aboutsummaryrefslogtreecommitdiff
path: root/libcody/buffer.cc
blob: 85c066fef715d54c3dd52ecb989c7605bb0f4a8d (plain)
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
// CODYlib		-*- mode:c++ -*-
// Copyright (C) 2020 Nathan Sidwell, nathan@acm.org
// License: Apache v2.0

// Cody
#include "internal.hh"
// C++
#include <algorithm>
// C
#include <cstring>
// OS
#include <unistd.h>
#include <cerrno>

// MessageBuffer code

// Lines consist of words and end with a NEWLINE (0xa) char
// Whitespace characters are TAB (0x9) and SPACE (0x20)
// Words consist of non-whitespace chars separated by whitespace.
// Multiple lines in one transaction are indicated by ending non-final
// lines with a SEMICOLON (0x3b) word, immediately before the NEWLINE
// Continuations with ; preceding it
// Words matching regexp [-+_/%.a-zA-Z0-9]+ need no quoting.
// Quoting with '...'
// Anything outside of [-+_/%.a-zA-Z0-9] needs quoting
// Anything outside of <= <space> or DEL or \' or \\ needs escaping.
// Escapes are \\, \', \n, \t, \_, everything else as \<hex><hex>?
// Spaces separate words, UTF8 encoding for non-ascii chars

namespace Cody {
namespace Detail {

static const char CONTINUE = S2C(u8";");

void MessageBuffer::BeginLine ()
{
  if (!buffer.empty ())
    {
      // Terminate the previous line with a continuation
      buffer.reserve (buffer.size () + 3);
      buffer.push_back (S2C(u8" "));
      buffer.push_back (CONTINUE);
      buffer.push_back (S2C(u8"\n"));
    }
  lastBol = buffer.size ();
}

// QUOTE means 'maybe quote', we search it for quote-needing chars

void MessageBuffer::Append (char const *str, bool quote, size_t len)
{
  if (len == ~size_t (0))
    len = strlen (str);

  if (!len && !quote)
    return;

  // We want to quote characters outside of [-+_A-Za-z0-9/%.], anything
  // that could remotely be shell-active.  UTF8 encoding for non-ascii.
  if (quote && len)
    {
      quote = false;
      // Scan looking for quote-needing characters.  We could just
      // append until we find one, but that's probably confusing
      for (size_t ix = len; ix--;)
	{
	  unsigned char c = (unsigned char)str[ix];
	  if (!((c >= S2C(u8"a") && c <= S2C(u8"z"))
		|| (c >= S2C(u8"A") && c <= S2C(u8"Z"))
		|| (c >= S2C(u8"0") && c <= S2C(u8"9"))
		|| c == S2C(u8"-") || c == S2C(u8"+") || c == S2C(u8"_")
		|| c == S2C(u8"/") || c == S2C(u8"%") || c == S2C(u8".")))
	    {
	      quote = true;
	      break;
	    }
	}
    }

  // Maximal length of appended string
  buffer.reserve (buffer.size () + len * (quote ? 3 : 1) + 2);

  if (quote)
    buffer.push_back (S2C(u8"'"));

  for (auto *end = str + len; str != end;)
    {
      auto *e = end;

      if (quote)
	// Look for next escape-needing char.  More relaxed than
	// the earlier needs-quoting check.
	for (e = str; e != end; ++e)
	  {
	    unsigned char c = (unsigned char)*e;
	    if (c < S2C(u8" ") || c == 0x7f
		|| c == S2C(u8"\\") || c == S2C(u8"'"))
	      break;
	  }
      buffer.insert (buffer.end (), str, e);
      str = e;

      if (str == end)
	break;

      buffer.push_back (S2C(u8"\\"));
      switch (unsigned char c = (unsigned char)*str++)
	{
	case S2C(u8"\t"):
	  c = S2C(u8"t");
	  goto append;

	case S2C(u8"\n"):
	  c = S2C(u8"n");
	  goto append;

	case S2C(u8"'"):
	case S2C(u8"\\"):
	append:
	  buffer.push_back (c);
	  break;

	default:
	  // Full-on escape.  Use 2 lower-case hex chars
	  for (unsigned shift = 8; shift;)
	    {
	      shift -= 4;

	      char nibble = (c >> shift) & 0xf;
	      nibble += S2C(u8"0");
	      if (nibble > S2C(u8"9"))
		nibble += S2C(u8"a") - (S2C(u8"9") + 1);
	      buffer.push_back (nibble);
	    }
	}
    }

  if (quote)
    buffer.push_back (S2C(u8"'"));
}

void MessageBuffer::Append (char c)
{
  buffer.push_back (c);
}

void MessageBuffer::AppendInteger (unsigned u)
{
  // Sigh, even though std::to_string is C++11, we support building on
  // gcc 4.8, which is a C++11 compiler lacking std::to_string.  so
  // have something horrible.
  std::string v (20, 0);
  size_t len = snprintf (const_cast<char *> (v.data ()), v.size (), "%u", u);
  v.erase (len);

