pacman/lib/libalpm/util.c

/*
 *  util.c
 *
 *  Copyright (c) 2006-2016 Pacman Development Team <pacman-dev@archlinux.org>
 *  Copyright (c) 2002-2006 by Judd Vinet <jvinet@zeroflux.org>
 *  Copyright (c) 2005 by Aurelien Foret <orelien@chez.com>
 *  Copyright (c) 2005 by Christian Hamar <krics@linuxforum.hu>
 *  Copyright (c) 2006 by David Kimpe <dnaku@frugalware.org>
 *  Copyright (c) 2005, 2006 by Miklos Vajna <vmiklos@frugalware.org>
 *
 *  This program is free software; you can redistribute it and/or modify
 *  it under the terms of the GNU General Public License as published by
 *  the Free Software Foundation; either version 2 of the License, or
 *  (at your option) any later version.
 *
 *  This program is distributed in the hope that it will be useful,
 *  but WITHOUT ANY WARRANTY; without even the implied warranty of
 *  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
 *  GNU General Public License for more details.
 *
 *  You should have received a copy of the GNU General Public License
 *  along with this program.  If not, see <http://www.gnu.org/licenses/>.
 */

#include <stdlib.h>
#include <unistd.h>
#include <ctype.h>
#include <dirent.h>
#include <time.h>
#include <errno.h>
#include <limits.h>
#include <sys/wait.h>
#include <fnmatch.h>
#include <poll.h>

/* libarchive */
#include <archive.h>
#include <archive_entry.h>

#ifdef HAVE_LIBSSL
#include <openssl/md5.h>
#include <openssl/sha.h>
#else
#include "md5.h"
#include "sha2.h"
#endif

/* libalpm */
#include "util.h"
#include "log.h"
#include "libarchive-compat.h"
#include "alpm.h"
#include "alpm_list.h"
#include "handle.h"
#include "trans.h"

#ifndef HAVE_STRSEP
/** Extracts tokens from a string.
 * Replaces strset which is not portable (missing on Solaris).
 * Copyright (c) 2001 by François Gouget <fgouget_at_codeweavers.com>
 * Modifies str to point to the first character after the token if one is
 * found, or NULL if one is not.
 * @param str string containing delimited tokens to parse
 * @param delim character delimiting tokens in str
 * @return pointer to the first token in str if str is not NULL, NULL if
 * str is NULL
 */
char *strsep(char **str, const char *delims)
{
	char *token;

	if(*str == NULL) {
		/* No more tokens */
		return NULL;
	}

	token = *str;
	while(**str != '\0') {
		if(strchr(delims, **str) != NULL) {
			**str = '\0';
			(*str)++;
			return token;
		}
		(*str)++;
	}
	/* There is no other token */
	*str = NULL;
	return token;
}
#endif

int _alpm_makepath(const char *path)
{
	return _alpm_makepath_mode(path, 0755);
}

/** Creates a directory, including parents if needed, similar to 'mkdir -p'.
 * @param path directory path to create
 * @param mode permission mode for created directories
 * @return 0 on success, 1 on error
 */
int _alpm_makepath_mode(const char *path, mode_t mode)
{
	char *ptr, *str;
	mode_t oldmask;
	int ret = 0;

	STRDUP(str, path, return 1);

	oldmask = umask(0000);

	for(ptr = str; *ptr; ptr++) {
		/* detect mid-path condition and zero length paths */
		if(*ptr != '/' || ptr == str || ptr[-1] == '/') {
			continue;
		}

		/* temporarily mask the end of the path */
		*ptr = '\0';

		if(mkdir(str, mode) < 0 && errno != EEXIST) {
			ret = 1;
			goto done;
		}

		/* restore path separator */
		*ptr = '/';
	}

	/* end of the string. add the full path. It will already exist when the path
	 * passed in has a trailing slash. */
	if(mkdir(str, mode) < 0 && errno != EEXIST) {
		ret = 1;
	}

done:
	umask(oldmask);
	free(str);
	return ret;
}

/** Copies a file.
 * @param src file path to copy from
 * @param dest file path to copy to
 * @return 0 on success, 1 on error
 */
int _alpm_copyfile(const char *src, const char *dest)
{
	char *buf;
	int in, out, ret = 1;
	ssize_t nread;
	struct stat st;

