1 |
# Copyright (c) 2001, 2002 by Intevation GmbH |
# Copyright (c) 2001, 2002, 2003 by Intevation GmbH |
2 |
# Authors: |
# Authors: |
3 |
# Bernhard Herzog <[email protected]> |
# Bernhard Herzog <[email protected]> |
4 |
|
# Jan-Oliver Wagner <[email protected]> |
5 |
# |
# |
6 |
# This program is free software under the GPL (>=v2) |
# This program is free software under the GPL (>=v2) |
7 |
# Read the file COPYING coming with Thuban for details. |
# Read the file COPYING coming with Thuban for details. |
13 |
__version__ = "$Revision$" |
__version__ = "$Revision$" |
14 |
|
|
15 |
import dbflib |
import dbflib |
|
from Thuban.common import * |
|
16 |
|
|
17 |
# the field types supported by a Table instance. |
# the field types supported by a Table instance. |
18 |
#FIELDTYPE_INT = "int" |
FIELDTYPE_INT = "int" |
19 |
#FIELDTYPE_STRING = "string" |
FIELDTYPE_STRING = "string" |
20 |
#FIELDTYPE_DOUBLE = "double" |
FIELDTYPE_DOUBLE = "double" |
21 |
|
|
22 |
|
|
23 |
# map the dbflib constants for the field types to our constants |
# map the dbflib constants for the field types to our constants |
25 |
dbflib.FTInteger: FIELDTYPE_INT, |
dbflib.FTInteger: FIELDTYPE_INT, |
26 |
dbflib.FTDouble: FIELDTYPE_DOUBLE} |
dbflib.FTDouble: FIELDTYPE_DOUBLE} |
27 |
|
|
|
class Table: |
|
28 |
|
|
29 |
|
class OldTableInterfaceMixin: |
30 |
|
|
31 |
|
"""Mixin to implement the old table interface using the new one""" |
32 |
|
|
33 |
|
def record_count(self): |
34 |
|
return self.NumRows() |
35 |
|
|
36 |
|
def field_count(self): |
37 |
|
return self.NumColumns() |
38 |
|
|
39 |
|
def field_info(self, field): |
40 |
|
"""Return a tuple (type, name, width, prec) for the field no. field |
41 |
|
|
42 |
|
type is the data type of the field, name the name, width the |
43 |
|
field width in characters and prec the decimal precision. width |
44 |
|
and prec will be zero if the information returned by the Column |
45 |
|
method doesn't provide values for them. |
46 |
|
""" |
47 |
|
col = self.Column(field) |
48 |
|
return (col.type, col.name, |
49 |
|
getattr(col, "width", 0), getattr(col, "prec", 0)) |
50 |
|
|
51 |
|
def field_info_by_name(self, col): |
52 |
|
try: |
53 |
|
return self.field_info(col) |
54 |
|
except KeyError: |
55 |
|
# FIXME: It may be that field_info raises other exceptions |
56 |
|
# when the name is not a valid column name. |
57 |
|
return None |
58 |
|
|
59 |
|
def field_range(self, fieldName): |
60 |
|
min, max = self.ValueRange(fieldName) |
61 |
|
return ((min, None), (max, None)) |
62 |
|
|
63 |
|
def GetUniqueValues(self, field): |
64 |
|
return self.UniqueValues(field) |
65 |
|
|
66 |
|
def read_record(self, r): |
67 |
|
return self.ReadRowAsDict(r) |
68 |
|
|
69 |
|
|
70 |
|
|
71 |
|
class DBFColumn: |
72 |
|
|
73 |
|
"""Description of a column in a DBFTable |
74 |
|
|
75 |
|
Instances have the following public attributes: |
76 |
|
|
77 |
|
name -- Name of the column |
78 |
|
type -- Type of the column (one of FIELDTYPE_STRING, FIELDTYPE_INT or\ |
79 |
|
FIELDTYPE_DOUBLE) |
80 |
|
index -- The index of the column |
81 |
|
width -- the width of the data in the column |
82 |
|
prec -- The precision of the data (only valid for type == FIELDTYPE_DOUBLE) |
83 |
""" |
""" |
|
Represent a table of data. |
|
84 |
|
|
85 |
Currently this is basically just a wrapper around dbflib. |
def __init__(self, name, type, width, prec, index): |
86 |
|
self.name = name |
87 |
|
self.type = type |
88 |
|
self.width = width |
89 |
|
self.prec = prec |
90 |
|
self.index = index |
91 |
|
|
92 |
|
|
93 |
|
class DBFTable(OldTableInterfaceMixin): |
94 |
|
|
95 |
|
""" |
96 |
|
Table interface for the data in a DBF file |
97 |
""" |
""" |
98 |
|
|
99 |
# Implementation strategy regarding writing to a DBF file: |
# Implementation strategy regarding writing to a DBF file: |
102 |
# important that Thuban can work with read-only files. Therefore the |
# important that Thuban can work with read-only files. Therefore the |
103 |
# DBF file is opened only for reading initially. Only when |
# DBF file is opened only for reading initially. Only when |
104 |
