pmacs3/mode_perl.py

import re, sets, string, sys
import color, commands, default, lex2, lex2_perl, method, mode2, regex, tab_perl
from point2 import Point

class Perl(mode2.Fundamental):
    def __init__(self, w):
        mode2.Fundamental.__init__(self, w)

        #self.tag_matching = True
        self.grammar      = lex2_perl.PerlGrammar()
        self.lexer        = lex2.Lexer(self.name(), self.grammar)

        self.add_action_and_bindings(PerlCheckSyntax(), ('C-c s',))
        self.add_action_and_bindings(PerlHashCleanup(), ('C-c h',))
        #self.add_action_and_bindings(PerlHashCleanup2(), ('C-c h',))
        self.add_action_and_bindings(PerlViewModulePerldoc(), ('C-c v',))
        self.add_action_and_bindings(PerlViewWordPerldoc(), ('C-c p',))
        self.add_action_and_bindings(PerlWrapLine(), ('M-q',))
        self.add_action_and_bindings(PerlGotoFunction(), ('C-c M-g',))
        self.add_action_and_bindings(PerlWhichFunction(), ('C-c w',))
        self.add_action_and_bindings(PerlListFunctions(), ('C-c W',))

        self.add_bindings('close-paren', (')',))
        self.add_bindings('close-brace', ('}',))
        self.add_bindings('close-bracket', (']',))

        self.default_color = color.build('default', 'default')

        self.colors = {
            # basic stuff
            'escaped':   color.build('magenta', 'default'),
            'null':      color.build('default', 'default'),
            'delimiter': color.build('default', 'default'),
            'sub':       color.build('cyan', 'default'),
            'number':    color.build('default', 'default'),
            'operator':  color.build('default', 'default'),
            'endblock':  color.build('red', 'default'),
            'keyword':   color.build('magenta', 'default'),
            'scalar':    color.build('yellow', 'default'),
            'array':     color.build('yellow', 'default'),
            'deref':     color.build('yellow', 'default'),
            'hash':      color.build('yellow', 'default'),
            'hash_key':  color.build('green', 'default'),
            'comment':   color.build('red', 'default'),
            'function':  color.build('cyan', 'default'),
            'builtin':   color.build('magenta', 'default'),
            'method':    color.build('cyan', 'default'),
            'bareword':  color.build('default', 'default'),
            'label':     color.build('cyan', 'default'),
            'package':   color.build('cyan', 'default'),
            'class':     color.build('cyan', 'default'),
            'use':       color.build('cyan', 'default'),
            'method':    color.build('cyan', 'default'),
        
            # heredoc
            'heredoc1.start': color.build('green', 'default'),
            'heredoc1.null':  color.build('green', 'default'),
            'heredoc1.end':   color.build('green', 'default'),
            'heredoc2.start': color.build('green', 'default'),
            'heredoc2.null':  color.build('green', 'default'),
            'heredoc2.end':   color.build('green', 'default'),
            'eval_heredoc.start': color.build('cyan', 'default'),
            'eval_heredoc.null':  color.build('cyan', 'default'),
            'eval_heredoc.end':   color.build('cyan', 'default'),
        
            # pod
            'pod.start': color.build('red', 'default'),
            'pod.null':  color.build('red', 'default'),
            'pod.entry': color.build('magenta', 'default'),
            'pod.end':   color.build('red', 'default'),
        
            # "" strings
            'string1.start':   color.build('green', 'default'),
            'string1.null':    color.build('green', 'default'),
            'string1.escaped': color.build('magenta', 'default'),
            'string1.deref':   color.build('yellow', 'default'),
            'string1.end':     color.build('green', 'default'),
        
            # '' strings
            'string2.start': color.build('green', 'default'),
            'string2.null':  color.build('green', 'default'),
            'string2.end':   color.build('green', 'default'),
        
            # `` strings
            'evalstring': color.build('cyan', 'default'),
        
            # quoted region
            'quoted':       color.build('cyan', 'default'),
            'quoted.start': color.build('cyan', 'default'),
            'quoted.null':  color.build('cyan', 'default'),
            'quoted.end':   color.build('cyan', 'default'),
        
            # match regex
            'match.start': color.build('cyan', 'default'),
            'match.end':   color.build('cyan', 'default'),
            'match.null':  color.build('cyan', 'default'),
        
            # replace regex
            'replace.start':  color.build('cyan', 'default'),
            'replace.middle': color.build('cyan', 'default'),
            'replace.end':    color.build('cyan', 'default'),
            'replace.null':   color.build('cyan', 'default'),
        
