diff options
Diffstat (limited to 'sal/osl/unx/file_url.cxx')
-rw-r--r-- | sal/osl/unx/file_url.cxx | 986 |
1 files changed, 986 insertions, 0 deletions
diff --git a/sal/osl/unx/file_url.cxx b/sal/osl/unx/file_url.cxx new file mode 100644 index 000000000..5e99d5306 --- /dev/null +++ b/sal/osl/unx/file_url.cxx @@ -0,0 +1,986 @@ +/* -*- Mode: C++; tab-width: 4; indent-tabs-mode: nil; c-basic-offset: 4 -*- */ +/* + * This file is part of the LibreOffice project. + * + * This Source Code Form is subject to the terms of the Mozilla Public + * License, v. 2.0. If a copy of the MPL was not distributed with this + * file, You can obtain one at http://mozilla.org/MPL/2.0/. + * + * This file incorporates work covered by the following license notice: + * + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed + * with this work for additional information regarding copyright + * ownership. The ASF licenses this file to you under the Apache + * License, Version 2.0 (the "License"); you may not use this file + * except in compliance with the License. You may obtain a copy of + * the License at http://www.apache.org/licenses/LICENSE-2.0 . + */ + +#include "file_url.hxx" + +#include "system.hxx" + +#include <algorithm> +#include <cassert> +#include <cstring> +#include <stdexcept> +#include <string_view> +#include <type_traits> +#include <limits.h> +#include <errno.h> +#include <strings.h> +#include <unistd.h> + +#include <o3tl/safeint.hxx> +#include <osl/file.hxx> +#include <osl/security.hxx> +#include <osl/socket.h> +#include <osl/diagnose.h> +#include <osl/thread.h> +#include <osl/process.h> + +#include <rtl/character.hxx> +#include <rtl/strbuf.hxx> +#include <rtl/uri.h> +#include <rtl/uri.hxx> +#include <rtl/ustring.hxx> +#include <rtl/ustrbuf.h> +#include <rtl/ustrbuf.hxx> +#include <rtl/textcvt.h> +#include <sal/log.hxx> + +#include <uri_internal.hxx> + +#include "file_error_transl.hxx" +#include "file_path_helper.hxx" + +#include "uunxapi.hxx" + +/** @file + + General note + + This file contains the part that handles File URLs. + + File URLs as scheme specific notion of URIs + (RFC2396) may be handled platform independent, but + will not in osl which is considered wrong. + Future version of osl should handle File URLs this + way. In rtl/uri there is already a URI parser etc. + so this code should be consolidated. + +*/ + +using namespace osl; + +namespace { + +// A slightly modified version of Pchar in rtl/source/uri.c, but without +// encoding slashes: +const sal_Bool uriCharClass[128] = { + false, false, false, false, false, false, false, false, + false, false, false, false, false, false, false, false, + false, false, false, false, false, false, false, false, + false, false, false, false, false, false, false, false, + false, true, false, false, true, false, true, true, // !"#$%&' + true, true, true, true, true, true, true, true, // ()*+,-./ + true, true, true, true, true, true, true, true, // 01234567 + true, true, true, false, false, true, false, false, // 89:;<=>? + true, true, true, true, true, true, true, true, // @ABCDEFG + true, true, true, true, true, true, true, true, // HIJKLMNO + true, true, true, true, true, true, true, true, // PQRSTUVW + true, true, true, false, false, false, false, true, // XYZ[\]^_ + false, true, true, true, true, true, true, true, // `abcdefg + true, true, true, true, true, true, true, true, // hijklmno + true, true, true, true, true, true, true, true, // pqrstuvw + true, true, true, false, false, false, true, false}; // xyz{|}~ + +} + +oslFileError SAL_CALL osl_getCanonicalName( rtl_uString* ustrFileURL, rtl_uString** pustrValidURL ) +{ + OSL_FAIL("osl_getCanonicalName not implemented"); + + rtl_uString_newFromString(pustrValidURL, ustrFileURL); + return osl_File_E_None; +} + +namespace { + + class UnicodeToTextConverter_Impl + { + rtl_UnicodeToTextConverter m_converter; + + UnicodeToTextConverter_Impl() + : m_converter (rtl_createUnicodeToTextConverter (osl_getThreadTextEncoding())) + {} + + ~UnicodeToTextConverter_Impl() + { + rtl_destroyUnicodeToTextConverter (m_converter); + } + public: + static UnicodeToTextConverter_Impl & getInstance() + { + static UnicodeToTextConverter_Impl g_theConverter; + return g_theConverter; + } + + sal_Size convert( + sal_Unicode const * pSrcBuf, sal_Size nSrcChars, char * pDstBuf, sal_Size nDstBytes, + sal_uInt32 nFlags, sal_uInt32 * pInfo, sal_Size * pSrcCvtChars) + { + OSL_ASSERT(m_converter != nullptr); + return rtl_convertUnicodeToText ( + m_converter, nullptr, pSrcBuf, nSrcChars, pDstBuf, nDstBytes, nFlags, pInfo, pSrcCvtChars); + } + }; + +bool convert(OUStringBuffer const & in, OStringBuffer * append) { + assert(append != nullptr); + for (sal_Size nConvert = in.getLength();;) { + auto const oldLen = append->getLength(); + auto n = std::min( + std::max(nConvert, sal_Size(PATH_MAX)), + sal_Size(std::numeric_limits<sal_Int32>::max() - oldLen)); + // approximation of required converted size + auto s = append->appendUninitialized(n); + sal_uInt32 info; + sal_Size converted; + //TODO: context, for reliable treatment of DESTBUFFERTOSMALL: + n = UnicodeToTextConverter_Impl::getInstance().convert( + in.getStr() + in.getLength() - nConvert, nConvert, s, n, + (RTL_UNICODETOTEXT_FLAGS_UNDEFINED_ERROR | RTL_UNICODETOTEXT_FLAGS_INVALID_ERROR + | RTL_UNICODETOTEXT_FLAGS_FLUSH), + &info, &converted); + if ((info & RTL_UNICODETOTEXT_INFO_ERROR) != 0) { + return false; + } + append->setLength(oldLen + n); + assert(converted <= nConvert); + nConvert -= converted; + assert((nConvert == 0) == ((info & RTL_UNICODETOTEXT_INFO_DESTBUFFERTOSMALL) == 0)); + if ((info & RTL_UNICODETOTEXT_INFO_DESTBUFFERTOSMALL) == 0) { + break; + } + } + return true; +} + +bool decodeFromUtf8(std::u16string_view text, OString * result) { + assert(result != nullptr); + auto p = text.data(); + auto const end = p + text.size(); + OUStringBuffer ubuf(static_cast<int>(text.size())); + OStringBuffer bbuf(PATH_MAX); + while (p < end) { + rtl::uri::detail::EscapeType t; + sal_uInt32 c = rtl::uri::detail::readUcs4(&p, end, true, RTL_TEXTENCODING_UTF8, &t); + switch (t) { + case rtl::uri::detail::EscapeNo: + if (c == '%') { + return false; + } + [[fallthrough]]; + case rtl::uri::detail::EscapeChar: + if (rtl::isSurrogate(c)) { + return false; + } + ubuf.appendUtf32(c); + break; + case rtl::uri::detail::EscapeOctet: + if (!convert(ubuf, &bbuf)) { + return false; + } + ubuf.setLength(0); + assert(c <= 0xFF); + bbuf.append(char(c)); + break; + } + } + if (!convert(ubuf, &bbuf)) { + return false; + } + *result = bbuf.makeStringAndClear(); + return true; +} + +template<typename T> oslFileError getSystemPathFromFileUrl( + OUString const & url, T * path, bool resolveHome) +{ + assert(path != nullptr); + // For compatibility with assumptions in other parts of the code base, + // assume