diff --git a/Doc/ref/ref2.tex b/Doc/ref/ref2.tex index e9fab58fd0d1da..c8ecb4f6dbfad1 100644 --- a/Doc/ref/ref2.tex +++ b/Doc/ref/ref2.tex @@ -7,11 +7,14 @@ \chapter{Lexical analysis\label{lexical}} \index{parser} \index{token} -Python uses the 7-bit \ASCII{} character set for program text and string -literals. 8-bit characters may be used in string literals and comments -but their interpretation is platform dependent; the proper way to -insert 8-bit characters in string literals is by using octal or -hexadecimal escape sequences. +Python uses the 7-bit \ASCII{} character set for program text. +\versionadded[An encoding declaration can be used to indicate that +string literals and comments use an encoding different from ASCII.]{2.3} +For compatibility with older versions, Python only warns if it finds +8-bit characters; those warnings should be corrected by either declaring +an explicit encoding, or using escape sequences if those bytes are binary +data, instead of characters. + The run-time character set depends on the I/O devices connected to the program but is generally a superset of \ASCII. @@ -69,6 +72,37 @@ \subsection{Comments\label{comments}} \index{hash character} +\subsection{Encoding declarations\label{encodings}} + +If a comment in the first or second line of the Python script matches +the regular expression "coding[=:]\s*([\w-_.]+)", this comment is +processed as an encoding declaration; the first group of this +expression names the encoding of the source code file. The recommended +forms of this expression are + +\begin{verbatim} +# -*- coding: -*- +\end{verbatim} + +which is recognized also by GNU Emacs, and + +\begin{verbatim} +# vim:fileencoding= +\end{verbatim} + +which is recognized by Bram Moolenar's VIM. In addition, if the first +bytes of the file are the UTF-8 signature ($'\xef\xbb\xbf'$), the +declared file encoding is UTF-8 (this is supported, among others, by +Microsoft's notepad.exe). + +If an encoding is declared, the encoding name must be recognized by +Python. % XXX there should be a list of supported encodings. +The encoding is used for all lexical analysis, in particular to find +the end of a string, and to interpret the contents of Unicode literals. +String literals are converted to Unicode for syntactical analysis, +then converted back to their original encoding before interpretation +starts. + \subsection{Explicit line joining\label{explicit-joining}} Two or more physical lines may be joined into logical lines using diff --git a/Grammar/Grammar b/Grammar/Grammar index 3f2b4ef1c60ed7..e52a544bc8426a 100644 --- a/Grammar/Grammar +++ b/Grammar/Grammar @@ -102,3 +102,6 @@ list_for: 'for' exprlist 'in' testlist_safe [list_iter] list_if: 'if' test [list_iter] testlist1: test (',' test)* + +# not used in grammar, but may appear in "node" passed from Parser to Compiler +encoding_decl: NAME diff --git a/Include/errcode.h b/Include/errcode.h index daa702f7e177c5..a8b1aaab6f7a06 100644 --- a/Include/errcode.h +++ b/Include/errcode.h @@ -25,6 +25,7 @@ extern "C" { #define E_OVERFLOW 19 /* Node had too many children */ #define E_TOODEEP 20 /* Too many indentation levels */ #define E_DEDENT 21 /* No matching outer block for dedent */ +#define E_DECODE 22 /* Error in decoding into Unicode */ #ifdef __cplusplus } diff --git a/Include/graminit.h