# write_record is called we try to open the DBF file for writing as |
# write_record is called we try to open the DBF file for writing as |
105 |
# well. If that succeeds the dbf read/write DBF file will be used |
# well. If that succeeds the read/write DBF file will be used for |
106 |
# for all IO afterwards. |
# all IO afterwards. |
107 |
# |
# |
108 |
# It's important to use the same DBF file object for both reading |
# It's important to use the same DBF file object for both reading |
109 |
# and writing to make sure that reading a records after writing |
# and writing to make sure that reading a records after writing |
117 |
# If true, self.dbf is open for writing. |
# If true, self.dbf is open for writing. |
118 |
self._writable = 0 |
self._writable = 0 |
119 |
|
|
120 |
def Destroy(self): |
# Create the column information objects |
121 |
self.dbf.close() |
self.columns = [] |
122 |
self.dbf = None |
self.column_map = {} |
123 |
|
for i in range(self.NumColumns()): |
124 |
|
ftype, name, width, prec = self.dbf.field_info(i) |
125 |
|
ftype = dbflib_fieldtypes[ftype] |
126 |
|
index = len(self.columns) |
127 |
|
col = DBFColumn(name, ftype, width, prec, index) |
128 |
|
self.columns.append(col) |
129 |
|
self.column_map[name] = col |
130 |
|
self.column_map[index] = col |
131 |
|
|
132 |
def record_count(self): |
def NumRows(self): |
133 |
"""Return the number of records""" |
"""Return the number of rows in the table""" |
134 |
return self.dbf.record_count() |
return self.dbf.record_count() |
135 |
|
|
136 |
def field_count(self): |
def NumColumns(self): |
137 |
"""Return the number of fields in a record""" |
"""Return the number of columns in the table""" |
138 |
return self.dbf.field_count() |
return self.dbf.field_count() |
139 |
|
|
140 |
def field_info(self, field): |
def Columns(self): |
141 |
"""Return a tuple (type, name, width, prec) for the field no. field |
"""Return the table's colum definitions |
142 |
|
|
143 |
type is the data type of the field, name the name, width the |
The return value is a sequence of DBFColumn instances, one for |
144 |
field width in characters and prec the decimal precision. |
each column. |
145 |
|
""" |
146 |
|
return self.columns |
147 |
|
|
148 |
|
def Column(self, col): |
149 |
|
"""Return information about the column given by its name or index |
150 |
|
|
151 |
|
The returned object is an instance of DBFColumn |
152 |
""" |
""" |
153 |
type, name, width, prec = self.dbf.field_info(field) |
return self.column_map[col] |
154 |
type = dbflib_fieldtypes[type] |
|
155 |
return type, name, width, prec |
def ReadRowAsDict(self, row): |
156 |
|
"""Return the entire row as a dictionary with column names as keys""" |
157 |
|
return self.dbf.read_record(row) |
158 |
|
|
159 |
def field_info_by_name(self, fieldName): |
def ReadValue(self, row, col): |
160 |
count = self.field_count() |
"""Return the value of the specified row and column |
161 |
|
|
162 |
for i in range(count): |
The col parameter may be the index of the column or its name. |
163 |
info = self.field_info(i) |
""" |
164 |
if info[1] == fieldName: |
return self.dbf.read_record(row)[self.column_map[col].name] |
|
return info |
|
165 |
|
|
166 |
return None |
def ValueRange(self, col): |
167 |
|
"""Return the minimum and maximum values of the values in the column |
168 |
|
|
169 |
def read_record(self, record): |
The return value is a tuple (min, max) unless the table is empty |
170 |
"""Return the record no. record as a dict mapping field names to values |
in which case the return value is None. |
171 |
""" |
""" |
172 |
return self.dbf.read_record(record) |
count = self.NumRows() |
173 |
|
|
174 |
|
if count == 0: |
175 |
|
return None |
176 |
|
|
177 |
|
min = max = self.ReadValue(0, col) |
178 |
|
for i in range(1, count): |
179 |
|
value = self.ReadValue(i, col) |
180 |
|
if value < min: |
181 |
|
min = value |
182 |
|
elif value > max: |
183 |
|
max = value |
184 |
|
|
185 |
|
return (min, max) |
186 |
|
|
187 |
|
def UniqueValues(self, col): |
188 |
|
"""Return a sorted list of all unique values in the column col""" |
189 |
|
dict = {} |
190 |
|
|
191 |
|
for i in range(self.NumRows()): |
192 |
|
