1 | #include "fitz-imp.h" |
2 | |
3 | #include <string.h> |
4 | |
5 | #define MIN_BOMB (100 << 20) |
6 | |
7 | /* |
8 | Read from a stream into a given data block. |
9 | |
10 | stm: The stream to read from. |
11 | |
12 | data: The data block to read into. |
13 | |
14 | len: The length of the data block (in bytes). |
15 | |
16 | Returns the number of bytes read. May throw exceptions. |
17 | */ |
18 | size_t |
19 | fz_read(fz_context *ctx, fz_stream *stm, unsigned char *buf, size_t len) |
20 | { |
21 | size_t count, n; |
22 | |
23 | count = 0; |
24 | do |
25 | { |
26 | n = fz_available(ctx, stm, len); |
27 | if (n > len) |
28 | n = len; |
29 | if (n == 0) |
30 | break; |
31 | |
32 | memcpy(buf, stm->rp, n); |
33 | stm->rp += n; |
34 | buf += n; |
35 | count += n; |
36 | len -= n; |
37 | } |
38 | while (len > 0); |
39 | |
40 | return count; |
41 | } |
42 | |
43 | static unsigned char skip_buf[4096]; |
44 | |
45 | /* |
46 | Read from a stream discarding data. |
47 | |
48 | stm: The stream to read from. |
49 | |
50 | len: The number of bytes to read. |
51 | |
52 | Returns the number of bytes read. May throw exceptions. |
53 | */ |
54 | size_t fz_skip(fz_context *ctx, fz_stream *stm, size_t len) |
55 | { |
56 | size_t count, l, total = 0; |
57 | |
58 | while (len) |
59 | { |
60 | l = len; |
61 | if (l > sizeof(skip_buf)) |
62 | l = sizeof(skip_buf); |
63 | count = fz_read(ctx, stm, skip_buf, l); |
64 | total += count; |
65 | if (count < l) |
66 | break; |
67 | len -= count; |
68 | } |
69 | return total; |
70 | } |
71 | |
72 | /* |
73 | Read all of a stream into a buffer. |
74 | |
75 | stm: The stream to read from |
76 | |
77 | initial: Suggested initial size for the buffer. |
78 | |
79 | Returns a buffer created from reading from the stream. May throw |
80 | exceptions on failure to allocate. |
81 | */ |
82 | fz_buffer * |
83 | fz_read_all(fz_context *ctx, fz_stream *stm, size_t initial) |
84 | { |
85 | return fz_read_best(ctx, stm, initial, NULL); |
86 | } |
87 | |
88 | /* |
89 | Attempt to read a stream into a buffer. If truncated |
90 | is NULL behaves as fz_read_all, sets a truncated flag in case of |
91 | error. |
92 | |
93 | stm: The stream to read from. |
94 | |
95 | initial: Suggested initial size for the buffer. |
96 | |
97 | truncated: Flag to store success/failure indication in. |
98 | |
99 | Returns a buffer created from reading from the stream. |
100 | */ |
101 | fz_buffer * |
102 | fz_read_best(fz_context *ctx, fz_stream *stm, size_t initial, int *truncated) |
103 | { |
104 | fz_buffer *buf = NULL; |
105 | int check_bomb = (initial > 0); |
106 | size_t n; |
107 | |
108 | fz_var(buf); |
109 | |
110 | if (truncated) |
111 | *truncated = 0; |
112 | |
113 | fz_try(ctx) |
114 | { |
115 | if (initial < 1024) |
116 | initial = 1024; |
117 | |
118 | buf = fz_new_buffer(ctx, initial+1); |
119 | |
120 | while (1) |
121 | { |
122 | if (buf->len == buf->cap) |
123 | fz_grow_buffer(ctx, buf); |
124 | |
125 | if (check_bomb && buf->len >= MIN_BOMB && buf->len / 200 > initial) |