b/Include/graminit.h index 50abda0d3000a8..1f2ab3eadcc177 100644 --- a/Include/graminit.h +++ b/Include/graminit.h @@ -65,3 +65,4 @@ #define list_for 320 #define list_if 321 #define testlist1 322 +#define encoding_decl 323 diff --git a/Makefile.pre.in b/Makefile.pre.in index 7e654789777e21..5ab5b146ddda7c 100644 --- a/Makefile.pre.in +++ b/Makefile.pre.in @@ -190,15 +190,15 @@ POBJS= \ Parser/node.o \ Parser/parser.o \ Parser/parsetok.o \ - Parser/tokenizer.o \ Parser/bitset.o \ Parser/metagrammar.o -PARSER_OBJS= $(POBJS) Parser/myreadline.o +PARSER_OBJS= $(POBJS) Parser/myreadline.o Parser/tokenizer.o PGOBJS= \ Objects/obmalloc.o \ Python/mysnprintf.o \ + Parser/tokenizer_pgen.o \ Parser/firstsets.o \ Parser/grammar.o \ Parser/pgen.o \ @@ -434,6 +434,8 @@ Parser/grammar.o: $(srcdir)/Parser/grammar.c \ $(srcdir)/Include/grammar.h Parser/metagrammar.o: $(srcdir)/Parser/metagrammar.c +Parser/tokenizer_pgen.o: $(srcdir)/Parser/tokenizer.c + Python/compile.o Python/symtable.o: $(GRAMMAR_H) diff --git a/Misc/NEWS b/Misc/NEWS index 77e8c62f52d8ad..0f06ff450f4d5c 100644 --- a/Misc/NEWS +++ b/Misc/NEWS @@ -6,6 +6,8 @@ Type/class unification and new-style classes Core and builtins +- Encoding declarations (PEP 263, phase 1) have been implemented. + - list.sort() has a new implementation. While cross-platform results may vary, and in data-dependent ways, this is much faster on many kinds of partially ordered lists than the previous implementation, diff --git a/Parser/parsetok.c b/Parser/parsetok.c index 5758fa749ba1d4..d70e2d64f83265 100644 --- a/Parser/parsetok.c +++ b/Parser/parsetok.c @@ -8,6 +8,7 @@ #include "parser.h" #include "parsetok.h" #include "errcode.h" +#include "graminit.h" int Py_TabcheckFlag; @@ -45,8 +46,8 @@ PyParser_ParseStringFlagsFilename(char *s, char *filename, return NULL; } + tok->filename = filename ? filename : ""; if (Py_TabcheckFlag || Py_VerboseFlag) { - tok->filename = filename ? filename : ""; tok->altwarning = (tok->filename != NULL); if (Py_TabcheckFlag >= 2) tok->alterror++; @@ -78,8 +79,8 @@ PyParser_ParseFileFlags(FILE *fp, char *filename, grammar *g, int start, err_ret->error = E_NOMEM; return NULL; } + tok->filename = filename; if (Py_TabcheckFlag || Py_VerboseFlag) { - tok->filename = filename; tok->altwarning = (filename != NULL); if (Py_TabcheckFlag >= 2) tok->alterror++; @@ -185,6 +186,13 @@ parsetok(struct tok_state *tok, grammar *g, int start, perrdetail *err_ret, err_ret->text[len] = '\0'; } } + } else if (tok->encoding != NULL) { + node* r = PyNode_New(encoding_decl); + r->n_str = tok->encoding; + r->n_nchildren = 1; + r->n_child = n; + tok->encoding = NULL; + n = r; } PyTokenizer_Free(tok); diff --git a/Parser/tokenizer.c b/Parser/tokenizer.c index b4e0fbf7e58d94..fffc19fa276c32 100644 --- a/Parser/tokenizer.c +++ b/Parser/tokenizer.c @@ -5,10 +5,19 @@ #include "pgenheaders.h" #include +#include #include "tokenizer.h" #include "errcode.h" +#ifndef PGEN +#include "unicodeobject.h" +#include "stringobject.h" +#include "fileobject.h" +#include "codecs.h" +#include "abstract.h" +#endif /* PGEN */ + extern char *PyOS_Readline(char *); /* Return malloc'ed string including trailing \n; empty malloc'ed string for EOF; @@ -114,9 +123,416 @@ tok_new(void) tok->alterror = 0; tok->alttabsize = 1; tok->altindstack[0] = 0; + tok->decoding_state = 