[3.12] gh-121130: Fix f-string format specifiers with debug expressions (GH-121150) (#122063)

Doc/library/ast.rst

@@ -302,9 +302,7 @@ Literals
                                 Name(id='a', ctx=Load())],
                             keywords=[]),
                         conversion=-1,
-                        format_spec=JoinedStr(
+                        format_spec=Constant(value='.3'))]))
-                            values=[
-                                Constant(value='.3')]))]))
 
 
 .. class:: List(elts, ctx)

Misc/NEWS.d/next/Core and Builtins/2024-06-29-10-46-14.gh-issue-121130.Rj66Xs.rst

@@ -0,0 +1,2 @@
+Fix f-strings with debug expressions in format specifiers. Patch by Pablo
+Galindo

Parser/tokenizer.h

@@ -14,133 +14,136 @@ extern "C" {
 #define MAXLEVEL 200        /* Max parentheses level */
 #define MAXFSTRINGLEVEL 150 /* Max f-string nesting level */
 
-enum decoding_state {
+enum decoding_state { STATE_INIT, STATE_SEEK_CODING, STATE_NORMAL };
-    STATE_INIT,
-    STATE_SEEK_CODING,
-    STATE_NORMAL
-};
 
 enum interactive_underflow_t {
-    /* Normal mode of operation: return a new token when asked in interactive mode */
+  /* Normal mode of operation: return a new token when asked in interactive mode
-    IUNDERFLOW_NORMAL,
+   */
-    /* Forcefully return ENDMARKER when asked for a new token in interactive mode. This
+  IUNDERFLOW_NORMAL,
-     * can be used to prevent the tokenizer to prompt the user for new tokens */
+  /* Forcefully return ENDMARKER when asked for a new token in interactive mode.
-    IUNDERFLOW_STOP,
+   * This can be used to prevent the tokenizer to prompt the user for new tokens
+   */
+  IUNDERFLOW_STOP,
 };
 
 struct token {
-    int level;
+  int level;
-    int lineno, col_offset, end_lineno, end_col_offset;
+  int lineno, col_offset, end_lineno, end_col_offset;
-    const char *start, *end;
+  const char *start, *end;
-    PyObject *metadata;
+  PyObject *metadata;
 };
 
 enum tokenizer_mode_kind_t {
-    TOK_REGULAR_MODE,
+  TOK_REGULAR_MODE,
-    TOK_FSTRING_MODE,
+  TOK_FSTRING_MODE,
 };
 
 #define MAX_EXPR_NESTING 3
 
 typedef struct _tokenizer_mode {
-    enum tokenizer_mode_kind_t kind;
+  enum tokenizer_mode_kind_t kind;
 
-    int curly_bracket_depth;
+  int curly_bracket_depth;
-    int curly_bracket_expr_start_depth;
+  int curly_bracket_expr_start_depth;
 
-    char f_string_quote;
+  char f_string_quote;
-    int f_string_quote_size;
+  int f_string_quote_size;
-    int f_string_raw;
+  int f_string_raw;
-    const char* f_string_start;
+  const char *f_string_start;
-    const char* f_string_multi_line_start;
+  const char *f_string_multi_line_start;
-    int f_string_line_start;
+  int f_string_line_start;
 
-    Py_ssize_t f_string_start_offset;
+  Py_ssize_t f_string_start_offset;
-    Py_ssize_t f_string_multi_line_start_offset;
+  Py_ssize_t f_string_multi_line_start_offset;
 
-    Py_ssize_t last_expr_size;
+  Py_ssize_t last_expr_size;
-    Py_ssize_t last_expr_end;
+  Py_ssize_t last_expr_end;
-    char* last_expr_buffer;
+  char *last_expr_buffer;
-    int f_string_debug;
+  int f_string_debug;
+  int in_format_spec;
 } tokenizer_mode;
 
