Files
jq/src/util.c
T
Muh Muhten 85caac4473 Work around fgets()'s lack of length reporting
Stop losing \0s from the end of files because we don't know how much
fgets() read into our buffer. If we prefill the buffer with anything
other than zeroes, we can actually tell the difference between the
terminator and unused space in the buffer.

This is possibly more expensive, but at least it should be correct.

Fixes #1504
2019-02-20 20:12:52 -06:00

468 lines
13 KiB
C

#ifdef HAVE_MEMMEM
#define _GNU_SOURCE
#endif
#include <sys/types.h>
#include <sys/stat.h>
#include <assert.h>
#include <errno.h>
#include <fcntl.h>
#include <limits.h>
#include <string.h>
#include <unistd.h>
#include <stdlib.h>
#include <stddef.h>
#ifdef HAVE_ALLOCA_H
# include <alloca.h>
#elif !defined alloca
# ifdef __GNUC__
# define alloca __builtin_alloca
# elif defined _MSC_VER
# include <malloc.h>
# define alloca _alloca
# elif !defined HAVE_ALLOCA
# ifdef __cplusplus
extern "C"
# endif
void *alloca (size_t);
# endif
#endif
#ifndef WIN32
#include <pwd.h>
#endif
#ifdef WIN32
#include <windows.h>
#include <processenv.h>
#include <shellapi.h>
#include <wchar.h>
#include <wtypes.h>
#endif
#include "util.h"
#include "jq.h"
#include "jv_alloc.h"
#ifdef WIN32
FILE *fopen(const char *fname, const char *mode) {
size_t sz = sizeof(wchar_t) * MultiByteToWideChar(CP_UTF8, 0, fname, -1, NULL, 0);
wchar_t *wfname = alloca(sz + 2); // +2 is not needed, but just in case
MultiByteToWideChar(CP_UTF8, 0, fname, -1, wfname, sz);
sz = sizeof(wchar_t) * MultiByteToWideChar(CP_UTF8, 0, mode, -1, NULL, 0);
wchar_t *wmode = alloca(sz + 2); // +2 is not needed, but just in case
MultiByteToWideChar(CP_UTF8, 0, mode, -1, wmode, sz);
return _wfopen(wfname, wmode);
}
#endif
#ifndef HAVE_MKSTEMP
int mkstemp(char *template) {
size_t len = strlen(template);
int tries=5;
int fd;
// mktemp() truncates template when it fails
char *s = alloca(len + 1);
assert(s != NULL);
strcpy(s, template);
do {
// Restore template
strcpy(template, s);
(void) mktemp(template);
fd = open(template, O_CREAT | O_EXCL | O_RDWR, 0600);
} while (fd == -1 && tries-- > 0);
return fd;
}
#endif
jv expand_path(jv path) {
assert(jv_get_kind(path) == JV_KIND_STRING);
const char *pstr = jv_string_value(path);
jv ret = path;
if (jv_string_length_bytes(jv_copy(path)) > 1 && pstr[0] == '~' && pstr[1] == '/') {
jv home = get_home();
if (jv_is_valid(home)) {
ret = jv_string_fmt("%s/%s",jv_string_value(home),pstr+2);
jv_free(home);
} else {
jv emsg = jv_invalid_get_msg(home);
ret = jv_invalid_with_msg(jv_string_fmt("Could not expand %s. (%s)", pstr, jv_string_value(emsg)));
jv_free(emsg);
}
jv_free(path);
}
return ret;
}
jv get_home() {
jv ret;
char *home = getenv("HOME");
if (!home) {
#ifndef WIN32
struct passwd* pwd = getpwuid(getuid());
if (pwd)
ret = jv_string(pwd->pw_dir);
else
ret = jv_invalid_with_msg(jv_string("Could not find home directory."));
#else
home = getenv("USERPROFILE");
if (!home) {
char *hd = getenv("HOMEDRIVE");
if (!hd) hd = "";
home = getenv("HOMEPATH");
if (!home) {
ret = jv_invalid_with_msg(jv_string("Could not find home directory."));
} else {
ret = jv_string_fmt("%s%s",hd,home);
}
} else {
ret = jv_string(home);
}
#endif
} else {
ret = jv_string(home);
}
return ret;
}
jv jq_realpath(jv path) {
int path_max;
char *buf = NULL;
#ifdef _PC_PATH_MAX
path_max = pathconf(jv_string_value(path),_PC_PATH_MAX);
#else
path_max = PATH_MAX;
#endif
if (path_max > 0) {
buf = jv_mem_alloc(path_max);
}
#ifdef WIN32
char *tmp = _fullpath(buf, jv_string_value(path), path_max);
#else
char *tmp = realpath(jv_string_value(path), buf);
#endif
if (tmp == NULL) {
free(buf);
return path;
}
jv_free(path);
path = jv_string(tmp);
free(tmp);
return path;
}
const void *_jq_memmem(const void *haystack, size_t haystacklen,
const void *needle, size_t needlelen) {
#ifdef HAVE_MEMMEM
return (const void*)memmem(haystack, haystacklen, needle, needlelen);
#else
const char *h = haystack;
const char *n = needle;
size_t hi, hi2, ni;
if (haystacklen < needlelen || haystacklen == 0)
return NULL;
for (hi = 0; hi < (haystacklen - needlelen + 1); hi++) {