0; + tok->decoding_erred = 0; + tok->read_coding_spec = 0; + tok->issued_encoding_warning = 0; + tok->encoding = NULL; + tok->decoding_readline = NULL; + tok->decoding_buffer = NULL; return tok; } +#ifdef PGEN + +static char * +decoding_fgets(char *s, int size, struct tok_state *tok) +{ + return fgets(s, size, tok->fp); +} + +static int +decoding_feof(struct tok_state *tok) +{ + return feof(tok->fp); +} + +static const char * +decode_str(const char *str, struct tok_state *tok) +{ + return str; +} + +#else /* PGEN */ + +static char * +error_ret(struct tok_state *tok) /* XXX */ +{ + tok->decoding_erred = 1; + if (tok->fp != NULL && tok->buf != NULL) /* see PyTokenizer_Free */ + PyMem_DEL(tok->buf); + tok->buf = NULL; + return NULL; /* as if it were EOF */ +} + +static char * +new_string(const char *s, int len) +{ + char* result = PyMem_NEW(char, len + 1); + if (result != NULL) { + memcpy(result, s, len); + result[len] = '\0'; + } + return result; +} + +static char * +get_normal_name(char *s) /* for utf-8 and latin-1 */ +{ + char buf[13]; + int i; + for (i = 0; i < 12; i++) { + int c = s[i]; + if (c == '\0') break; + else if (c == '_') buf[i] = '-'; + else buf[i] = tolower(c); + } + buf[i] = '\0'; + if (strcmp(buf, "utf-8") == 0 || + strncmp(buf, "utf-8-", 6) == 0) return "utf-8"; + else if (strcmp(buf, "latin-1") == 0 || + strcmp(buf, "iso-8859-1") == 0 || + strcmp(buf, "iso-latin-1") == 0 || + strncmp(buf, "latin-1-", 8) == 0 || + strncmp(buf, "iso-8859-1-", 11) == 0 || + strncmp(buf, "iso-latin-1-", 12) == 0) return "iso-8859-1"; + else return s; +} + +/* Return the coding spec in S, or NULL if none is found. */ + +static char * +get_coding_spec(const char *s, int size) +{ + int i; + for (i = 0; i < size - 6; i++) { /* XXX inefficient search */ + const char* t = s + i; + if (strncmp(t, "coding", 6) == 0) { + const char* begin = NULL; + t += 6; + if (t[0] != ':' && t[0] != '=') + continue; + do { + t++; + } while (t[0] == '\x20' || t[0] == '\t'); + + begin = t; + while (isalnum(t[0]) || t[0] == '-' || t[0] == '_' || + t[0] == '.') + t++; + + if (begin < t) { + char* r = new_string(begin, t - begin); + char* q = get_normal_name(r); + if (r != q) { + assert(strlen(r) >= strlen(q)); + strcpy(r, q); + } + return r; + } + } + } + return NULL; +} + +/* Check whether the line contains a coding spec. If it does, + invoke the set_readline function for the new encoding. + This function receives the tok_state and the new encoding. + Return 1 on success, 0 on failure. */ + +static int +check_coding_spec(const char* line, int size, struct tok_state *tok, + int set_readline(struct tok_state *, const char *)) +{ + int r = 1; + char* cs = get_coding_spec(line, size); + if (cs != NULL) { + tok->read_coding_spec = 1; + if (tok->encoding == NULL) { + assert(tok->decoding_state == 1); /* raw */ + if (strcmp(cs, "utf-8") == 0 || + strcmp(cs, "iso-8859-1") == 0) { + tok->encoding = cs; + } else { + r = set_readline(tok, cs); + if (r) { + tok->encoding = cs; + tok->decoding_state = -1; + } + } + } else { /* then, compare cs with BOM */ + r = (strcmp(tok->encoding, cs) == 0); + PyMem_DEL(cs); + } + } + return r; +} + +/* See whether the file starts with a BOM. If it does, + invoke the set_readline function with the new encoding. + Return 1 on success, 0 on failure. */ + +static int +check_bom(int