Thuban/Model/table.py

# Copyright (c) 2001, 2002, 2003 by Intevation GmbH
# Authors:
# Bernhard Herzog <[email protected]>
# Jan-Oliver Wagner <[email protected]>
#
# This program is free software under the GPL (>=v2)
# Read the file COPYING coming with Thuban for details.

"""
Classes for handling tables of data.
"""

__version__ = "$Revision$"

import dbflib

# the field types supported by a Table instance.
FIELDTYPE_INT = "int"
FIELDTYPE_STRING = "string"
FIELDTYPE_DOUBLE = "double"


# map the dbflib constants for the field types to our constants
dbflib_fieldtypes = {dbflib.FTString: FIELDTYPE_STRING,
                     dbflib.FTInteger: FIELDTYPE_INT,
                     dbflib.FTDouble: FIELDTYPE_DOUBLE}

class MemoryTable:

    """Quite simple table implementation that operates on a list of tuples.
    All of the data are kept in the memory."""

    def __init__(self, fields, data):
        """Initialize the MemoryTable

        Parameters:
        fields -- List of (name, field_type) pairs
        data -- List of tuples, one for each row of data
        """
        self.fields = fields
        self.data = data

    def field_count(self):
        return len(self.fields)

    def field_info(self, index):
        name, type = self.fields[index]
        return (type, name)

    def record_count(self):
        return len(self.data)

    def read_record(self, index):
        return dict([(self.fields[i][0], self.data[index][i])
                      for i in range(len(self.fields))])

    def write_record(self, record, values):
        # TODO: Check for correct lenght and perhaps also
        # for correct types in case values is a tuple. How to report problems?
        # TODO: Allow values to be a dictionary and write the single
        # fields that are specified.
        self.data[record] = values


class DBFTable:

    """
    Table interface for the data in a DBF file
    """

    # Implementation strategy regarding writing to a DBF file:
    #
    # Most of the time Thuban only needs to read from a table and it is
    # important that Thuban can work with read-only files. Therefore the
    # DBF file is opened only for reading initially. Only when
    # write_record is called we try to open the DBF file for writing as
    # well. If that succeeds the read/write DBF file will be used for
    # all IO afterwards.
    #
    # It's important to use the same DBF file object for both reading
    # and writing to make sure that reading a records after writing
    # returns the new values. With two separate objects this wouldn't
    # work because a DBF file object buffers some data

    def __init__(self, filename):
        self.filename = filename
        self.dbf = dbflib.DBFFile(filename)

        # If true, self.dbf is open for writing.
        self._writable = 0

    def Destroy(self):
        self.dbf.close()
        self.dbf = None

    def record_count(self):
        """Return the number of records"""
        return self.dbf.record_count()

    def field_count(self):
        """Return the number of fields in a record"""
        return self.dbf.field_count()

    def field_info(self, field):
        """Return a tuple (type, name, width, prec) for the field no. field

        type is the data type of the field, name the name, width the
        field width in characters and prec the decimal precision.
        """
        type, name, width, prec = self.dbf.field_info(field)
        type = dbflib_fieldtypes[type]
        return type, name, width, prec

    def field_info_by_name(self, fieldName):
        count = self.field_count()

        for i in range(count):
            info = self.field_info(i)
            if info[1] == fieldName:
                return info

        return None

    def field_range(self, fieldName):
        """Finds the first occurences of the minimum and maximum values 
        in the table for the given field.

        This assumes that the standard comparison operators (<, >, etc.)
        will work for the given data.

        Returns a tuple ((min, rec), (max, rec)) where:
            min is the minimum value
            max is the maximum value
            rec is the record number where the value was found. One
                should check that the record number of min is not
                the same as the record number of max.

        Returns None if there are no records

        """


        count = self.record_count()

        if count == 0:
            return None

        rec = self.read_record(0)

        min = rec[fieldName]
        min_rec = 0

        max = rec[fieldName]
        max_rec = 0

        for i in range(1, count):
            rec = self.read_record(i)
            data = rec[fieldName]

            if data < min: 
                min = data
                min_rec = rec
            elif data > max: 
                max = data
                max_rec = rec

        return ((min, min_rec), (max, max_rec))

    def GetUniqueValues(self, fieldName):
        """Return a list of all unique entries in the table for the given
        field name.
        """

        dict = {}

        for i in range(0, self.record_count()):
            rec = self.read_record(i)
            data = rec[fieldName]

            if not dict.has_key(data):
                dict[data] = 0

        return dict.keys()

    def read_record(self, record):
        """Return the record no. record as a dict mapping field names to values
        """
        return self.dbf.read_record(record)

    def write_record(self, record, values):
        """Write the values into the record

        The values parameter may either be a dictionary or a sequence.

        If it's a dictionary the keys must be the names of the fields
        and their value must have a suitable type. Only the fields
        actually contained in the dictionary are written. Fields for
        which there's no item in the dict are not modified.