 /* Tokenizer state */
 struct tok_state {
-    /* Input state; buf <= cur <= inp <= end */
+  /* Input state; buf <= cur <= inp <= end */
-    /* NB an entire line is held in the buffer */
+  /* NB an entire line is held in the buffer */
-    char *buf;          /* Input buffer, or NULL; malloc'ed if fp != NULL or readline != NULL */
+  char *buf; /* Input buffer, or NULL; malloc'ed if fp != NULL or readline !=
-    char *cur;          /* Next character in buffer */
+                NULL */
-    char *inp;          /* End of data in buffer */
+  char *cur; /* Next character in buffer */
-    int fp_interactive; /* If the file descriptor is interactive */
+  char *inp; /* End of data in buffer */
-    char *interactive_src_start; /* The start of the source parsed so far in interactive mode */
+  int fp_interactive;          /* If the file descriptor is interactive */
-    char *interactive_src_end; /* The end of the source parsed so far in interactive mode */
+  char *interactive_src_start; /* The start of the source parsed so far in
-    const char *end;    /* End of input buffer if buf != NULL */
+                                  interactive mode */
-    const char *start;  /* Start of current token if not NULL */
+  char *interactive_src_end;   /* The end of the source parsed so far in
-    int done;           /* E_OK normally, E_EOF at EOF, otherwise error code */
+                                  interactive mode */
-    /* NB If done != E_OK, cur must be == inp!!! */
+  const char *end;             /* End of input buffer if buf != NULL */
-    FILE *fp;           /* Rest of input; NULL if tokenizing a string */
+  const char *start;           /* Start of current token if not NULL */
-    int tabsize;        /* Tab spacing */
+  int done; /* E_OK normally, E_EOF at EOF, otherwise error code */
-    int indent;         /* Current indentation index */
+  /* NB If done != E_OK, cur must be == inp!!! */
-    int indstack[MAXINDENT];            /* Stack of indents */
+  FILE *fp;                /* Rest of input; NULL if tokenizing a string */
-    int atbol;          /* Nonzero if at begin of new line */
+  int tabsize;             /* Tab spacing */
-    int pendin;         /* Pending indents (if > 0) or dedents (if < 0) */
+  int indent;              /* Current indentation index */
-    const char *prompt, *nextprompt;          /* For interactive prompting */
+  int indstack[MAXINDENT]; /* Stack of indents */
-    int lineno;         /* Current line number */
+  int atbol;               /* Nonzero if at begin of new line */
-    int first_lineno;   /* First line of a single line or multi line string
+  int pendin;              /* Pending indents (if > 0) or dedents (if < 0) */
-                           expression (cf. issue 16806) */
+  const char *prompt, *nextprompt; /* For interactive prompting */
-    int starting_col_offset; /* The column offset at the beginning of a token */
+  int lineno;                      /* Current line number */
-    int col_offset;     /* Current col offset */
+  int first_lineno;        /* First line of a single line or multi line string
-    int level;          /* () [] {} Parentheses nesting level */
+                              expression (cf. issue 16806) */
-            /* Used to allow free continuations inside them */
+  int starting_col_offset; /* The column offset at the beginning of a token */
-    char parenstack[MAXLEVEL];
+  int col_offset;          /* Current col offset */
-    int parenlinenostack[MAXLEVEL];
+  int level;               /* () [] {} Parentheses nesting level */
-    int parencolstack[MAXLEVEL];
+                           /* Used to allow free continuations inside them */
-    PyObject *filename;
+  char parenstack[MAXLEVEL];
-    /* Stuff for checking on different tab sizes */
+  int parenlinenostack[MAXLEVEL];
-    int altindstack[MAXINDENT];         /* Stack of alternate indents */
+  int parencolstack[MAXLEVEL];
-    /* Stuff for PEP 0263 */
+  PyObject *filename;
-    enum decoding_state decoding_state;
+  /* Stuff for checking on different tab sizes */
-    int decoding_erred;         /* whether erred in decoding  */
+  int altindstack[MAXINDENT]; /* Stack of alternate indents */
-    char *encoding;         /* Source encoding. */
+  /* Stuff for PEP 0263 */
-    int cont_line;          /* whether we are in a continuation line. */
+  enum decoding_state decoding_state;
-    const char* line_start;     /* pointer to start of current line */
+  int decoding_erred;           /* whether erred in decoding  */
-    const char* multi_line_start; /* pointer to start of first line of
+  char *encoding;               /* Source encoding. */
-                                     a single line or multi line string
+  int cont_line;                /* whether we are in a continuation line. */
-                                     expression (cf. issue 16806) */
+  const char *line_start;       /* pointer to start of current line */
-    PyObject *decoding_readline; /* open(...).readline */
+  const char *multi_line_start; /* pointer to start of first line of
-    PyObject *decoding_buffer;
+                                   a single line or multi line string
-    PyObject *readline;     /* readline() function */
+                                   expression (cf. issue 16806) */
-    const char* enc;        /* Encoding for the current str. */
+  PyObject *decoding_readline;  /* open(...).readline */
-    char* str;          /* Source string being tokenized (if tokenizing from a string)*/
+  PyObject *decoding_buffer;
-    char* input;       /* Tokenizer's newline translated copy of the string. */
+  PyObject *readline; /* readline() function */
+  const char *enc;    /* Encoding for the current str. */
+  char *str;   /* Source string being tokenized (if tokenizing from a string)*/
+  char *input; /* Tokenizer's newline translated copy of the string. */
 
