cosmopolitan/third_party/python/Parser/grammar.c

276 lines
8.4 KiB
C
Raw Normal View History

/*-*- mode:c;indent-tabs-mode:nil;c-basic-offset:4;tab-width:8;coding:utf-8 -*-│
vi: set net ft=c ts=4 sts=4 sw=4 fenc=utf-8 :vi
Python 3
https://docs.python.org/3/license.html │
*/
Undiamond Python headers This change gets the Python codebase into a state where it conforms to the conventions of this codebase. It's now possible to include headers from Python, without worrying about ordering. Python has traditionally solved that problem by "diamonding" everything in Python.h, but that's problematic since it means any change to any Python header invalidates all the build artifacts. Lastly it makes tooling not work. Since it is hard to explain to Emacs when I press C-c C-h to add an import line it shouldn't add the header that actually defines the symbol, and instead do follow the nonstandard Python convention. Progress has been made on letting Python load source code from the zip executable structure via the standard C library APIs. System calss now recognizes zip!FILENAME alternative URIs as equivalent to zip:FILENAME since Python uses colon as its delimiter. Some progress has been made on embedding the notice license terms into the Python object code. This is easier said than done since Python has an extremely complicated ownership story. - Some termios APIs have been added - Implement rewinddir() dirstream API - GetCpuCount() API added to Cosmopolitan Libc - More bugs in Cosmopolitan Libc have been fixed - zipobj.com now has flags for mangling the path - Fixed bug a priori with sendfile() on certain BSDs - Polyfill F_DUPFD and F_DUPFD_CLOEXEC across platforms - FIOCLEX / FIONCLEX now polyfilled for fast O_CLOEXEC changes - APE now supports a hybrid solution to no-self-modify for builds - Many BSD-only magnums added, e.g. O_SEARCH, O_SHLOCK, SF_NODISKIO
2021-08-12 07:42:14 +00:00
#include "third_party/python/Include/grammar.h"
#include "third_party/python/Include/objimpl.h"
#include "third_party/python/Include/pgenheaders.h"
#include "third_party/python/Include/pyerrors.h"
#include "third_party/python/Include/token.h"
/* clang-format off */
extern int Py_DebugFlag;
grammar *
newgrammar(int start)
{
grammar *g;
g = (grammar *)PyObject_MALLOC(sizeof(grammar));
if (g == NULL)
Py_FatalError("no mem for new grammar");
g->g_ndfas = 0;
g->g_dfa = NULL;
g->g_start = start;
g->g_ll.ll_nlabels = 0;
g->g_ll.ll_label = NULL;
g->g_accel = 0;
return g;
}
void
freegrammar(grammar *g)
{
int i;
for (i = 0; i < g->g_ndfas; i++) {
free(g->g_dfa[i].d_name);
for (int j = 0; j < g->g_dfa[i].d_nstates; j++)
PyObject_FREE(g->g_dfa[i].d_state[j].s_arc);
PyObject_FREE(g->g_dfa[i].d_state);
}
PyObject_FREE(g->g_dfa);
for (i = 0; i < g->g_ll.ll_nlabels; i++)
free(g->g_ll.ll_label[i].lb_str);
PyObject_FREE(g->g_ll.ll_label);
PyObject_FREE(g);
}
dfa *
adddfa(grammar *g, int type, const char *name)
{
dfa *d;
g->g_dfa = (dfa *)PyObject_REALLOC(g->g_dfa,
sizeof(dfa) * (g->g_ndfas + 1));
if (g->g_dfa == NULL)
Py_FatalError("no mem to resize dfa in adddfa");
d = &g->g_dfa[g->g_ndfas++];
d->d_type = type;
d->d_name = strdup(name);
d->d_nstates = 0;
d->d_state = NULL;
d->d_initial = -1;
d->d_first = NULL;
return d; /* Only use while fresh! */
}
int
addstate(dfa *d)
{
state *s;
d->d_state = (state *)PyObject_REALLOC(d->d_state,
sizeof(state) * (d->d_nstates + 1));
if (d->d_state == NULL)
Py_FatalError("no mem to resize state in addstate");
s = &d->d_state[d->d_nstates++];
s->s_narcs = 0;
s->s_arc = NULL;
s->s_lower = 0;
s->s_upper = 0;
s->s_accel = NULL;
s->s_accept = 0;
return Py_SAFE_DOWNCAST(s - d->d_state, intptr_t, int);
}
void
addarc(dfa *d, int from, int to, int lbl)
{
state *s;
arc *a;
assert(0 <= from && from < d->d_nstates);
assert(0 <= to && to < d->d_nstates);
s = &d->d_state[from];
s->s_arc = (arc *)PyObject_REALLOC(s->s_arc, sizeof(arc) * (s->s_narcs + 1));
if (s->s_arc == NULL)
Py_FatalError("no mem to resize arc list in addarc");
a = &s->s_arc[s->s_narcs++];
a->a_lbl = lbl;
a->a_arrow = to;
}
int
addlabel(labellist *ll, int type, const char *str)
{
int i;
label *lb;
for (i = 0; i < ll->ll_nlabels; i++) {
if (ll->ll_label[i].lb_type == type &&
strcmp(ll->ll_label[i].lb_str, str) == 0)
return i;
}
ll->ll_label = (label *)PyObject_REALLOC(ll->ll_label,
sizeof(label) * (ll->ll_nlabels + 1));
if (ll->ll_label == NULL)
Py_FatalError("no mem to resize labellist in addlabel");
lb = &ll->ll_label[ll->ll_nlabels++];
lb->lb_type = type;
lb->lb_str = strdup(str);
if (Py_DebugFlag)
printf("Label @ %8p, %d: %s\n", ll, ll->ll_nlabels,
PyGrammar_LabelRepr(lb));
return Py_SAFE_DOWNCAST(lb - ll->ll_label, intptr_t, int);
}
/* Same, but rather dies than adds */
int
findlabel(labellist *ll, int type, const char *str)
{
int i;
for (i = 0; i < ll->ll_nlabels; i++) {
if (ll->ll_label[i].lb_type == type /*&&
strcmp(ll->ll_label[i].lb_str, str) == 0*/)
return i;
}
fprintf(stderr, "Label %d/'%s' not found\n", type, str);
Py_FatalError("grammar.c:findlabel()");
/* Py_FatalError() is declared with __attribute__((__noreturn__)).
