1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
|
## Copyright (C) 2015-2025 Philip Nienhuis
## Copyright (C) 2018-2025 Matthew Parkan
##
## This program is free software; you can redistribute it and/or modify it
## under the terms of the GNU General Public License as published by
## the Free Software Foundation; either version 3 of the License, or
## (at your option) any later version.
##
## This program is distributed in the hope that it will be useful,
## but WITHOUT ANY WARRANTY; without even the implied warranty of
## MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
## GNU General Public License for more details.
##
## You should have received a copy of the GNU General Public License
## along with this program. If not, see <http://www.gnu.org/licenses/>.
## -*- texinfo -*-
## @deftypefn {Function File} [@var{status}] = dbfwrite (@var{fname}, @var{data})
## Write data in a cell array to a dbf (xBase) file, provisionally dBase III+.
##
## @var{fname} must be a valid file name, optionally with '.dbf' suffix.
## @var{data} should be a cell array of which the top row contains column
## names (character strings, each max. 10 characters; longer column names will
## be truncated). Each column must contain only one class of data, except of
## course the top entry (the column header). Integers interspersed in double
## type colums will be written as doubles. Data types that can be written are
## character (text string), numeric (integer and float, the latter with 6
## decimal places), and logical.
##
## Output argument @var{status} is 1 if the file was written successfully, -1 if
## one or more data columns were skipped, 0 otherwise. If 0 the incomplete file
## will be deleted as well.
##
## Provisionally only dBase v. III+ files without memos can be written.
##
## @seealso{dbfread}
## @end deftypefn
## Authors: Philip Nienhuis <prnienhuis@users.sf.net>,whos
## Matthew Parkan <matthew.parkan@gmail.com>
## Created: 2014-12-24
function [status] = dbfwrite (fname, data)
status = 0;
## Input validation
if (! ischar (fname))
error ("dbfwrite: file name expected for argument #1\n");
elseif (! iscell (data))
error ("dbfwrite: cell array expected for argument #2\n");
elseif (! iscellstr (data (1, :)))
error ("dbfwrite: column header titles (text) expected on first row of data\n");
endif
## Column headers length cannot exceed 10 characters
toolong = [];
for ii=1:size (data, 2)
title = data{1, ii};
if (length (title) > 10)
toolong = [ toolong, ii ];
data(1, ii) = title(1:10);
endif
endfor
if (! isempty (toolong))
## Truncate headers if required and check for uniqueness
warning ("dbfwrite: one or more column header(s) > 10 characters - truncated\n");
fmt = [repmat("%d ", 1, numel (toolong))];
printf ("Applies to column(s): %s\n", sprintf (fmt, toolong));
if (numel (unique (data(1, :))) < numel (data(1, :)))
error ("dbfwrite: column headers aren't unique - please fix data\n");
endif
endif
## Assess nr of records
## Data contains header row. Data toprow = 2
nrecs = size (data, 1) - 1;
tr = 2;
## Check file name
[pth, fnm, ext] = fileparts (fname);
if (isempty (ext))
fname = [fname ".dbf"];
elseif (! strcmpi (ext, ".dbf"))
error ("dbfwrite: file name should have a '.dbf' suffix\n");
endif
## Try to open file
fid = fopen (fname, "w", "ieee-le");
if (fid < 0)
error ("dbfwrite: could not open file %s\n", fname);
endif
unwind_protect
## Start writing header
## Provisionally assume dbase III+ w/o memos
fwrite (fid, 3, "uint8");
## Date of last update (YYMMDD), with YY the number of years since 1900
t = now;
