pmacs3/mode/awk.py

import commands
import color, mode, tab
from lex import Grammar, PatternRule, RegionRule
from mode.python import StringGrammar2
from tab import Marker

class RegexGrammar(Grammar):
    rules = [
        PatternRule(r'escaped', r'\\.'),
        PatternRule(r'data', r'[^\\/]+'),
    ]

class AwkGrammar(Grammar):
    rules = [
        PatternRule(r'comment', r'#.*\n$'),
        PatternRule(r'spaces', r' +'),
        RegionRule(r'awk_regex', r'/(?! )', RegexGrammar, r'/'),
        PatternRule(r'awk_global', r'(?:TEXTDOMAIN|SUBSEP|RLENGTH|RSTART|RT|RS|PROCINFO|ORS|OFS|OFMT|NR|NF|LINT|IGNORECASE|FS|FNR|FILENAME|FIELDWIDTHS|ERRNO|ENVIRON|CONVFMT|BINMODE|ARGV|ARGIND|ARGC)(?![a-zA-Z0-9_])'),
        PatternRule(r'delimiter', r'(?:[\{\}()\[\]?:;,]|=(?!=)|\+=|-=|\*=|/=|\%=|\^=)'),
        PatternRule(r'keyword', r'(?:BEGIN|END|if|else|while|do|for|break|continue|delete|exit)(?![a-zA-Z0-9_])'),
        PatternRule(r'builtin', r'(?:close|getline|nextfile|next|printf|print|system|fflush|atan2|cos|exp|int|log|rand|sin|sqrt|srand|asorti|asort|gensub|gsub|index|length|match|split|sprintf|strtonum|substr|sub|tolower|toupper|mktime|strftime|systime|and|compl|lshift|or|xor|rshift|bindtextdomain|dcgettext|dcngettext|function|extension)(?![a-zA-Z0-9_])'),

        PatternRule(r'awk_field', r'\$\d*'),

        PatternRule(r'number', r'-?0x[0-9A-Fa-f]+'),
        PatternRule(r'number', r'-?0[0-7]*'),
        PatternRule(r'number', r'-?[0-9]+\.?[0-9]*'),
        PatternRule(r'number', r'-?\.[0-9]+'),

        PatternRule(r'unop', r'!(?![=~])|--|\+\+'),
        PatternRule(r'binop', r'(?:&&|\|\||<=|>=|!=|!~|==|\^|%|[-~/+*<>])'),
        RegionRule(r'string', r'"', StringGrammar2, r'"'),
        PatternRule(r'awk_function', r'[a-zA-Z_][a-zA-Z0-9_]*(?=\()'),
        PatternRule(r'awk_identifier', r'[a-zA-Z_][a-zA-Z0-9_]*'),
        
        PatternRule(r'continuation', r'\\\n$'),
        PatternRule(r'eol', r'\n'),
    ]

class AwkTabber(tab.StackTabber):
    open_tokens  = {'{': '}', '(': ')', '[': ']'}
    close_tokens = {'}': '{', ')': '(', ']': '['}
    def __init__(self, m):
        self.mode  = m
        self.name  = m.name()
        self.lines = {}
        self._reset()
    def region_added(self, p, newlines):
        self.lines = {}
    def region_removed(self, p1, p2):
        self.lines = {}
    def get_level(self, y):
        if y not in self.lines:
            self._calc_level(y)
        return self.lines.get(y)

    def _calc_level(self, y):
        target = y
        while not self._is_base(y) and y > 0:
            y -= 1
        self._reset()
        while y <= target:
            self._save_curr_level()
            self._handle_tokens(y)
            y += 1

    def _is_base(self, y):
        if y == 0:
            return True

        t = self._get_tokens(y)[0]
        if t.fqname() == 'awk_regex.start':
            return True
        elif t.name in ('awk_field', 'awk_global'):
            return True
        elif t.name == 'keyword' and t.string in ('BEGIN', 'END'):
            return True
        else:
            return False
        
    def _reset(self):
        self.record     = {}
        self.stack      = []
        self.markers    = self.stack
        self.curr_level = 0
    def _get_curr_level(self):
        if self.stack:
            return self.stack[-1].level
        else:
            return 0
    def _get_next_level(self):
        return self._get_curr_level() + self.mode.tabwidth
    def _save_curr_level(self):
        self.curr_level = self._get_curr_level()

