pokedex/db/load.py

   1 """CSV to database or vice versa."""
   2 import csv
   3 import fnmatch
   4 import os.path
   5 import sys
   6
   7 import sqlalchemy.sql.util
   8 import sqlalchemy.types
   9
  10 from pokedex.db import metadata
  11 import pokedex.db.tables as tables
  12 from pokedex.defaults import get_default_csv_dir
  13 from pokedex.db.dependencies import find_dependent_tables
  14
  15
  16 def _get_table_names(metadata, patterns):
  17     """Returns a list of table names from the given metadata.  If `patterns`
  18     exists, only tables matching one of the patterns will be returned.
  19     """
  20     if patterns:
  21         table_names = set()
  22         for pattern in patterns:
  23             if '.' in pattern or '/' in pattern:
  24                 # If it looks like a filename, pull out just the table name
  25                 _, filename = os.path.split(pattern)
  26                 table_name, _ = os.path.splitext(filename)
  27                 pattern = table_name
  28
  29             table_names.update(fnmatch.filter(metadata.tables.keys(), pattern))
  30     else:
  31         table_names = metadata.tables.keys()
  32
  33     return list(table_names)
  34
  35 def _get_verbose_prints(verbose):
  36     """If `verbose` is true, returns three functions: one for printing a
  37     starting message, one for printing an interim status update, and one for
  38     printing a success or failure message when finished.
  39
  40     If `verbose` is false, returns no-op functions.
  41     """
  42
  43     if not verbose:
  44         # Return dummies
  45         def dummy(*args, **kwargs):
  46             pass
  47
  48         return dummy, dummy, dummy
  49
  50     ### Okay, verbose == True; print stuff
  51
  52     def print_start(thing):
  53         # Truncate to 66 characters, leaving 10 characters for a success
  54         # or failure message
  55         truncated_thing = thing[:66]
  56
  57         # Also, space-pad to keep the cursor in a known column
  58         num_spaces = 66 - len(truncated_thing)
  59
  60         print "%s...%s" % (truncated_thing, ' ' * num_spaces),
  61         sys.stdout.flush()
  62
  63     if sys.stdout.isatty():
  64         # stdout is a terminal; stupid backspace tricks are OK.
  65         # Don't use print, because it always adds magical spaces, which
  66         # makes backspace accounting harder
  67
  68         backspaces = [0]
  69         def print_status(msg):
  70             # Overwrite any status text with spaces before printing
  71             sys.stdout.write('\b' * backspaces[0])
  72             sys.stdout.write(' ' * backspaces[0])
  73             sys.stdout.write('\b' * backspaces[0])
  74             sys.stdout.write(msg)
  75             sys.stdout.flush()
  76             backspaces[0] = len(msg)
  77
  78         def print_done(msg='ok'):
  79             # Overwrite any status text with spaces before printing
  80             sys.stdout.write('\b' * backspaces[0])
  81             sys.stdout.write(' ' * backspaces[0])
  82             sys.stdout.write('\b' * backspaces[0])
  83             sys.stdout.write(msg + "\n")
  84             sys.stdout.flush()
  85             backspaces[0] = 0
  86
  87     else:
  88         # stdout is a file (or something); don't bother with status at all
  89         def print_status(msg):
  90             pass
  91
  92         def print_done(msg='ok'):
  93             print msg
  94
  95     return print_start, print_status, print_done
  96
  97
  98 def load(session, tables=[], directory=None, drop_tables=False, verbose=False, safe=True, recursive=False):
  99     """Load data from CSV files into the given database session.
 100
 101     Tables are created automatically.
 102
 103     `session`
 104         SQLAlchemy session to use.
 105
 106     `tables`
 107         List of tables to load.  If omitted, all tables are loaded.
 108
 109     `directory`
 110         Directory the CSV files reside in.  Defaults to the `pokedex` data
 111         directory.
 112
 113     `drop_tables`
 114         If set to True, existing `pokedex`-related tables will be dropped.
 115
 116     `verbose`
 117         If set to True, status messages will be printed to stdout.
 118
 119     `safe`
 120         If set to False, load can be faster, but can corrupt the database if
 121         it crashes or is interrupted.
