),
    "violet":  (0.50, 0.00, 1.00),
    "indigo":  (0.00, 0.50, 1.00)}


def parse_color(val, dflt=None):
    """!Parses the string "val" as a GRASS colour, which can be either one of
    the named colours or an R:G:B tuple e.g. 255:255:255. Returns an
    (r,g,b) triple whose components are floating point values between 0
    and 1. Example:

    @code
    >>> parse_color("red")
    (1.0, 0.0, 0.0)
    >>> parse_color("255:0:0")
    (1.0, 0.0, 0.0)

    @endcode

    @param val color value
    @param dflt default color value

    @return tuple RGB
    """
    if val in named_colors:
        return named_colors[val]

    vals = val.split(':')
    if len(vals) == 3:
        return tuple(float(v) / 255 for v in vals)

    return dflt

# check GRASS_OVERWRITE


def overwrite():
    """!Return True if existing files may be overwritten"""
    owstr = 'GRASS_OVERWRITE'
    return owstr in os.environ and os.environ[owstr] != '0'

# check GRASS_VERBOSE


def verbosity():
    """!Return the verbosity level selected by GRASS_VERBOSE"""
    vbstr = os.getenv('GRASS_VERBOSE')
    if vbstr:
        return int(vbstr)
    else:
        return 2

## various utilities, not specific to GRASS

# basename inc. extension stripping


def basename(path, ext=None):
    """!Remove leading directory components and an optional extension
    from the specified path

    @param path path
    @param ext extension
    """
    name = os.path.basename(path)
    if not ext:
        return name
    fs = name.rsplit('.', 1)
    if len(fs) > 1 and fs[1].lower() == ext:
        name = fs[0]
    return name

def find_program(pgm, *args):
    """!Attempt to run a program, with optional arguments.
    You must call the program in a way that will return a successful
    exit code. For GRASS modules this means you need to pass it some
    valid CLI option, like "--help". For other programs a common
    valid do-little option is "--version".
    
    Example:

    @code
    >>> grass.find_program('r.sun', 'help')
    True
    >>> grass.find_program('gdalwarp', '--version')
    True
    @endcode

    @param pgm program name
    @param args list of arguments

    @return False if the attempt failed due to a missing executable
            or non-zero return code
    @return True otherwise
    """
    nuldev = file(os.devnull, 'w+')
    try:
        call([pgm] + list(args), stdin = nuldev, stdout = nuldev, stderr = nuldev)
        found = True
    except:
        found = False
    nuldev.close()

    return found

# try to remove a file, without complaints


def try_remove(path):
    """!Attempt to remove a file; no exception is generated if the
    attempt fails.

    @param path path to file to remove
    """
    try:
        os.remove(path)
    except:
        pass

# try to remove a directory, without complaints


def try_rmdir(path):
    """!Attempt to remove a directory; no exception is generated if the
    attempt fails.

    @param path path to directory to remove
    """
    try:
        os.rmdir(path)
    except:
        shutil.rmtree(path, ignore_errors=True)


def float_or_dms(s):
    """!Convert DMS to float.

    @param s DMS value

    @return float value
    """
    return sum(float(x) / 60 ** n for (n, x) in enumerate(s.split(':')))

# interface to g.mapsets


def mapsets(search_path=False):
    """!List available mapsets

    @param search_path True to list mapsets only in search path

    @return list of mapsets
    """
    if search_path:
        flags = 'p'
    else:
        flags = 'l'
    mapsets = read_command('g.mapsets',
                           flags=flags,
                           sep='newline',
                           quiet=True)
    if not mapsets:
        fatal(_("Unable to list mapsets"))

    return mapsets.splitlines()

# interface to `g.proj -c`


def create_location(dbase, location, epsg=None, proj4=None, filename=None,
                    wkt=None, datum=None, datum_trans=None, desc=None, overwrite=False):
    """!Create new location

    Raise ScriptError on error.

    @param dbase path to GRASS database
    @param location location name to create
    @param epsg if given create new location based on EPSG code
    @param proj4 if given create new location based on Proj4 definition
    @param filename if given create new location based on georeferenced file
    @param wkt if given create new location based on WKT definition (path to PRJ file)
    @param datum GRASS format datum code
    @param datum_trans datum transformation parameters (used for epsg and proj4)
    @param desc description of the location (creates MYNAME file)
    @param overwrite True to overwrite location if exists (WARNING: ALL DATA from existing location ARE DELETED!)
    """
    gisdbase = None
    if epsg or proj4 or filename or wkt:
        # FIXME: changing GISDBASE mid-session is not background-job safe
        gisdbase = gisenv()['GISDBASE']
        run_command('g.gisenv', set='GISDBASE=%s' % dbase)
    # create dbase if not exists
    if not os.path.exists(dbase):
            os.mkdir(dbase)

