/
edit-input.cpp
474 lines (461 loc) · 14.7 KB
/
edit-input.cpp
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
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
//===-- runtime/edit-input.cpp --------------------------------------------===//
//
// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
// See https://llvm.org/LICENSE.txt for license information.
// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
//
//===----------------------------------------------------------------------===//
#include "edit-input.h"
#include "flang/Common/real.h"
#include "flang/Common/uint128.h"
#include <algorithm>
namespace Fortran::runtime::io {
static bool EditBOZInput(IoStatementState &io, const DataEdit &edit, void *n,
int base, int totalBitSize) {
std::optional<int> remaining;
std::optional<char32_t> next{io.PrepareInput(edit, remaining)};
common::UnsignedInt128 value{0};
for (; next; next = io.NextInField(remaining)) {
char32_t ch{*next};
if (ch == ' ' || ch == '\t') {
continue;
}
int digit{0};
if (ch >= '0' && ch <= '1') {
digit = ch - '0';
} else if (base >= 8 && ch >= '2' && ch <= '7') {
digit = ch - '0';
} else if (base >= 10 && ch >= '8' && ch <= '9') {
digit = ch - '0';
} else if (base == 16 && ch >= 'A' && ch <= 'Z') {
digit = ch + 10 - 'A';
} else if (base == 16 && ch >= 'a' && ch <= 'z') {
digit = ch + 10 - 'a';
} else {
io.GetIoErrorHandler().SignalError(
"Bad character '%lc' in B/O/Z input field", ch);
return false;
}
value *= base;
value += digit;
}
// TODO: check for overflow
std::memcpy(n, &value, totalBitSize >> 3);
return true;
}
// Prepares input from a field, and consumes the sign, if any.
// Returns true if there's a '-' sign.
static bool ScanNumericPrefix(IoStatementState &io, const DataEdit &edit,
std::optional<char32_t> &next, std::optional<int> &remaining) {
next = io.PrepareInput(edit, remaining);
bool negative{false};
if (next) {
negative = *next == '-';
if (negative || *next == '+') {
io.GotChar();
io.SkipSpaces(remaining);
next = io.NextInField(remaining);
}
}
return negative;
}
bool EditIntegerInput(
IoStatementState &io, const DataEdit &edit, void *n, int kind) {
RUNTIME_CHECK(io.GetIoErrorHandler(), kind >= 1 && !(kind & (kind - 1)));
switch (edit.descriptor) {
case DataEdit::ListDirected:
case 'G':
case 'I':
break;
case 'B':
return EditBOZInput(io, edit, n, 2, kind << 3);
case 'O':
return EditBOZInput(io, edit, n, 8, kind << 3);
case 'Z':
return EditBOZInput(io, edit, n, 16, kind << 3);
default:
io.GetIoErrorHandler().SignalError(IostatErrorInFormat,
"Data edit descriptor '%c' may not be used with an INTEGER data item",
edit.descriptor);
return false;
}
std::optional<int> remaining;
std::optional<char32_t> next;
bool negate{ScanNumericPrefix(io, edit, next, remaining)};
common::UnsignedInt128 value;
for (; next; next = io.NextInField(remaining)) {
char32_t ch{*next};
if (ch == ' ' || ch == '\t') {
if (edit.modes.editingFlags & blankZero) {
ch = '0'; // BZ mode - treat blank as if it were zero
} else {
continue;
}
}
int digit{0};
if (ch >= '0' && ch <= '9') {
digit = ch - '0';
} else {
io.GetIoErrorHandler().SignalError(
"Bad character '%lc' in INTEGER input field", ch);
return false;
}
value *= 10;
value += digit;
}
if (negate) {
value = -value;
}
std::memcpy(n, &value, kind);
return true;
}
// Parses a REAL input number from the input source as a normalized
// fraction into a supplied buffer -- there's an optional '-', a
// decimal point, and at least one digit. The adjusted exponent value
// is returned in a reference argument. The returned value is the number
// of characters that (should) have been written to the buffer -- this can
// be larger than the buffer size and can indicate overflow. Replaces
// blanks with zeroes if appropriate.
static int ScanRealInput(char *buffer, int bufferSize, IoStatementState &io,
const DataEdit &edit, int &exponent) {
std::optional<int> remaining;
std::optional<char32_t> next;
int got{0};
std::optional<int> decimalPoint;
auto Put{[&](char ch) -> void {
if (got < bufferSize) {
buffer[got] = ch;
}
++got;
}};
if (ScanNumericPrefix(io, edit, next, remaining)) {
Put('-');
}
if (!next) { // empty field means zero
Put('0');
return got;
}
char32_t decimal = edit.modes.editingFlags & decimalComma ? ',' : '.';
char32_t first{*next >= 'a' && *next <= 'z' ? *next + 'A' - 'a' : *next};
if (first == 'N' || first == 'I') {
// NaN or infinity - convert to upper case
// Subtle: a blank field of digits could be followed by 'E' or 'D',
for (; next &&
((*next >= 'a' && *next <= 'z') || (*next >= 'A' && *next <= 'Z'));
next = io.NextInField(remaining)) {
if (*next >= 'a' && *next <= 'z') {
Put(*next - 'a' + 'A');
} else {
Put(*next);
}
}
if (next && *next == '(') { // NaN(...)
