Skip to content
This repository has been archived by the owner on Apr 7, 2024. It is now read-only.

Commit

Permalink
rewrite command line parser
Browse files Browse the repository at this point in the history
  • Loading branch information
@guillermooo
guillermooo committed Sep 22, 2012
1 parent 14a9cda commit f9cf2e8
Show file tree
Hide file tree
Showing 10 changed files with 1,156 additions and 1,088 deletions.
4 changes: 3 additions & 1 deletion CHANGELOG.txt
View file
Original file line number Diff line number Diff line change
@@ -1,2 +1,4 @@
12.9.10
22.9.12
* rewrite command line parser
12.9.12
* implemented basic "/" and "?" commands
168 changes: 23 additions & 145 deletions ex_command_parser.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -6,6 +6,7 @@
from itertools import takewhile

import ex_error
from vex.parsers import rangeparser


# Data used to parse strings into ex commands and map them to an actual
Expand All @@ -24,106 +25,14 @@

# Holds a parsed ex command data.
# TODO: elaborate on params info.
EX_CMD = namedtuple('ex_command', 'name command forced range args parse_errors')
EX_CMD = namedtuple('ex_command', 'name command forced args parse_errors new_range can_have_range')

# TODO: Allow escapes in search-based addresses, as in POSTFIX_ADDRESS.
# Address that can only appear in a prefix range (before a command).
PREFIX_ADDRESS = r'[.$%]|(?:/.*?(?<!\\)/|\?.*?(?<!\\)\?){1,2}|[+-]?\d+|[\'][a-zA-Z0-9<>]'
# Address that can only appear after a command.
POSTFIX_ADDRESS = r'[.$]|(?:/.*?(?<!\\)/|\?.*?(?<!\\)\?){1,2}|[+-]?\d+|[\'][a-zA-Z0-9<>]'
ADDRESS_OFFSET = r'[-+]\d+'
ADDRESS_SEPARATOR = r'[,;]'
# Can only appear standalone.
OPENENDED_SEARCH_ADDRESS = r'^[/?].*'

# Can appear as prefix ranges or standalone.
INCOMPLETE_RANGE_SEPARATOR = ','

# Matches ranges preceding commands.
# TODO: +100,-100del should be valid ranges too.
EX_PREFIX_RANGE = re.compile(
r'''(?x)
# Ranges missing a member, like 10, and ,10
^(?:
(?P<incomplete>
(?:
(?P<inc_laddress>%(address)s)
(?P<inc_loffset>%(address_offset)s)*
(?P<suf_alt_separator>%(alt_separator)s)
)|
(?:
(?P<pref_alt_separator>%(alt_separator)s)
(?P<inc_raddress>%(address)s)
(?P<inc_roffset>%(address_offset)s)*
)
)|
# A left address...
(?P<laddress>%(address)s)
# with optional offsets...
(?P<loffset>%(address_offset)s)*
# and an optional right address...
(?:
# (which includes the address separator)
(?P<separator>%(address_separator)s)
(?P<raddress>%(address)s)
# with optional offsets.
(?P<roffset>%(address_offset)s)*
)?
)
# We need to make sure that we match up to the separator, which
# comes before the actual ex command. As far as I can tell, ex commands always
# start with A-Za-z or !.
(?=[a-zA-Z]|!)
''' % {'address': PREFIX_ADDRESS,
'address_separator': ADDRESS_SEPARATOR,
'address_offset': ADDRESS_OFFSET,
'alt_separator': INCOMPLETE_RANGE_SEPARATOR,}
)

# Matches ranges that stand alone, without being followed by anything. They
# simply represent an address to move the caret to.
EX_STANDALONE_RANGE = re.compile(
r'''(?x)
^(?:
# A full range consisting of...
(?P<incomplete>
(?:
(?P<inc_laddress>%(address)s)
(?P<inc_loffset>%(address_offset)s)*
(?P<suf_alt_separator>%(alt_separator)s)
)|
(?:
(?P<pref_alt_separator>%(alt_separator)s)
(?P<inc_raddress>%(address)s)
(?P<inc_roffset>%(address_offset)s)*
)
)|
(?:
# a left address...
(?P<laddress>%(address)s)
# optionally followed by offsets...
(?P<loffset>%(address_offset)s)*
# and an optional right address...
(?:
# (including the address separator)
(?P<separator>%(address_separator)s)
(?P<raddress>%(address)s)
# and any number of offsets...
(?P<roffset>%(address_offset)s)*
)?
)|
# or an openended search-based address.
(?P<openended>%(openended)s)
)$
''' % {'address': PREFIX_ADDRESS,
'address_separator': ADDRESS_SEPARATOR,
'address_offset': ADDRESS_OFFSET,
'openended': OPENENDED_SEARCH_ADDRESS,
'alt_separator': INCOMPLETE_RANGE_SEPARATOR,}
)