	MALLOC(buf, (size_t)ALPM_BUFFER_SIZE, return 1);

	OPEN(in, src, O_RDONLY | O_CLOEXEC);
	do {
		out = open(dest, O_WRONLY | O_CREAT | O_BINARY | O_CLOEXEC, 0000);
	} while(out == -1 && errno == EINTR);
	if(in < 0 || out < 0) {
		goto cleanup;
	}

	if(fstat(in, &st) || fchmod(out, st.st_mode)) {
		goto cleanup;
	}

	/* do the actual file copy */
	while((nread = read(in, buf, ALPM_BUFFER_SIZE)) > 0 || errno == EINTR) {
		ssize_t nwrite = 0;
		if(nread < 0) {
			continue;
		}
		do {
			nwrite = write(out, buf + nwrite, nread);
			if(nwrite >= 0) {
				nread -= nwrite;
			} else if(errno != EINTR) {
				goto cleanup;
			}
		} while(nread > 0);
	}
	ret = 0;

cleanup:
	free(buf);
	if(in >= 0) {
		close(in);
	}
	if(out >= 0) {
		close(out);
	}
	return ret;
}

/** Trim trailing newlines from a string (if any exist).
 * @param str a single line of text
 * @param len size of str, if known, else 0
 * @return the length of the trimmed string
 */
size_t _alpm_strip_newline(char *str, size_t len)
{
	if(*str == '\0') {
		return 0;
	}
	if(len == 0) {
		len = strlen(str);
	}
	while(len > 0 && str[len - 1] == '\n') {
		len--;
	}
	str[len] = '\0';

	return len;
}

/* Compression functions */

/** Open an archive for reading and perform the necessary boilerplate.
 * This takes care of creating the libarchive 'archive' struct, setting up
 * compression and format options, opening a file descriptor, setting up the
 * buffer size, and performing a stat on the path once opened.
 * On error, no file descriptor is opened, and the archive pointer returned
 * will be set to NULL.
 * @param handle the context handle
 * @param path the path of the archive to open
 * @param buf space for a stat buffer for the given path
 * @param archive pointer to place the created archive object
 * @param error error code to set on failure to open archive
 * @return -1 on failure, >=0 file descriptor on success
 */
int _alpm_open_archive(alpm_handle_t *handle, const char *path,
		struct stat *buf, struct archive **archive, alpm_errno_t error)
{
	int fd;
	size_t bufsize = ALPM_BUFFER_SIZE;
	errno = 0;

	if((*archive = archive_read_new()) == NULL) {
		RET_ERR(handle, ALPM_ERR_LIBARCHIVE, -1);
	}

	_alpm_archive_read_support_filter_all(*archive);
	archive_read_support_format_all(*archive);

	_alpm_log(handle, ALPM_LOG_DEBUG, "opening archive %s\n", path);
	OPEN(fd, path, O_RDONLY | O_CLOEXEC);
	if(fd < 0) {
		_alpm_log(handle, ALPM_LOG_ERROR,
				_("could not open file %s: %s\n"), path, strerror(errno));
		goto error;
	}

	if(fstat(fd, buf) != 0) {
		_alpm_log(handle, ALPM_LOG_ERROR,
				_("could not stat file %s: %s\n"), path, strerror(errno));
		goto error;
	}
#ifdef HAVE_STRUCT_STAT_ST_BLKSIZE
	if(buf->st_blksize > ALPM_BUFFER_SIZE) {
		bufsize = buf->st_blksize;
	}
#endif

	if(archive_read_open_fd(*archive, fd, bufsize) != ARCHIVE_OK) {
		_alpm_log(handle, ALPM_LOG_ERROR, _("could not open file %s: %s\n"),
				path, archive_error_string(*archive));
		goto error;
	}

	return fd;

error:
	_alpm_archive_read_free(*archive);
	*archive = NULL;
	if(fd >= 0) {
		close(fd);
	}
	RET_ERR(handle, error, -1);
}

