pokedex/db/load.py

   1 """CSV to database or vice versa."""
   2 import csv
   3 import os.path
   4 import pkg_resources
   5 import re
   6 import sys
   7
   8 from sqlalchemy.orm.attributes import instrumentation_registry
   9 import sqlalchemy.sql.util
  10 import sqlalchemy.types
  11
  12 from pokedex.db import metadata
  13 import pokedex.db.tables as tables
  14
  15
  16 def _wildcard_char_to_regex(char):
  17     """Converts a single wildcard character to the regex equivalent."""
  18
  19     if char == '?':
  20         return '.?'
  21     elif char == '*':
  22         return '.*'
  23     else:
  24         return re.escape(char)
  25
  26 def _wildcard_glob_to_regex(glob):
  27     """Converts a single wildcard glob to a regex STRING."""
  28
  29     # If it looks like a filename, make it not one
  30     if '.' in glob or '/' in glob:
  31         _, filename = os.path.split(glob)
  32         table_name, _ = os.path.splitext(filename)
  33         glob = table_name
  34
  35     return u''.join(map(_wildcard_char_to_regex, glob))
  36
  37 def _wildcards_to_regex(strings):
  38     """Converts a list of wildcard globs to a single regex object."""
  39
  40     regex_parts = map(_wildcard_glob_to_regex, strings)
  41
  42     regex = '^(?:' + '|'.join(regex_parts) + ')$'
  43
  44     return re.compile(regex)
  45
  46
  47 def _get_verbose_prints(verbose):
  48     """If `verbose` is true, returns three functions: one for printing a
  49     starting message, one for printing an interim status update, and one for
  50     printing a success or failure message when finished.
  51
  52     If `verbose` is false, returns no-op functions.
  53     """
  54
  55     if not verbose:
  56         # Return dummies
  57         def dummy(*args, **kwargs):
  58             pass
  59
  60         return dummy, dummy, dummy
  61
  62     ### Okay, verbose == True; print stuff
  63
  64     def print_start(thing):
  65         # Truncate to 66 characters, leaving 10 characters for a success
  66         # or failure message
  67         truncated_thing = thing[0:66]
  68
  69         # Also, space-pad to keep the cursor in a known column
  70         num_spaces = 66 - len(truncated_thing)
  71
  72         print "%s...%s" % (truncated_thing, ' ' * num_spaces),
  73         sys.stdout.flush()
  74
  75     if sys.stdout.isatty():
  76         # stdout is a terminal; stupid backspace tricks are OK.
  77         # Don't use print, because it always adds magical spaces, which
  78         # makes backspace accounting harder
  79
  80         backspaces = [0]
  81         def print_status(msg):
  82             # Overwrite any status text with spaces before printing
  83             sys.stdout.write('\b' * backspaces[0])
  84             sys.stdout.write(' ' * backspaces[0])
  85             sys.stdout.write('\b' * backspaces[0])
  86             sys.stdout.write(msg)
  87             sys.stdout.flush()
  88             backspaces[0] = len(msg)
  89
  90         def print_done(msg='ok'):
  91             # Overwrite any status text with spaces before printing
  92             sys.stdout.write('\b' * backspaces[0])
  93             sys.stdout.write(' ' * backspaces[0])
  94             sys.stdout.write('\b' * backspaces[0])
  95             sys.stdout.write(msg + "\n")
  96             sys.stdout.flush()
  97             backspaces[0] = 0
  98
  99     else:
 100         # stdout is a file (or something); don't bother with status at all
 101         def print_status(msg):
 102             pass
 103
 104         def print_done(msg='ok'):
 105             print msg
 106
 107     return print_start, print_status, print_done
 108
 109
 110 def load(session, tables=[], directory=None, drop_tables=False, verbose=False):
 111     """Load data from CSV files into the given database session.
 112
 113     Tables are created automatically.
 114
 115     `session`
 116         SQLAlchemy session to use.
 117
 118     `tables`
 119         List of tables to load.  If omitted, all tables are loaded.
 120
 121     `directory`
 122         Directory the CSV files reside in.  Defaults to the `pokedex` data
 123         directory.
 124
 125     `drop_tables`
 126         If set to True, existing `pokedex`-related tables will be dropped.
 127
 128     `verbose`
 129         If set to True, status messages will be printed to stdout.
 130     """
 131
 132     # First take care of verbosity
 133     print_start, print_status, print_done = _get_verbose_prints(verbose)
 134
 135
 136     if not directory:
 137         directory = pkg_resources.resource_filename('pokedex', 'data/csv')
 138
 139     if tables:
 140         regex = _wildcards_to_regex(tables)
 141         table_names = filter(regex.match, metadata.tables.keys())
 142     else:
 143         table_names = metadata.tables.keys()
 144
 145     table_objs = [metadata.tables[name] for name in table_names]
 146     table_objs = sqlalchemy.sql.util.sort_tables(table_objs)
 147
 148
 149     # Drop all tables if requested
 150     if drop_tables:
 151         print_start('Dropping tables')
 152         for table in reversed(table_objs):
 153             table.drop(checkfirst=True)
 154         print_done()
 155
 156     for table in table_objs:
 157         table.create()
 158     connection = session.connection()
 159
 160     # Okay, run through the tables and actually load the data now
 161     for table_obj in table_objs:
 162         table_name = table_obj.name
 163         insert_stmt = table_obj.insert()
 164
 165         print_start(table_name)
 166
 167         try:
 168             csvpath = "%s/%s.csv" % (directory, table_name)
 169             csvfile = open(csvpath, 'rb')
 170         except IOError:
 171             # File doesn't exist; don't load anything!
 172             print_done('missing?')