            # translate regex
            'translate.start':  color.build('magenta', 'default'),
            'translate.middle': color.build('magenta', 'default'),
            'translate.end':    color.build('magenta', 'default'),
            'translate.null':   color.build('magenta', 'default'),
        }

#        self.colors = {
#            'heredoc':                     color.build('green', 'default'),
#            'endblock':                    color.build('red', 'default'),
#            'pod':                         color.build('red', 'default'),
#            'comment':                     color.build('red', 'default'),
#            'string1':                     color.build('green', 'default'),
#            'string2':                     color.build('green', 'default'),
#            'evalstring':                  color.build('cyan', 'default'),
#            'default string':              color.build('green', 'default'),
#            'keyword':                     color.build('magenta', 'default'),
#            'length scalar':               color.build('yellow', 'default'),
#            'system scalar':               color.build('yellow', 'default'),
#            'system array':                color.build('yellow', 'default'),
#            'scalar':                      color.build('yellow', 'default'),
#            'dereference':                 color.build('yellow', 'default'),
#            'array':                       color.build('yellow', 'default'),
#            'hash':                        color.build('yellow', 'default'),
#            'hash bareword index':         color.build('green', 'default'),
#            'quoted region':               color.build('cyan', 'default'),
#            'match regex':                 color.build('cyan', 'default'),
#            'replace regex':               color.build('cyan', 'default'),
#            'literal hash bareword index': color.build('green', 'default'),
#            'interpolated scalar':         color.build('yellow', 'default'),
#            'interpolated system scalar':  color.build('yellow', 'default'),
#            'interpolated array':          color.build('yellow', 'default'),
#            'interpolated system array':   color.build('yellow', 'default'),
#            'interpolated hash':           color.build('yellow', 'default'),
#            'label':                       color.build('cyan', 'default'),
#            'package':                     color.build('cyan', 'default'),
#            'use':                         color.build('cyan', 'default'),
#            'method':                      color.build('cyan', 'default'),
#            'methodref':                   color.build('cyan', 'default'),
#            'method declaration':          color.build('cyan', 'default'),
#            'instance method':             color.build('cyan', 'default'),
#            'static method':               color.build('cyan', 'default'),
#            'built-in method':             color.build('magenta', 'default'),
#            'bareword method':             color.build('cyan', 'default'),
#            #'bareword':                   color.build('yellow', 'magenta'),
#            'bizzaro':                     color.build('magenta', 'green')
#        }

        #self.highlighter.lex_buffer()
        #self.get_regions()
        #self.tabber = tab_perl.PerlTabber(self)
        #self.functions = None

    def name(self):
        return "Perl"

    def build_function_map(self):
        b = self.window.buffer
        self.functions = {}
        for i in range(0, len(b.lines)):
            m = regex.perl_function.match(b.lines[i])
            if m:
                self.functions[m.group(1)] = i

    def get_functions(self):
        if self.functions is None:
            self.build_function_map()
        return self.functions

    def get_function_names(self):
        functions = self.get_functions()
        pairs = [[functions[key], key] for key in functions]
        pairs.sort()
        names = [x[1] for x in pairs]
        return names
    
class PerlWrapLine(method.Method):
    '''Wrap lines, comments, POD'''
    margin = 80
    comment_re = re.compile('^( *)(#+)( *)([^ ].*)$')
    def _execute(self, w, **vargs):
        pcursor = w.physical_cursor()
        r = w.get_region(pcursor)
        if r is None:
            return

        t = r[4]
        if t == 'pod':
            assert False, 'POD: %s' % repr(r)
        elif t == 'comment':
            self._wrap_comment(w)
        else:
            return

    def _wrap_comment(self, w):
        l = w.logical_cursor()
        m = self.comment_re.match(w.buffer.lines[l.y])
        if not m:
            assert False, 'no match oh geez'

        pad = m.group(1) + m.group(2) + m.group(3)
        data = m.group(4) + ' '

        start = l.y
        end = l.y + 1
        
        while end < len(w.buffer.lines):
            m = self.comment_re.match(w.buffer.lines[end])
            if m:
                data += m.group(4) + ' '
                end += 1
            else:
                break

        words = [word for word in data.split() if word]

        lines = [pad]
        for word in words:
            if len(lines[-1]) == len(pad):
                lines[-1] += word
            elif len(lines[-1]) + 1 + len(word) <= self.margin:
                lines[-1] += ' ' + word
            else:
                lines.append(pad + word)

