diff options
Diffstat (limited to 'pygments/lexers/sql.py')
| -rw-r--r-- | pygments/lexers/sql.py | 78 |
1 files changed, 63 insertions, 15 deletions
diff --git a/pygments/lexers/sql.py b/pygments/lexers/sql.py index 7507c0fc..d789052d 100644 --- a/pygments/lexers/sql.py +++ b/pygments/lexers/sql.py @@ -34,7 +34,7 @@ The ``tests/examplefiles`` contains a few test files with data to be parsed by these lexers. - :copyright: Copyright 2006-2017 by the Pygments team, see AUTHORS. + :copyright: Copyright 2006-2019 by the Pygments team, see AUTHORS. :license: BSD, see LICENSE for details. """ @@ -59,7 +59,14 @@ line_re = re.compile('.*?\n') language_re = re.compile(r"\s+LANGUAGE\s+'?(\w+)'?", re.IGNORECASE) -do_re = re.compile(r'\bDO\b', re.IGNORECASE) +do_re = re.compile(r'\bDO\b', re.IGNORECASE) + +# Regular expressions for analyse_text() +name_between_bracket_re = re.compile(r'\[[a-zA-Z_]\w*\]') +name_between_backtick_re = re.compile(r'`[a-zA-Z_]\w*`') +tsql_go_re = re.compile(r'\bgo\b', re.IGNORECASE) +tsql_declare_re = re.compile(r'\bdeclare\s+@', re.IGNORECASE) +tsql_variable_re = re.compile(r'@[a-zA-Z_]\w*\b') def language_callback(lexer, match): @@ -82,7 +89,7 @@ def language_callback(lexer, match): lexer.text[max(0, match.start()-25):match.start()])) if m: l = lexer._get_lexer('plpgsql') - + # 1 = $, 2 = delimiter, 3 = $ yield (match.start(1), String, match.group(1)) yield (match.start(2), String.Delimiter, match.group(2)) @@ -155,7 +162,7 @@ class PostgresLexer(PostgresBase, RegexLexer): (r'\s+', Text), (r'--.*\n?', Comment.Single), (r'/\*', Comment.Multiline, 'multiline-comments'), - (r'(' + '|'.join(s.replace(" ", "\s+") + (r'(' + '|'.join(s.replace(" ", r"\s+") for s in DATATYPES + PSEUDO_TYPES) + r')\b', Name.Builtin), (words(KEYWORDS, suffix=r'\b'), Keyword), @@ -308,14 +315,7 @@ class PostgresConsoleLexer(Lexer): # and continue until the end of command is detected curcode = '' insertions = [] - while 1: - try: - line = next(lines) - except StopIteration: - # allow the emission of partially collected items - # the repl loop will be broken below - break - + for line in lines: # Identify a shell prompt in case of psql commandline example if line.startswith('$') and not curcode: lexer = get_lexer_by_name('console', **self.options) @@ -346,8 +346,7 @@ class PostgresConsoleLexer(Lexer): # Emit the output lines out_token = Generic.Output - while 1: - line = next(lines) + for line in lines: mprompt = re_prompt.match(line) if mprompt is not None: # push the line back to have it processed by the prompt @@ -363,6 +362,8 @@ class PostgresConsoleLexer(Lexer): yield (mmsg.start(2), out_token, mmsg.group(2)) else: yield (0, out_token, line) + else: + return class SqlLexer(RegexLexer): @@ -480,6 +481,9 @@ class SqlLexer(RegexLexer): ] } + def analyse_text(text): + return 0.01 + class TransactSqlLexer(RegexLexer): """ @@ -499,7 +503,7 @@ class TransactSqlLexer(RegexLexer): tokens = { 'root': [ (r'\s+', Whitespace), - (r'--(?m).*?$\n?', Comment.Single), + (r'(?m)--.*?$\n?', Comment.Single), (r'/\*', Comment.Multiline, 'multiline-comments'), (words(_tsql_builtins.OPERATORS), Operator), (words(_tsql_builtins.OPERATOR_WORDS, suffix=r'\b'), Operator.Word), @@ -536,6 +540,33 @@ class TransactSqlLexer(RegexLexer): ] } + def analyse_text(text): + rating = 0 + if tsql_declare_re.search(text): + # Found T-SQL variable declaration. + rating = 1.0 + else: + name_between_backtick_count = len( + name_between_backtick_re.findall((text))) + name_between_bracket_count = len( + name_between_bracket_re.findall(text)) + # We need to check if there are any names using + # backticks or brackets, as otherwise both are 0 + # and 0 >= 2 * 0, so we would always assume it's true + dialect_name_count = name_between_backtick_count + name_between_bracket_count + if dialect_name_count >= 1 and name_between_bracket_count >= 2 * name_between_backtick_count: + # Found at least twice as many [name] as `name`. + rating += 0.5 + elif name_between_bracket_count > name_between_backtick_count: + rating += 0.2 + elif name_between_bracket_count > 0: + rating += 0.1 + if tsql_variable_re.search(text) is not None: + rating += 0.1 + if tsql_go_re.search(text) is not None: + rating += 0.1 + return rating + class MySqlLexer(RegexLexer): """ @@ -609,6 +640,23 @@ class MySqlLexer(RegexLexer): ] } + def analyse_text(text): + rating = 0 + name_between_backtick_count = len( + name_between_backtick_re.findall((text))) + name_between_bracket_count = len( + name_between_bracket_re.findall(text)) + # Same logic as above in the TSQL analysis + dialect_name_count = name_between_backtick_count + name_between_bracket_count + if dialect_name_count >= 1 and name_between_backtick_count >= 2 * name_between_bracket_count: + # Found at least twice as many `name` as [name]. + rating += 0.5 + elif name_between_backtick_count > name_between_bracket_count: + rating += 0.2 + elif name_between_backtick_count > 0: + rating += 0.1 + return rating + class SqliteConsoleLexer(Lexer): """ |