 122
 123     `recursive`
 124         If set to True, load all dependent tables too.
 125     """
 126
 127     # First take care of verbosity
 128     print_start, print_status, print_done = _get_verbose_prints(verbose)
 129
 130
 131     if directory is None:
 132         directory = get_default_csv_dir()
 133
 134     # XXX why isn't this done in command_load
 135     table_names = _get_table_names(metadata, tables)
 136     table_objs = [metadata.tables[name] for name in table_names]
 137
 138     if recursive:
 139         table_objs.extend(find_dependent_tables(table_objs))
 140
 141     table_objs = sqlalchemy.sql.util.sort_tables(table_objs)
 142
 143     # SQLite speed tweaks
 144     if not safe and session.connection().dialect.name == 'sqlite':
 145         session.connection().execute("PRAGMA synchronous=OFF")
 146         session.connection().execute("PRAGMA journal_mode=OFF")
 147
 148     # Drop all tables if requested
 149     if drop_tables:
 150         print_start('Dropping tables')
 151         for n, table in enumerate(reversed(table_objs)):
 152             table.drop(checkfirst=True)
 153             print_status('%s/%s' % (n, len(table_objs)))
 154         print_done()
 155
 156     print_start('Creating tables')
 157     for n, table in enumerate(table_objs):
 158         table.create()
 159         print_status('%s/%s' % (n, len(table_objs)))
 160     print_done()
 161     connection = session.connection()
 162
 163     # Okay, run through the tables and actually load the data now
 164     for table_obj in table_objs:
 165         table_name = table_obj.name
 166         insert_stmt = table_obj.insert()
 167
 168         print_start(table_name)
 169
 170         try:
 171             csvpath = "%s/%s.csv" % (directory, table_name)
 172             csvfile = open(csvpath, 'rb')
 173         except IOError:
 174             # File doesn't exist; don't load anything!
 175             print_done('missing?')
 176             continue
 177
 178         csvsize = os.stat(csvpath).st_size
 179
 180         reader = csv.reader(csvfile, lineterminator='\n')
 181         column_names = [unicode(column) for column in reader.next()]
 182
 183         if not safe and session.connection().dialect.name == 'postgresql':
 184             """
 185             Postgres' CSV dialect works with our data, if we mark the not-null
 186             columns with FORCE NOT NULL.
 187             COPY is only allowed for DB superusers. If you're not one, use safe
 188             loading (pokedex load -S).
 189             """
 190             session.commit()
 191             not_null_cols = [c for c in column_names if not table_obj.c[c].nullable]
 192             if not_null_cols:
 193                 force_not_null = 'FORCE NOT NULL ' + ','.join('"%s"' % c for c in not_null_cols)
 194             else:
 195                 force_not_null = ''
 196             command = "COPY %(table_name)s (%(columns)s) FROM '%(csvpath)s' CSV HEADER %(force_not_null)s"
 197             session.connection().execute(
 198                 command % dict(
 199                     table_name=table_name,
 200                     csvpath=csvpath,
 201                     columns=','.join('"%s"' % c for c in column_names),
 202                     force_not_null=force_not_null,
 203                 )
 204             )
 205             session.commit()
 206             print_done()
 207             continue
 208
 209         # Self-referential tables may contain rows with foreign keys of other
 210         # rows in the same table that do not yet exist.  Pull these out and add
 211         # them to the session last
 212         # ASSUMPTION: Self-referential tables have a single PK called "id"
 213         deferred_rows = []  # ( row referring to id, [foreign ids we need] )
 214         seen_ids = set()    # primary keys we've seen
 215
 216         # Fetch foreign key columns that point at this table, if any
 217         self_ref_columns = []
 218         for column in table_obj.c:
 219             if any(x.references(table_obj) for x in column.foreign_keys):
 220                 self_ref_columns.append(column)
 221
 222         new_rows = []
 223         def insert_and_commit():
 224             if not new_rows:
 225                 return
 226             session.connection().execute(insert_stmt, new_rows)
 227             session.commit()
 228             new_rows[:] = []
 229
 230             progress = "%d%%" % (100 * csvfile.tell() // csvsize)
 231             print_status(progress)
 232
 233         for csvs in reader:
 234             row_data = {}
 235
 236             for column_name, value in zip(column_names, csvs):
 237                 column = table_obj.c[column_name]
 238                 if column.nullable and value == '':
 239                     # Empty string in a nullable column really means NULL
 240                     value = None
 241                 elif isinstance(column.type, sqlalchemy.types.Boolean):
 242                     # Boolean values are stored as string values 0/1, but both
 243                     # of those evaluate as true; SQLA wants True/False
 244                     if value == '0':
 245                         value = False
 246                     else:
 247                         value = True
 248                 else:
 249                     # Otherwise, unflatten from bytes
 250                     value = value.decode('utf-8')
 251
 252                 # nb: Dictionaries flattened with ** have to have string keys
 253                 row_data[ str(column_name) ] = value
 254
 255             # May need to stash this row and add it later if it refers to a
 256             # later row in this table
 257             if self_ref_columns:
 258                 foreign_ids = set(row_data[x.name] for x in self_ref_columns)
 259                 foreign_ids.discard(None)  # remove NULL ids
 260
 261                 if not foreign_ids:
 262                     # NULL key.  Remember this row and add as usual.
