1 | /* $Id: parse-bourne-shell-like.cpp 106061 2024-09-16 14:03:52Z vboxsync $ */
|
---|
2 | /** @file
|
---|
3 | * IPRT - RTParseShell - Bourne-like Shell Parser.
|
---|
4 | */
|
---|
5 |
|
---|
6 | /*
|
---|
7 | * Copyright (C) 2022-2024 Oracle and/or its affiliates.
|
---|
8 | *
|
---|
9 | * This file is part of VirtualBox base platform packages, as
|
---|
10 | * available from https://www.virtualbox.org.
|
---|
11 | *
|
---|
12 | * This program is free software; you can redistribute it and/or
|
---|
13 | * modify it under the terms of the GNU General Public License
|
---|
14 | * as published by the Free Software Foundation, in version 3 of the
|
---|
15 | * License.
|
---|
16 | *
|
---|
17 | * This program is distributed in the hope that it will be useful, but
|
---|
18 | * WITHOUT ANY WARRANTY; without even the implied warranty of
|
---|
19 | * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
|
---|
20 | * General Public License for more details.
|
---|
21 | *
|
---|
22 | * You should have received a copy of the GNU General Public License
|
---|
23 | * along with this program; if not, see <https://www.gnu.org/licenses>.
|
---|
24 | *
|
---|
25 | * The contents of this file may alternatively be used under the terms
|
---|
26 | * of the Common Development and Distribution License Version 1.0
|
---|
27 | * (CDDL), a copy of it is provided in the "COPYING.CDDL" file included
|
---|
28 | * in the VirtualBox distribution, in which case the provisions of the
|
---|
29 | * CDDL are applicable instead of those of the GPL.
|
---|
30 | *
|
---|
31 | * You may elect to license modified versions of this file under the
|
---|
32 | * terms and conditions of either the GPL or the CDDL or both.
|
---|
33 | *
|
---|
34 | * SPDX-License-Identifier: GPL-3.0-only OR CDDL-1.0
|
---|
35 | */
|
---|
36 |
|
---|
37 |
|
---|
38 | /*********************************************************************************************************************************
|
---|
39 | * Header Files *
|
---|
40 | *********************************************************************************************************************************/
|
---|
41 | #include "internal/iprt.h"
|
---|
42 | #include <iprt/parseshell.h>
|
---|
43 |
|
---|
44 | #include <iprt/file.h>
|
---|
45 | #include <iprt/mem.h>
|
---|
46 | #include <iprt/string.h>
|
---|
47 |
|
---|
48 |
|
---|
49 | /*********************************************************************************************************************************
|
---|
50 | * Structures and Typedefs *
|
---|
51 | *********************************************************************************************************************************/
|
---|
52 | typedef struct RTPARSESRCFILE
|
---|
53 | {
|
---|
54 | /** For looking up by the filename string (case sensitive). */
|
---|
55 | RTSTRSPACECORE ByName;
|
---|
56 | /** For looking up by the file ID. */
|
---|
57 | AVLU32NODECORE ById;
|
---|
58 | /** The filename. */
|
---|
59 | char szFilename[RT_FLEXIBLE_ARRAY];
|
---|
60 | } RTPARSESRCFILE;
|
---|
61 | typedef RTPARSESRCFILE *PRTPARSESRCFILE;
|
---|
62 |
|
---|
63 | typedef struct RTPARSESRCMGR
|
---|
64 | {
|
---|
65 | /** String space managing the filenames. */
|
---|
66 | RTSTRSPACE hStrSpace;
|
---|
67 | /** By ID AVL tree root. */
|
---|
68 | AVLU32TREE IdTreeRoot;
|
---|
69 | /** The next filename ID. */
|
---|
70 | uint32_t idNext;
|
---|
71 | } RTPARSESRCMGR;
|
---|
72 | typedef RTPARSESRCMGR *PRTPARSESRCMGR;
|
---|
73 |
|
---|
74 |
|
---|
75 |
|
---|
76 | /**
|
---|
77 | * Instance data for a bourne-like shell parser.
