From 26cca8067bf5306e372c0e90036d832c5021fd90 Mon Sep 17 00:00:00 2001 From: Pablo Galindo Salgado Date: Sat, 26 Mar 2022 16:29:02 +0000 Subject: [PATCH] bpo-47117: Don't crash if we fail to decode characters when the tokenizer buffers are uninitialized (GH-32129) Automerge-Triggered-By: GH:pablogsal --- .../2022-03-26-15-45-57.bpo-47117.60W6GQ.rst | 2 ++ Parser/pegen_errors.c | 9 +++++++-- 2 files changed, 9 insertions(+), 2 deletions(-) create mode 100644 Misc/NEWS.d/next/Core and Builtins/2022-03-26-15-45-57.bpo-47117.60W6GQ.rst diff --git a/Misc/NEWS.d/next/Core and Builtins/2022-03-26-15-45-57.bpo-47117.60W6GQ.rst b/Misc/NEWS.d/next/Core and Builtins/2022-03-26-15-45-57.bpo-47117.60W6GQ.rst new file mode 100644 index 00000000000..5098ed86d07 --- /dev/null +++ b/Misc/NEWS.d/next/Core and Builtins/2022-03-26-15-45-57.bpo-47117.60W6GQ.rst @@ -0,0 +1,2 @@ +Fix a crash if we fail to decode characters in interactive mode if the +tokenizer buffers are uninitialized. Patch by Pablo Galindo. diff --git a/Parser/pegen_errors.c b/Parser/pegen_errors.c index 0be9df0ae55..48969967963 100644 --- a/Parser/pegen_errors.c +++ b/Parser/pegen_errors.c @@ -248,7 +248,12 @@ get_error_line_from_tokenizer_buffers(Parser *p, Py_ssize_t lineno) assert((p->tok->fp == NULL && p->tok->str != NULL) || p->tok->fp == stdin); char *cur_line = p->tok->fp_interactive ? p->tok->interactive_src_start : p->tok->str; - assert(cur_line != NULL); + if (cur_line == NULL) { + assert(p->tok->fp_interactive); + // We can reach this point if the tokenizer buffers for interactive source have not been + // initialized because we failed to decode the original source with the given locale. + return PyUnicode_FromStringAndSize("", 0); + } Py_ssize_t relative_lineno = p->starting_lineno ? lineno - p->starting_lineno + 1 : lineno; const char* buf_end = p->tok->fp_interactive ? p->tok->interactive_src_end : p->tok->inp; @@ -311,7 +316,7 @@ _PyPegen_raise_error_known_location(Parser *p, PyObject *errtype, goto error; } - if (p->tok->fp_interactive) { + if (p->tok->fp_interactive && p->tok->interactive_src_start != NULL) { error_line = get_error_line_from_tokenizer_buffers(p, lineno); } else if (p->start_rule == Py_file_input) {