that anything starting with a slash is a system path instead of a + // (relative) file URL (except if it starts with two slashes, in which case + // it is a relative URL with an authority component): + if (url.isEmpty() + || (url[0] == '/' && (url.getLength() == 1 || url[1] != '/'))) + { + return osl_File_E_INVAL; + } + // Check for non file scheme: + sal_Int32 i = 0; + if (rtl::isAsciiAlpha(url[0])) { + for (sal_Int32 j = 1; j != url.getLength(); ++j) { + auto c = url[j]; + if (c == ':') { + if (rtl_ustr_ascii_compareIgnoreAsciiCase_WithLengths( + url.pData->buffer, j, + RTL_CONSTASCII_STRINGPARAM("file")) + != 0) + { + return osl_File_E_INVAL; + } + i = j + 1; + break; + } + if (!rtl::isAsciiAlphanumeric(c) && c != '+' && c != '-' + && c != '.') + { + break; + } + } + } + // Handle query or fragment: + if (url.indexOf('?', i) != -1 || url.indexOf('#', i) != -1) + return osl_File_E_INVAL; + // Handle authority, supporting a host of "localhost", "127.0.0.1", or the exact value (e.g., + // not supporting an additional final dot, for simplicity) reported by osl_getLocalHostname + // (and, in each case, ignoring case of ASCII letters): + if (url.getLength() - i >= 2 && url[i] == '/' && url[i + 1] == '/') + { + i += 2; + sal_Int32 j = url.indexOf('/', i); + if (j == -1) + j = url.getLength(); + if (j != i + && (rtl_ustr_ascii_compareIgnoreAsciiCase_WithLengths( + url.pData->buffer + i, j - i, + RTL_CONSTASCII_STRINGPARAM("localhost")) + != 0) + && (rtl_ustr_ascii_compareIgnoreAsciiCase_WithLengths( + url.pData->buffer + i, j - i, + RTL_CONSTASCII_STRINGPARAM("127.0.0.1")) + != 0)) + { + OUString hostname; + if (osl_getLocalHostname(&hostname.pData) != osl_Socket_Ok + || (rtl_ustr_compareIgnoreAsciiCase_WithLength( + url.pData->buffer + i, j - i, hostname.getStr(), hostname.getLength()) + != 0)) + { + return osl_File_E_INVAL; + } + } + i = j; + } + // Handle empty path: + if (i == url.getLength()) + { + *path = "/"; + return osl_File_E_None; + } + // Path must not contain %2F: + if (url.indexOf("%2F", i) != -1 || url.indexOf("%2f", i) != -1) + return osl_File_E_INVAL; + + if constexpr (std::is_same_v<T, rtl::OString>) { + if (!decodeFromUtf8(std::u16string_view(url).substr(i), path)) { + return osl_File_E_INVAL; + } + } else if constexpr (std::is_same_v<T, rtl::OUString>) { + *path = rtl::Uri::decode( + url.copy(i), rtl_UriDecodeWithCharset, RTL_TEXTENCODING_UTF8); + } else { + static_assert(std::is_same_v<T, rtl::OString> || std::is_same_v<T, rtl::OUString>); + } + // Path must not contain %2F: + if (path->indexOf('\0') != -1) + return osl_File_E_INVAL; + + // Handle ~ notation: + if (resolveHome && path->getLength() >= 2 && (*path)[1] == '~') + { + sal_Int32 j = path->indexOf('/', 2); + if (j == -1) + j = path->getLength(); + + if (j == 2) + { + OUString home; + if (!osl::Security().getHomeDir(home)) + { + SAL_WARN("sal.file", "osl::Security::getHomeDir failed"); + return osl_File_E_INVAL; + } + + i = url.indexOf('/', i + 1); + + if (i == -1) + i = url.getLength(); + else + ++i; + + //TODO: cheesy way of ensuring home's path ends in slash: + if (!home.isEmpty() && home[home.getLength() - 1] != '/') + home += "/"; + try + { + home = rtl::Uri::convertRelToAbs(home, url.copy(i)); + } + catch (rtl::MalformedUriException & e) + { + SAL_WARN("sal.file", "rtl::MalformedUriException " << e.getMessage()); + return osl_File_E_INVAL; + } + return