util_text.py

#!/usr/bin/env python
# -*- coding: utf-8 -*-
#
# Copyright (c) 2012-2018 Snowflake Computing Inc. All right reserved.
#
import logging
import re

import ijson

COMMENT_PATTERN_RE = re.compile(r'^\s*\-\-')
EMPTY_LINE_RE = re.compile(r'^\s*$')

_logger = logging.getLogger(__name__)


def split_statements(buf, remove_comments=False):
    """
    Splits a stream into SQL statements (ends with a semicolon) or
    commands (!...)
    :param buf: Unicode data stream
    :param remove_comments: True removes all comments
    :return: yields a SQL statement or a command
    """
    in_quote = False
    ch_quote = None
    in_comment = False
    in_double_dollars = False

    line = buf.readline()
    if isinstance(line, bytes):
        raise TypeError("Input data must not be binary type.")

    statement = []
    while line != '':
        col = 0
        col0 = 0
        len_line = len(line)
        while True:
            if col >= len_line:
                if col0 < col:
                    if not in_comment and not in_quote \
                            and not in_double_dollars:
                        statement.append((line[col0:col], True))
                        if len(statement) == 1 and statement[0][0] == '':
                            statement = []
                        break
                    elif not in_comment and in_quote:
                        statement.append((line[col0:col], True))
                    elif not remove_comments:
                        statement.append((line[col0:col], False))
                break
            elif in_comment:
                if line[col:].startswith("*/"):
                    in_comment = False
                    if not remove_comments:
                        statement.append((line[col0:col + 2], False))
                    col += 2
                    col0 = col
                else:
                    col += 1
            elif in_double_dollars:
                if line[col:].startswith("$$"):
                    in_double_dollars = False
                    statement.append((line[col0:col + 2], False))
                    col += 2
                    col0 = col
                else:
                    col += 1
            elif in_quote:
                if line[col] == ch_quote:
                    if col < len_line - 1 and line[col + 1] != ch_quote or \
                                    col == len_line - 1:
                        # exits quote
                        in_quote = False
                        statement.append((line[col0:col + 1], True))
                        col += 1
                        col0 = col
                    else:
                        # escaped quote and still in quote
                        col += 2
                else:
                    col += 1
            else:
                if line[col] in ("'", '"'):
                    in_quote = True
                    ch_quote = line[col]
                    col += 1
                elif line[col] in (' ', '\t'):
                    statement.append((line[col0:col + 1], True))
                    col += 1
                    col0 = col
                elif line[col:].startswith("--"):
                    statement.append((line[col0:col], True))
                    if not remove_comments:
                        # keep the comment
                        statement.append((line[col:], False))
                    col = len_line + 1
                    col0 = col
                elif line[col:].startswith("/*") and \
                        not line[col0:].startswith("file://"):
                    if not remove_comments:
                        statement.append((line[col0:col + 2], False))
                    col += 2
                    col0 = col
                    in_comment = True
                elif line[col:].startswith("$$"):
                    statement.append((line[col0:col + 2], True))
                    col += 2
                    col0 = col
                    in_double_dollars = True
                elif line[col] == ';':
                    statement.append((line[col0:col + 1], True))
                    col += 1
                    try:
                        if line[col] == '>':
                            col += 1
                            statement[-1] = (statement[-1][0] + '>',
                                             statement[-1][1])
                    except IndexError:
                        pass
                    if COMMENT_PATTERN_RE.match(line[col:]) or \
                            EMPTY_LINE_RE.match(line[col:]):
                        if not remove_comments:
                            # keep the comment
                            statement.append((line[col:], False))
                        col = len_line
                    while col < len_line and line[col] in (' ', '\t'):
                        col += 1
                    yield _concatenate_statements(statement)
                    col0 = col
                    statement = []
                elif col == 0 and line[col] == '!':  # command
                    if len(statement) > 0:
                        yield _concatenate_statements(statement)
                        statement = []
                    yield line.rstrip(';').strip(), False
                    break
                else:
                    col += 1
        line = buf.readline()

    if len(statement) > 0:
        yield _concatenate_statements(statement)


def _concatenate_statements(statement_list):
    """
    concatenate statements

    is_put_or_get is set to True if the statement is PUT or GET otherwise
    False for valid statement. None is set if the statement is empty or
    comment only.
    :return: a statement, is_put_or_get
    """
    valid_statement_list = []
    is_put_or_get = None
    for text, is_statement in statement_list:
        valid_statement_list.append(text)
        if is_put_or_get is None and is_statement and len(text.strip()) >= 3:
            is_put_or_get = text[:3].upper() in ('PUT', 'GET')
    return u''.join(valid_statement_list).strip(), is_put_or_get


def split_rows_from_stream(stream):
    """
    Splits into rows from a stream object. Generator.
    """
    row = []
    in_row = False
    for prefix, event, value in ijson.parse(stream):
        if prefix == '':
            continue
        if in_row:
            if event == 'end_array':
                yield row
                row = []
                in_row = False
            else:
                row.append(value)
        elif event == 'start_array':
            in_row = True


def construct_hostname(region, account):
    """
    Constructs hostname from region and account
    """
    if region:
        if account.find(u'.') > 0:
            account = account[0:account.find(u'.')]
        host = u'{0}.{1}.snowflakecomputing.com'.format(account, region)
    else:
        host = u'{0}.snowflakecomputing.com'.format(account)
    return host