for (ni = 0, hi2 = hi; ni < needlelen; ni++, hi2++) {
if (h[hi2] != n[ni])
goto not_this;
}
return &h[hi];
not_this:
continue;
}
return NULL;
#endif /* !HAVE_MEMMEM */
}
struct jq_util_input_state {
jq_util_msg_cb err_cb;
void *err_cb_data;
jv_parser *parser;
FILE* current_input;
char **files;
int nfiles;
int curr_file;
int failures;
jv slurped;
char buf[4096];
size_t buf_valid_len;
jv current_filename;
size_t current_line;
};
static void fprinter(void *data, const char *fname) {
fprintf((FILE *)data, "jq: error: Could not open file %s: %s\n", fname, strerror(errno));
}
// If parser == NULL -> RAW
jq_util_input_state *jq_util_input_init(jq_util_msg_cb err_cb, void *err_cb_data) {
if (err_cb == NULL) {
err_cb = fprinter;
err_cb_data = stderr;
}
jq_util_input_state *new_state = jv_mem_alloc(sizeof(*new_state));
memset(new_state, 0, sizeof(*new_state));
new_state->err_cb = err_cb;
new_state->err_cb_data = err_cb_data;
new_state->parser = NULL;
new_state->current_input = NULL;
new_state->files = NULL;
new_state->nfiles = 0;
new_state->curr_file = 0;
new_state->slurped = jv_invalid();
new_state->buf[0] = 0;
new_state->buf_valid_len = 0;
new_state->current_filename = jv_invalid();
new_state->current_line = 0;
return new_state;
}
void jq_util_input_set_parser(jq_util_input_state *state, jv_parser *parser, int slurp) {
assert(!jv_is_valid(state->slurped));
state->parser = parser;
if (parser == NULL && slurp)
state->slurped = jv_string("");
else if (slurp)
state->slurped = jv_array();
else
state->slurped = jv_invalid();
}
void jq_util_input_free(jq_util_input_state **state) {
jq_util_input_state *old_state = *state;
*state = NULL;
if (old_state == NULL)
return;
if (old_state->parser != NULL)
jv_parser_free(old_state->parser);
for (int i = 0; i < old_state->nfiles; i++)
free(old_state->files[i]);
free(old_state->files);
jv_free(old_state->slurped);
jv_free(old_state->current_filename);
jv_mem_free(old_state);
}
void jq_util_input_add_input(jq_util_input_state *state, const char *fname) {
state->files = jv_mem_realloc(state->files, (state->nfiles + 1) * sizeof(state->files[0]));
state->files[state->nfiles++] = jv_mem_strdup(fname);
}
int jq_util_input_errors(jq_util_input_state *state) {
return state->failures;
}
static const char *next_file(jq_util_input_state *state) {
if (state->curr_file < state->nfiles)
return state->files[state->curr_file++];
return NULL;
}
static int jq_util_input_read_more(jq_util_input_state *state) {
if (!state->current_input || feof(state->current_input) || ferror(state->current_input)) {
if (state->current_input && ferror(state->current_input)) {
// System-level input error on the stream. It will be closed (below).
// TODO: report it. Can't use 'state->err_cb()' as it is hard-coded for
// 'open' related problems.
fprintf(stderr,"Input error: %s\n", strerror(errno));
}
if (state->current_input) {
if (state->current_input == stdin) {
clearerr(stdin); // perhaps we can read again; anyways, we don't fclose(stdin)
} else {
fclose(state->current_input);
}
state->current_input = NULL;
jv_free(state->current_filename);
state->current_filename = jv_invalid();
state->current_line = 0 ;
}
const char *f = next_file(state);
if (f != NULL) {
if (!strcmp(f, "-")) {
state->current_input = stdin;
state->current_filename = jv_string("<stdin>");
} else {
state->current_input = fopen(f, "r");
state->current_filename = jv_string(f);
if (!state->current_input) {
state->err_cb(state->err_cb_data, f);
state->failures++;
}
}
state->current_line = 0;
}
}
state->buf[0] = 0;
state->buf_valid_len = 0;
if (state->current_input) {
char *res;
memset(state->buf, 0xff, sizeof(state->buf));
while (!(res = fgets(state->buf, sizeof(state->buf), state->current_input)) &&
ferror(state->current_input) && errno == EINTR)
clearerr(state->current_input);
if (res == NULL) {
state->buf[0] = 0;
if (ferror(state->current_input))
state->failures++;
} else {
const char *p = memchr(state->buf, '\n', sizeof(state->buf));
if (p != NULL)
state->current_line++;
if (p == NULL && state->parser != NULL) {
/*
* There should be no NULs in JSON texts (but JSON text
* sequences are another story).