/** Unpack a specific file in an archive.
 * @param handle the context handle
 * @param archive the archive to unpack
 * @param prefix where to extract the files
 * @param filename a file within the archive to unpack
 * @return 0 on success, 1 on failure
 */
int _alpm_unpack_single(alpm_handle_t *handle, const char *archive,
		const char *prefix, const char *filename)
{
	alpm_list_t *list = NULL;
	int ret = 0;
	if(filename == NULL) {
		return 1;
	}
	list = alpm_list_add(list, (void *)filename);
	ret = _alpm_unpack(handle, archive, prefix, list, 1);
	alpm_list_free(list);
	return ret;
}

/** Unpack a list of files in an archive.
 * @param handle the context handle
 * @param path the archive to unpack
 * @param prefix where to extract the files
 * @param list a list of files within the archive to unpack or NULL for all
 * @param breakfirst break after the first entry found
 * @return 0 on success, 1 on failure
 */
int _alpm_unpack(alpm_handle_t *handle, const char *path, const char *prefix,
		alpm_list_t *list, int breakfirst)
{
	int ret = 0;
	mode_t oldmask;
	struct archive *archive;
	struct archive_entry *entry;
	struct stat buf;
	int fd, cwdfd;

	fd = _alpm_open_archive(handle, path, &buf, &archive, ALPM_ERR_PKG_OPEN);
	if(fd < 0) {
		return 1;
	}

	oldmask = umask(0022);

	/* save the cwd so we can restore it later */
	OPEN(cwdfd, ".", O_RDONLY | O_CLOEXEC);
	if(cwdfd < 0) {
		_alpm_log(handle, ALPM_LOG_ERROR, _("could not get current working directory\n"));
	}

	/* just in case our cwd was removed in the upgrade operation */
	if(chdir(prefix) != 0) {
		_alpm_log(handle, ALPM_LOG_ERROR, _("could not change directory to %s (%s)\n"),
				prefix, strerror(errno));
		ret = 1;
		goto cleanup;
	}

	while(archive_read_next_header(archive, &entry) == ARCHIVE_OK) {
		const char *entryname;
		mode_t mode;

		entryname = archive_entry_pathname(entry);

		/* If specific files were requested, skip entries that don't match. */
		if(list) {
			char *entry_prefix = strdup(entryname);
			char *p = strstr(entry_prefix,"/");
			if(p) {
				*(p + 1) = '\0';
			}
			char *found = alpm_list_find_str(list, entry_prefix);
			free(entry_prefix);
			if(!found) {
				if(archive_read_data_skip(archive) != ARCHIVE_OK) {
					ret = 1;
					goto cleanup;
				}
				continue;
			} else {
				_alpm_log(handle, ALPM_LOG_DEBUG, "extracting: %s\n", entryname);
			}
		}

		mode = archive_entry_mode(entry);
		if(S_ISREG(mode)) {
			archive_entry_set_perm(entry, 0644);
		} else if(S_ISDIR(mode)) {
			archive_entry_set_perm(entry, 0755);
		}