        If it's a sequence, all fields must be present in the right
        order.
        """
        if not self._writable:
            new_dbf = dbflib.DBFFile(self.filename, "r+b")
            self.dbf.close()
            self.dbf = new_dbf
            self._writable = 1
        self.dbf.write_record(record, values)
        self.dbf.commit()


# Temporary backwards compatibility
Table = DBFTable
1	# Copyright (c) 2001, 2002, 2003 by Intevation GmbH
2	# Authors:
3	# Bernhard Herzog <[email protected]>
4	# Jan-Oliver Wagner <[email protected]>
5	#
6	# This program is free software under the GPL (>=v2)
7	# Read the file COPYING coming with Thuban for details.
8
9	"""
10	Classes for handling tables of data.
11	"""
12
13	__version__ = "$Revision$"
14
15	import dbflib
16
17	# the field types supported by a Table instance.
18	FIELDTYPE_INT = "int"
19	FIELDTYPE_STRING = "string"
20	FIELDTYPE_DOUBLE = "double"
21
22
23	# map the dbflib constants for the field types to our constants
24	dbflib_fieldtypes = {dbflib.FTString: FIELDTYPE_STRING,
25	dbflib.FTInteger: FIELDTYPE_INT,
26	dbflib.FTDouble: FIELDTYPE_DOUBLE}
27
28	class MemoryTable:
29
30	"""Quite simple table implementation that operates on a list of tuples.
31	All of the data are kept in the memory."""
32
33	def __init__(self, fields, data):
34	"""Initialize the MemoryTable
35
36	Parameters:
37	fields -- List of (name, field_type) pairs
38	data -- List of tuples, one for each row of data
39	"""
40	self.fields = fields
41	self.data = data
42
43	def field_count(self):
44	return len(self.fields)
45
46	def field_info(self, index):
47	name, type = self.fields[index]
48	return (type, name)
49
50	def record_count(self):
51	return len(self.data)
52
53	def read_record(self, index):
54	return dict([(self.fields[i][0], self.data[index][i])
55	for i in range(len(self.fields))])
56
57	def write_record(self, record, values):
58	# TODO: Check for correct lenght and perhaps also
59	# for correct types in case values is a tuple. How to report problems?
60	# TODO: Allow values to be a dictionary and write the single
61	# fields that are specified.
62	self.data[record] = values
63
64
65	class DBFTable:
66
67	"""
68	Table interface for the data in a DBF file
69	"""
70
71	# Implementation strategy regarding writing to a DBF file:
72	#
73	# Most of the time Thuban only needs to read from a table and it is
74	# important that Thuban can work with read-only files. Therefore the
75	# DBF file is opened only for reading initially. Only when
76	# write_record is called we try to open the DBF file for writing as
77	# well. If that succeeds the read/write DBF file will be used for
78	# all IO afterwards.
79	#
80	# It's important to use the same DBF file object for both reading
81	# and writing to make sure that reading a records after writing
82	# returns the new values. With two separate objects this wouldn't
83	# work because a DBF file object buffers some data
84
85	def __init__(self, filename):
86	self.filename = filename
87	self.dbf = dbflib.DBFFile(filename)
88
89	# If true, self.dbf is open for writing.
90	self._writable = 0
91
92	def Destroy(self):
93	self.dbf.close()
94	self.dbf = None
95
96	def record_count(self):
97	"""Return the number of records"""
98	return self.dbf.record_count()
99
100	def field_count(self):
101	"""Return the number of fields in a record"""
102	return self.dbf.field_count()
103
104	def field_info(self, field):
105	"""Return a tuple (type, name, width, prec) for the field no. field
106
107	type is the data type of the field, name the name, width the
108	field width in characters and prec the decimal precision.
109	"""
110	type, name, width, prec = self.dbf.field_info(field)
111	type = dbflib_fieldtypes[type]
112	return type, name, width, prec
113
114	def field_info_by_name(self, fieldName):
115	count = self.field_count()
116
117	for i in range(count):
118	info = self.field_info(i)
119	if info[1] == fieldName:
120	return info
121
122	return None
123
124	def field_range(self, fieldName):
125	"""Finds the first occurences of the minimum and maximum values
126	in the table for the given field.
127
128	This assumes that the standard comparison operators (<, >, etc.)
129	will work for the given data.
130
131	Returns a tuple ((min, rec), (max, rec)) where:
132	min is the minimum value
133	max is the maximum value
134	rec is the record number where the value was found. One
135	should check that the record number of min is not
136	the same as the record number of max.
137
138	Returns None if there are no records
139
140	"""
141
142
143	count = self.record_count()
144
145	if count == 0:
146	return None
147
148	rec = self.read_record(0)
149
150	min = rec[fieldName]
151	min_rec = 0
152
153	max = rec[fieldName]
154	max_rec = 0
155
156	for i in range(1, count):
157	rec = self.read_record(i)
158	data = rec[fieldName]
159
160	if data < min:
161	min = data
162	min_rec = rec
163	elif data > max:
164	max = data
165	max_rec = rec
166
167	return ((min, min_rec), (max, max_rec))
168
169	def GetUniqueValues(self, fieldName):
170	"""Return a list of all unique entries in the table for the given
171	field name.
172	"""
173
174	dict = {}
175
176	for i in range(0, self.record_count()):
177	rec = self.read_record(i)
178	data = rec[fieldName]
179
180	if not dict.has_key(data):
181	dict[data] = 0
182
183	return dict.keys()
184
185	def read_record(self, record):
186	"""Return the record no. record as a dict mapping field names to values
187	"""
188	return self.dbf.read_record(record)
189
190	def write_record(self, record, values):
191	"""Write the values into the record
192
193	The values parameter may either be a dictionary or a sequence.
194
195	If it's a dictionary the keys must be the names of the fields
196	and their value must have a suitable type. Only the fields
197	actually contained in the dictionary are written. Fields for
198	which there's no item in the dict are not modified.
199
200	If it's a sequence, all fields must be present in the right
201	order.
202	"""
203	if not self._writable:
204	new_dbf = dbflib.DBFFile(self.filename, "r+b")
205	self.dbf.close()
206	self.dbf = new_dbf
207	self._writable = 1
208	self.dbf.write_record(record, values)
209	self.dbf.commit()
210
211
212
213	# Temporary backwards compatibility
214	Table = DBFTable
Name	Value
svn:eol-style	native
svn:keywords	Author Date Id Revision