Rietveld Code Review Tool
Help | Bug tracker | Discussion group | Source code | Sign in
(9)

Side by Side Diff: Include/tokenizer.h

Issue 3353: make built-in tokenizer available via Python C API
Patch Set: Created 4 years, 10 months ago
Left:
Right:
Use n/p to move between diff chunks; N/P to move between comments. Please Sign in to add in-line comments.
Jump to:
View unified diff | Download patch
OLDNEW
(Empty)
1 #ifndef Py_TOKENIZER_H
2 #define Py_TOKENIZER_H
3 #ifdef __cplusplus
4 extern "C" {
5 #endif
6
7 #include "object.h"
8
9 /* Tokenizer interface */
10
11 #include "token.h" /* For token types */
Nick Coghlan 2015/04/14 18:25:39 This suggests the token #defines will also need to
12
13 typedef struct PyTokenizer_State PyTokenizer_State;
Nick Coghlan 2015/04/14 18:25:39 As a now public API, this should be included from
14
15 #ifndef Py_LIMITED_API
16 #define MAXINDENT 100 /* Max indentation level */
Nick Coghlan 2015/04/14 18:25:39 This needs a PYTOKENIZER_* prefix.
17
18 enum decoding_state {
19 STATE_INIT,
20 STATE_RAW,
21 STATE_NORMAL /* have a codec associated with input */
Nick Coghlan 2015/04/14 18:25:39 PYTOKENIZER_* prefix
22 };
23
24 struct PyTokenizer_State {
25 /* Input state; buf <= cur <= inp <= end */
26 /* NB an entire line is held in the buffer */
27 char *buf; /* Input buffer, or NULL; malloc'ed if fp != NULL */
28 char *cur; /* Next character in buffer */
29 char *inp; /* End of data in buffer */
30 char *end; /* End of input buffer if buf != NULL */
31 char *start; /* Start of current token if not NULL */
32 int done; /* E_OK normally, E_EOF at EOF, otherwise error code */
33 /* NB If done != E_OK, cur must be == inp!!! */
34 FILE *fp; /* Rest of input; NULL if tokenizing a string */
35 int tabsize; /* Tab spacing */
36 int indent; /* Current indentation index */
37 int indstack[MAXINDENT]; /* Stack of indents */
38 int atbol; /* Nonzero if at begin of new line */
39 int pendin; /* Pending indents (if > 0) or dedents (if < 0) */
40 const char *prompt, *nextprompt; /* For interactive prompting */
41 int lineno; /* Current line number */
42 int level; /* () [] {} Parentheses nesting level */
43 /* Used to allow free continuations inside them */
44 /* Stuff for checking on different tab sizes */
45 #ifndef PGEN
46 /* pgen doesn't have access to Python codecs, it cannot decode the input
47 filename. The bytes filename might be kept, but it is only used by
48 indenterror() and it is not really needed: pgen only compiles one file
49 (Grammar/Grammar). */
50 PyObject *filename;
51 #endif
52 int altwarning; /* Issue warning if alternate tabs don't match */
53 int alterror; /* Issue error if alternate tabs don't match */
54 int alttabsize; /* Alternate tab spacing */
55 int altindstack[MAXINDENT]; /* Stack of alternate indents */
56 /* Stuff for PEP 0263 */
57 enum decoding_state decoding_state;
58 int decoding_erred; /* whether erred in decoding */
59 int read_coding_spec; /* whether 'coding:...' has been read */
60 char *encoding; /* Source encoding. */
61 int cont_line; /* whether we are in a continuation line. */
62 const char* line_start; /* pointer to start of current line */
63 #ifndef PGEN
64 PyObject *decoding_readline; /* open(...).readline */
65 PyObject *decoding_buffer;
66 #endif
67 const char* enc; /* Encoding for the current str. */
68 const char* str;
69 const char* input; /* Tokenizer's newline translated copy of the string. */
70 };
71 #endif
72
73 PyAPI_FUNC(PyTokenizer_State *)PyTokenizer_FromString(const char *, int);
74 PyAPI_FUNC(PyTokenizer_State *)PyTokenizer_FromUTF8(const char *, int);
75 PyAPI_FUNC(PyTokenizer_State *)PyTokenizer_FromFile(FILE *, const char*,
76 const char *, const char *);
77 PyAPI_FUNC(void) PyTokenizer_Free(PyTokenizer_State *);
78 PyAPI_FUNC(int) PyTokenizer_Get(PyTokenizer_State *, char **, char **);
79
80 #ifdef __cplusplus
81 }
82 #endif
83 #endif /* !Py_TOKENIZER_H */
OLDNEW
« Doc/c-api/tokenizer.rst ('K') | « Doc/c-api/utilities.rst ('k') | Makefile.pre.in » ('j') | no next file with comments »

RSS Feeds Recent Issues | This issue
This is Rietveld 894c83f36cb7+