value = self.ReadValue(i, col) |
193 |
|
dict[value] = 0 |
194 |
|
|
195 |
|
values = dict.keys() |
196 |
|
values.sort() |
197 |
|
return values |
198 |
|
|
199 |
|
|
200 |
|
# DBF specific interface parts. |
201 |
|
|
202 |
|
def Destroy(self): |
203 |
|
self.dbf.close() |
204 |
|
self.dbf = None |
205 |
|
|
206 |
def write_record(self, record, values): |
def write_record(self, record, values): |
207 |
"""Write the values into the record |
"""Write the values into the record |
224 |
self.dbf.write_record(record, values) |
self.dbf.write_record(record, values) |
225 |
self.dbf.commit() |
self.dbf.commit() |
226 |
|
|
227 |
|
|
228 |
|
|
229 |
|
# Temporary backwards compatibility |
230 |
|
Table = DBFTable |
231 |
|
|
232 |
|
|
233 |
|
|
234 |
|
class MemoryColumn: |
235 |
|
|
236 |
|
def __init__(self, name, type, index): |
237 |
|
self.name = name |
238 |
|
self.type = type |
239 |
|
self.index = index |
240 |
|
|
241 |
|
class MemoryTable(OldTableInterfaceMixin): |
242 |
|
|
243 |
|
"""Very simple table implementation that operates on a list of tuples""" |
244 |
|
|
245 |
|
def __init__(self, fields, data): |
246 |
|
"""Initialize the MemoryTable |
247 |
|
|
248 |
|
Parameters: |
249 |
|
fields -- List of (name, field_type) pairs |
250 |
|
data -- List of tuples, one for each row of data |
251 |
|
""" |
252 |
|
self.data = data |
253 |
|
|
254 |
|
# Create the column information objects |
255 |
|
self.columns = [] |
256 |
|
self.column_map = {} |
257 |
|
for name, ftype in fields: |
258 |
|
index = len(self.columns) |
259 |
|
col = MemoryColumn(name, ftype, index) |
260 |
|
self.columns.append(col) |
261 |
|
self.column_map[name] = col |
262 |
|
self.column_map[index] = col |
263 |
|
|
264 |
|
def NumColumns(self): |
265 |
|
"""Return the number of columns in the table""" |
266 |
|
return len(self.columns) |
267 |
|
|
268 |
|
def Column(self, col): |
269 |
|
"""Return information about the column given by its name or index |
270 |
|
|
271 |
|
The returned object is an instance of MemoryColumn. |
272 |
|
""" |
273 |
|
return self.column_map[col] |
274 |
|
|
275 |
|
def Columns(self): |
276 |
|
"""Return the table's colum definitions |
277 |
|
|
278 |
|
The return value is a sequence of MemoryColumn instances, one |
279 |
|
for each column. |
280 |
|
""" |
281 |
|
return self.columns |
282 |
|
|
283 |
|
def NumRows(self): |
284 |
|
"""Return the number of rows in the table""" |
285 |
|
return len(self.data) |
286 |
|
|
287 |
|
def ReadValue(self, row, col): |
288 |
|
"""Return the value of the specified row and column |
289 |
|
|
290 |
|
The col parameter may be the index of the column or its name. |
291 |
|
""" |
292 |
|
return self.data[row][self.column_map[col].index] |
293 |
|
|
294 |
|
def ReadRowAsDict(self, index): |
295 |
|
"""Return the entire row as a dictionary with column names as keys""" |
296 |
|
return dict([(col.name, self.data[index][col.index]) |
297 |
|
for col in self.columns]) |
298 |
|
|
299 |
|
def ValueRange(self, col): |
300 |
|
"""Return the minimum and maximum values of the values in the column |
301 |
|
|
302 |
|
The return value is a tuple (min, max) unless the table is empty |
303 |
|
in which case the return value is None. |
304 |
|
""" |
305 |
|
|
306 |
|
index = self.column_map[col].index |
307 |
|
values = [row[index] for row in self.data] |
308 |
|
if not values: |
309 |
|
return None |
310 |
|
|
311 |
|
return min(values), max(values) |
312 |
|
|
313 |
|
def UniqueValues(self, col): |
314 |
|
"""Return a sorted list of all unique values in the column col""" |
315 |
|
dict = {} |
316 |
|
|
317 |
|
for i in range(self.NumRows()): |
318 |
|
value = self.ReadValue(i, col) |
319 |
|
dict[value] = 0 |
320 |
|
|
321 |
|
values = dict.keys() |
322 |
|
values.sort() |
323 |
|
return values |
324 |
|
|
325 |
|
|
326 |
|
def write_record(self, record, values): |
327 |
|
# TODO: Check for correct lenght and perhaps also |
328 |
|
# for correct types in case values is a tuple. How to report problems? |
329 |
|
# TODO: Allow values to be a dictionary and write the single |
330 |
|
# fields that are specified. |
331 |
|
self.data[record] = values |