126 | fz_throw(ctx, FZ_ERROR_GENERIC, "compression bomb detected" ); |
127 | |
128 | n = fz_read(ctx, stm, buf->data + buf->len, buf->cap - buf->len); |
129 | if (n == 0) |
130 | break; |
131 | |
132 | buf->len += n; |
133 | } |
134 | } |
135 | fz_catch(ctx) |
136 | { |
137 | if (fz_caught(ctx) == FZ_ERROR_TRYLATER) |
138 | { |
139 | fz_drop_buffer(ctx, buf); |
140 | fz_rethrow(ctx); |
141 | } |
142 | if (truncated) |
143 | { |
144 | *truncated = 1; |
145 | } |
146 | else |
147 | { |
148 | fz_drop_buffer(ctx, buf); |
149 | fz_rethrow(ctx); |
150 | } |
151 | } |
152 | |
153 | return buf; |
154 | } |
155 | |
156 | /* |
157 | Read a line from stream into the buffer until either a |
158 | terminating newline or EOF, which it replaces with a null byte ('\0'). |
159 | |
160 | Returns buf on success, and NULL when end of file occurs while no characters |
161 | have been read. |
162 | */ |
163 | char * |
164 | fz_read_line(fz_context *ctx, fz_stream *stm, char *mem, size_t n) |
165 | { |
166 | char *s = mem; |
167 | int c = EOF; |
168 | while (n > 1) |
169 | { |
170 | c = fz_read_byte(ctx, stm); |
171 | if (c == EOF) |
172 | break; |
173 | if (c == '\r') { |
174 | c = fz_peek_byte(ctx, stm); |
175 | if (c == '\n') |
176 | fz_read_byte(ctx, stm); |
177 | break; |
178 | } |
179 | if (c == '\n') |
180 | break; |
181 | *s++ = c; |
182 | n--; |
183 | } |
184 | if (n) |
185 | *s = '\0'; |
186 | return (s == mem && c == EOF) ? NULL : mem; |
187 | } |
188 | |
189 | /* |
190 | return the current reading position within a stream |
191 | */ |
192 | int64_t |
193 | fz_tell(fz_context *ctx, fz_stream *stm) |
194 | { |
195 | return stm->pos - (stm->wp - stm->rp); |
196 | } |
197 | |
198 | /* |
199 | Seek within a stream. |
200 | |
201 | stm: The stream to seek within. |
202 | |
203 | offset: The offset to seek to. |
204 | |
205 | whence: From where the offset is measured (see fseek). |
206 | */ |
207 | void |
208 | fz_seek(fz_context *ctx, fz_stream *stm, int64_t offset, int whence) |
209 | { |
210 | stm->avail = 0; /* Reset bit reading */ |
211 | if (stm->seek) |
212 | { |
213 | if (whence == 1) |
214 | { |
215 | offset += fz_tell(ctx, stm); |
216 | whence = 0; |
217 | } |
218 | stm->seek(ctx, stm, offset, whence); |
219 | stm->eof = 0; |
220 | } |
221 | else if (whence != 2) |
222 | { |
223 | if (whence == 0) |
224 | offset -= fz_tell(ctx, stm); |
225 | if (offset < 0) |
226 | fz_warn(ctx, "cannot seek backwards" ); |
227 | /* dog slow, but rare enough */ |
228 | while (offset-- > 0) |
229 | { |
230 | if (fz_read_byte(ctx, stm) == EOF) |
231 | { |
232 | fz_warn(ctx, "seek failed" ); |
233 | break; |
234 | } |
235 | } |
236 | } |
237 | else |
238 | fz_warn(ctx, "cannot seek" ); |
239 | } |
240 | |
241 | /* |
242 | Read all the contents of a file into a buffer. |
243 | */ |
244 | fz_buffer * |
245 | fz_read_file(fz_context *ctx, const char *filename) |
246 | { |
247 | fz_stream *stm; |
248 | fz_buffer *buf = NULL; |
249 | |
250 | fz_var(buf); |