-    int type_comments;      /* Whether to look for type comments */
+  int type_comments; /* Whether to look for type comments */
 
-    /* async/await related fields (still needed depending on feature_version) */
+  /* async/await related fields (still needed depending on feature_version) */
-    int async_hacks;     /* =1 if async/await aren't always keywords */
+  int async_hacks;      /* =1 if async/await aren't always keywords */
-    int async_def;        /* =1 if tokens are inside an 'async def' body. */
+  int async_def;        /* =1 if tokens are inside an 'async def' body. */
-    int async_def_indent; /* Indentation level of the outermost 'async def'. */
+  int async_def_indent; /* Indentation level of the outermost 'async def'. */
-    int async_def_nl;     /* =1 if the outermost 'async def' had at least one
+  int async_def_nl;     /* =1 if the outermost 'async def' had at least one
-                             NEWLINE token after it. */
+                           NEWLINE token after it. */
-    /* How to proceed when asked for a new token in interactive mode */
+  /* How to proceed when asked for a new token in interactive mode */
-    enum interactive_underflow_t interactive_underflow;
+  enum interactive_underflow_t interactive_underflow;
-    int report_warnings;
+  int report_warnings;
-    // TODO: Factor this into its own thing
+  // TODO: Factor this into its own thing
-    tokenizer_mode tok_mode_stack[MAXFSTRINGLEVEL];
+  tokenizer_mode tok_mode_stack[MAXFSTRINGLEVEL];
-    int tok_mode_stack_index;
+  int tok_mode_stack_index;
-    int tok_extra_tokens;
+  int tok_extra_tokens;
-    int comment_newline;
+  int comment_newline;
-    int implicit_newline;
+  int implicit_newline;
 #ifdef Py_DEBUG
-    int debug;
+  int debug;
 #endif
 };
 
 extern struct tok_state *_PyTokenizer_FromString(const char *, int, int);
 extern struct tok_state *_PyTokenizer_FromUTF8(const char *, int, int);
-extern struct tok_state *_PyTokenizer_FromReadline(PyObject*, const char*, int, int);
+extern struct tok_state *_PyTokenizer_FromReadline(PyObject *, const char *,
-extern struct tok_state *_PyTokenizer_FromFile(FILE *, const char*,
+                                                   int, int);
-                                              const char *, const char *);
+extern struct tok_state *_PyTokenizer_FromFile(FILE *, const char *,
+                                               const char *, const char *);
 extern void _PyTokenizer_Free(struct tok_state *);
 extern void _PyToken_Free(struct token *);
 extern void _PyToken_Init(struct token *);