getSystemPathFromFileUrl(home, path, false); + } + // FIXME: replace ~user with user's home directory + return osl_File_E_INVAL; + } + return osl_File_E_None; +} + +} + +oslFileError SAL_CALL osl_getSystemPathFromFileURL( rtl_uString *ustrFileURL, rtl_uString **pustrSystemPath ) +{ + OUString path; + oslFileError e; + try + { + e = getSystemPathFromFileUrl( + OUString::unacquired(&ustrFileURL), &path, true); + } + catch (std::length_error &) + { + e = osl_File_E_RANGE; + } + + if (e == osl_File_E_None) + rtl_uString_assign(pustrSystemPath, path.pData); + + return e; +} + +oslFileError SAL_CALL osl_getFileURLFromSystemPath( rtl_uString *ustrSystemPath, rtl_uString **pustrFileURL ) +{ + static const sal_Unicode pDoubleSlash[2] = { '/', '/' }; + + rtl_uString *pTmp = nullptr; + sal_Int32 nIndex; + + if( ustrSystemPath->length == 0 ) + return osl_File_E_INVAL; + + /* temporary hack: if already file url, return ustrSystemPath */ + + if( rtl_ustr_ascii_shortenedCompare_WithLength( ustrSystemPath->buffer, ustrSystemPath->length,"file:", 5 ) == 0 ) + return osl_File_E_INVAL; + + /* check if system path starts with ~ or ~user and replace it with the appropriate home dir */ + if( ustrSystemPath->buffer[0] == '~' ) + { + /* check if another user is specified */ + if( ( ustrSystemPath->length == 1 ) || + ( ustrSystemPath->buffer[1] == '/' ) ) + { + /* osl_getHomeDir returns file URL */ + oslSecurity pSecurity = osl_getCurrentSecurity(); + osl_getHomeDir( pSecurity , &pTmp ); + osl_freeSecurityHandle( pSecurity ); + + if (!pTmp) + return osl_File_E_INVAL; + + /* remove "file://" prefix */ + rtl_uString_newFromStr_WithLength( &pTmp, pTmp->buffer + 7, pTmp->length - 7 ); + + /* replace '~' in original string */ + rtl_uString_newReplaceStrAt( &pTmp, ustrSystemPath, 0, 1, pTmp ); + } + else + { + /* FIXME: replace ~user with users home directory */ + return osl_File_E_INVAL; + } + } + + /* check if initial string contains repeated '/' characters */ + nIndex = rtl_ustr_indexOfStr_WithLength( ustrSystemPath->buffer, ustrSystemPath->length, pDoubleSlash, 2 ); + if( nIndex != -1 ) + { + sal_Int32 nSrcIndex; + sal_Int32 nDeleted = 0; + + /* if pTmp is not already allocated, copy ustrSystemPath for modification */ + if( pTmp == nullptr ) + rtl_uString_newFromString( &pTmp, ustrSystemPath ); + + /* adapt index to pTmp */ + nIndex += pTmp->length - ustrSystemPath->length; + + /* replace repeated '/' characters with a single '/' */ + for( nSrcIndex = nIndex + 1; nSrcIndex < pTmp->length; nSrcIndex++ ) + { + if( (pTmp->buffer[nSrcIndex] == '/') && (pTmp->buffer[nIndex] == '/') ) + nDeleted++; + else + pTmp->buffer[++nIndex] = pTmp->buffer[nSrcIndex]; + } + + /* adjust length member */ + pTmp->length -= nDeleted; + } + + if( pTmp == nullptr ) + rtl_uString_assign( &pTmp, ustrSystemPath ); + + /* file URLs must be URI encoded */ + rtl_uriEncode( pTmp, uriCharClass, rtl_UriEncodeIgnoreEscapes, RTL_TEXTENCODING_UTF8, pustrFileURL ); + + rtl_uString_release( pTmp ); + + /* absolute urls should start with 'file://' */ + if( (*pustrFileURL)->buffer[0] == '/' ) + { + rtl_uString *pProtocol = nullptr; + + rtl_uString_newFromAscii( &pProtocol, "file://" ); + rtl_uString_newConcat( pustrFileURL, pProtocol, *pustrFileURL ); + rtl_uString_release( pProtocol ); + } + + return osl_File_E_None; +} + +/* + * relative URLs are not accepted + */ +oslFileError osl_getSystemPathFromFileURL_Ex( + rtl_uString *ustrFileURL, rtl_uString **pustrSystemPath) +{ + rtl_uString* temp = nullptr; + oslFileError osl_error = osl_getSystemPathFromFileURL(ustrFileURL, &temp); + + if (osl_error == osl_File_E_None) + { + if (temp->buffer[0] == '/') + { + *pustrSystemPath = temp; + } + else + { + rtl_uString_release(temp); + osl_error = osl_File_E_INVAL; + } + } + + return osl_error; +} + +namespace +{ + + /** Helper function, return a pointer to the final '\0' + of a string + */ + + sal_Unicode* ustrtoend(sal_Unicode* pStr) + { + return (pStr + rtl_ustr_getLength(pStr)); + } + + sal_Unicode* ustrchrcat(const sal_Unicode chr, sal_Unicode* d) + { + sal_Unicode* p = ustrtoend(d); + *p++ = chr; + *p = 0; + return d; + } + + bool _islastchr(sal_Unicode* pStr, sal_Unicode Chr) + { + sal_Unicode* p = ustrtoend(pStr); + if (p > pStr) + p--; + return (*p == Chr); + } + + /** + Remove the last part of a path, a path that has + only a '/' or no '/' at all will be returned + unmodified + */ + + sal_Unicode* _rmlastpathtoken(sal_Unicode* aPath) + { + /* we may always skip -2 because we + may at least stand on a '/' but + either there is no other character + before this '/' or it's another + character than the '/' + */ + sal_Unicode* p = ustrtoend(aPath) - 2; + + /* move back to the next path separator + or to the start of the string */ + while ((p > aPath) && (*p != '/')) + p--; + + if (p >= aPath) + { + if (*p == '/') + { + p++; + *p = '\0'; + } + else + { + *p = '\0'; + } + } + + return aPath; + } + + oslFileError _osl_resolvepath( + /*inout*/ sal_Unicode* path, + /*inout*/ bool* failed) + { + oslFileError ferr = osl_File_E_None; + + if (!*failed) + { + char unresolved_path[PATH_MAX]; + if (!UnicodeToText(unresolved_path, sizeof(unresolved_path), path, rtl_ustr_getLength(path))) + return oslTranslateFileError(ENAMETOOLONG); + + char resolved_path[PATH_MAX]; + if (realpath(unresolved_path, resolved_path)) + { + if (!TextToUnicode(resolved_path, strlen(resolved_path), path, PATH_MAX)) + return oslTranslateFileError(ENAMETOOLONG); + } + else + { + if (EACCES == errno || ENOTDIR == errno || ENOENT == errno) + *failed = true; + else + ferr = oslTranslateFileError(errno); + } + } + + return ferr; + } + + /** + Works even with non existing paths. The resulting path must not exceed + PATH_MAX else osl_File_E_NAMETOOLONG is the result + */ + + oslFileError osl_getAbsoluteFileURL_impl_(const OUString& unresolved_path, OUString& resolved_path) + { + /* the given unresolved path must not exceed PATH_MAX */ + if (unresolved_path.getLength() >= (PATH_MAX - 2)) + return oslTranslateFileError(ENAMETOOLONG); + + sal_Unicode path_resolved_so_far[PATH_MAX]; + const sal_Unicode* punresolved = unresolved_path.getStr(); + sal_Unicode* presolvedsf = path_resolved_so_far; + + /* reserve space for leading '/' and trailing '\0' + do not exceed this limit */ + sal_Unicode* sentinel = path_resolved_so_far + PATH_MAX - 2; + + /* if realpath fails with error ENOTDIR, EACCES or ENOENT + we will not call it again, because _osl_realpath should also + work with non existing directories etc. */ + bool realpath_failed = false; + oslFileError ferr; + + path_resolved_so_far[0] = '\0'; + + while (*punresolved != '\0') + { + /* ignore '/.' , skip one part back when '/..' */ + if ((*punresolved == '.') && (*presolvedsf == '/')) + { + if (*(punresolved + 1) == '\0') + { + punresolved++; + continue; + } + if (*(punresolved + 1) == '/') + { + punresolved += 2; + continue; + } + if ((*(punresolved + 1) == '.') && (*(punresolved + 2) == '\0' || (*(punresolved + 2) == '/'))) + { + _rmlastpathtoken(path_resolved_so_far); + + presolvedsf = ustrtoend(path_resolved_so_far) - 1; + + if (*(punresolved + 2) == '/') + punresolved += 3; + else + punresolved += 2; + + continue; + } + + /* a file or directory name may start with '.' */ + if ((presolvedsf = ustrtoend(path_resolved_so_far)) > sentinel) + return oslTranslateFileError(ENAMETOOLONG); + + ustrchrcat(*punresolved++, path_resolved_so_far); + + if (*punresolved == '\0' && !realpath_failed) + { + ferr = _osl_resolvepath( + path_resolved_so_far, + &realpath_failed); + + if (ferr != osl_File_E_None) + return ferr; + } + } + else if (*punresolved == '/') + { + if ((presolvedsf = ustrtoend(path_resolved_so_far)) > sentinel) + return oslTranslateFileError(ENAMETOOLONG); + + ustrchrcat(*punresolved++, path_resolved_so_far); + + if (!realpath_failed) + { + ferr = _osl_resolvepath( + path_resolved_so_far, + &realpath_failed); + + if (ferr != osl_File_E_None) + return ferr; + + if (!_islastchr(path_resolved_so_far, '/')) + { + if ((presolvedsf = ustrtoend(path_resolved_so_far)) > sentinel) + return oslTranslateFileError(ENAMETOOLONG); + + ustrchrcat('/', path_resolved_so_far); + } + } + } + else // any other character + { + if ((presolvedsf = ustrtoend(path_resolved_so_far)) > sentinel) + return oslTranslateFileError(ENAMETOOLONG); + + ustrchrcat(*punresolved++, path_resolved_so_far); + + if (*punresolved == '\0' && !realpath_failed) + { + ferr = _osl_resolvepath( + path_resolved_so_far, + &realpath_failed); + + if (ferr != osl_File_E_None) + return ferr; + } + } + } + + sal_Int32 len = rtl_ustr_getLength(path_resolved_so_far); + + OSL_ASSERT(len < PATH_MAX); + + resolved_path = OUString(path_resolved_so_far, len); + + return osl_File_E_None; + } + +} + +oslFileError osl_getAbsoluteFileURL( + rtl_uString* ustrBaseDirURL, + rtl_uString* ustrRelativeURL, + rtl_uString** pustrAbsoluteURL) +{ + /* Work around the below call to getSystemPathFromFileURL rejecting input + that starts with "/" (for whatever reason it behaves that way; but + changing that would start to break lots of tests at least) */ + OUString relUrl(ustrRelativeURL); + if (relUrl.startsWith("//")) + relUrl = "file:" + relUrl; + else if (relUrl.startsWith("/")) + relUrl = "file://" + relUrl; + + OUString unresolved_path; + + FileBase::RC frc = FileBase::getSystemPathFromFileURL(relUrl, unresolved_path); + if (frc != FileBase::E_None) + return oslFileError(frc); + + if (systemPathIsRelativePath(unresolved_path)) + { + OUString base_path; + oslFileError rc = osl_getSystemPathFromFileURL_Ex(ustrBaseDirURL, &base_path.pData); + if (rc != osl_File_E_None) + return rc; + + unresolved_path = systemPathMakeAbsolutePath(base_path, unresolved_path); + } + + OUString resolved_path; + oslFileError rc = osl_getAbsoluteFileURL_impl_(unresolved_path, resolved_path); + if (rc == osl_File_E_None) + { + rc = osl_getFileURLFromSystemPath(resolved_path.pData, pustrAbsoluteURL); + OSL_ASSERT(osl_File_E_None == rc); + } + + return rc; +} + +namespace osl::detail { + + /** + No separate error code if unicode to text conversion or getenv fails because for the + caller there is no difference why a file could not be found in $PATH + */ + bool find_in_PATH(const OUString& file_path, OUString& result) + { + bool