    # check if location already exists
    if os.path.exists(os.path.join(dbase, location)):
        if not overwrite:
            warning(_("Location <%s> already exists. Operation canceled.") % location)
            return
        else:
            warning(_("Location <%s> already exists and will be overwritten") % location)
            shutil.rmtree(os.path.join(dbase, location))
    
    kwargs = dict()
    if datum:
        kwargs['datum'] = datum
    if datum_trans:
        kwargs['datum_trans'] = datum_trans

    if epsg:
        ps = pipe_command('g.proj', quiet=True, flags='t', epsg=epsg,
                          location=location, stderr=PIPE, **kwargs)
    elif proj4:
        ps = pipe_command('g.proj', quiet=True, flags='t', proj4=proj4,
                          location=location, stderr=PIPE, **kwargs)
    elif filename:
        ps = pipe_command('g.proj', quiet=True, georef=filename,
                          location=location, stderr=PIPE)
    elif wkt:
        ps = pipe_command('g.proj', quiet=True, wkt=wkt, location=location,
                          stderr=PIPE)
    else:
        _create_location_xy(dbase, location)

    if epsg or proj4 or filename or wkt:
        error = ps.communicate()[1]
        run_command('g.gisenv', set='GISDBASE=%s' % gisdbase)

        if ps.returncode != 0 and error:
            raise ScriptError(repr(error))

    try:
        fd = codecs.open(os.path.join(dbase, location, 'PERMANENT', 'MYNAME'),
                         encoding='utf-8', mode='w')
        if desc:
            fd.write(desc + os.linesep)
        else:
            fd.write(os.linesep)
        fd.close()
    except OSError as e:
        raise ScriptError(repr(e))


def _create_location_xy(database, location):
    """!Create unprojected location

    Raise ScriptError on error.

    @param database GRASS database where to create new location
    @param location location name
    """
    cur_dir = os.getcwd()
    try:
        os.chdir(database)
        os.mkdir(location)
        os.mkdir(os.path.join(location, 'PERMANENT'))

        # create DEFAULT_WIND and WIND files
        regioninfo = ['proj:       0',
                      'zone:       0',
                      'north:      1',
                      'south:      0',
                      'east:       1',
                      'west:       0',
                      'cols:       1',
                      'rows:       1',
                      'e-w resol:  1',
                      'n-s resol:  1',
                      'top:        1',
                      'bottom:     0',
                      'cols3:      1',
                      'rows3:      1',
                      'depths:     1',
                      'e-w resol3: 1',
                      'n-s resol3: 1',
                      't-b resol:  1']

        defwind = open(os.path.join(location,
                                    "PERMANENT", "DEFAULT_WIND"), 'w')
        for param in regioninfo:
            defwind.write(param + '%s' % os.linesep)
        defwind.close()

        shutil.copy(os.path.join(location, "PERMANENT", "DEFAULT_WIND"),
                    os.path.join(location, "PERMANENT", "WIND"))

        os.chdir(cur_dir)
    except OSError as e:
        raise ScriptError(repr(e))

# interface to g.version


def version():
    """!Get GRASS version as dictionary

    @code
    print version()

    {'proj4': '4.8.0', 'geos': '3.3.5', 'libgis_revision': '52468',
     'libgis_date': '2012-07-27 22:53:30 +0200 (Fri, 27 Jul 2012)',
     'version': '7.0.svn', 'date': '2012', 'gdal': '2.0dev',
     'revision': '53670'}
    @endcode
    """
    data = parse_command('g.version', flags='rge')
    for k, v in data.iteritems():
        data[k.strip()] = v.replace('"', '').strip()

    return data

# get debug_level
_debug_level = None


def debug_level():
    global _debug_level
    if _debug_level is not None:
        return _debug_level
    _debug_level = 0
    if find_program('g.gisenv', '--help'):
        _debug_level = int(gisenv().get('DEBUG', 0))


def legal_name(s):
    """!Checks if the string contains only allowed characters.

    This is the Python implementation of G_legal_filename() function.

    @note It is not clear when to use this function.
    """
    if not s or s[0] == '.':
        warning(_("Illegal filename <%s>. Cannot be 'NULL' or start with " \
                  "'.'.") % s)
        return False

    illegal = [c
               for c in s
               if c in '/"\'@,=*~' or c <= ' ' or c >= '\177']
    if illegal:
        illegal = ''.join(sorted(set(illegal)))
        warning(_("Illegal filename <%(s)s>. <%(il)s> not allowed.\n") % {
        's': s, 'il': illegal})
        return False

    return True


if __name__ == '__main__':
    import doctest
    doctest.testmod()