|
---|
78 | */
|
---|
79 | typedef struct RTPARSESHELLINT
|
---|
80 | {
|
---|
81 | /** Magic value (RTPARSESHELLINT_MAGIC). */
|
---|
82 | uint32_t uMagic;
|
---|
83 | /** Reference count. */
|
---|
84 | uint32_t volatile cRefs;
|
---|
85 | /** Parser flags, RTPARSESHELL_F_XXX. */
|
---|
86 | uint64_t fFlags;
|
---|
87 | /** Source name manager. */
|
---|
88 | RTPARSESRCMGR SrcMgr;
|
---|
89 | /** Parser instance name (for logging/whatever). */
|
---|
90 | char szName[RT_FLEXIBLE_ARRAY];
|
---|
91 | } RTPARSESHELLINT;
|
---|
92 |
|
---|
93 | /** Magic value for RTPARSESHELLINT. */
|
---|
94 | #define RTPARSESHELLINT_MAGIC UINT32_C(0x18937566)
|
---|
95 |
|
---|
96 |
|
---|
97 | typedef enum RTPARSESHELLTOKEN
|
---|
98 | {
|
---|
99 | RTPARSESHELLTOKEN_INVALID = 0,
|
---|
100 | RTPARSESHELLTOKEN_EOF,
|
---|
101 | RTPARSESHELLTOKEN_NEWLINE, /**< Newline (\n) */
|
---|
102 | RTPARSESHELLTOKEN_NOT, /**< ! cmd */
|
---|
103 | RTPARSESHELLTOKEN_BACKGROUND, /**< cmd & */
|
---|
104 | RTPARSESHELLTOKEN_AND, /**< cmd1 && cmd2 */
|
---|
105 | RTPARSESHELLTOKEN_PIPE, /**< cmd1 | cmd2 */
|
---|
106 | RTPARSESHELLTOKEN_OR, /**< cmd1 || cmd2 */
|
---|
107 | RTPARSESHELLTOKEN_SEMICOLON, /**< cmd ; */
|
---|
108 | RTPARSESHELLTOKEN_END_CASE /**< pattern) cmd ;; */
|
---|
109 | };
|
---|
110 |
|
---|
111 | /**
|
---|
112 | * The parser state.
|
---|
113 | */
|
---|
114 | typedef struct RTPARSESHELLSTATE
|
---|
115 | {
|
---|
116 | /** Pointer to the buffer of bytes to parse. */
|
---|
117 | const char *pchSrc;
|
---|
118 | /** Number of characters (bytes) to parse starting with pchSrc. */
|
---|
119 | size_t cchSrc;
|
---|
120 | /** Current source offset. */
|
---|
121 | size_t offSrc;
|
---|
122 | /** Offset relative to pchSrc of the start of the current line. */
|
---|
123 | size_t offSrcLineStart;
|
---|
124 | /** Current line number. */
|
---|
125 | uint32_t iLine;
|
---|
126 | /** The source file ID. */
|
---|
127 | uint32_t idSrcFile;
|
---|
128 | /** The parser instance. */
|
---|
129 | PRTPARSESHELLSTATE pParser;
|
---|
130 | } RTPARSESHELLSTATE;
|
---|
131 | typedef RTPARSESHELLSTATE *PRTPARSESHELLSTATE;
|
---|
132 |
|
---|
133 |
|
---|
134 |
|
---|
135 | static void rtParseSrcMgrInit(PRTPARSESRCMGR pSrcMgr)
|
---|
136 | {
|
---|
137 | pSrcMgr->hStrSpace = NULL;
|
---|
138 | pSrcMgr->IdTreeRoot = NULL;
|
---|
139 | pSrcMgr->idNext = 1;
|
---|
140 | }
|
---|
141 |
|
---|
142 |
|
---|
143 | /**
|
---|
144 | * @callback_method_impl{AVLU32CALLBACK, Worker for rtParseSrcMgrDelete.}
|
---|
145 | */
|
---|
146 | static DECLCALLBACK(int) AVLU32CALLBACK,(PAVLU32NODECORE pNode, void *pvUser)
|
---|
147 | {
|
---|
148 | RT_NOREF(pvUser);
|
---|
149 | PRTPARSESRCFILE pSrcFile = RT_FROM_MEMBER(pNode, RTPARSESRCFILE, ById);
|
---|
150 | RTMemFree(pSrcFile);
|
---|
151 | }
|
---|
152 |
|
---|
153 |
|
---|
154 | static void rtParseSrcMgrDelete(PRTPARSESRCMGR pSrcMgr)
|
---|
155 | {
|
---|
156 | RTAvlU32Destroy(pSrcMgr->IdTreeRoot, rtParseSrcMgrDeleteSrcFile, NULL);
|
---|
157 | pSrcMgr->hStrSpace = NULL;
|
---|
158 | pSrcMgr->IdTreeRoot = NULL;
|
---|
159 | pSrcMgr->idNext = 1;
|
---|
160 | }
|
---|
161 |
|
---|
162 |
|
---|
163 | static int32_t rtParseSrcMgrFilenameToIdx(PRTPARSESRCMGR pSrcMgr, const char *pszFilename)
|
---|
164 | {
|
---|
165 | /*
|
---|
166 | * Look it up.