 173             continue
 174
 175         csvsize = os.stat(csvpath).st_size
 176
 177         reader = csv.reader(csvfile, lineterminator='\n')
 178         column_names = [unicode(column) for column in reader.next()]
 179
 180         # Self-referential tables may contain rows with foreign keys of other
 181         # rows in the same table that do not yet exist.  Pull these out and add
 182         # them to the session last
 183         # ASSUMPTION: Self-referential tables have a single PK called "id"
 184         deferred_rows = []  # ( row referring to id, [foreign ids we need] )
 185         seen_ids = {}       # primary key we've seen => 1
 186
 187         # Fetch foreign key columns that point at this table, if any
 188         self_ref_columns = []
 189         for column in table_obj.c:
 190             if any(_.references(table_obj) for _ in column.foreign_keys):
 191                 self_ref_columns.append(column)
 192
 193         new_rows = []
 194         def insert_and_commit():
 195             session.connection().execute(insert_stmt, new_rows)
 196             session.commit()
 197             new_rows[:] = []
 198
 199             progress = "{0}%".format(100 * csvfile.tell() // csvsize)
 200             print_status(progress)
 201
 202         for csvs in reader:
 203             row_data = {}
 204
 205             for column_name, value in zip(column_names, csvs):
 206                 column = table_obj.c[column_name]
 207                 if column.nullable and value == '':
 208                     # Empty string in a nullable column really means NULL
 209                     value = None
 210                 elif isinstance(column.type, sqlalchemy.types.Boolean):
 211                     # Boolean values are stored as string values 0/1, but both
 212                     # of those evaluate as true; SQLA wants True/False
 213                     if value == '0':
 214                         value = False
 215                     else:
 216                         value = True
 217                 else:
 218                     # Otherwise, unflatten from bytes
 219                     value = value.decode('utf-8')
 220
 221                 # nb: Dictionaries flattened with ** have to have string keys
 222                 row_data[ str(column_name) ] = value
 223
 224             # May need to stash this row and add it later if it refers to a
 225             # later row in this table
 226             if self_ref_columns:
 227                 foreign_ids = [row_data[_.name] for _ in self_ref_columns]
 228                 foreign_ids = [_ for _ in foreign_ids if _]  # remove NULL ids
 229
 230                 if not foreign_ids:
 231                     # NULL key.  Remember this row and add as usual.
 232                     seen_ids[row_data['id']] = 1
 233
 234                 elif all(_ in seen_ids for _ in foreign_ids):
 235                     # Non-NULL key we've already seen.  Remember it and commit
 236                     # so we know the old row exists when we add the new one
 237                     insert_and_commit()
 238                     seen_ids[row_data['id']] = 1
 239
 240                 else:
 241                     # Non-NULL future id.  Save this and insert it later!
 242                     deferred_rows.append((row_data, foreign_ids))
 243                     continue
 244
 245             # Insert row!
 246             new_rows.append(row_data)
 247
 248             # Remembering some zillion rows in the session consumes a lot of
 249             # RAM.  Let's not do that.  Commit every 1000 rows
 250             if len(new_rows) >= 1000:
 251                 insert_and_commit()
 252
 253         insert_and_commit()
 254
 255         # Attempt to add any spare rows we've collected
 256         for row_data, foreign_ids in deferred_rows:
 257             if not all(_ in seen_ids for _ in foreign_ids):
 258                 # Could happen if row A refers to B which refers to C.
 259                 # This is ridiculous and doesn't happen in my data so far
 260                 raise ValueError("Too many levels of self-reference!  "
 261                                  "Row was: " + str(row))
 262
 263             session.connection().execute(
 264                 insert_stmt.values(**row_data)
 265             )
 266             seen_ids[row_data['id']] = 1
 267         session.commit()
 268
 269         print_done()
 270
 271
 272
 273 def dump(session, tables=[], directory=None, verbose=False):
 274     """Dumps the contents of a database to a set of CSV files.  Probably not
 275     useful to anyone besides a developer.
 276
 277     `session`
 278         SQLAlchemy session to use.
 279
 280     `tables`
 281         List of tables to dump.  If omitted, all tables are dumped.
 282
 283     `directory`
 284         Directory the CSV files should be put in.  Defaults to the `pokedex`
 285         data directory.
 286
 287     `verbose`
 288         If set to True, status messages will be printed to stdout.
 289     """
 290
 291     # First take care of verbosity
 292     print_start, print_status, print_done = _get_verbose_prints(verbose)
 293
 294
 295     if not directory:
 296         directory = pkg_resources.resource_filename('pokedex', 'data/csv')
 297
 298     if tables:
 299         regex = _wildcards_to_regex(tables)
 300         table_names = filter(regex.match, metadata.tables.keys())
 301     else:
 302         table_names = metadata.tables.keys()
 303
 304     table_names.sort()
 305
 306
 307     for table_name in table_names:
 308         print_start(table_name)
 309         table = metadata.tables[table_name]
 310
 311         writer = csv.writer(open("%s/%s.csv" % (directory, table_name), 'wb'),
 312                             lineterminator='\n')
 313         columns = [col.name for col in table.columns]
 314         writer.writerow(columns)
 315
 316         primary_key = table.primary_key
 317         for row in session.query(table).order_by(*primary_key).all():
 318             csvs = []
 319             for col in columns:
 320                 # Convert Pythony values to something more universal
 321                 val = getattr(row, col)
 322                 if val == None:
 323                     val = ''
 324                 elif val == True:
 325                     val = '1'
 326                 elif val == False:
 327                     val = '0'
 328                 else:
 329                     val = unicode(val).encode('utf-8')
 330
 331                 csvs.append(val)
 332
 333             writer.writerow(csvs)
 334
 335         print_done()