        # remove the old text and add the new
        start_p = Point(0, start)
        end_p = Point(len(w.buffer.lines[end-1]), end-1)
        w.kill(start_p, end_p)
        w.insert(start_p, '\n'.join(lines))

class PerlCheckSyntax(method.Method):
    '''Check the syntax of a perl file'''
    def _args(self):
        return [method.Argument("lib", type=type(""), prompt="Location of lib: ",
                                default=default.build_constant("."))]
    def _execute(self, window, **vargs):
        a = vargs['lib']
        cmd = "perl -c -I '%s' '%s'" % (a, window.buffer.path)
        (status, output) = commands.getstatusoutput(cmd)
        if status == 0:
            window.application.set_error("Syntax OK")
            window.application.data_buffer("*Perl-Check-Syntax*", output, switch_to=False)
        else:
            window.application.data_buffer("*Perl-Check-Syntax*", output)

class PerlViewModulePerldoc(method.Method):
    '''View documentation about this file using perldoc'''
    def _execute(self, w, **vargs):
        cmd = "perldoc -t -T '%s'" % w.buffer.path
        (status, output) = commands.getstatusoutput(cmd)
        w.application.data_buffer("*Perldoc*", output, switch_to=True)

class PerlViewWordPerldoc(method.Method):
    '''View documentation about a package or function using perldoc'''
    def _execute(self, w, **vargs):
        cursor = w.logical_cursor()
        line = w.buffer.lines[cursor.y]

        word_chars = string.letters + string.digits + '_:'

        if line[cursor.x] not in word_chars:
            w.application.set_error('error: no word selected')
            return

        start = cursor.x
        while start > 0 and line[start - 1] in word_chars:
            start -= 1

        end = cursor.x + 1
        while end < len(line) - 1 and line[end] in word_chars:
            end += 1

        word = line[start:end]
        w.application.set_error('the current word is: %r' % word)

        ok = False
        data = ''

        perl_word_re = re.compile('^[a-zA-Z_][a-zA-Z_0-9]*(?:::[a-zA-Z_][a-zA-Z0-9]*)*$')
        if not perl_word_re.match(word):
            w.application.set_error('invalid word: %r' % word)
            return

        if '::' in word:
            # we are probably dealing with a package
            parts = word.split('::')
            while len(parts) > 0:
                newword = '::'.join(parts)
                cmd = "perldoc -t -T '%s'" % newword
                (status, data) = commands.getstatusoutput(cmd)
                if status == 0:
                    word = newword
                    ok = True
                    break
                parts.pop(-1)
        elif ':' in word:
            w.application.set_error('invalid word2222: %r' % word)
            return
        else:
            cmd = "perldoc -t -T -f '%s'" % word
            (status, data) = commands.getstatusoutput(cmd)
            if status == 0:
                ok = True
            else:
                cmd = "perldoc -t -T -f '%s'" % word
                (status, data) = commands.getstatusoutput(cmd)
                ok = status == 0

        if not ok:
            w.application.set_error('nothing found for %r' % word)
        else:
            w.application.data_buffer("*Perldoc*", data, switch_to=True)
            w.application.set_error('displaying documentation for %r' % word)

class PerlGotoFunction(method.Method):
    '''Jump to a function defined in this module'''
    def _args(self):
        return [method.Argument("name", type=type(""), datatype="perlfunction",
                                prompt="Goto Function: ")]
    def _execute(self, w, **vargs):
        name = vargs['name']
        functions = w.mode.get_functions()
        if name in functions:
            number = functions[name]
            p = Point(0, number)
            w.goto(p)
        else:
            w.application.set_error("Function %r was not found" % name)

class PerlListFunctions(method.Method):
    '''Show the user all functions defined in this module'''
    def _execute(self, w, **vargs):
        names = w.mode.get_function_names()
        output = "\n".join(names) + "\n"
        w.application.data_buffer("*Perl-List-Functions*", output, switch_to=True)

class PerlWhichFunction(method.Method):
    '''Show the user what function they are in'''
    def _execute(self, w, **vargs):
        cursor = w.logical_cursor()
        i = cursor.y
        name = None
        while i >= 0 and name is None:
            line = w.buffer.lines[i]
            m = regex.perl_function.match(line)
            if m:
                name = m.group(1)
            else:
                i -= 1
        if name is None:
            w.application.set_error("None");
        else:
            w.application.set_error("line %d: %s" % (i, name))

class PerlHashCleanup(method.Method):
    '''Correctly align assignment blocks and literal hashes'''
    def _execute(self, window, **vargs):
        cursor = window.logical_cursor()
        b = window.buffer

        # so this is where we will store the groups that we find
        groups_by_line = {}

        # the regex we will try
        regexes = [regex.perl_hash_cleanup,
                   regex.perl_assign_cleanup]

        # if we aren't in a hash, inform the user and exit
        line = b.lines[cursor.y]
        myregex = None
        for r in regexes:
            if r.match(line):
                myregex = r

        if myregex is None:
            raise Exception, "Not a perl hash line"

        groups_by_line[cursor.y] = myregex.match(line).groups()
        