get_char(struct tok_state *), + void unget_char(int, struct tok_state *), + int set_readline(struct tok_state *, const char *), + struct tok_state *tok) +{ + int ch = get_char(tok); + tok->decoding_state = 1; + if (ch == EOF) { + return 1; + } else if (ch == 0xEF) { + ch = get_char(tok); if (ch != 0xBB) goto NON_BOM; + ch = get_char(tok); if (ch != 0xBF) goto NON_BOM; +#if 0 + /* Disable support for UTF-16 BOMs until a decision + is made whether this needs to be supported. */ + } else if (ch == 0xFE) { + ch = get_char(tok); if (ch != 0xFF) goto NON_BOM; + if (!set_readline(tok, "utf-16-be")) return 0; + tok->decoding_state = -1; + } else if (ch == 0xFF) { + ch = get_char(tok); if (ch != 0xFE) goto NON_BOM; + if (!set_readline(tok, "utf-16-le")) return 0; + tok->decoding_state = -1; +#endif + } else { + unget_char(ch, tok); + return 1; + } + tok->encoding = new_string("utf-8", 5); /* resulting is in utf-8 */ + return 1; + NON_BOM: + /* any token beginning with '\xEF', '\xFE', '\xFF' is a bad token */ + unget_char(0xFF, tok); /* XXX this will cause a syntax error */ + return 1; +} + +/* Read a line of text from TOK into S, using the stream in TOK. + Return NULL on failure, else S. */ + +static char * +fp_readl(char *s, int size, struct tok_state *tok) +{ + PyObject* utf8; + PyObject* buf = tok->decoding_buffer; + if (buf == NULL) { + buf = PyObject_CallObject(tok->decoding_readline, NULL); + if (buf == NULL) return error_ret(tok); + } else { + tok->decoding_buffer = NULL; + } + utf8 = PyUnicode_AsUTF8String(buf); + Py_DECREF(buf); + if (utf8 == NULL) return error_ret(tok); + else { + const char* str = PyString_AsString(utf8); + assert(strlen(str) < size); /* XXX */ + strcpy(s, str); + Py_DECREF(utf8); + if (s[0] == '\0') return NULL; /* EOF */ + return s; + } +} + +/* Set the readline function for TOK to a StreamReader's + readline function. The StreamReader is named ENC. + + This function is called from check_bom and check_coding_spec. + + ENC is usually identical to the future value of tok->encoding, + except for the (currently unsupported) case of UTF-16. + + Return 1 on success, 0 on failure. */ + +static int +fp_setreadl(struct tok_state *tok, const char* enc) +{ + PyObject *reader, *stream, *readline; + + stream = PyFile_FromFile(tok->fp, tok->filename, "rb", NULL); + if (stream == NULL) return 0; + + reader = PyCodec_StreamReader(enc, stream, NULL); + Py_DECREF(stream); + if (reader == NULL) return 0; + + readline = PyObject_GetAttrString(reader, "readline"); + Py_DECREF(reader); + if (readline == NULL) return 0; + + tok->decoding_readline = readline; + return 1; +} + +/* Fetch the next byte from TOK. */ + +static int fp_getc(struct tok_state *tok) { + return getc(tok->fp); +} + +/* Unfetch the last byte back into TOK. */ + +static void fp_ungetc(int c, struct tok_state *tok) { + ungetc(c, tok->fp); +} + +/* Read a line of input from TOK. Determine encoding + if necessary. */ + +static char * +decoding_fgets(char *s, int size, struct tok_state *tok) +{ + char *line; + int warn = 0, badchar = 0; + for (;;) + if (tok->decoding_state < 0) { + /* We already have a codec associated with + this input. */ + line = fp_readl(s, size, tok); + break; + } else if (tok->decoding_state > 0) { + /* We want a 'raw' read. */ + line = Py_UniversalNewlineFgets(s, size, + tok->fp, NULL); + warn = 1; + break; + } else { + /* We have not yet determined the encoding. + If an encoding is found, use the file-pointer + reader functions from now on. */ + if (!check_bom(fp_getc, fp_ungetc, fp_setreadl, tok)) + return error_ret(tok); + assert(tok->decoding_state != 0); + } + if (line != NULL && tok->lineno < 2 && !tok->read_coding_spec) { + if (!check_coding_spec(line, strlen(line), tok, fp_setreadl)) { + return error_ret(tok); + } + } +#ifndef PGEN + if (warn && line && !tok->issued_encoding_warning && !tok->encoding) { + unsigned char *c; + for (c = line; *c; c++) + if (*c > 127) { + badchar = *c; + break; + } + } + if (badchar) { + char buf[200]; + sprintf(buf, "Non-ASCII character '\\x%.2x', " + "but no declared encoding", badchar); + PyErr_WarnExplicit(PyExc_DeprecationWarning, + buf, tok->filename, tok->lineno, + NULL, NULL); + tok->issued_encoding_warning = 1; + } +#endif + return line; +} + +static int +decoding_feof(struct tok_state *tok) +{ + if (tok->decoding_state >= 0) { + return feof(tok->fp); + } else { + PyObject* buf = tok->decoding_buffer; + if (buf == NULL) { + buf = PyObject_CallObject(tok->decoding_readline, NULL); + if (buf == NULL) { + error_ret(tok); + return 1; + } else { + tok->decoding_buffer = buf; + } + } + return PyObject_Length(buf) == 0; + } +} + +/* Fetch a byte from TOK, using the string buffer. */ + +static int buf_getc(struct tok_state *tok) { + return *tok->str++; +} + +/* Unfetch a byte from TOK, using the string buffer. */ + +static void buf_ungetc(int c, struct tok_state *tok) { + tok->str--; + assert(*tok->str == c); /* tok->cur may point to read-only segment */ +} + +/* Set the readline function for TOK to ENC. For the string-based + tokenizer, this means to just record the encoding. */ + +static int buf_setreadl(struct tok_state *tok, const char* enc) { + tok->enc = enc; + return 1; +} + +/* Return a UTF-8 encoding Python string object from the + C byte string STR, which is encoded with ENC. */ + +static PyObject * +translate_into_utf8(const char* str, const char* enc) { + PyObject *utf8; + PyObject* buf = PyUnicode_Decode(str, strlen(str), enc, NULL); + if (buf == NULL) + return NULL; + utf8 = PyUnicode_AsUTF8String(buf); + Py_DECREF(buf); + return utf8; +} + +/* Decode a byte string STR for use as the buffer of TOK. + Look for encoding declarations inside STR, and record them + inside TOK. */ + +static const char * +decode_str(const char *str, struct tok_state *tok) +{ + PyObject* utf8 = NULL; + const char *s; + int lineno = 0; + tok->enc = NULL; + tok->str = str; + if (!check_bom(buf_getc, buf_ungetc, buf_setreadl, tok)) + return NULL; + str = tok->str; /* string after BOM if any */ + assert(r); + if (tok->enc != NULL) { + utf8 = translate_into_utf8(str, tok->enc); + if (utf8 == NULL) + return NULL; + str = PyString_AsString(utf8); + } + for (s = str;; s++) { + if (*s == '\0') break; + else if (*s == '\n') { + lineno++; + if (lineno == 2) break; + } + } + tok->enc = NULL; + if (!check_coding_spec(str, s - str, tok, buf_setreadl)) + return NULL; + if (tok->enc != NULL) { + assert(utf8 == NULL); + utf8 = translate_into_utf8(str, tok->enc); + if (utf8 == NULL) + return NULL; + str = PyString_AsString(utf8); + } + assert(tok->decoding_buffer == NULL); + tok->decoding_buffer = utf8; /* CAUTION */ + return str; +} + +#endif /* PGEN */ /* Set up tokenizer for string */ @@ -126,6 +542,9 @@ PyTokenizer_FromString(char *str) struct tok_state *tok = tok_new(); if (tok == NULL) return NULL; + str = (char *)decode_str(str, tok); + if (str == NULL) + return NULL; tok->buf = tok->cur = tok->end = tok->inp = str; return tok; } @@ -157,6 +576,10 @@ PyTokenizer_FromFile(FILE *fp, char *ps1, char *ps2) void PyTokenizer_Free(struct tok_state *tok) { + if (tok->encoding != NULL) + PyMem_DEL(tok->encoding); + Py_XDECREF(tok->decoding_readline); + Py_XDECREF(tok->decoding_buffer); if (tok->fp != NULL && tok->buf != NULL) PyMem_DEL(tok->buf); PyMem_DEL(tok); @@ -246,8 +669,8 @@ tok_nextc(register struct tok_state *tok) } tok->end = tok->buf + BUFSIZ; } - if (Py_UniversalNewlineFgets(tok->buf, (int)(tok->end - tok->buf), - tok->fp, NULL) == NULL) { + if (decoding_fgets(tok->buf, (int)(tok->end - tok->buf), + tok) == NULL) { tok->done = E_EOF; done = 1; } @@ -259,7 +682,7 @@ tok_nextc(register struct tok_state *tok) } else { cur = tok->cur - tok->buf; - if (feof(tok->fp)) { + if (decoding_feof(tok)) { tok->done = E_EOF; done = 1; } @@ -285,9 +708,9 @@ tok_nextc(register struct tok_state *tok) tok->end = tok->buf + newsize; tok->start = curstart < 0 ? NULL : tok->buf + curstart; - if (Py_UniversalNewlineFgets(tok->inp, + if (decoding_fgets(tok->inp, (int)(tok->end - tok->inp), - tok->fp, NULL) == NULL) { + tok) == NULL) { /* Last line does not end in \n, fake one */ strcpy(tok->inp, "\n"); @@ -506,9 +929,8 @@ indenterror(struct tok_state *tok) /* Get next token, after space stripping etc. */ -int -PyTokenizer_Get(register struct tok_state *tok, char **p_start, - char **p_end) +static int +tok_get(register struct tok_state *tok, char **p_start, char **p_end) { register int c; int blankline; @@ -915,6 +1337,16 @@ PyTokenizer_Get(register struct tok_state *tok, char **p_start, return PyToken_OneChar(c); } +int +PyTokenizer_Get(struct tok_state *tok, char **p_start, char **p_end) +{ + int result = tok_get(tok, p_start, p_end); + if (tok->decoding_erred) { + result = ERRORTOKEN; + tok->done = E_DECODE; + } + return result; +} #ifdef Py_DEBUG diff --git a/Parser/tokenizer.h b/Parser/tokenizer.h index 8fded37ad998cb..9782666fd82450 100644 --- a/Parser/tokenizer.h +++ b/Parser/tokenizer.h @@ -4,6 +4,7 @@ extern "C" { #endif +#include "object.h" /* Tokenizer interface */ @@ -38,6 +39,16 @@ struct tok_state { int alterror; /* Issue error if alternate tabs don't match */ int alttabsize; /* Alternate tab spacing */ int altindstack[MAXINDENT]; /* Stack of alternate indents */ + /* Stuff for PEP 0263 */ + int decoding_state; /* -1:decoding, 0:init, 1:raw */ + int decoding_erred; /* whether erred in decoding */ + int read_coding_spec; /* whether 'coding:...' has been read */ + int issued_encoding_warning; /* whether non-ASCII warning was issued */ + char *encoding; + PyObject *decoding_readline; /* codecs.open(...).readline */ + PyObject *decoding_buffer; + const char* enc; + const char* str; }; extern struct tok_state *PyTokenizer_FromString(char *); diff --git a/Parser/tokenizer_pgen.c b/Parser/tokenizer_pgen.c new file mode 100644 index 00000000000000..9cb8492d6a647b --- /dev/null +++ b/Parser/tokenizer_pgen.c @@ -0,0 +1,2 @@ +#define PGEN +#include "tokenizer.c" diff --git a/Python/compile.c b/Python/compile.c