		/* Extract the archive entry. */
		int readret = archive_read_extract(archive, entry, 0);
		if(readret == ARCHIVE_WARN) {
			/* operation succeeded but a non-critical error was encountered */
			_alpm_log(handle, ALPM_LOG_WARNING, _("warning given when extracting %s (%s)\n"),
					entryname, archive_error_string(archive));
		} else if(readret != ARCHIVE_OK) {
			_alpm_log(handle, ALPM_LOG_ERROR, _("could not extract %s (%s)\n"),
					entryname, archive_error_string(archive));
			ret = 1;
			goto cleanup;
		}

		if(breakfirst) {
			break;
		}
	}

cleanup:
	umask(oldmask);
	_alpm_archive_read_free(archive);
	close(fd);
	if(cwdfd >= 0) {
		if(fchdir(cwdfd) != 0) {
			_alpm_log(handle, ALPM_LOG_ERROR,
					_("could not restore working directory (%s)\n"), strerror(errno));
		}
		close(cwdfd);
	}

	return ret;
}

/** Determine if there are files in a directory.
 * @param handle the context handle
 * @param path the full absolute directory path
 * @param full_count whether to return an exact count of files
 * @return a file count if full_count is != 0, else >0 if directory has
 * contents, 0 if no contents, and -1 on error
 */
ssize_t _alpm_files_in_directory(alpm_handle_t *handle, const char *path,
		int full_count)
{
	ssize_t files = 0;
	struct dirent *ent;
	DIR *dir = opendir(path);

	if(!dir) {
		if(errno == ENOTDIR) {
			_alpm_log(handle, ALPM_LOG_DEBUG, "%s was not a directory\n", path);
		} else {
			_alpm_log(handle, ALPM_LOG_DEBUG, "could not read directory %s\n",
					path);
		}
		return -1;
	}
	while((ent = readdir(dir)) != NULL) {
		const char *name = ent->d_name;

		if(strcmp(name, ".") == 0 || strcmp(name, "..") == 0) {
			continue;
		}

		files++;

		if(!full_count) {
			break;
		}
	}

	closedir(dir);
	return files;
}

static int should_retry(int errnum)
{
	return errnum == EAGAIN
/* EAGAIN may be the same value as EWOULDBLOCK (POSIX.1) - prevent GCC warning */
#if EAGAIN != EWOULDBLOCK
	|| errnum == EWOULDBLOCK
#endif
	|| errnum == EINTR;
}

static int _alpm_chroot_write_to_child(alpm_handle_t *handle, int fd,
		char *buf, ssize_t *buf_size, ssize_t buf_limit,
		_alpm_cb_io out_cb, void *cb_ctx)
{
	ssize_t nwrite;
	struct sigaction newaction, oldaction;

	if(*buf_size == 0) {
		/* empty buffer, ask the callback for more */
		if((*buf_size = out_cb(buf, buf_limit, cb_ctx)) == 0) {
			/* no more to write, close the pipe */
			return -1;
		}
	}

	/* ignore SIGPIPE in case the pipe has been closed */
	newaction.sa_handler = SIG_IGN;
	sigemptyset(&newaction.sa_mask);
	newaction.sa_flags = 0;
	sigaction(SIGPIPE, &newaction, &oldaction);

	nwrite = write(fd, buf, *buf_size);

	/* restore previous SIGPIPE handler */
	sigaction(SIGPIPE, &oldaction, NULL);

	if(nwrite != -1) {
		/* write was successful, remove the written data from the buffer */
		*buf_size -= nwrite;
		memmove(buf, buf + nwrite, *buf_size);
	} else if(should_retry(errno)) {
		/* nothing written, try again later */
	} else {
		_alpm_log(handle, ALPM_LOG_ERROR,
				_("unable to write to pipe (%s)\n"), strerror(errno));
		return -1;
	}

	return 0;
}

static void _alpm_chroot_process_output(alpm_handle_t *handle, const char *line)
{
	alpm_event_scriptlet_info_t event = {
		.type = ALPM_EVENT_SCRIPTLET_INFO,
		.line = line
	};
	alpm_logaction(handle, "ALPM-SCRIPTLET", "%s", line);
	EVENT(handle, &event);
}

static int _alpm_chroot_read_from_child(alpm_handle_t *handle, int fd,
		char *buf, ssize_t *buf_size, ssize_t buf_limit)
{
	ssize_t space = buf_limit - *buf_size - 2; /* reserve 2 for "\n\0" */
	ssize_t nread = read(fd, buf + *buf_size, space);
	if(nread > 0) {
		char *newline = memchr(buf + *buf_size, '\n', nread);
		*buf_size += nread;
		if(newline) {
			while(newline) {
				size_t linelen = newline - buf + 1;
				char old = buf[linelen];
				buf[linelen] = '\0';
				_alpm_chroot_process_output(handle, buf);
				buf[linelen] = old;

