diff options
Diffstat (limited to 'contrib/tools/python3/Python/fileutils.c')
| -rw-r--r-- | contrib/tools/python3/Python/fileutils.c | 274 |
1 files changed, 214 insertions, 60 deletions
diff --git a/contrib/tools/python3/Python/fileutils.c b/contrib/tools/python3/Python/fileutils.c index 45de2b891d9..81be9d5e8ab 100644 --- a/contrib/tools/python3/Python/fileutils.c +++ b/contrib/tools/python3/Python/fileutils.c @@ -2,8 +2,11 @@ #include "pycore_fileutils.h" // fileutils definitions #include "pycore_runtime.h" // _PyRuntime #include "osdefs.h" // SEP -#include <locale.h> + #include <stdlib.h> // mbstowcs() +#ifdef HAVE_UNISTD_H +# include <unistd.h> // getcwd() +#endif #ifdef MS_WINDOWS # include <malloc.h> @@ -19,7 +22,7 @@ extern int winerror_to_errno(int); #endif #ifdef HAVE_LANGINFO_H -#include <langinfo.h> +# include <langinfo.h> // nl_langinfo(CODESET) #endif #ifdef HAVE_SYS_IOCTL_H @@ -27,12 +30,12 @@ extern int winerror_to_errno(int); #endif #ifdef HAVE_NON_UNICODE_WCHAR_T_REPRESENTATION -#error #include <iconv.h> +# error #include <iconv.h> // iconv_open() #endif #ifdef HAVE_FCNTL_H -#include <fcntl.h> -#endif /* HAVE_FCNTL_H */ +# include <fcntl.h> // fcntl(F_GETFD) +#endif #ifdef O_CLOEXEC /* Does open() support the O_CLOEXEC flag? Possible values: @@ -105,14 +108,14 @@ _Py_device_encoding(int fd) #else if (_PyRuntime.preconfig.utf8_mode) { _Py_DECLARE_STR(utf_8, "utf-8"); - return Py_NewRef(&_Py_STR(utf_8)); + return &_Py_STR(utf_8); } return _Py_GetLocaleEncodingObject(); #endif } -static size_t +static int is_valid_wide_char(wchar_t ch) { #ifdef HAVE_NON_UNICODE_WCHAR_T_REPRESENTATION @@ -520,15 +523,7 @@ decode_current_locale(const char* arg, wchar_t **wstr, size_t *wlen, break; } - if (converted == INCOMPLETE_CHARACTER) { - /* Incomplete character. This should never happen, - since we provide everything that we have - - unless there is a bug in the C library, or I - misunderstood how mbrtowc works. */ - goto decode_error; - } - - if (converted == DECODE_ERROR) { + if (converted == DECODE_ERROR || converted == INCOMPLETE_CHARACTER) { if (!surrogateescape) { goto decode_error; } @@ -1499,7 +1494,7 @@ set_inheritable(int fd, int inheritable, int raise, int *atomic_flag_works) #else #if defined(HAVE_SYS_IOCTL_H) && defined(FIOCLEX) && defined(FIONCLEX) - if (ioctl_works != 0 && raise != 0) { + if (raise != 0 && _Py_atomic_load_int_relaxed(&ioctl_works) != 0) { /* fast-path: ioctl() only requires one syscall */ /* caveat: raise=0 is an indicator that we must be async-signal-safe * thus avoid using ioctl() so we skip the fast-path. */ @@ -1509,7 +1504,9 @@ set_inheritable(int fd, int inheritable, int raise, int *atomic_flag_works) request = FIOCLEX; err = ioctl(fd, request, NULL); if (!err) { - ioctl_works = 1; + if (_Py_atomic_load_int_relaxed(&ioctl_works) == -1) { + _Py_atomic_store_int_relaxed(&ioctl_works, 1); + } return 0; } @@ -1536,7 +1533,7 @@ set_inheritable(int fd, int inheritable, int raise, int *atomic_flag_works) with EACCES. While FIOCLEX is safe operation it may be unavailable because ioctl was denied altogether. This can be the case on Android. */ - ioctl_works = 0; + _Py_atomic_store_int_relaxed(&ioctl_works, 0); } /* fallback to fcntl() if ioctl() does not work */ } @@ -2292,6 +2289,99 @@ PathCchCombineEx(wchar_t *buffer, size_t bufsize, const