        # find the beginning of this hash block
        start = 0
        i = cursor.y - 1
        while i >= 0:
            line = b.lines[i]
            m = myregex.match(line)
            if not m:
                start = i + 1
                break
            else:
                groups_by_line[i] = m.groups()
            i -= 1

        # find the end of this hash block
        end = len(b.lines) - 1
        i = cursor.y + 1
        while i < len(b.lines):
            line = b.lines[i]
            m = myregex.match(line)
            if not m:
                end = i - 1
                break
            else:
                groups_by_line[i] = m.groups()
            i += 1
    
        # assume that the least indented line is correct
        indent_w = min([len(groups_by_line[k][0]) for k in groups_by_line])

        # find the longest hash key to base all the other padding on
        key_w = max([len(groups_by_line[k][1]) for k in groups_by_line])

        # for each line, format it correctly
        keys = groups_by_line.keys()
        keys.sort()
        data = ''
        for i in keys:
            indent_pad = ' ' * indent_w
            key = groups_by_line[i][1]
            sep = groups_by_line[i][3]
            value = groups_by_line[i][5]
            key_pad = ' ' * (key_w - len(key))
            data += indent_pad + key + key_pad + ' ' + sep + ' ' + value + '\n'

        # remove the old text and add the new
        start_p = Point(0, start)
        end_p = Point(0, end + 1)
        window.kill(start_p, end_p)
        window.insert(start_p, data)

class PerlHashCleanup2(method.Method):
    '''Correctly align assignment blocks and literal hashes'''
    def process_line2(self, line_regions, sep=None, indent=None):
        (pre_toks, sep_tok, post_toks)  = ([], None, [])
        ok = False
        before = True
        for r in line_regions:
            (start, end, attr, s, name) = r
            if name == "":
                continue
            elif before:
                if len(pre_toks) == 0:
                    pre_toks.append(r)
                elif (name == "delimiter" and s == sep or
                      (sep is None and (s == "=" or s == "=>"))):
                    sep_tok = r
                    before = False
                else:
                    pre_toks.append(r)
            else:
                post_toks.append(r)
                ok = True

        if ok:
            return (True, sep_tok[3], (pre_toks, sep_tok, post_toks))
        else:
            return (False, "", ([], None, []))
        
    def _execute(self, w, **vargs):
        cursor = w.logical_cursor()
        b = w.buffer

        # so this is where we will store the groups that we find
        groups_by_line = {}

        all_regions = w.mode.get_regions()
        line_regions = all_regions[cursor.y]
        (ok, sep, group) = self.process_line2(line_regions)

        if not ok:
            raise Exception, "Not a perl hash line"
        groups_by_line[cursor.y] = group
        
        # find the beginning of this hash block
        start = cursor.y
        while start >= 0:
            (ok2, sep2, group2) = self.process_line2(all_regions[start - 1], sep)
            if not ok2:
                break
            start -= 1
            groups_by_line[start] = group2

        # find the end of this hash block
        end = cursor.y
        while end < len(b.lines) - 1:
            (ok2, sep2, group2) = self.process_line2(all_regions[end + 1], sep)
            if not ok2:
                break
            end += 1
            groups_by_line[end] = group2
    
        # find the minimum indented line
        indent_w = None
        for k in groups_by_line:
            x = groups_by_line[k][0][0].start
            if indent_w is None or x < indent_w:
                indent_w = x

        # find the max key length
        key_w = None
        for k in groups_by_line:
            x = groups_by_line[k][0][-1].end - groups_by_line[k][0][0].start
            if key_w is None or x > key_w:
                key_w = x

        # for each line, format it correctly
        keys = groups_by_line.keys()
        keys.sort()
        data = ''
        for i in keys:
            line = ' ' * indent_w
            l = groups_by_line[i][0][0].start
            for t in groups_by_line[i][0]:
                line += ' ' * max(0, t.start - l)
                line += t.value
                l = t.end
            line += ' ' * max(0, key_w - l + groups_by_line[i][0][0].start)
            line += ' ' + groups_by_line[i][1].value + ' '
            l = groups_by_line[i][2][0].start
            for t in groups_by_line[i][2]:
                line += ' ' * max(0, t.start - l)
                line += t.value
                l = t.end
            data += line + '\n'

        # remove the old text and add the new
        start_p = Point(0, start)
        end_p = Point(0, end + 1)
        w.kill(start_p, end_p)
        w.insert(start_p, data)