  AppendWord (v);
}

int MessageBuffer::Write (int fd) noexcept
{
  size_t limit = buffer.size () - lastBol;
  ssize_t count = write (fd, &buffer.data ()[lastBol], limit);

  int err = 0;
  if (count < 0)
    err = errno;
  else
    {
      lastBol += count;
      if (size_t (count) != limit)
	err = EAGAIN;
    }

  if (err != EAGAIN && err != EINTR)
    {
      // Reset for next message
      buffer.clear ();
      lastBol = 0;
    }

  return err;
}

int MessageBuffer::Read (int fd) noexcept
{
  constexpr size_t blockSize = 200;

  size_t lwm = buffer.size ();
  size_t hwm = buffer.capacity ();
  if (hwm - lwm < blockSize / 2)
    hwm += blockSize;
  buffer.resize (hwm);

  auto iter = buffer.begin () + lwm;
  ssize_t count = read (fd, &*iter, hwm - lwm);
  buffer.resize (lwm + (count >= 0 ? count : 0));

  if (count < 0)
    return errno;

  if (!count)
    // End of file
    return -1;

  bool more = true;
  for (;;)
    {
      auto newline = std::find (iter, buffer.end (), S2C(u8"\n"));
      if (newline == buffer.end ())
	break;
      more = newline != buffer.begin () && newline[-1] == CONTINUE;
      iter = newline + 1;
	
      if (iter == buffer.end ())
	break;

      if (!more)
	{
	  // There is no continuation, but there are chars after the
	  // newline.  Truncate the buffer and return an error
	  buffer.resize (iter - buffer.begin ());
	  return EINVAL;
	}
    }

  return more ? EAGAIN : 0;
}

int MessageBuffer::Lex (std::vector<std::string> &result)
{
  result.clear ();

  if (IsAtEnd ())
    return ENOENT;

  Assert (buffer.back () == S2C(u8"\n"));

  auto iter = buffer.begin () + lastBol;

  for (std::string *word = nullptr;;)
    {
      char c = *iter;

      ++iter;
      if (c == S2C(u8" ") || c == S2C(u8"\t"))
	{
	  word = nullptr;
	  continue;
	}

      if (c == S2C(u8"\n"))
	break;

      if (c == CONTINUE)
	{
	  // Line continuation
	  if (word || *iter != S2C(u8"\n"))
	    goto malformed;
	  ++iter;
	  break;
	}

      if (c <= S2C(u8" ") || c >= 0x7f)
	goto malformed;

      if (!word)
	{
	  result.emplace_back ();
	  word = &result.back ();
	}

      if (c == S2C(u8"'"))
	{
	  // Quoted word
	  for (;;)
	    {
	      c = *iter;

	      if (c == S2C(u8"\n"))
		{
		malformed:;
		  result.clear ();
		  iter = std::find (iter, buffer.end (), S2C(u8"\n"));
		  auto back = iter;
		  if (back[-1] == CONTINUE  && back[-2] == S2C(u8" "))
		    // Smells like a line continuation
		    back -= 2;
		  result.emplace_back (&buffer[lastBol],
				       back - buffer.begin () - lastBol);
		  ++iter;
		  lastBol = iter - buffer.begin ();
		  return EINVAL;
		}

	      if (c < S2C(u8" ") || c >= 0x7f)
		goto malformed;

	      ++iter;
	      if (c == S2C(u8"'"))
		break;

	      if (c == S2C(u8"\\"))
		// escape
		switch (c = *iter)
		  {
		    case S2C(u8"\\"):
		    case S2C(u8"'"):
		      ++iter;
		      break;

		    case S2C(u8"n"):
		      c = S2C(u8"\n");
		      ++iter;
		      break;

		    case S2C(u8"_"):
		      // We used to escape SPACE as \_, so accept that
		      c = S2C(u8" ");
		      ++iter;
		      break;

		    case S2C(u8"t"):
		      c = S2C(u8"\t");
		      ++iter;
		      break;

		    default:
		      {
			unsigned v = 0;
			for (unsigned nibble = 0; nibble != 2; nibble++)
			  {
			    c = *iter;
			    if (c < S2C(u8"0"))
			      {
				if (!nibble)
				  goto malformed;
				break;
			      }
			    else if (c <= S2C(u8"9"))
			      c -= S2C(u8"0");
			    else if (c < S2C(u8"a"))
			      {
				if (!nibble)
				  goto malformed;
				break;
			      }
			    else if (c <= S2C(u8"f"))
			      c -= S2C(u8"a") - 10;
			    else
			      {
				if (!nibble)
				  goto malformed;
				break;
			      }
			    ++iter;
			    v = (v << 4) | c;
			  }
			c = v;
		      }
		  }
	      word->push_back (c);
	    }
	}
      else
	// Unquoted character
	word->push_back (c);
    }
  lastBol = iter - buffer.begin ();
  if (result.empty ())
    return ENOENT;

  return 0;
}

void MessageBuffer::LexedLine (std::string &str)
{
  if (lastBol)
    {
      size_t pos = lastBol - 1;
      for (; pos; pos--)
	if (buffer[pos-1] == S2C(u8"\n"))
	  break;

      size_t end = lastBol - 1;
      if (buffer[end-1] == CONTINUE && buffer[end-2] == S2C(u8" "))
	// Strip line continuation
	end -= 2;
      str.append (&buffer[pos], end - pos);
    }
}
} // Detail
} // Cody