wchar_t *dirname, #endif /* defined(MS_WINDOWS_GAMES) && !defined(MS_WINDOWS_DESKTOP) */ +void +_Py_skiproot(const wchar_t *path, Py_ssize_t size, Py_ssize_t *drvsize, + Py_ssize_t *rootsize) +{ + assert(drvsize); + assert(rootsize); +#ifndef MS_WINDOWS +#define IS_SEP(x) (*(x) == SEP) + *drvsize = 0; + if (!IS_SEP(&path[0])) { + // Relative path, e.g.: 'foo' + *rootsize = 0; + } + else if (!IS_SEP(&path[1]) || IS_SEP(&path[2])) { + // Absolute path, e.g.: '/foo', '///foo', '////foo', etc. + *rootsize = 1; + } + else { + // Precisely two leading slashes, e.g.: '//foo'. Implementation defined per POSIX, see + // https://pubs.opengroup.org/onlinepubs/9699919799/basedefs/V1_chap04.html#tag_04_13 + *rootsize = 2; + } +#undef IS_SEP +#else + const wchar_t *pEnd = size >= 0 ? &path[size] : NULL; +#define IS_END(x) (pEnd ? (x) == pEnd : !*(x)) +#define IS_SEP(x) (*(x) == SEP || *(x) == ALTSEP) +#define SEP_OR_END(x) (IS_SEP(x) || IS_END(x)) + if (IS_SEP(&path[0])) { + if (IS_SEP(&path[1])) { + // Device drives, e.g. \\.\device or \\?\device + // UNC drives, e.g. \\server\share or \\?\UNC\server\share + Py_ssize_t idx; + if (path[2] == L'?' && IS_SEP(&path[3]) && + (path[4] == L'U' || path[4] == L'u') && + (path[5] == L'N' || path[5] == L'n') && + (path[6] == L'C' || path[6] == L'c') && + IS_SEP(&path[7])) + { + idx = 8; + } + else { + idx = 2; + } + while (!SEP_OR_END(&path[idx])) { + idx++; + } + if (IS_END(&path[idx])) { + *drvsize = idx; + *rootsize = 0; + } + else { + idx++; + while (!SEP_OR_END(&path[idx])) { + idx++; + } + *drvsize = idx; + if (IS_END(&path[idx])) { + *rootsize = 0; + } + else { + *rootsize = 1; + } + } + } + else { + // Relative path with root, e.g. \Windows + *drvsize = 0; + *rootsize = 1; + } + } + else if (!IS_END(&path[0]) && path[1] == L':') { + *drvsize = 2; + if (IS_SEP(&path[2])) { + // Absolute drive-letter path, e.g. X:\Windows + *rootsize = 1; + } + else { + // Relative path with drive, e.g. X:Windows + *rootsize = 0; + } + } + else { + // Relative path, e.g. Windows + *drvsize = 0; + *rootsize = 0; + } +#undef SEP_OR_END +#undef IS_SEP +#undef IS_END +#endif +} + // The caller must ensure "buffer" is big enough. static int join_relfile(wchar_t *buffer, size_t bufsize, @@ -2408,49 +2498,40 @@ _Py_normpath_and_size(wchar_t *path, Py_ssize_t size, Py_ssize_t *normsize) #endif #define SEP_OR_END(x) (IS_SEP(x) || IS_END(x)) - // Skip leading '.\' - if (p1[0] == L'.' && IS_SEP(&p1[1])) { - path = &path[2]; - while (IS_SEP(path) && !IS_END(path)) { - path++; - } - p1 = p2 = minP2 = path; - lastC = SEP; - } -#ifdef MS_WINDOWS - // Skip past drive segment and update minP2 - else if (p1[0] && p1[1] == L':') { - *p2++ = *p1++; - *p2++ = *p1++; - minP2 = p2; - lastC = L':'; - } - // Skip past all \\-prefixed paths, including \\?\, \\.\, - // and network paths, including the first segment. - else if (IS_SEP(&p1[0]) && IS_SEP(&p1[1])) { - int sepCount = 2; - *p2++ = SEP; - *p2++ = SEP; - p1 += 2; - for (; !IS_END(p1) && sepCount; ++p1) { - if (IS_SEP(p1)) { - --sepCount; - *p2++ = lastC = SEP; - } else { - *p2++ = lastC = *p1; + Py_ssize_t drvsize, rootsize; + _Py_skiproot(path, size, &drvsize, &rootsize); + if (drvsize || rootsize) { + // Skip past root and update minP2 + p1 = &path[drvsize + rootsize]; +#ifndef ALTSEP + p2 = p1; +#else + for (; p2 < p1; ++p2) { + if (*p2 == ALTSEP) { + *p2 = SEP; } } +#endif minP2 = p2 - 1; + lastC = *minP2; +#ifdef MS_WINDOWS + if (lastC != SEP) { + minP2++; + } +#endif } -#else - // Skip past two leading SEPs - else if (IS_SEP(&p1[0]) && IS_SEP(&p1[1]) && !IS_SEP(&p1[2])) { - *p2++ = *p1++; - *p2++ = *p1++; - minP2 = p2 - 1; // Absolute path has SEP at minP2 - lastC = SEP; + if (p1[0] == L'.' && SEP_OR_END(&p1[1])) { + // Skip leading '.