GCC emits a warning without "return 0;" (compiler bug!), but Clang is
smarter and emits a warning on the return... */
#ifndef __clang__
return 0; /* Make gcc -Wall happy */
#endif
}
/* Forward */
static void translabel(grammar *, label *);
void
translatelabels(grammar *g)
{
int i;
#ifdef Py_DEBUG
printf("Translating labels ...\n");
#endif
/* Don't translate EMPTY */
for (i = EMPTY+1; i < g->g_ll.ll_nlabels; i++)
translabel(g, &g->g_ll.ll_label[i]);
}
static void
translabel(grammar *g, label *lb)
{
int i;
if (Py_DebugFlag)
printf("Translating label %s ...\n", PyGrammar_LabelRepr(lb));
if (lb->lb_type == NAME) {
for (i = 0; i < g->g_ndfas; i++) {
if (strcmp(lb->lb_str, g->g_dfa[i].d_name) == 0) {
if (Py_DebugFlag)
printf(
"Label %s is non-terminal %d.\n",
lb->lb_str,
g->g_dfa[i].d_type);
lb->lb_type = g->g_dfa[i].d_type;
free(lb->lb_str);
lb->lb_str = NULL;
return;
}
}
for (i = 0; i < (int)N_TOKENS; i++) {
if (strcmp(lb->lb_str, _PyParser_TokenNames[i]) == 0) {
if (Py_DebugFlag)
printf("Label %s is terminal %d.\n",
lb->lb_str, i);
lb->lb_type = i;
free(lb->lb_str);
lb->lb_str = NULL;
return;
}
}
printf("Can't translate NAME label '%s'\n", lb->lb_str);
return;
}
if (lb->lb_type == STRING) {
if (isalpha(Py_CHARMASK(lb->lb_str[1])) ||
lb->lb_str[1] == '_') {
char *p;
char *src;
char *dest;
size_t name_len;
if (Py_DebugFlag)
printf("Label %s is a keyword\n", lb->lb_str);
lb->lb_type = NAME;
src = lb->lb_str + 1;
p = strchr(src, '\'');
if (p)
name_len = p - src;
else
name_len = strlen(src);
dest = (char *)malloc(name_len + 1);
if (!dest) {
printf("Can't alloc dest '%s'\n", src);
return;
}
strncpy(dest, src, name_len);
dest[name_len] = '\0';
free(lb->lb_str);
lb->lb_str = dest;
}
else if (lb->lb_str[2] == lb->lb_str[0]) {
int type = (int) PyToken_OneChar(lb->lb_str[1]);
if (type != OP) {
lb->lb_type = type;
free(lb->lb_str);
lb->lb_str = NULL;
}
else
printf("Unknown OP label %s\n",
lb->lb_str);
}
else if (lb->lb_str[2] && lb->lb_str[3] == lb->lb_str[0]) {
int type = (int) PyToken_TwoChars(lb->lb_str[1],
lb->lb_str[2]);
if (type != OP) {
lb->lb_type = type;
free(lb->lb_str);
lb->lb_str = NULL;
}
else
printf("Unknown OP label %s\n",
lb->lb_str);
}
else if (lb->lb_str[2] && lb->lb_str[3] && lb->lb_str[4] == lb->lb_str[0]) {
int type = (int) PyToken_ThreeChars(lb->lb_str[1],
lb->lb_str[2],
lb->lb_str[3]);
if (type != OP) {
lb->lb_type = type;
free(lb->lb_str);
lb->lb_str = NULL;
}
else
printf("Unknown OP label %s\n",
lb->lb_str);
}
else
printf("Can't translate STRING label %s\n",
lb->lb_str);
}
else
printf("Can't translate label '%s'\n",
PyGrammar_LabelRepr(lb));
}