*/
state->buf_valid_len = strlen(state->buf);
} else if (p == NULL && feof(state->current_input)) {
size_t i;
/*
* XXX We don't know how many bytes we've read!
*
* We can't use getline() because there need not be any newlines
* in the input. The only entirely correct choices are: use
* fgetc() or fread(). Using fread() will complicate buffer
* management here.
*
* For now we check how much fgets() read by scanning backwards for the
* terminating '\0'. This only works because we previously memset our
* buffer with something nonzero.
*/
for (p = state->buf, i = sizeof(state->buf) - 1; i > 0; i--) {
if (state->buf[i] == '\0')
break;
}
state->buf_valid_len = i;
} else if (p == NULL) {
state->buf_valid_len = sizeof(state->buf) - 1;
} else {
state->buf_valid_len = (p - state->buf) + 1;
}
}
}
return state->curr_file == state->nfiles &&
(!state->current_input || feof(state->current_input) || ferror(state->current_input));
}
jv jq_util_input_next_input_cb(jq_state *jq, void *data) {
return jq_util_input_next_input((jq_util_input_state *)data);
}
// Return the current_filename:current_line
jv jq_util_input_get_position(jq_state *jq) {
jq_input_cb cb = NULL;
void *cb_data = NULL;
jq_get_input_cb(jq, &cb, &cb_data);
assert(cb == jq_util_input_next_input_cb);
if (cb != jq_util_input_next_input_cb)
return jv_invalid_with_msg(jv_string("Invalid jq_util_input API usage"));
jq_util_input_state *s = (jq_util_input_state *)cb_data;
// We can't assert that current_filename is a string because if
// the error was a JSON parser error then we may not have set
// current_filename yet.
if (jv_get_kind(s->current_filename) != JV_KIND_STRING)
return jv_string("<unknown>");
jv v = jv_string_fmt("%s:%lu", jv_string_value(s->current_filename), (unsigned long)s->current_line);
return v;
}
jv jq_util_input_get_current_filename(jq_state* jq) {
jq_input_cb cb=NULL;
void *cb_data=NULL;
jq_get_input_cb(jq, &cb, &cb_data);
if (cb != jq_util_input_next_input_cb)
return jv_invalid_with_msg(jv_string("Unknown input filename"));
jq_util_input_state *s = (jq_util_input_state *)cb_data;
jv v = jv_copy(s->current_filename);
return v;
}
jv jq_util_input_get_current_line(jq_state* jq) {
jq_input_cb cb=NULL;
void *cb_data=NULL;
jq_get_input_cb(jq, &cb, &cb_data);
if (cb != jq_util_input_next_input_cb)
return jv_invalid_with_msg(jv_string("Unknown input line number"));
jq_util_input_state *s = (jq_util_input_state *)cb_data;
jv v = jv_number(s->current_line);
return v;
}
// Blocks to read one more input from stdin and/or given files
// When slurping, it returns just one value
jv jq_util_input_next_input(jq_util_input_state *state) {
int is_last = 0;
int has_more = 0;
jv value = jv_invalid(); // need more input
do {
if (state->parser == NULL) {
// Raw input
is_last = jq_util_input_read_more(state);
if (state->buf_valid_len == 0)
continue;
if (jv_is_valid(state->slurped)) {
// Slurped raw input
state->slurped = jv_string_concat(state->slurped, jv_string_sized(state->buf, state->buf_valid_len));
} else {
if (!jv_is_valid(value))
value = jv_string("");
if (state->buf[state->buf_valid_len-1] == '\n') {
// whole line
state->buf[state->buf_valid_len-1] = 0;
return jv_string_concat(value, jv_string_sized(state->buf, state->buf_valid_len-1));
}
value = jv_string_concat(value, jv_string_sized(state->buf, state->buf_valid_len));
state->buf[0] = '\0';
state->buf_valid_len = 0;
}
} else {
if (jv_parser_remaining(state->parser) == 0) {
is_last = jq_util_input_read_more(state);
if (is_last && state->buf_valid_len == 0)
value = jv_invalid();
jv_parser_set_buf(state->parser, state->buf, state->buf_valid_len, !is_last);
}
value = jv_parser_next(state->parser);
if (jv_is_valid(state->slurped)) {
// When slurping an input that doesn't have a trailing newline,
// we might have more than one value on the same line, so let's check
// to see if we have more data to parse.
has_more = jv_parser_remaining(state->parser);
if (jv_is_valid(value)) {
state->slurped = jv_array_append(state->slurped, value);
value = jv_invalid();
} else if (jv_invalid_has_msg(jv_copy(value)))
return value; // Not slurped parsed input
} else if (jv_is_valid(value) || jv_invalid_has_msg(jv_copy(value))) {
return value;
}
}
} while (!is_last || has_more);
if (jv_is_valid(state->slurped)) {
value = state->slurped;
state->slurped = jv_invalid();
}
return value;
}