# Matches addresses after commands, like :copy10.
#
# ** IMPORTANT **
# Vim's documentation on valid addresses is wrong. For postfixed addresses,
# as in :copy10,20, only the left end is parsed and used; the rest is discarded
Expand Down Expand Up @@ -193,12 +102,12 @@
('map', 'map'): ex_cmd_data(
command='ex_map',
invocations=(),
error_on=()
error_on=(ex_error.ERR_NO_RANGE_ALLOWED,)
),
('abbreviate', 'ab'): ex_cmd_data(
command='ex_abbreviate',
invocations=(),
error_on=()
error_on=(ex_error.ERR_NO_RANGE_ALLOWED,)
),
('quit', 'q'): ex_cmd_data(
command='ex_quit',
Expand All @@ -209,7 +118,8 @@
('qall', 'qa'): ex_cmd_data(
command='ex_quit_all',
invocations=(),
error_on=(ex_error.ERR_TRAILING_CHARS,)
error_on=(ex_error.ERR_TRAILING_CHARS,
ex_error.ERR_NO_RANGE_ALLOWED,)
),
# TODO: add invocations
('wq', 'wq'): ex_cmd_data(
Expand Down Expand Up @@ -360,6 +270,13 @@
error_on=(ex_error.ERR_TRAILING_CHARS,
ex_error.ERR_NO_RANGE_ALLOWED,)
),
('split', 'sp'): ex_cmd_data(
command='ex_split',
invocations=(re.compile(r'^$',),
),
error_on=(ex_error.ERR_TRAILING_CHARS,
ex_error.ERR_NO_RANGE_ALLOWED,)
),
(':', ':'): ex_cmd_data(
command='ex_goto',
invocations=(),
Expand Down Expand Up @@ -388,63 +305,24 @@ def find_command(cmd_name):
return partial_matches[0]


def is_only_range(cmd_line):
# Make sure we match a range standing alone (not followed by a command).
return EX_STANDALONE_RANGE.search(cmd_line) and \
not EX_PREFIX_RANGE.search(cmd_line)


def get_cmd_line_range(cmd_line):
try:
start, end = EX_PREFIX_RANGE.search(cmd_line).span()
except AttributeError:
return None
return cmd_line[start:end]


def is_valid_command_name(cmd_name):
return (cmd_name[0].isalpha() or cmd_name[0] in r"&")


def extract_command_name(cmd_line):
if cmd_line[0] in ':!':
return cmd_line[0]
if cmd_line:
return ''.join(takewhile(lambda c: is_valid_command_name(c), cmd_line))


def parse_command(cmd):
cmd_name = cmd.strip()
if len(cmd_name) > 1:
cmd_name = cmd_name[1:]
elif not cmd_name == ':':
return None

parser = rangeparser.CommandLineParser(cmd[1:])
r_ = parser.parse_cmd_line()

if is_only_range(cmd_name):
range_ = cmd_name
cmd_name = ':'
else:
range_ = get_cmd_line_range(cmd_name)
if range_:
cmd_name = cmd_name[len(range_):]

# FIXME: is this needed?
if not (cmd_name.startswith(('!', ':')) or
is_valid_command_name(cmd_name[0])):
return

command = extract_command_name(cmd_name)
args = cmd_name[len(command):]

bang = args.startswith('!')
if bang:
args = args[1:]

command = r_['commands'][0]['cmd']
bang = r_['commands'][0]['forced']
args = r_['commands'][0]['args']
cmd_data = find_command(command)
if not cmd_data:
return
cmd_data = EX_COMMANDS[cmd_data]
can_have_range = ex_error.ERR_NO_RANGE_ALLOWED not in cmd_data.error_on

cmd_args = {}
for pattern in cmd_data.invocations:
Expand All @@ -463,7 +341,7 @@ def parse_command(cmd):
parse_errors.append(ex_error.ERR_NO_BANG_ALLOWED)
if err == ex_error.ERR_TRAILING_CHARS and args:
parse_errors.append(ex_error.ERR_TRAILING_CHARS)
if err == ex_error.ERR_NO_RANGE_ALLOWED and range_:
if err == ex_error.ERR_NO_RANGE_ALLOWED and r_['range']['text_range']:
parse_errors.append(ex_error.ERR_NO_RANGE_ALLOWED)
if err == ex_error.ERR_INVALID_RANGE and not cmd_args:
parse_errors.append(ex_error.ERR_INVALID_RANGE)
Expand All @@ -473,7 +351,7 @@ def parse_command(cmd):
return EX_CMD(name=command,
command=cmd_data.command,
forced=bang,
range=range_,
args=cmd_args,
parse_errors=parse_errors
)
parse_errors=parse_errors,
new_range=r_['range'],
can_have_range=can_have_range,)
Loading

0 comments on commit f9cf2e8

Please sign in to comment.