 263                     seen_ids.add(row_data['id'])
 264
 265                 elif foreign_ids.issubset(seen_ids):
 266                     # Non-NULL key we've already seen.  Remember it and commit
 267                     # so we know the old row exists when we add the new one
 268                     insert_and_commit()
 269                     seen_ids.add(row_data['id'])
 270
 271                 else:
 272                     # Non-NULL future id.  Save this and insert it later!
 273                     deferred_rows.append((row_data, foreign_ids))
 274                     continue
 275
 276             # Insert row!
 277             new_rows.append(row_data)
 278
 279             # Remembering some zillion rows in the session consumes a lot of
 280             # RAM.  Let's not do that.  Commit every 1000 rows
 281             if len(new_rows) >= 1000:
 282                 insert_and_commit()
 283
 284         insert_and_commit()
 285
 286         # Attempt to add any spare rows we've collected
 287         for row_data, foreign_ids in deferred_rows:
 288             if not foreign_ids.issubset(seen_ids):
 289                 # Could happen if row A refers to B which refers to C.
 290                 # This is ridiculous and doesn't happen in my data so far
 291                 raise ValueError("Too many levels of self-reference!  "
 292                                  "Row was: " + str(row))
 293
 294             session.connection().execute(
 295                 insert_stmt.values(**row_data)
 296             )
 297             seen_ids.add(row_data['id'])
 298         session.commit()
 299
 300         print_done()
 301
 302     # SQLite check
 303     if session.connection().dialect.name == 'sqlite':
 304         session.connection().execute("PRAGMA integrity_check")
 305
 306
 307
 308 def dump(session, tables=[], directory=None, verbose=False):
 309     """Dumps the contents of a database to a set of CSV files.  Probably not
 310     useful to anyone besides a developer.
 311
 312     `session`
 313         SQLAlchemy session to use.
 314
 315     `tables`
 316         List of tables to dump.  If omitted, all tables are dumped.
 317
 318     `directory`
 319         Directory the CSV files should be put in.  Defaults to the `pokedex`
 320         data directory.
 321
 322     `verbose`
 323         If set to True, status messages will be printed to stdout.
 324     """
 325
 326     # First take care of verbosity
 327     print_start, print_status, print_done = _get_verbose_prints(verbose)
 328
 329
 330     if not directory:
 331         directory = get_default_csv_dir()
 332
 333     table_names = _get_table_names(metadata, tables)
 334     table_names.sort()
 335
 336
 337     for table_name in table_names:
 338         print_start(table_name)
 339         table = metadata.tables[table_name]
 340
 341         writer = csv.writer(open("%s/%s.csv" % (directory, table_name), 'wb'),
 342                             lineterminator='\n')
 343         columns = [col.name for col in table.columns]
 344         writer.writerow(columns)
 345
 346         primary_key = table.primary_key
 347         for row in session.query(table).order_by(*primary_key).all():
 348             csvs = []
 349             for col in columns:
 350                 # Convert Pythony values to something more universal
 351                 val = getattr(row, col)
 352                 if val == None:
 353                     val = ''
 354                 elif val == True:
 355                     val = '1'
 356                 elif val == False:
 357                     val = '0'
 358                 else:
 359                     val = unicode(val).encode('utf-8')
 360
 361                 csvs.append(val)
 362
 363             writer.writerow(csvs)
 364
 365         print_done()