251 | |
252 | stm = fz_open_file(ctx, filename); |
253 | fz_try(ctx) |
254 | { |
255 | buf = fz_read_all(ctx, stm, 0); |
256 | } |
257 | fz_always(ctx) |
258 | { |
259 | fz_drop_stream(ctx, stm); |
260 | } |
261 | fz_catch(ctx) |
262 | { |
263 | fz_rethrow(ctx); |
264 | } |
265 | |
266 | return buf; |
267 | } |
268 | |
269 | /* |
270 | fz_read_[u]int(16|24|32|64)(_le)? |
271 | |
272 | Read a 16/32/64 bit signed/unsigned integer from stream, |
273 | in big or little-endian byte orders. |
274 | |
275 | Throws an exception if EOF is encountered. |
276 | */ |
277 | uint16_t fz_read_uint16(fz_context *ctx, fz_stream *stm) |
278 | { |
279 | uint32_t a = fz_read_byte(ctx, stm); |
280 | uint32_t b = fz_read_byte(ctx, stm); |
281 | uint32_t x = (a<<8) | (b); |
282 | if (a == EOF || b == EOF) |
283 | fz_throw(ctx, FZ_ERROR_GENERIC, "premature end of file in int16" ); |
284 | return x; |
285 | } |
286 | |
287 | uint32_t fz_read_uint24(fz_context *ctx, fz_stream *stm) |
288 | { |
289 | uint32_t a = fz_read_byte(ctx, stm); |
290 | uint32_t b = fz_read_byte(ctx, stm); |
291 | uint32_t c = fz_read_byte(ctx, stm); |
292 | uint32_t x = (a<<16) | (b<<8) | (c); |
293 | if (a == EOF || b == EOF || c == EOF) |
294 | fz_throw(ctx, FZ_ERROR_GENERIC, "premature end of file in int24" ); |
295 | return x; |
296 | } |
297 | |
298 | uint32_t fz_read_uint32(fz_context *ctx, fz_stream *stm) |
299 | { |
300 | uint32_t a = fz_read_byte(ctx, stm); |
301 | uint32_t b = fz_read_byte(ctx, stm); |
302 | uint32_t c = fz_read_byte(ctx, stm); |
303 | uint32_t d = fz_read_byte(ctx, stm); |
304 | uint32_t x = (a<<24) | (b<<16) | (c<<8) | (d); |
305 | if (a == EOF || b == EOF || c == EOF || d == EOF) |
306 | fz_throw(ctx, FZ_ERROR_GENERIC, "premature end of file in int32" ); |
307 | return x; |
308 | } |
309 | |
310 | uint64_t fz_read_uint64(fz_context *ctx, fz_stream *stm) |
311 | { |
312 | uint64_t a = fz_read_byte(ctx, stm); |
313 | uint64_t b = fz_read_byte(ctx, stm); |
314 | uint64_t c = fz_read_byte(ctx, stm); |
315 | uint64_t d = fz_read_byte(ctx, stm); |
316 | uint64_t e = fz_read_byte(ctx, stm); |
317 | uint64_t f = fz_read_byte(ctx, stm); |
318 | uint64_t g = fz_read_byte(ctx, stm); |
319 | uint64_t h = fz_read_byte(ctx, stm); |
320 | uint64_t x = (a<<56) | (b<<48) | (c<<40) | (d<<32) | (e<<24) | (f<<16) | (g<<8) | (h); |
321 | if (a == EOF || b == EOF || c == EOF || d == EOF || e == EOF || f == EOF || g == EOF || h == EOF) |
322 | fz_throw(ctx, FZ_ERROR_GENERIC, "premature end of file in int64" ); |
323 | return x; |
324 | } |
325 | |
326 | uint16_t fz_read_uint16_le(fz_context *ctx, fz_stream *stm) |
327 | { |
328 | uint32_t a = fz_read_byte(ctx, stm); |
329 | uint32_t b = fz_read_byte(ctx, stm); |
330 | uint32_t x = (a) | (b<<8); |
331 | if (a == EOF || b == EOF) |
332 | fz_throw(ctx, FZ_ERROR_GENERIC, "premature end of file in int16" ); |
333 | return x; |
334 | } |
335 | |
336 | uint32_t fz_read_uint24_le(fz_context *ctx, fz_stream *stm) |
337 | { |
