metrixpp/releases/metrixplusplus-1.0.63/ext/std/code/java.py

#
#    Metrix++, Copyright 2009-2013, Metrix++ Project
#    Link: http://metrixplusplus.sourceforge.net
#    
#    This file is a part of Metrix++ Tool.
#    
#    Metrix++ is free software: you can redistribute it and/or modify
#    it under the terms of the GNU General Public License as published by
#    the Free Software Foundation, version 3 of the License.
#    
#    Metrix++ is distributed in the hope that it will be useful,
#    but WITHOUT ANY WARRANTY; without even the implied warranty of
#    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
#    GNU General Public License for more details.
#    
#    You should have received a copy of the GNU General Public License
#    along with Metrix++.  If not, see <http://www.gnu.org/licenses/>.
#


import re
import binascii
import logging

import core.api

class Plugin(core.api.Plugin, core.api.Parent, core.api.IParser, core.api.IConfigurable, core.api.ICode):
    
    def declare_configuration(self, parser):
        parser.add_option("--std.code.java.files", default="*.java",
                         help="Enumerates filename extensions to match Java files [default: %default]")
    
    def configure(self, options):
        self.files = options.__dict__['std.code.java.files'].split(',')
        self.files.sort() # sorted list goes to properties
        
    def initialize(self):
        # trigger version property set
        core.api.Plugin.initialize(self)
        db_loader = self.get_plugin_loader().get_database_loader()
        prev_ext = db_loader.set_property(self.get_name() + ":files", ','.join(self.files))
        if prev_ext != ','.join(self.files):
            self.is_updated = True

        self.get_plugin_loader().register_parser(self.files, self)
        
    def process(self, parent, data, is_updated):
        is_updated = is_updated or self.is_updated
        count_mismatched_brackets = 0
        if is_updated == True:
            count_mismatched_brackets = JavaCodeParser().run(data)
        self.notify_children(data, is_updated)
        return count_mismatched_brackets
            
class JavaCodeParser(object):
    
    regex_cpp = re.compile(r'''
                   //(?=\n|\r|\r\n)                                   # Match Java style comments (empty comment line)
                |  //.*?(?=\n|\r|\r\n)                                # Match Java style comments
                                                                      # NOTE: end of line is NOT consumed
                                                                      # NOTE: it is slightly different in C++
                | /\*\*/                                              # Match C style comments (empty comment line)
                                                                      # NOTE: it is slightly different in C++
                | /\*.*?\*/                                           # Match C style comments
                                                                      # NOTE: it is slightly different in C++
                | \'(?:\\.|[^\\\'])*\'                                # Match quoted strings
                | "(?:\\.|[^\\"])*"                                   # Match double quoted strings
                | (?P<fn_name>([_$a-zA-Z][_$a-zA-Z0-9]*))\s*[(]       # Match function
                                                                      # NOTE: Java may include $ in the name
                                                                      # LIMITATION: if there are comments after function name
                                                                      # and before '(', it is not detected
                | ((?P<block_type>class|interface)                    # Match class or namespace
                    (?P<block_name>(\s+[_$a-zA-Z][_$a-zA-Z0-9]*)))
                                                                      # NOTE: noname instances are impossible in Java
                                                                      # LIMITATION: if there are comments between keyword and name,
                                                                      # it is not detected
                | [{};]                                               # Match block start/end and statement separator
                                                                      # NOTE: C++ parser includes processing of <> and : 
                                                                      #       to handle template definitions, it is easier in Java
                | ((?:\n|\r|\r\n)\s*(?:\n|\r|\r\n))                   # Match double empty line
            ''',
            re.DOTALL | re.MULTILINE | re.VERBOSE
        )

    regex_ln = re.compile(r'(\n)|(\r)|(\r\n)')

    def run(self, data):
        self.__init__() # Go to initial state if it is called twice
        return self.parse(data)
        
    def finalize_block(self, text, block, block_end):
        space_match = re.match('^\s*', text[block['start']:block_end], re.MULTILINE)
        block['start'] += space_match.end() # trim spaces at the beginning
        block['end'] = block_end

        start_pos = block['start']
        crc32 = 0
        for child in block['children']:
            # exclude children
            crc32 = binascii.crc32(text[start_pos:child['start']], crc32)
            start_pos = child['end']
        block['checksum'] = binascii.crc32(text[start_pos:block['end']], crc32) & 0xffffffff # to match python 3
        
    def add_lines_data(self, text, blocks):
        def add_lines_data_rec(self, text, blocks):
            for each in blocks:
                # add line begin
                self.total_current += len(self.regex_ln.findall(text, self.total_last_pos, each['start']))
                each['line_begin'] = self.total_current
                self.total_last_pos = each['start']
                # process enclosed
                add_lines_data_rec(self, text, each['children'])
                # add line end
                self.total_current += len(self.regex_ln.findall(text, self.total_last_pos, each['end']))
                each['line_end'] = self.total_current
                self.total_last_pos = each['end']
        self.total_last_pos = 0
        self.total_current = 1
        add_lines_data_rec(self, text, blocks)