while (next && *next != ')') {
next = io.NextInField(remaining);
}
}
exponent = 0;
} else if (first == decimal || (first >= '0' && first <= '9') ||
first == 'E' || first == 'D' || first == 'Q') {
Put('.'); // input field is normalized to a fraction
auto start{got};
bool bzMode{(edit.modes.editingFlags & blankZero) != 0};
for (; next; next = io.NextInField(remaining)) {
char32_t ch{*next};
if (ch == ' ' || ch == '\t') {
if (bzMode) {
ch = '0'; // BZ mode - treat blank as if it were zero
} else {
continue;
}
}
if (ch == '0' && got == start && !decimalPoint) {
// omit leading zeroes before the decimal
} else if (ch >= '0' && ch <= '9') {
Put(ch);
} else if (ch == decimal && !decimalPoint) {
// the decimal point is *not* copied to the buffer
decimalPoint = got - start; // # of digits before the decimal point
} else {
break;
}
}
if (got == start) {
Put('0'); // emit at least one digit
}
if (next &&
(*next == 'e' || *next == 'E' || *next == 'd' || *next == 'D' ||
*next == 'q' || *next == 'Q')) {
// Optional exponent letter. Blanks are allowed between the
// optional exponent letter and the exponent value.
io.SkipSpaces(remaining);
next = io.NextInField(remaining);
}
// The default exponent is -kP, but the scale factor doesn't affect
// an explicit exponent.
exponent = -edit.modes.scale;
if (next &&
(*next == '-' || *next == '+' || (*next >= '0' && *next <= '9') ||
(bzMode && (*next == ' ' || *next == '\t')))) {
bool negExpo{*next == '-'};
if (negExpo || *next == '+') {
next = io.NextInField(remaining);
}
for (exponent = 0; next; next = io.NextInField(remaining)) {
if (*next >= '0' && *next <= '9') {
exponent = 10 * exponent + *next - '0';
} else if (bzMode && (*next == ' ' || *next == '\t')) {
exponent = 10 * exponent;
} else {
break;
}
}
if (negExpo) {
exponent = -exponent;
}
}
if (decimalPoint) {
exponent += *decimalPoint;
} else {
// When no decimal point (or comma) appears in the value, the 'd'
// part of the edit descriptor must be interpreted as the number of
// digits in the value to be interpreted as being to the *right* of
// the assumed decimal point (13.7.2.3.2)
exponent += got - start - edit.digits.value_or(0);
}
} else {
// TODO: hex FP input
exponent = 0;
return 0;
}
// Consume the trailing ')' of a list-directed or NAMELIST complex
// input value.
if (edit.descriptor == DataEdit::ListDirectedImaginaryPart) {
if (next && (*next == ' ' || *next == '\t')) {
next = io.NextInField(remaining);
}
if (!next) { // NextInField fails on separators like ')'
next = io.GetCurrentChar();
if (next && *next == ')') {
io.HandleRelativePosition(1);
}
}
} else if (remaining) {
while (next && (*next == ' ' || *next == '\t')) {
next = io.NextInField(remaining);
}
if (next) {
return 0; // error: unused nonblank character in fixed-width field
}
}
return got;
}
template <int KIND>
bool EditCommonRealInput(IoStatementState &io, const DataEdit &edit, void *n) {
constexpr int binaryPrecision{common::PrecisionOfRealKind(KIND)};
static constexpr int maxDigits{
common::MaxDecimalConversionDigits(binaryPrecision)};
static constexpr int bufferSize{maxDigits + 18};
char buffer[bufferSize];
int exponent{0};
int got{ScanRealInput(buffer, maxDigits + 2, io, edit, exponent)};
if (got >= maxDigits + 2) {
io.GetIoErrorHandler().Crash("EditCommonRealInput: buffer was too small");
return false;
}
if (got == 0) {
io.GetIoErrorHandler().SignalError("Bad REAL input value");
return false;
}
bool hadExtra{got > maxDigits};
if (exponent != 0) {
got += std::snprintf(&buffer[got], bufferSize - got, "e%d", exponent);
}
buffer[got] = '\0';
const char *p{buffer};
decimal::ConversionToBinaryResult<binaryPrecision> converted{
decimal::ConvertToBinary<binaryPrecision>(p, edit.modes.round)};
if (hadExtra) {
converted.flags = static_cast<enum decimal::ConversionResultFlags>(
converted.flags | decimal::Inexact);
}
// TODO: raise converted.flags as exceptions?