|
---|
167 | */
|
---|
168 | RTSTRSPACECORE pStrCore = RTStrSpaceGet(pSrcMgr->hStrSpace, pszFilename);
|
---|
169 | if (pStrCore)
|
---|
170 | {
|
---|
171 | PRTPARSESRCFILE pSrcFile = RT_FROM_MEMBER(pNode, RTPARSESRCFILE, ByName);
|
---|
172 | AssertMsg(pSrcFile->ById.Key > 0 && pSrcFile->ById.Key < UINT16_MAX, ("%#x\n", pSrcFile->ById.Key));
|
---|
173 | return pSrcFile->ById.Key;
|
---|
174 | }
|
---|
175 |
|
---|
176 | /*
|
---|
177 | * Add it.
|
---|
178 | */
|
---|
179 | uint32_t const idFile = pSrcMgr->idNext;
|
---|
180 | AssertReturn(idFile < UINT16_MAX, VERR_TOO_MANY_OPEN_FILES);
|
---|
181 |
|
---|
182 | size_t const cchFilename = strlen(pszFilename);
|
---|
183 | PRTPARSESRCFILE pSrcFile = (PRTPARSESRCFILE)RTMemAllocZVar(RT_UOFFSETOF_DYN(RTPARSESRCFILE, szFilename[cchFilename + 1]));
|
---|
184 | if (pSrcFile)
|
---|
185 | {
|
---|
186 | memcpy(&pSrcFile->szFilename[0], pszFilename, cchFilename);
|
---|
187 | if (RTStrSpaceInsert(pSrcMgr->hStrSpace, &pSrcFile->ByName))
|
---|
188 | {
|
---|
189 | pSrcFile->ById.Key = idFile;
|
---|
190 | if (RTAvlU32Insert(&pSrcMgr->IdTreeRoot, &pSrcFile->ById))
|
---|
191 | {
|
---|
192 | pSrcMgr->idNext = idFile + 1;
|
---|
193 | return (int32_t)idFile;
|
---|
194 | }
|
---|
195 | AssertFailed();
|
---|
196 | RTStrSpaceRemove(pSrcMgr->hStrSpace, &pSrcFile->szFilename);
|
---|
197 | }
|
---|
198 | else
|
---|
199 | AssertFailed();
|
---|
200 | RTMemFree(pSrcFile);
|
---|
201 | return VERR_INTERNAL_ERROR_2;
|
---|
202 | }
|
---|
203 | return VERR_NO_MEMORY;
|
---|
204 | }
|
---|
205 |
|
---|
206 |
|
---|
207 | RTDECL(int) RTParseShellCreate(PRTPARSESHELL phParser, uint64_t fFlags, const char *pszName)
|
---|
208 | {
|
---|
209 | /*
|
---|
210 | * Validate input.
|
---|
211 | */
|
---|
212 | AssertPtrReturn(phParser, VERR_INVALID_POINTER);
|
---|
213 | *phParser = NIL_RTPARSESHELL;
|
---|
214 |
|
---|
215 | AssertPtrReturn(!(fFlags & ~0), VERR_INVALID_FLAGS);
|
---|
216 |
|
---|
217 | AssertPtrReturn(pszName, VERR_INVALID_POINTER);
|
---|
218 | size_t const cchName = strlen(pszName);
|
---|
219 | AssertReturn(cchName > 0 && cchName < 1024, VERR_OUT_OF_RANGE);
|
---|
220 |
|
---|
221 | /*
|
---|
222 | * Allocate and initialize the parser instance.