\' + lastC = *++p1; +#ifdef ALTSEP + if (lastC == ALTSEP) { + lastC = SEP; + } +#endif + while (IS_SEP(p1)) { + p1++; + } } -#endif /* MS_WINDOWS */ /* if pEnd is specified, check that. Else, check for null terminator */ for (; !IS_END(p1); ++p1) { @@ -2875,9 +2956,9 @@ done: * non-opened fd in the middle. * 2b. If fdwalk(3) isn't available, just do a plain close(2) loop. */ -#ifdef __FreeBSD__ +#ifdef HAVE_CLOSEFROM # define USE_CLOSEFROM -#endif /* __FreeBSD__ */ +#endif /* HAVE_CLOSEFROM */ #ifdef HAVE_FDWALK # define USE_FDWALK @@ -2919,7 +3000,7 @@ _Py_closerange(int first, int last) #ifdef USE_CLOSEFROM if (last >= sysconf(_SC_OPEN_MAX)) { /* Any errors encountered while closing file descriptors are ignored */ - closefrom(first); + (void)closefrom(first); } else #endif /* USE_CLOSEFROM */ @@ -2940,3 +3021,76 @@ _Py_closerange(int first, int last) #endif /* USE_FDWALK */ _Py_END_SUPPRESS_IPH } + + +#ifndef MS_WINDOWS +// Ticks per second used by clock() and times() functions. +// See os.times() and time.process_time() implementations. +int +_Py_GetTicksPerSecond(long *ticks_per_second) +{ +#if defined(HAVE_SYSCONF) && defined(_SC_CLK_TCK) + long value = sysconf(_SC_CLK_TCK); + if (value < 1) { + return -1; + } + *ticks_per_second = value; +#elif defined(HZ) + assert(HZ >= 1); + *ticks_per_second = HZ; +#else + // Magic fallback value; may be bogus + *ticks_per_second = 60; +#endif + return 0; +} +#endif + + +/* Check if a file descriptor is valid or not. + Return 0 if the file descriptor is invalid, return non-zero otherwise. */ +int +_Py_IsValidFD(int fd) +{ +/* dup() is faster than fstat(): fstat() can require input/output operations, + whereas dup() doesn't. There is a low risk of EMFILE/ENFILE at Python + startup. Problem: dup() doesn't check if the file descriptor is valid on + some platforms. + + fcntl(fd, F_GETFD) is even faster, because it only checks the process table. + It is preferred over dup() when available, since it cannot fail with the + "too many open files" error (EMFILE). + + bpo-30225: On macOS Tiger, when stdout is redirected to a pipe and the other + side of the pipe is closed, dup(1) succeed, whereas fstat(1, &st) fails with + EBADF. FreeBSD has similar issue (bpo-32849). + + Only use dup() on Linux where dup() is enough to detect invalid FD + (bpo-32849). +*/ + if (fd < 0) { + return 0; + } +#if defined(F_GETFD) && ( \ + defined(__linux__) || \ + defined(__APPLE__) || \ + (defined(__wasm__) && !defined(__wasi__))) + return fcntl(fd, F_GETFD) >= 0; +#elif defined(__linux__) + int fd2 = dup(fd); + if (fd2 >= 0) { + close(fd2); + } + return (fd2 >= 0); +#elif defined(MS_WINDOWS) + HANDLE hfile; + _Py_BEGIN_SUPPRESS_IPH + hfile = (HANDLE)_get_osfhandle(fd); + _Py_END_SUPPRESS_IPH + return (hfile != INVALID_HANDLE_VALUE + && GetFileType(hfile) != FILE_TYPE_UNKNOWN); +#else + struct stat st; + return (fstat(fd, &st) == 0); +#endif +} |