				*buf_size -= linelen;
				memmove(buf, buf + linelen, *buf_size);
				newline = memchr(buf, '\n', *buf_size);
			}
		} else if(nread == space) {
			/* we didn't read a full line, but we're out of space */
			strcpy(buf + *buf_size, "\n");
			_alpm_chroot_process_output(handle, buf);
			*buf_size = 0;
		}
	} else if(nread == 0) {
		/* end-of-file */
		if(*buf_size) {
			strcpy(buf + *buf_size, "\n");
			_alpm_chroot_process_output(handle, buf);
		}
		return -1;
	} else if(should_retry(errno)) {
		/* nothing read, try again */
	} else {
		/* read error */
		if(*buf_size) {
			strcpy(buf + *buf_size, "\n");
			_alpm_chroot_process_output(handle, buf);
		}
		_alpm_log(handle, ALPM_LOG_ERROR,
				_("unable to read from pipe (%s)\n"), strerror(errno));
		return -1;
	}
	return 0;
}

/** Execute a command with arguments in a chroot.
 * @param handle the context handle
 * @param cmd command to execute
 * @param argv arguments to pass to cmd
 * @param stdin_cb callback to provide input to the chroot on stdin
 * @param stdin_ctx context to be passed to @a stdin_cb
 * @return 0 on success, 1 on error
 */
int _alpm_run_chroot(alpm_handle_t *handle, const char *cmd, char *const argv[],
		_alpm_cb_io stdin_cb, void *stdin_ctx)
{
	pid_t pid;
	int child2parent_pipefd[2], parent2child_pipefd[2];
	int cwdfd;
	int retval = 0;

	/* save the cwd so we can restore it later */
	OPEN(cwdfd, ".", O_RDONLY | O_CLOEXEC);
	if(cwdfd < 0) {
		_alpm_log(handle, ALPM_LOG_ERROR, _("could not get current working directory\n"));
	}

	/* just in case our cwd was removed in the upgrade operation */
	if(chdir(handle->root) != 0) {
		_alpm_log(handle, ALPM_LOG_ERROR, _("could not change directory to %s (%s)\n"),
				handle->root, strerror(errno));
		goto cleanup;
	}

	_alpm_log(handle, ALPM_LOG_DEBUG, "executing \"%s\" under chroot \"%s\"\n",
			cmd, handle->root);

	/* Flush open fds before fork() to avoid cloning buffers */
	fflush(NULL);

	if(pipe(child2parent_pipefd) == -1) {
		_alpm_log(handle, ALPM_LOG_ERROR, _("could not create pipe (%s)\n"), strerror(errno));
		retval = 1;
		goto cleanup;
	}

	if(pipe(parent2child_pipefd) == -1) {
		_alpm_log(handle, ALPM_LOG_ERROR, _("could not create pipe (%s)\n"), strerror(errno));
		retval = 1;
		goto cleanup;
	}

	/* fork- parent and child each have separate code blocks below */
	pid = fork();
	if(pid == -1) {
		_alpm_log(handle, ALPM_LOG_ERROR, _("could not fork a new process (%s)\n"), strerror(errno));
		retval = 1;
		goto cleanup;
	}

	if(pid == 0) {
		/* this code runs for the child only (the actual chroot/exec) */
		close(0);
		close(1);
		close(2);
		while(dup2(child2parent_pipefd[1], 1) == -1 && errno == EINTR);
		while(dup2(child2parent_pipefd[1], 2) == -1 && errno == EINTR);
		while(dup2(parent2child_pipefd[0], 0) == -1 && errno == EINTR);
		close(parent2child_pipefd[0]);
		close(parent2child_pipefd[1]);
		close(child2parent_pipefd[0]);
		close(child2parent_pipefd[1]);
		if(cwdfd >= 0) {
			close(cwdfd);
		}