|
---|
223 | */
|
---|
224 | RTPARSESHELLINT *pThis = (RTPARSESHELLINT *)RTMemAllocZVar(RT_UOFFSETOF_DYN(RTPARSESHELLINT, szName[cchName + 1]));
|
---|
225 | AssertReturn(pThis, VERR_NO_MEMORY);
|
---|
226 |
|
---|
227 | pThis->uMagic = RTPARSESHELLINT_MAGIC;
|
---|
228 | pThis->cRefs = 1;
|
---|
229 | pThis->fFlags = fFlags;
|
---|
230 | memcpy(pThis->szName, pszName, cchName);
|
---|
231 |
|
---|
232 | rtParseSrcMgrInit(&pThis->SrcMgr);
|
---|
233 |
|
---|
234 | /*
|
---|
235 | * Return success.
|
---|
236 | */
|
---|
237 | *phParser = pThis;
|
---|
238 | return VINF_SUCCESS;
|
---|
239 | }
|
---|
240 |
|
---|
241 |
|
---|
242 | RTDECL(uint32_t) RTParseShellRelease(RTPARSESHELL hParser)
|
---|
243 | {
|
---|
244 | if (hParser == NIL_RTPARSESHELL)
|
---|
245 | return 0;
|
---|
246 |
|
---|
247 | RTPARSESHELLINT * const pThis = hParser;
|
---|
248 | AssertPtrReturn(pThis, UINT32_MAX);
|
---|
249 | AssertPtrReturn(pThis->uMagic == RTPARSESHELLINT_MAGIC, UINT32_MAX);
|
---|
250 |
|
---|
251 | uint32_t const cRefs = ASMAtomicDecU32(&pThis->cRefs);
|
---|
252 | Assert(cRefs < _4K);
|
---|
253 | if (cRefs == 0)
|
---|
254 | {
|
---|
255 | pThis->uMagic = ~RTPARSESHELLINT_MAGIC;
|
---|
256 | pThis->szName[0] = '~';
|
---|
257 | rtParseSrcMgrDelete(&pThis->SrcMgr);
|
---|
258 | RTMemFree(pThis);
|
---|
259 | }
|
---|
260 | return cRefs;
|
---|
261 | }
|
---|
262 |
|
---|
263 |
|
---|
264 | RTDECL(uint32_t) RTParseShellRetain(RTPARSESHELL hParser)
|
---|
265 | {
|
---|
266 | RTPARSESHELLINT * const pThis = hParser;
|
---|
267 | AssertPtrReturn(pThis, UINT32_MAX);
|
---|
268 | AssertPtrReturn(pThis->uMagic == RTPARSESHELLINT_MAGIC, UINT32_MAX);
|
---|
269 |
|
---|
270 | uint32_t const cRefs = ASMAtomicIncU32(&pThis->cRefs);
|
---|
271 | Assert(cRefs > 1);
|
---|
272 | Assert(cRefs < _4K);
|
---|
273 | return cRefs;
|
---|
274 | }
|
---|
275 |
|
---|
276 |
|
---|
277 | static rtParseShellGetToken(P RTPARSESHELLSTATE pState)
|
---|
278 | {
|
---|
279 |
|
---|
280 | }
|
---|
281 |
|
---|
282 |
|
---|
283 | static int rtParseShellStringInt(RTPARSESHELLINT *pThis, const char *pchString, size_t cchString,
|
---|
284 | const char *pszFilename, uint32_t iLine, uint32_t fFlags,
|
---|
285 | PRTLISTANCHOR pNodeList, PRTERRINFO pErrInfo)
|
---|
286 | {
|
---|
287 | /*
|
---|
288 | * Add the source file to the source manager.
|
---|
289 | */
|
---|
290 | int32_t const idFile = rtParseSrcMgrFilenameToIdx(&pThis->SrcMgr, pszFilename);
|
---|
291 | AssertReturn(idFile > 0, idFile);
|
---|
292 |
|
---|
293 | /*
|
---|
294 | *
|
---|
295 | */
|
---|
296 |
|
---|
297 |
|
---|
298 | }
|
---|
299 |
|
---|
300 |
|
---|
301 | static void rtParseTreeDestroyList(PRTLISTANCHOR pNodeList)
|
---|
302 | {
|
---|
303 | RT_NOREF(pNodeList);
|
---|
304 | }
|
---|
305 |
|
---|
306 |
|
---|
307 |
|
---|
308 | RTDECL(int) RTParseShellString(RTPARSESHELL hParser, const char *pchString, size_t cchString,
|
---|
309 | const char *pszFilename, uint32_t iStartLine, uint32_t fFlags,
|
---|
310 | PRTLISTANCHOR pNodeList, PRTERRINFO pErrInfo)
|
---|
311 | {
|
---|
312 | /*
|
---|
313 | * Validate input.
