source: tracdarcs/tracdarcs/dbutil.py @ 168

# -*- coding: utf-8 -*-
#
# Copyright (C) 2005 Edgewall Software
# Copyright (C) 2006 K.S.Sreeram <sreeram@tachyontech.net>
# Copyright (C) 2007-2010 Lele Gaifax <lele@metapensiero.it>
#
# This software is licensed as described in the file COPYING, which
# you should have received as part of this distribution. The terms
# are also available at http://trac.edgewall.com/license.html.
#
# This software consists of voluntary contributions made by many
# individuals. For the exact contribution history, see the revision
# history and logs, available at http://projects.edgewall.com/trac/.
#
# Author: K.S.Sreeram <sreeram@tachyontech.net>

import re
from trac import __version__ as tracVersion

# these values are used in the 'node_type' field in the
# 'darcs_nodes' table.
NODE_FILE_TYPE = 'F'
NODE_DIR_TYPE = 'D'

# these values are used in the 'change' field in the
# 'darcs_node_changes' table
CHANGE_ADDED = 'A'
CHANGE_REMOVED = 'R'
CHANGE_MOVED = 'M'
CHANGE_EDITED = 'E'
CHANGE_MOVED_EDITED = 'ME'

IS_TRAC_0_10_X = not not re.match(r'^0\.10', tracVersion)
IS_TRAC_0_11_X = not not re.match(r'^0\.11', tracVersion)
IS_TRAC_0_12_OR_BETTER = not not re.match(r'^0\.1[2-9]', tracVersion)

def query_nodes_for_revision(repo_id, rev=None):
    '''
    Returns a query which represents the nodes available as of
    revision 'rev'. If 'rev' is None, then the latest revision is
    used. Returns a set of 'node_id,rev,path,parent_id' values.
    '''

    if rev is None:
        revfilter = "dn.remove_rev IS NULL"
    else:
        revfilter = ("dnc2.rev <= %d"
                     " AND ((dn.remove_rev IS NULL) "
                     "      OR (dn.remove_rev > %d))" % (rev,rev))
    q = """SELECT dnc.node_id, dnc.rev, dnc.path, dnc.parent_id
           FROM darcs_node_changes AS dnc
           WHERE dnc.repo_id = %s
             AND dnc.rev = (SELECT max(dnc2.rev)
                            FROM darcs_node_changes AS dnc2, darcs_nodes AS dn
                            WHERE dnc2.repo_id = dnc.repo_id
                              AND dnc2.node_id = dnc.node_id
                              AND dn.repo_id = dnc2.repo_id
                              AND dn.node_id = dnc2.node_id
                              AND %s)""" % (repo_id, revfilter)
    return q

def get_node_type(db, repo_id, node_id):
    c = db.cursor()
    c.execute('SELECT node_type FROM darcs_nodes '
              'WHERE repo_id = %s AND node_id = %s', (repo_id, node_id))
    return c.fetchone()[0]

def get_prev_path_rev(db, repo_id, node_id, rev):
    c = db.cursor()
    c.execute('SELECT path,rev FROM darcs_node_changes '
              'WHERE repo_id = %s AND node_id = %s AND rev < %s '
              'ORDER BY rev DESC LIMIT 1', (repo_id, node_id, rev))
    path,rev = c.fetchone()
    return path,rev

def get_repository_id(db, path):
    if IS_TRAC_0_10_X or IS_TRAC_0_11_X:
        return ''

    c = db.cursor()
    c.execute('SELECT r1.id '
              'FROM repository r1, repository r2 '
              'WHERE r1.id = r2.id '
              '  AND r1.name = %s  AND r1.value = %s '
              '  AND r2.name = %s  AND r2.value = %s '
              'LIMIT 1', ('dir', path,
                          'type', 'darcs'))
    row = c.fetchone()
    if row is not None:
        return row[0]
    return ''

if __debug__:
    class TimedCursor(object):
        "Wrapper around a database cursor that keeps track of execution timings."

        # Map a given statement to its stats, four values respectively
        # number of executions, fastest run, slowest run, total time.
        _timings = {}

        def __init__(self, cursor):
            self.cursor = cursor

        def execute(self, stmt, *args):
            from time import time

            t0 = time()
            self.cursor.execute(stmt, *args)
            t1 = time()
            usec = (t1-t0) * 1e6
            old = self._timings.setdefault(stmt, [0, None, 0, 0])
            old[0] += 1
            if old[1] is None or old[1] > usec:
                old[1] = usec
            if old[2] < usec:
                old[2] = usec
            old[3] += usec

        def fetchone(self):
            return self.cursor.fetchone()

        def fetchall(self):
            return self.cursor.fetchall()

        def __iter__(self):
            return iter(self.cursor)

        @classmethod
        def report(klass, log):
            def trepr(usec):
                if usec < 1000:
                    return "%.*g microsecs" % (3, usec)
                else:
                    msec = usec / 1000
                    if msec < 1000:
                        return "%.*g millisecs" % (3, msec)
                    else:
                        sec = msec / 1000
                        return "%.*g seconds" % (3, sec)

            all = list(klass._timings.items())
            all.sort(key=lambda item: item[1][3]/item[1][0])
            log.debug("Executed %d different statements, ordered by average execution time:",
                      len(all))
            for stmt, stats in all:
                nr, mint, maxt, tott = stats
                if nr>1:
                    avg = tott / nr
                    log.debug("%s: %d executions in %s, fastest run %s, slowest %s, average %s",
                              stmt, nr, trepr(tott), trepr(mint), trepr(maxt), trepr(avg))
                else:
                    log.debug("%s: executed once in %s", stmt, trepr(tott))

    class TimedDB(object):
        "Simple wrapper around a connection uses TimedCursors and print a final report."

        def __init__(self, db, log):
            self.db = db
            self.log = log

        def __del__(self):
            TimedCursor.report(self.log)

        def commit(self):
            return self.db.commit()

        def rollback(self):
            return self.db.rollback()

        def cursor(self):
            return TimedCursor(self.db.cursor())
else:
    def TimedDB(db, log):
        return db