		/* use fprintf instead of _alpm_log to send output through the parent */
		if(chroot(handle->root) != 0) {
			fprintf(stderr, _("could not change the root directory (%s)\n"), strerror(errno));
			exit(1);
		}
		if(chdir("/") != 0) {
			fprintf(stderr, _("could not change directory to %s (%s)\n"),
					"/", strerror(errno));
			exit(1);
		}
		umask(0022);
		execv(cmd, argv);
		/* execv only returns if there was an error */
		fprintf(stderr, _("call to execv failed (%s)\n"), strerror(errno));
		exit(1);
	} else {
		/* this code runs for the parent only (wait on the child) */
		int status;
		char obuf[PIPE_BUF]; /* writes <= PIPE_BUF are guaranteed atomic */
		char ibuf[LINE_MAX];
		ssize_t olen = 0, ilen = 0;
		nfds_t nfds = 2;
		struct pollfd fds[2], *child2parent = &(fds[0]), *parent2child = &(fds[1]);

		child2parent->fd = child2parent_pipefd[0];
		child2parent->events = POLLIN;
		fcntl(child2parent->fd, F_SETFL, O_NONBLOCK);
		close(child2parent_pipefd[1]);
		close(parent2child_pipefd[0]);

		if(stdin_cb) {
			parent2child->fd = parent2child_pipefd[1];
			parent2child->events = POLLOUT;
			fcntl(parent2child->fd, F_SETFL, O_NONBLOCK);
		} else {
			parent2child->fd = -1;
			parent2child->events = 0;
			close(parent2child_pipefd[1]);
		}

#define STOP_POLLING(p) do { close(p->fd); p->fd = -1; } while(0)

		while((child2parent->fd != -1 || parent2child->fd != -1)
				&& poll(fds, nfds, -1) > 0) {
			if(child2parent->revents & POLLIN) {
				if(_alpm_chroot_read_from_child(handle, child2parent->fd,
							ibuf, &ilen, sizeof(ibuf)) != 0) {
					/* we encountered end-of-file or an error */
					STOP_POLLING(child2parent);
				}
			} else if(child2parent->revents) {
				/* anything but POLLIN indicates an error */
				STOP_POLLING(child2parent);
			}
			if(parent2child->revents & POLLOUT) {
				if(_alpm_chroot_write_to_child(handle, parent2child->fd, obuf, &olen,
							sizeof(obuf), stdin_cb, stdin_ctx) != 0) {
					STOP_POLLING(parent2child);
				}
			} else if(parent2child->revents) {
				/* anything but POLLOUT indicates an error */
				STOP_POLLING(parent2child);
			}
		}
		/* process anything left in the input buffer */
		if(ilen) {
			/* buffer would have already been flushed if it had a newline */
			strcpy(ibuf + ilen, "\n");
			_alpm_chroot_process_output(handle, ibuf);
		}

#undef STOP_POLLING

		if(parent2child->fd != -1) {
			close(parent2child->fd);
		}
		if(child2parent->fd != -1) {
			close(child2parent->fd);
		}

		while(waitpid(pid, &status, 0) == -1) {
			if(errno != EINTR) {
				_alpm_log(handle, ALPM_LOG_ERROR, _("call to waitpid failed (%s)\n"), strerror(errno));
				retval = 1;
				goto cleanup;
			}
		}