|
---|
314 | */
|
---|
315 | AssertPtrReturn(pNodeList, VERR_INVALID_POINTER);
|
---|
316 | RTListInit(pNodeList);
|
---|
317 |
|
---|
318 | RTPARSESHELLINT * const pThis = hParser;
|
---|
319 | AssertPtrReturn(pThis, UINT32_MAX);
|
---|
320 | AssertPtrReturn(pThis->uMagic == RTPARSESHELLINT_MAGIC, UINT32_MAX);
|
---|
321 |
|
---|
322 | AssertPtrReturn(pchString, VERR_INVALID_POINTER);
|
---|
323 | AssertPtrReturn(pszFilename, VERR_INVALID_POINTER);
|
---|
324 | AssertPtrReturn(fFlags & ~(UINT32_C(0)), VERR_INVALID_FLAGS);
|
---|
325 |
|
---|
326 | /*
|
---|
327 | * Call common function for doing the parsing.
|
---|
328 | */
|
---|
329 | int rc = rtParseShellStringInt(pThis, pchString, cchString, pszFilename, iStartLine, fFlags, pNodeList, pErrInfo);
|
---|
330 | if (RT_FAILURE(rc))
|
---|
331 | rtParseTreeDestroyList(pNodeList);
|
---|
332 | return rc;
|
---|
333 | }
|
---|
334 |
|
---|
335 |
|
---|
336 | RTDECL(uint32_t) RTParseShellFile(RTPARSESHELL hParser, const char *pszFilename, uint32_t fFlags,
|
---|
337 | PRTLISTANCHOR pNodeList, PRTERRINFO pErrInfo)
|
---|
338 | {
|
---|
339 | /*
|
---|
340 | * Validate input.
|
---|
341 | */
|
---|
342 | AssertPtrReturn(pNodeList, VERR_INVALID_POINTER);
|
---|
343 | RTListInit(pNodeList);
|
---|
344 |
|
---|
345 | RTPARSESHELLINT * const pThis = hParser;
|
---|
346 | AssertPtrReturn(pThis, UINT32_MAX);
|
---|
347 | AssertPtrReturn(pThis->uMagic == RTPARSESHELLINT_MAGIC, UINT32_MAX);
|
---|
348 |
|
---|
349 | AssertPtrReturn(pchString, VERR_INVALID_POINTER);
|
---|
350 | AssertPtrReturn(pszFilename, VERR_INVALID_POINTER);
|
---|
351 | AssertPtrReturn(fFlags & ~(UINT32_C(0)), VERR_INVALID_FLAGS);
|
---|
352 |
|
---|
353 | /*
|
---|
354 | * Read the file into memory and hand it over to the common parser function.
|
---|
355 | */
|
---|
356 | void *pvFile = NULL;
|
---|
357 | size_t cbFile = 0;
|
---|
358 | int rc = RTFileReadAll(pszFilename, &pvFile, &cbFile)
|
---|
359 | if (RT_SUCCESS(rc))
|
---|
360 | {
|
---|
361 | rc = RTStrValidateEncodingEx((char *)pvFile, cbFile, RTSTR_VALIDATE_ENCODING_EXACT_LENGTH);
|
---|
362 | if (RT_SUCCESS(rc))
|
---|
363 | {
|
---|
364 | /*
|
---|
365 | * Call common function for doing the parsing.
|
---|
366 | */
|
---|
367 | rc = rtParseShellStringInt(pThis, pchString, cchString, pszFilename, 1 /*iStartLine*/, fFlags, pNodeList, pErrInfo);
|
---|
368 | if (RT_FAILURE(rc))
|
---|
369 | rtParseTreeDestroyList(pNodeList);
|
---|
370 | }
|
---|
371 | else
|
---|
372 | rc = RTErrInfoSetF(pErrInfo, rc, "RTStrValidateEncodingEx failed on %s: %Rrc", pszFilename, rc);
|
---|
373 | RTFileReadAllFree(pvFile, cbFile);
|
---|
374 | }
|
---|
375 | return rc;
|
---|
376 | }
|
---|
377 |
|
---|