upd = datevec(t) - [1900, 0, 0, 0, 0, 0];
fwrite (fid, uint8(upd(1:3)), "uint8");
## Number of records in the table
fwrite (fid, nrecs, "uint32");
## The next two uint16 fields are to be written later, just fill temporarily
pos_lhdr = ftell(fid);
fwrite (fid, 0, "uint32");
## Another place holder, write enough to allow next fseek to succeed
fwrite (fid, uint32 (zeros (1, 7)), "uint32");
## Write record descriptors
nfields = size (data, 2);
fldtyp = "";
fldlngs = {};
reclen = 1; ## "Erased" byte first
fseek (fid, 32, "bof");
RR = zeros (32, nfields, "uint8");
colskipped = 0;
for ii=1:nfields
decpl = 0;
recdesc = sprintf ("%d", uint32 (zeros (1, 8)));
recdesc(1:10) = strjust (sprintf ("%10s", data{1, ii}), "left"); ## Field name
## Be strict on mixing char and numeric; this implies logicals
## interspersed in numeric type column won't be accepted either
if (all (cellfun (@isnumeric, data(tr:end, ii), "uni", 1)))
## We're lax on interspersed integers, they'll be cast to double
if (isinteger ([data{tr:end, ii}]) ||
all ([data{tr:end, ii}] - floor([data{tr:end, ii}]) < eps))
ftype = "N";
decpl = 0;
else
ftype = "F";
## ML compatibility for .dbf/.shp file: 6 decimal places
decpl = 6;
endif
fldlng = 20;
elseif (all (cellfun (@ischar, data(tr:end, ii), "uni", 1)))
ftype = "C";
fldlng = max (cellfun (@(x) length(x), data(tr:end, ii)));
elseif (all (cellfun (@islogical, (data(tr:end, ii)), "uni", 1)))
ftype = "L";
fldlng = 1;
else
warning (["dbfwrite: heterogeneous data types in column %d ('%s'), ", ...
"skipped.\n"], ii, data{1, ii});
RR(:, end) = [];
nfields--;
colskipped = 1;
continue ;
## unwind_protect_cleanup takes care of closing & wiping file
endif
recdesc(12) = ftype; ## Field type
fldtyp = [ fldtyp ftype ];
recdesc(17) = uint8 (fldlng); ## Field length
recdesc(18) = uint8 (decpl); ## Decimal places
recdesc(32) = "\0"; ## Fill to byte# 32
RR(:, ii) = recdesc';
reclen += fldlng;
fldlngs = [ fldlngs; sprintf("%d", fldlng) ];
endfor
fwrite (fid, RR, "char");
## Write header record terminator
fwrite (fid, 13, "uint8");
## Remember position
fpos_data = ftell (fid);
## Write missing data in header
fseek (fid, pos_lhdr, "bof");
fwrite (fid, fpos_data, "uint16");
fwrite (fid, reclen, "uint16");
## Write data2
fseek (fid, fpos_data, "bof");
## Determine data record format. "Erased byte" is first char of format
fmt = "\0";
for j = 1:nfields
switch fldtyp(j)
case "C" ## character
txt = ["%", fldlngs{j}, "s"];
case "N" ## numeric
txt = ["%" fldlngs{j} "d"];
case "L" ## logical
txt = ["%", fldlngs{j}, "c"];
case "F" ## float
txt = ["%" fldlngs{j} "f"];
case "D" ## date; currently inactive
## txt = sprintf (["%" fldlngs{jj} "s"], data{ii, jj});
otherwise
end
fmt = [fmt, txt]; ## append format
end
## Convert boolean attributes to Y/N characters
str_logical = {"N", "Y"};
for jj = find (fldtyp == "L")
data(2:end, jj) = str_logical (double ([data{2:end, jj}] + 1))';
end
## Write data in ~100 MB chunks to avoid overflow. First find an optimal
## chunk size as max. nr. of records in a chunk= (= nr.of rows in data)
chunk_sz = floor (1e8 / reclen);
for ii=1 : chunk_sz : nrecs
## Reshape chunk of data matrix
T = [data(ii+1:min (ii+chunk_sz, nrecs+1), :)'(:)];
blob = sprintf (fmt, T{:});
## Write blob to file
fwrite (fid, blob, "char");
endfor
status = 1;
unwind_protect_cleanup
fclose (fid);
if (! status)
printf ("dbfwrite: removing incomplete file %s.\n", fname);
unlink (fname);
elseif (colskipped)
status = -1;
endif
end_unwind_protect
endfunction
|