    def add_regions(self, data, blocks):
        # Note: data.add_region() internals depend on special ordering of regions
        # in order to identify enclosed regions efficiently
        def add_regions_rec(self, data, blocks):
            def get_type_id(data, named_type):
                if named_type == "function":
                    return data.get_region_types().FUNCTION
                elif named_type == "class":
                    return data.get_region_types().CLASS
                elif named_type == "interface":
                    return data.get_region_types().INTERFACE
                elif named_type == "__global__":
                    return data.get_region_types().GLOBAL
                else:
                    assert(False)
            for each in blocks:
                data.add_region(each['name'], each['start'], each['end'],
                                each['line_begin'], each['line_end'], each['cursor'],
                                get_type_id(data, each['type']), each['checksum'])
                add_regions_rec(self, data, each['children'])
        add_regions_rec(self, data, blocks)
        
    def parse(self, data):
        
        def reset_next_block(start):
            return {'name':'', 'start':start, 'cursor':0, 'type':''}
        
        count_mismatched_brackets = 0
        
        text = data.get_content()
        indent_current = 0;
        
        blocks = [{'name':'__global__', 'start':0, 'cursor':0, 'type':'__global__', 'indent_start':indent_current, 'children':[]}]
        curblk = 0
        
        next_block = reset_next_block(0)
        
        cursor_last_pos = 0
        cursor_current = 1
        
        for m in re.finditer(self.regex_cpp, text):
            # Comment
            if text[m.start()] == '/':
                data.add_marker(m.start(), m.end(), data.get_marker_types().COMMENT)
                if text[m.start():m.end()].startswith("//\n"):
                    print text[m.start():m.end()]
            
            # String
            elif text[m.start()] == '"' or text[m.start()] == '\'':
                data.add_marker(m.start() + 1, m.end() - 1, data.get_marker_types().STRING)
            
            # Statement end
            elif text[m.start()] == ';':
                # Reset next block name and start
                next_block['name'] = ""
                next_block['start'] = m.end() # potential region start

            # Double end line
            elif text[m.start()] == '\n' or text[m.start()] == '\r':
                # Reset next block start, if has not been named yet
                if next_block['name'] == "":
                    next_block['start'] = m.end() # potential region start

            # Block start...
            elif text[m.start()] == '{':
                # shift indent right
                indent_current += 1
                
                # ... if name detected previously
                if next_block['name'] != '': # - Start of enclosed block
                    blocks.append({'name':next_block['name'],
                                   'start':next_block['start'],
                                   'cursor':next_block['cursor'],
                                   'type':next_block['type'],
                                   'indent_start':indent_current,
                                   'children':[]})
                    next_block = reset_next_block(m.end())
                    curblk += 1
                # ... reset next block start, otherwise
                else: # - unknown type of block start
                    next_block['start'] = m.end() # potential region start
            
            # Block end...
            elif text[m.start()] == '}':
                # ... if indent level matches the start
                if blocks[curblk]['indent_start'] == indent_current:
                    next_block = reset_next_block(m.end())
                    if curblk == 0:
                        logging.warning("Non-matching closing bracket '}' detected: " + data.get_path() + ":" +
                                        str(cursor_current + len(self.regex_ln.findall(text, cursor_last_pos, m.start()))))
                        count_mismatched_brackets += 1
                        continue
                    
                    self.finalize_block(text, blocks[curblk], m.end())
                    assert(blocks[curblk]['type'] != '__global__')
                    
                    curblk -= 1
                    assert(curblk >= 0)
                    blocks[curblk]['children'].append(blocks.pop())

                # shift indent left
                indent_current -= 1
                if indent_current < 0:
                    logging.warning("Non-matching closing bracket '}' detected")
                    count_mismatched_brackets += 1
                    indent_current = 0

            # Potential class, interface
            elif m.group('block_type') != None:
                if next_block['name'] == "":
                    # - 'name'
                    next_block['name'] = m.group('block_name').strip()
                    # - 'cursor'
                    cursor_current += len(self.regex_ln.findall(text, cursor_last_pos, m.start('block_name')))
                    cursor_last_pos = m.start('block_name')
                    next_block['cursor'] = cursor_current
                    # - 'type'
                    next_block['type'] = m.group('block_type').strip()
                    # - 'start' detected earlier

            # Potential function name detected...
            elif m.group('fn_name') != None:
                # ... if outside of a function
                #     (do not detect functions enclosed directly in a function, i.e. without classes)
                # ... and other name before has not been matched 
                if blocks[curblk]['type'] != 'function' and (next_block['name'] == ""):
                    # - 'name'
                    next_block['name'] = m.group('fn_name').strip()
                    # - 'cursor'
                    cursor_current += len(self.regex_ln.findall(text, cursor_last_pos, m.start('fn_name')))
                    cursor_last_pos = m.start('fn_name')
                    # NOTE: cursor could be collected together with line_begin, line_end,
                    # but we keep it here separately for easier debugging of file parsing problems
                    next_block['cursor'] = cursor_current
                    # - 'type'
                    next_block['type'] = 'function'
                    # - 'start' detected earlier
            else:
                assert(len("Unknown match by regular expression") == 0)

        while indent_current > 0:
            # log all
            logging.warning("Non-matching opening bracket '{' detected")
            count_mismatched_brackets += 1
            indent_current -= 1

        for (ind, each) in enumerate(blocks):
            each = each # used
            block = blocks[len(blocks) - 1 - ind]
            self.finalize_block(text, block, len(text))

        self.add_lines_data(text, blocks)
        self.add_regions(data, blocks)
        
        return count_mismatched_brackets