bpo-41215: Don't use NULL by default in the PEG parser keyword list (GH-21355)

pablogsal · web-flow · commit 1ac0cbca369f · 2020-07-06T12:31:16.000-07:00
Automerge-Triggered-By: @lysnikolaou
diff --git a/Misc/NEWS.d/next/Core and Builtins/2020-07-06-18-36-33.bpo-41215.vFGFIz.rst b/Misc/NEWS.d/next/Core and Builtins/2020-07-06-18-36-33.bpo-41215.vFGFIz.rst
@@ -0,0 +1,2 @@
+Use non-NULL default values in the PEG parser keyword list to overcome a bug that was preventing
+Python from being properly compiled when using the XLC compiler. Patch by Pablo Galindo.
diff --git a/Parser/parser.c b/Parser/parser.c
@@ -9,8 +9,8 @@ extern int Py_DebugFlag;
 #endif
 static const int n_keyword_lists = 9;
 static KeywordToken *reserved_keywords[] = {
-    NULL,
-    NULL,
+    (KeywordToken[]) {{NULL, -1}},
+    (KeywordToken[]) {{NULL, -1}},
     (KeywordToken[]) {
         {"if", 510},
         {"in", 518},
diff --git a/Parser/pegen.c b/Parser/pegen.c
@@ -525,10 +525,13 @@ _PyPegen_dummy_name(Parser *p, ...)
 static int
 _get_keyword_or_name_type(Parser *p, const char *name, int name_len)
 {
-    if (name_len >= p->n_keyword_lists || p->keywords[name_len] == NULL) {
+    assert(name_len != 0);
+    if (name_len >= p->n_keyword_lists ||
+        p->keywords[name_len] == NULL ||
+        p->keywords[name_len]->type == -1) {
         return NAME;
     }
-    for (KeywordToken *k = p->keywords[name_len]; k->type != -1; k++) {
+    for (KeywordToken *k = p->keywords[name_len]; k != NULL && k->type != -1; k++) {
         if (strncmp(k->str, name, name_len) == 0) {
             return k->type;
         }
diff --git a/Tools/peg_generator/pegen/c_generator.py b/Tools/peg_generator/pegen/c_generator.py
@@ -440,7 +440,7 @@ def _setup_keywords(self) -> None:
             num_groups = max(groups) + 1 if groups else 1
             for keywords_length in range(num_groups):
                 if keywords_length not in groups.keys():
-                    self.print("NULL,")
+                    self.print("(KeywordToken[]) {{NULL, -1}},")
                 else:
                     self.print("(KeywordToken[]) {")
                     with self.indent():

Original file line number	Diff line number	Diff line change
`@@ -0,0 +1,2 @@`
	`1`	`+Use non-NULL default values in the PEG parser keyword list to overcome a bug that was preventing`
	`2`	`+Python from being properly compiled when using the XLC compiler. Patch by Pablo Galindo.`
Original file line number	Diff line number	Diff line change
`@@ -525,10 +525,13 @@ _PyPegen_dummy_name(Parser *p, ...)`
`525`	`525`	`static int`
`526`	`526`	`_get_keyword_or_name_type(Parser p, const char name, int name_len)`
`527`	`527`	`{`
`528`		`- if (name_len >= p->n_keyword_lists \|\| p->keywords[name_len] == NULL) {`
	`528`	`+ assert(name_len != 0);`
	`529`	`+ if (name_len >= p->n_keyword_lists \|\|`
	`530`	`+ p->keywords[name_len] == NULL \|\|`
	`531`	`+ p->keywords[name_len]->type == -1) {`
`529`	`532`	`return NAME;`
`530`	`533`	`}`
`531`		`- for (KeywordToken *k = p->keywords[name_len]; k->type != -1; k++) {`
	`534`	`+ for (KeywordToken *k = p->keywords[name_len]; k != NULL && k->type != -1; k++) {`
`532`	`535`	`if (strncmp(k->str, name, name_len) == 0) {`
`533`	`536`	`return k->type;`
`534`	`537`	`}`