index 3a0948e0c4ecbd..512b5a339dbb39 100644 --- a/Python/compile.c +++ b/Python/compile.c @@ -485,6 +485,7 @@ struct compiling { int c_closure; /* Is nested w/freevars? */ struct symtable *c_symtable; /* pointer to module symbol table */ PyFutureFeatures *c_future; /* pointer to module's __future__ */ + char *c_encoding; /* source encoding (a borrowed reference) */ }; static int @@ -1181,6 +1182,23 @@ parsenumber(struct compiling *co, char *s) } } +static PyObject * +decode_utf8(char **sPtr, char *end, char* encoding) +{ + PyObject *u, *v; + char *s, *t; + t = s = *sPtr; + /* while (s < end && *s != '\\') s++; */ /* inefficient for u".." */ + while (s < end && (*s & 0x80)) s++; + *sPtr = s; + u = PyUnicode_DecodeUTF8(t, s - t, NULL); + if (u == NULL) + return NULL; + v = PyUnicode_AsEncodedString(u, encoding, NULL); + Py_DECREF(u); + return v; +} + static PyObject * parsestr(struct compiling *com, char *s) { @@ -1193,6 +1211,8 @@ parsestr(struct compiling *com, char *s) int first = *s; int quote = first; int rawmode = 0; + char* encoding = ((com == NULL) ? NULL : com->c_encoding); + int need_encoding; int unicode = 0; if (isalpha(quote) || quote == '_') { @@ -1230,28 +1250,101 @@ parsestr(struct compiling *com, char *s) } #ifdef Py_USING_UNICODE if (unicode || Py_UnicodeFlag) { + PyObject *u, *w; + if (encoding == NULL) { + buf = s; + u = NULL; + } else if (strcmp(encoding, "iso-8859-1") == 0) { + buf = s; + u = NULL; + } else { + /* "\XX" may become "\u005c\uHHLL" (12 bytes) */ + u = PyString_FromStringAndSize((char *)NULL, len * 4); + if (u == NULL) + return NULL; + p = buf = PyString_AsString(u); + end = s + len; + while (s < end) { + if (*s == '\\') { + *p++ = *s++; + if (*s & 0x80) { + strcpy(p, "u005c"); + p += 5; + } + } + if (*s & 0x80) { /* XXX inefficient */ + char *r; + int rn, i; + w = decode_utf8(&s, end, "utf-16-be"); + if (w == NULL) { + Py_DECREF(u); + return NULL; + } + r = PyString_AsString(w); + rn = PyString_Size(w); + assert(rn % 2 == 0); + for (i = 0; i < rn; i += 2) { + sprintf(p, "\\u%02x%02x", + r[i + 0] & 0xFF, + r[i + 1] & 0xFF); + p += 6; + } + Py_DECREF(w); + } else { + *p++ = *s++; + } + } + len = p - buf; + } if (rawmode) - v = PyUnicode_DecodeRawUnicodeEscape( - s, len, NULL); + v = PyUnicode_DecodeRawUnicodeEscape(buf, len, NULL); else - v = PyUnicode_DecodeUnicodeEscape( - s, len, NULL); + v = PyUnicode_DecodeUnicodeEscape(buf, len, NULL); + Py_XDECREF(u); if (v == NULL) PyErr_SyntaxLocation(com->c_filename, com->c_lineno); return v; } #endif - if (rawmode || strchr(s, '\\') == NULL) - return PyString_FromStringAndSize(s, len); - v = PyString_FromStringAndSize((char *)NULL, len); + need_encoding = (encoding != NULL && + strcmp(encoding, "utf-8") != 0 && + strcmp(encoding, "iso-8859-1") != 0); + if (rawmode || strchr(s, '\\') == NULL) { + if (need_encoding) { + PyObject* u = PyUnicode_DecodeUTF8(s, len, NULL); + if (u == NULL) + return NULL; + v = PyUnicode_AsEncodedString(u, encoding, NULL); + Py_DECREF(u); + return v; + } else { + return PyString_FromStringAndSize(s, len); + } + } + v = PyString_FromStringAndSize((char *)NULL, /* XXX 4 is enough? */ + need_encoding ? len * 4 : len); if (v == NULL) return NULL; p = buf = PyString_AsString(v); end = s + len; while (s < end) { if (*s != '\\') { - *p++ = *s++; + ORDINAL: + if (need_encoding && (*s & 0x80)) { + char *r; + int rn; + PyObject* w = decode_utf8(&s, end, encoding); + if (w == NULL) + return NULL; + r = PyString_AsString(w); + rn = PyString_Size(w); + memcpy(p, r, rn); + p += rn; + Py_DECREF(w); + } else { + *p++ = *s++; + } continue; } s++; @@ -1320,8 +1413,8 @@ parsestr(struct compiling *com, char *s) #endif default: *p++ = '\\'; - *p++ = s[-1]; - break; + s--; + goto ORDINAL; } } _PyString_Resize(&v, (int)(p - buf)); @@ -4149,6 +4242,12 @@ jcompile(node *n, char *filename, struct compiling *base, PyCodeObject *co; if (!com_init(&sc, filename)) return NULL; + if (TYPE(n) == encoding_decl) { + sc.c_encoding = STR(n); + n = CHILD(n, 0); + } else { + sc.c_encoding = NULL; + } if (base) { sc.c_private = base->c_private; sc.c_symtable = base->c_symtable; @@ -4157,6 +4256,10 @@ jcompile(node *n, char *filename, struct compiling *base, || (sc.c_symtable->st_cur->ste_type == TYPE_FUNCTION)) sc.c_nested = 1; sc.c_flags |= base->c_flags & PyCF_MASK; + if (base->c_encoding != NULL) { + assert(sc.c_encoding == NULL); + sc.c_encoding = base->c_encoding; + } } else { sc.c_private = NULL; sc.c_future = PyNode_Future(n, filename); diff --git a/Python/graminit.c b/Python/graminit.c index ef7d4677b5f370..98bad94bc7c6bc 100644 --- a/Python/graminit.c +++ b/Python/graminit.c @@ -1463,7 +1463,17 @@ static state states_66[2] = { {1, arcs_66_0}, {2, arcs_66_1}, }; -static dfa dfas[67] = { +static arc arcs_67_0[1] = { + {12, 1}, +}; +static arc arcs_67_1[1] = { + {0, 1}, +}; +static state states_67[2] = { + {1, arcs_67_0}, + {1, arcs_67_1}, +}; +static dfa dfas[68] = { {256, "single_input", 0, 3, states_0, "\004\030\001\000\000\000\124\360\213\011\162\000\002\000\140\210\244\005\001"}, {257, "file_input", 0, 2, states_1, @@ -1598,8 +1608,10 @@ static dfa dfas[67] = { "\000\000\000\000\000\000\000\000\000\000\002\000\000\000\000\000\000\000\000"}, {322, "testlist1", 0, 2, states_66, "\000\020\001\000\000\000\000\000\000\000\000\000\002\000\140\210\244\005\000"}, + {323, "encoding_decl", 0, 2, states_67, + "\000\020\000\000\000\000\000\000\000\000\000\000\000\000\000\000\000\000\000"}, }; -static label labels[148] = { +static label labels[149] = { {0, "EMPTY"}, {256, 0}, {4, 0}, @@ -1748,10 +1760,11 @@ static label labels[148] = { {318, 0}, {319, 0}, {321, 0}, + {323, 0}, }; grammar _PyParser_Grammar = { - 67, + 68, dfas, - {148, labels}, + {149, labels}, 256 }; diff --git a/Python/pythonrun.c b/Python/pythonrun.c index b1fde29bd285be..006ff083d25274 100644 --- a/Python/pythonrun.c +++ b/Python/pythonrun.c @@ -1221,6 +1221,7 @@ static void err_input(perrdetail *err) { PyObject *v, *w, *errtype; + PyObject* u = NULL; char *msg = NULL; errtype = PyExc_SyntaxError; v = Py_BuildValue("(ziiz)", err->filename, @@ -1272,12 +1273,24 @@ err_input(perrdetail *err) errtype = PyExc_IndentationError; msg = "too many levels of indentation"; break; + case E_DECODE: { /* XXX */ + PyThreadState* tstate = PyThreadState_Get(); + PyObject* value = tstate->curexc_value; + if (value != NULL) { + u = PyObject_Repr(value); + if (u != NULL) { + msg = PyString_AsString(u); + break; + } + } + } default: fprintf(stderr, "error=%d\n", err->error); msg = "unknown parsing error"; break; } w = Py_BuildValue("(sO)", msg, v); + Py_XDECREF(u); Py_XDECREF(v); PyErr_SetObject(errtype, w); Py_XDECREF(w);