338 | uint32_t a = fz_read_byte(ctx, stm); |
339 | uint32_t b = fz_read_byte(ctx, stm); |
340 | uint32_t c = fz_read_byte(ctx, stm); |
341 | uint32_t x = (a) | (b<<8) | (c<<16); |
342 | if (a == EOF || b == EOF || c == EOF) |
343 | fz_throw(ctx, FZ_ERROR_GENERIC, "premature end of file in int24" ); |
344 | return x; |
345 | } |
346 | |
347 | uint32_t fz_read_uint32_le(fz_context *ctx, fz_stream *stm) |
348 | { |
349 | uint32_t a = fz_read_byte(ctx, stm); |
350 | uint32_t b = fz_read_byte(ctx, stm); |
351 | uint32_t c = fz_read_byte(ctx, stm); |
352 | uint32_t d = fz_read_byte(ctx, stm); |
353 | uint32_t x = (a) | (b<<8) | (c<<16) | (d<<24); |
354 | if (a == EOF || b == EOF || c == EOF || d == EOF) |
355 | fz_throw(ctx, FZ_ERROR_GENERIC, "premature end of file in int32" ); |
356 | return x; |
357 | } |
358 | |
359 | uint64_t fz_read_uint64_le(fz_context *ctx, fz_stream *stm) |
360 | { |
361 | uint64_t a = fz_read_byte(ctx, stm); |
362 | uint64_t b = fz_read_byte(ctx, stm); |
363 | uint64_t c = fz_read_byte(ctx, stm); |
364 | uint64_t d = fz_read_byte(ctx, stm); |
365 | uint64_t e = fz_read_byte(ctx, stm); |
366 | uint64_t f = fz_read_byte(ctx, stm); |
367 | uint64_t g = fz_read_byte(ctx, stm); |
368 | uint64_t h = fz_read_byte(ctx, stm); |
369 | uint64_t x = (a) | (b<<8) | (c<<16) | (d<<24) | (e<<32) | (f<<40) | (g<<48) | (h<<56); |
370 | if (a == EOF || b == EOF || c == EOF || d == EOF || e == EOF || f == EOF || g == EOF || h == EOF) |
371 | fz_throw(ctx, FZ_ERROR_GENERIC, "premature end of file in int64" ); |
372 | return x; |
373 | } |
374 | |
375 | int16_t fz_read_int16(fz_context *ctx, fz_stream *stm) { return (int16_t)fz_read_uint16(ctx, stm); } |
376 | int32_t fz_read_int32(fz_context *ctx, fz_stream *stm) { return (int32_t)fz_read_uint32(ctx, stm); } |
377 | int64_t fz_read_int64(fz_context *ctx, fz_stream *stm) { return (int64_t)fz_read_uint64(ctx, stm); } |
378 | |
379 | int16_t fz_read_int16_le(fz_context *ctx, fz_stream *stm) { return (int16_t)fz_read_uint16_le(ctx, stm); } |
380 | int32_t fz_read_int32_le(fz_context *ctx, fz_stream *stm) { return (int32_t)fz_read_uint32_le(ctx, stm); } |
381 | int64_t fz_read_int64_le(fz_context *ctx, fz_stream *stm) { return (int64_t)fz_read_uint64_le(ctx, stm); } |
382 | |
383 | /* |
384 | Read a null terminated string from the stream into |
385 | a buffer of a given length. The buffer will be null terminated. |
386 | Throws on failure (including the failure to fit the entire string |
387 | including the terminator into the buffer). |
388 | */ |
389 | void fz_read_string(fz_context *ctx, fz_stream *stm, char *buffer, int len) |
390 | { |
391 | int c; |
392 | do |
393 | { |
394 | if (len <= 0) |
395 | fz_throw(ctx, FZ_ERROR_GENERIC, "Buffer overrun reading null terminated string" ); |
396 | |
397 | c = fz_read_byte(ctx, stm); |
398 | if (c == EOF) |
399 | fz_throw(ctx, FZ_ERROR_GENERIC, "EOF reading null terminated string" ); |
400 | *buffer++ = c; |
401 | len--; |
402 | } |
403 | while (c != 0); |
404 | } |
405 | |