antlr4 python 3 print or dump tokens from plsql grammar

问题

I am using antlr4 in Python, to read the following grammar :

https://github.com/antlr/grammars-v4/tree/master/plsql

file grants.sql just has "begin select 'bob' from dual; end;"

simple code to print lisp like tree

from antlr4 import *
from PlSqlLexer import PlSqlLexer
from PlSqlParser import PlSqlParser
from PlSqlParserListener import PlSqlParserListener

input = FileStream('grants.sql')
lexer = PlSqlLexer(input)

stream = CommonTokenStream(lexer)
parser = PlSqlParser(stream)
tree = parser.sql_script()

print ("Tree " + tree.toStringTree(recog=parser));

Output is thus:

Tree (sql_script (unit_statement (anonymous_block BEGIN (seq_of_statements (statement (sql_statement (data_manipulation_language_statements (select_statement (subquery (subquery_basic_elements (query_block SELECT (selected_element (select_list_elements (expressions (expression (logical_expression (multiset_expression (relational_expression (compound_expression (concatenation (model_expression (unary_expression (atom (constant (quoted_string 'bob')))))))))))))) (from_clause FROM (table_ref_list (table_ref (table_ref_aux (table_ref_aux_internal (dml_table_expression_clause (tableview_name (identifier (id_expression (regular_id DUAL))))))))))))))))) ;) END ;)) )

I'd like to be able to have python code that lists the above not in a lisp like statement but lists all the rules and tokens.. i.e

.sql_script
1. ..unit_statement
2. ...anonymous_block
3. ....BEGIN

etc etc

Can someone supply python code that does this or give me some hints. Gratefully appreciated.

回答1:

Here's a start:

from antlr4 import *
from antlr4.tree.Tree import TerminalNodeImpl
from PlSqlLexer import PlSqlLexer
from PlSqlParser import PlSqlParser

# Generate the lexer nad parser like this:
#
#   java -jar antlr-4.7.1-complete.jar -Dlanguage=Python3 *.g4
#
def main():
    lexer = PlSqlLexer(InputStream("SELECT * FROM TABLE_NAME"))
    parser = PlSqlParser(CommonTokenStream(lexer))
    tree = parser.sql_script()
    traverse(tree, parser.ruleNames)

def traverse(tree, rule_names, indent = 0):
    if tree.getText() == "<EOF>":
        return
    elif isinstance(tree, TerminalNodeImpl):
        print("{0}TOKEN='{1}'".format("  " * indent, tree.getText()))
    else:
        print("{0}{1}".format("  " * indent, rule_names[tree.getRuleIndex()]))
        for child in tree.children:
            traverse(child, rule_names, indent + 1)

if __name__ == '__main__':
    main()

which prints:

sql_script
  unit_statement
    data_manipulation_language_statements
      select_statement
        subquery
          subquery_basic_elements
            query_block
              TOKEN='SELECT'
              TOKEN='*'
              from_clause
                TOKEN='FROM'
                table_ref_list
                  table_ref
                    table_ref_aux
                      table_ref_aux_internal
                        dml_table_expression_clause
                          tableview_name
                            identifier
                              id_expression
                                regular_id
                                  TOKEN='TABLE_NAME'

Note that for the lexer and parser to work properly, I added the following Python classes:

# PlSqlBaseLexer.py
from antlr4 import *

class PlSqlBaseLexer(Lexer):

    def IsNewlineAtPos(self, pos):
        la = self._input.LA(pos)
        return la == -1 or la == '\n'

and:

# PlSqlBaseParser.py
from antlr4 import *

class PlSqlBaseParser(Parser):

    _isVersion10 = False
    _isVersion12 = True

    def isVersion10(self):
        return self._isVersion10

    def isVersion12(self):
        return self._isVersion12

    def setVersion10(self, value):
        self._isVersion10 = value

    def setVersion12(self, value):
        self._isVersion12 = value

which I placed in the same folder as the generated Python classes. I also needed to and the import statement from PlSqlBaseLexer import PlSqlBaseLexer in the generated PlSqlLexer.py class, and fix the import statement in PlSqlParser.py from from ./PlSqlBaseParser import PlSqlBaseParser to from PlSqlBaseParser import PlSqlBaseParser.

Note that running the demo is rather slow. Unless you have a hard requirement to do this in Python, I recommend going with the (much!) faster Java or C# target instead.

来源：https://stackoverflow.com/questions/52673751/antlr4-python-3-print-or-dump-tokens-from-plsql-grammar

标签

python

parsing

tree

token

antlr4