*reinterpret_cast<decimal::BinaryFloatingPointNumber<binaryPrecision> *>(n) =
converted.binary;
return true;
}
template <int KIND>
bool EditRealInput(IoStatementState &io, const DataEdit &edit, void *n) {
constexpr int binaryPrecision{common::PrecisionOfRealKind(KIND)};
switch (edit.descriptor) {
case DataEdit::ListDirected:
case DataEdit::ListDirectedRealPart:
case DataEdit::ListDirectedImaginaryPart:
case 'F':
case 'E': // incl. EN, ES, & EX
case 'D':
case 'G':
return EditCommonRealInput<KIND>(io, edit, n);
case 'B':
return EditBOZInput(
io, edit, n, 2, common::BitsForBinaryPrecision(binaryPrecision));
case 'O':
return EditBOZInput(
io, edit, n, 8, common::BitsForBinaryPrecision(binaryPrecision));
case 'Z':
return EditBOZInput(
io, edit, n, 16, common::BitsForBinaryPrecision(binaryPrecision));
default:
io.GetIoErrorHandler().SignalError(IostatErrorInFormat,
"Data edit descriptor '%c' may not be used for REAL input",
edit.descriptor);
return false;
}
}
// 13.7.3 in Fortran 2018
bool EditLogicalInput(IoStatementState &io, const DataEdit &edit, bool &x) {
switch (edit.descriptor) {
case DataEdit::ListDirected:
case 'L':
case 'G':
break;
default:
io.GetIoErrorHandler().SignalError(IostatErrorInFormat,
"Data edit descriptor '%c' may not be used for LOGICAL input",
edit.descriptor);
return false;
}
std::optional<int> remaining;
std::optional<char32_t> next{io.PrepareInput(edit, remaining)};
if (next && *next == '.') { // skip optional period
next = io.NextInField(remaining);
}
if (!next) {
io.GetIoErrorHandler().SignalError("Empty LOGICAL input field");
return false;
}
switch (*next) {
case 'T':
case 't':
x = true;
break;
case 'F':
case 'f':
x = false;
break;
default:
io.GetIoErrorHandler().SignalError(
"Bad character '%lc' in LOGICAL input field", *next);
return false;
}
if (remaining) { // ignore the rest of the field
io.HandleRelativePosition(*remaining);
} else if (edit.descriptor == DataEdit::ListDirected) {
while (io.NextInField(remaining)) { // discard rest of field
}
}
return true;
}
// See 13.10.3.1 paragraphs 7-9 in Fortran 2018
static bool EditDelimitedCharacterInput(
IoStatementState &io, char *x, std::size_t length, char32_t delimiter) {
bool result{true};
while (true) {
auto ch{io.GetCurrentChar()};
if (!ch) {
if (io.AdvanceRecord()) {
continue;
} else {
result = false; // EOF in character value
break;
}
}
io.HandleRelativePosition(1);
if (*ch == delimiter) {
auto next{io.GetCurrentChar()};
if (next && *next == delimiter) {
// Repeated delimiter: use as character value
io.HandleRelativePosition(1);
} else {
break; // closing delimiter
}
}
if (length > 0) {
*x++ = *ch;
--length;
}
}
std::fill_n(x, length, ' ');
return result;
}
static bool EditListDirectedDefaultCharacterInput(
IoStatementState &io, char *x, std::size_t length) {
auto ch{io.GetCurrentChar()};
if (ch && (*ch == '\'' || *ch == '"')) {
io.HandleRelativePosition(1);
return EditDelimitedCharacterInput(io, x, length, *ch);
}
// Undelimited list-directed character input: stop at a value separator
// or the end of the current record.
std::optional<int> remaining{length};
for (std::optional<char32_t> next{io.NextInField(remaining)}; next;
next = io.NextInField(remaining)) {
switch (*next) {
case ' ':
case '\t':
case ',':
case ';':
case '/':
remaining = 0; // value separator: stop
break;
default:
*x++ = *next;
--length;
}
}
std::fill_n(x, length, ' ');
return true;
}
bool EditDefaultCharacterInput(
IoStatementState &io, const DataEdit &edit, char *x, std::size_t length) {
switch (edit.descriptor) {
case DataEdit::ListDirected:
return EditListDirectedDefaultCharacterInput(io, x, length);
case 'A':
case 'G':
break;
default:
io.GetIoErrorHandler().SignalError(IostatErrorInFormat,
"Data edit descriptor '%c' may not be used with a CHARACTER data item",
edit.descriptor);
return false;
}
std::optional<int> remaining{length};
if (edit.width && *edit.width > 0) {
remaining = *edit.width;
}
// When the field is wider than the variable, we drop the leading
// characters. When the variable is wider than the field, there's
// trailing padding.
std::int64_t skip{*remaining - static_cast<std::int64_t>(length)};
for (std::optional<char32_t> next{io.NextInField(remaining)}; next;
next = io.NextInField(remaining)) {
if (skip > 0) {
--skip;
io.GotChar(-1);
} else {
*x++ = *next;
--length;
}
}
std::fill_n(x, length, ' ');
return true;
}
template bool EditRealInput<2>(IoStatementState &, const DataEdit &, void *);
template bool EditRealInput<3>(IoStatementState &, const DataEdit &, void *);
template bool EditRealInput<4>(IoStatementState &, const DataEdit &, void *);
template bool EditRealInput<8>(IoStatementState &, const DataEdit &, void *);
template bool EditRealInput<10>(IoStatementState &, const DataEdit &, void *);
// TODO: double/double
template bool EditRealInput<16>(IoStatementState &, const DataEdit &, void *);
} // namespace Fortran::runtime::io