bfound = false; + OUString path("PATH"); + OUString env_path; + + if (osl_getEnvironment(path.pData, &env_path.pData) == osl_Process_E_None) + bfound = osl::searchPath(file_path, env_path, result); + + return bfound; + } +} + +namespace +{ + /** + No separate error code if unicode to text conversion or getcwd fails because for the + caller there is no difference why a file could not be found in CDW + */ + bool find_in_CWD(const OUString& file_path, OUString& result) + { + bool bfound = false; + OUString cwd_url; + + if (osl_getProcessWorkingDir(&cwd_url.pData) == osl_Process_E_None) + { + OUString cwd; + FileBase::getSystemPathFromFileURL(cwd_url, cwd); + bfound = osl::searchPath(file_path, cwd, result); + } + return bfound; + } + + bool find_in_searchPath(const OUString& file_path, rtl_uString* search_path, OUString& result) + { + return (search_path && osl::searchPath(file_path, OUString(search_path), result)); + } + +} + +oslFileError osl_searchFileURL(rtl_uString* ustrFilePath, rtl_uString* ustrSearchPath, rtl_uString** pustrURL) +{ + OSL_PRECOND(ustrFilePath && pustrURL, "osl_searchFileURL: invalid parameter"); + + FileBase::RC rc; + OUString file_path; + + // try to interpret search path as file url else assume it's a system path list + rc = FileBase::getSystemPathFromFileURL(ustrFilePath, file_path); + if (rc == FileBase::E_INVAL) + file_path = ustrFilePath; + else if (rc != FileBase::E_None) + return oslFileError(rc); + + bool bfound = false; + OUString result; + + if (find_in_searchPath(file_path, ustrSearchPath, result) || + osl::detail::find_in_PATH(file_path, result) || + find_in_CWD(file_path, result)) + { + OUString resolved; + + if (osl::realpath(result, resolved)) + { + oslFileError osl_error = osl_getFileURLFromSystemPath(resolved.pData, pustrURL); + SAL_WARN_IF(osl_File_E_None != osl_error, "sal.file", "osl_getFileURLFromSystemPath failed"); + bfound = true; + } + } + return bfound ? osl_File_E_None : osl_File_E_NOENT; +} + +oslFileError FileURLToPath(char * buffer, size_t bufLen, rtl_uString* ustrFileURL) +{ + OString strSystemPath; + oslFileError osl_error = osl::detail::convertUrlToPathname( + OUString::unacquired(&ustrFileURL), &strSystemPath); + + if(osl_error != osl_File_E_None) + return osl_error; + + osl_systemPathRemoveSeparator(strSystemPath.pData); + + if (o3tl::make_unsigned(strSystemPath.getLength()) >= bufLen) { + return osl_File_E_OVERFLOW; + } + std::strcpy(buffer, strSystemPath.getStr()); + + return osl_error; +} + +int UnicodeToText( char * buffer, size_t bufLen, const sal_Unicode * uniText, sal_Int32 uniTextLen ) +{ + sal_uInt32 nInfo = 0; + sal_Size nSrcChars = 0; + + sal_Size nDestBytes = UnicodeToTextConverter_Impl::getInstance().convert ( + uniText, uniTextLen, buffer, bufLen, + OUSTRING_TO_OSTRING_CVTFLAGS | RTL_UNICODETOTEXT_FLAGS_FLUSH, &nInfo, &nSrcChars); + + if( nInfo & RTL_UNICODETOTEXT_INFO_DESTBUFFERTOSMALL ) + { + errno = EOVERFLOW; + return 0; + } + + /* ensure trailing '\0' */ + buffer[nDestBytes] = '\0'; + return nDestBytes; +} + +namespace +{ + class TextToUnicodeConverter_Impl + { + rtl_TextToUnicodeConverter m_converter; + + TextToUnicodeConverter_Impl() + : m_converter (rtl_createTextToUnicodeConverter (osl_getThreadTextEncoding())) + {} + + ~TextToUnicodeConverter_Impl() + { + rtl_destroyTextToUnicodeConverter (m_converter); + } + + public: + static TextToUnicodeConverter_Impl & getInstance() + { + static TextToUnicodeConverter_Impl g_theConverter; + return g_theConverter; + } + + sal_Size convert( + char const * pSrcBuf, sal_Size nSrcBytes, sal_Unicode * pDstBuf, sal_Size nDstChars, + sal_uInt32 nFlags, sal_uInt32 * pInfo, sal_Size * pSrcCvtBytes) + { + OSL_ASSERT(m_converter != nullptr); + return rtl_convertTextToUnicode ( + m_converter, nullptr, pSrcBuf, nSrcBytes, pDstBuf, nDstChars, nFlags, pInfo, pSrcCvtBytes); + } + }; +} + +int TextToUnicode( + const char* text, + size_t text_buffer_size, + sal_Unicode* unic_text, + sal_Int32 unic_text_buffer_size) +{ + sal_uInt32 nInfo = 0; + sal_Size nSrcChars = 0; + + sal_Size nDestBytes = TextToUnicodeConverter_Impl::getInstance().convert( + text, text_buffer_size, unic_text, unic_text_buffer_size, + OSTRING_TO_OUSTRING_CVTFLAGS | RTL_TEXTTOUNICODE_FLAGS_FLUSH, &nInfo, &nSrcChars); + + if (nInfo & RTL_TEXTTOUNICODE_INFO_DESTBUFFERTOOSMALL) + { + errno = EOVERFLOW; + return 0; + } + + /* ensure trailing '\0' */ + unic_text[nDestBytes] = '\0'; + return nDestBytes; +} + +oslFileError osl::detail::convertUrlToPathname(OUString const & url, OString * pathname) { + assert(pathname != nullptr); + oslFileError e; + try { + e = getSystemPathFromFileUrl(url, pathname, true); + } catch (std::length_error &) { + e = osl_File_E_RANGE; + } + if (e == osl_File_E_None && !pathname->startsWith("/")) { + e = osl_File_E_INVAL; + } + return e; +} + +oslFileError osl::detail::convertPathnameToUrl(OString const & pathname, OUString * url) { + assert(url != nullptr); + OUStringBuffer buf(10+pathname.getLength()); + buf.append("file:"); + if (pathname.startsWith("/")) { + buf.append("//"); + // so if pathname should ever start with "//" that isn't mistaken for an authority + // component + } + for (sal_Size convert = pathname.getLength();;) { + auto n = std::max(convert, sal_Size(PATH_MAX)); // approximation of required converted size + OUStringBuffer ubuf(static_cast<int>(n)); + auto s = ubuf.appendUninitialized(n); + sal_uInt32 info; + sal_Size converted; + //TODO: context, for reliable treatment of DESTBUFFERTOSMALL: + n = TextToUnicodeConverter_Impl::getInstance().convert( + pathname.getStr() + pathname.getLength() - convert, convert, s, n, + (RTL_TEXTTOUNICODE_FLAGS_UNDEFINED_ERROR | RTL_TEXTTOUNICODE_FLAGS_MBUNDEFINED_ERROR + | RTL_TEXTTOUNICODE_FLAGS_INVALID_ERROR | RTL_TEXTTOUNICODE_FLAGS_FLUSH), + &info, &converted); + ubuf.setLength(n); + buf.append( + rtl::Uri::encode( + ubuf.makeStringAndClear(), uriCharClass, rtl_UriEncodeIgnoreEscapes, + RTL_TEXTENCODING_UTF8)); + assert(converted <= convert); + convert -= converted; + if ((info & RTL_TEXTTOUNICODE_INFO_ERROR) != 0) { + assert(convert > 0); + //TODO: see writeEscapeOctet in sal/rtl/uri.cxx + buf.append("%"); + unsigned char c = pathname[pathname.getLength() - convert]; + assert(c >= 0x80); + static sal_Unicode const aHex[16] + = { 0x30, 0x31, 0x32, 0x33, 0x34, 0x35, 0x36, 0x37, 0x38, 0x39, + 0x41, 0x42, 0x43, 0x44, 0x45, 0x46 }; /* '0'--'9', 'A'--'F' */ + buf.append(aHex[c >> 4]); + buf.append(aHex[c & 15]); + --convert; + continue; + } + assert((convert == 0) == ((info & RTL_TEXTTOUNICODE_INFO_DESTBUFFERTOOSMALL) == 0)); + if ((info & RTL_TEXTTOUNICODE_INFO_DESTBUFFERTOOSMALL) == 0) { + break; + } + } + *url = buf.makeStringAndClear(); + return osl_File_E_None; +} + +/* vim:set shiftwidth=4 softtabstop=4 expandtab: */ |