    def _match(self, *names):
        return self.stack and self.stack[-1].name in names
    def _nomatch(self, *names):
        return self.stack and self.stack[-1].name not in names
    def _pop(self, *names):
        if self._match(*names):
            self.stack.pop()
    def _pop_while(self, *names):
        while self._match(*names):
            self.stack.pop()
    def _pop_until(self, *names):
        while self._nomatch(*names):
            self.stack.pop()

    def _append(self, name, level):
        self.stack.append(Marker(name, level))
    def _append_unless(self, name, level):
        if self._nomatch(name):
            self.stack.append(Marker(name, level))

    def _get_tokens(self, y):
        return self.mode.window.buffer.highlights[self.name].tokens[y]
    def _handle_tokens(self, y):
        tokens = self._get_tokens(y)
        assert tokens
        start = int(self._is_indent(tokens[0]))
        end   = len(tokens) - 1

        while end > 0 and self._is_ignored(tokens[end]):
            end -= 1

        for i in range(0, end + 1 - start):
            token  = tokens[start + i]
            if self._is_ignored(token):
                pass
            elif self._is_close_token(token):
                self._handle_close_token(y, tokens, start, end, i, token)
            elif self._is_open_token(token):
                self._handle_open_token(y, tokens, start, end, i, token)
            else:
                self._handle_other_token(y, tokens, start, end, i, token)
        self.lines[y] = self.curr_level
        self.record[y] = tuple(self.stack)

    def _is_indent(self, token):
        return token.name == 'spaces'
    def _is_ignored(self, token):
        return token.name in ('spaces', 'eol', 'comment')

    def _is_close_token(self, token):
        return token.name == 'delimiter' and token.string in self.close_tokens
    def _handle_close_token(self, y, tokens, start, end, i, token):
        if not self.stack:
            raise Exception, "unmatched %r, line %d" % (token.string, y)
        while True:
            marker = self.stack[-1]
            if marker.name in ('control', 'continue'):
                self.stack.pop()
            elif marker.name in self.open_tokens:
                if self.open_tokens[marker.name] == token.string:
                    self.stack.pop()
                    break
                else:
                    raise Exception, "mismatched %r, line %d (expected %r)" % \
                        (token.string, y, d[marker.name])
            else:
                raise Exception, "what? %r" % marker.name
        if i == 0:
            self._save_curr_level()

    def _is_open_token(self, token):
        return token.name == 'delimiter' and token.string in self.open_tokens
    def _handle_open_token(self, y, tokens, start, end, i, token):
        if i == 0 and self.stack and self.stack[-1].name == 'continue':
            self.stack.pop()
        if token.string == '{':
            self._pop_while('continue', 'control')
        if i == end - start:
            level = self._get_next_level()
        else:
            level = tokens[i + 1].x
        self._append(token.string, level)

    def _handle_other_token(self, y, tokens, start, end, i, token):
        name, s = token.name, token.string
        if i + start == end:
            self._pop_while('continue', 'control')
        
        if name == 'continuation':
            self._append_unless('continue', self._get_next_level())
        elif name == 'keyword' and s in ('if', 'else', 'while', 'do', 'for'):
            if i == start:
                self._save_curr_level()
            self._append_unless('control', self._get_next_level())

class Awk(mode.Fundamental):
    tabbercls   = AwkTabber
    modename   = 'awk'
    extensions = ['.awk']
    grammar    = AwkGrammar
    colors     = {
        'awk_global':      ('yellow', 'default', 'bold'),
        'awk_function':    ('magenta', 'default', 'bold'),
        'awk_field':       ('yellow', 'default', 'bold'),
        'awk_identifier':  ('yellow', 'default', 'bold'),
        'awk_regex.start': ('cyan', 'default', 'bold'),
        'awk_regex.null':  ('cyan', 'default', 'bold'),
        'awk_regex.data':  ('cyan', 'default', 'bold'),
        'awk_regex.end':   ('cyan', 'default', 'bold'),
    }

install = Awk.install