		/* check the return status, make sure it is 0 (success) */
		if(WIFEXITED(status)) {
			_alpm_log(handle, ALPM_LOG_DEBUG, "call to waitpid succeeded\n");
			if(WEXITSTATUS(status) != 0) {
				_alpm_log(handle, ALPM_LOG_ERROR, _("command failed to execute correctly\n"));
				retval = 1;
			}
		} else if(WIFSIGNALED(status) != 0) {
			char *signal_description = strsignal(WTERMSIG(status));
			/* strsignal can return NULL on some (non-Linux) platforms */
			if(signal_description == NULL) {
				signal_description = _("Unknown signal");
			}
			_alpm_log(handle, ALPM_LOG_ERROR, _("command terminated by signal %d: %s\n"),
						WTERMSIG(status), signal_description);
			retval = 1;
		}
	}

cleanup:
	if(cwdfd >= 0) {
		if(fchdir(cwdfd) != 0) {
			_alpm_log(handle, ALPM_LOG_ERROR,
					_("could not restore working directory (%s)\n"), strerror(errno));
		}
		close(cwdfd);
	}

	return retval;
}

/** Run ldconfig in a chroot.
 * @param handle the context handle
 * @return 0 on success, 1 on error
 */
int _alpm_ldconfig(alpm_handle_t *handle)
{
	char line[PATH_MAX];

	_alpm_log(handle, ALPM_LOG_DEBUG, "running ldconfig\n");

	snprintf(line, PATH_MAX, "%setc/ld.so.conf", handle->root);
	if(access(line, F_OK) == 0) {
		snprintf(line, PATH_MAX, "%s%s", handle->root, LDCONFIG);
		if(access(line, X_OK) == 0) {
			char arg0[32];
			char *argv[] = { arg0, NULL };
			strcpy(arg0, "ldconfig");
			return _alpm_run_chroot(handle, LDCONFIG, argv, NULL, NULL);
		}
	}

	return 0;
}

/** Helper function for comparing strings using the alpm "compare func"
 * signature.
 * @param s1 first string to be compared
 * @param s2 second string to be compared
 * @return 0 if strings are equal, positive int if first unequal character
 * has a greater value in s1, negative if it has a greater value in s2
 */
int _alpm_str_cmp(const void *s1, const void *s2)
{
	return strcmp(s1, s2);
}

/** Find a filename in a registered alpm cachedir.
 * @param handle the context handle
 * @param filename name of file to find
 * @return malloced path of file, NULL if not found
 */
char *_alpm_filecache_find(alpm_handle_t *handle, const char *filename)
{
	char path[PATH_MAX];
	char *retpath;
	alpm_list_t *i;
	struct stat buf;

	/* Loop through the cache dirs until we find a matching file */
	for(i = handle->cachedirs; i; i = i->next) {
		snprintf(path, PATH_MAX, "%s%s", (char *)i->data,
				filename);
		if(stat(path, &buf) == 0 && S_ISREG(buf.st_mode)) {
			retpath = strdup(path);
			_alpm_log(handle, ALPM_LOG_DEBUG, "found cached pkg: %s\n", retpath);
			return retpath;
		}
	}
	/* package wasn't found in any cachedir */
	return NULL;
}

/** Check the alpm cachedirs for existence and find a writable one.
 * If no valid cache directory can be found, use /tmp.
 * @param handle the context handle
 * @return pointer to a writable cache directory.
 */
const char *_alpm_filecache_setup(alpm_handle_t *handle)
{
	struct stat buf;
	alpm_list_t *i;
	char *cachedir;
	const char *tmpdir;

	/* Loop through the cache dirs until we find a usable directory */
	for(i = handle->cachedirs; i; i = i->next) {
		cachedir = i->data;
		if(stat(cachedir, &buf) != 0) {
			/* cache directory does not exist.... try creating it */
			_alpm_log(handle, ALPM_LOG_WARNING, _("no %s cache exists, creating...\n"),
					cachedir);
			if(_alpm_makepath(cachedir) == 0) {
				_alpm_log(handle, ALPM_LOG_DEBUG, "using cachedir: %s\n", cachedir);
				return cachedir;
			}
		} else if(!S_ISDIR(buf.st_mode)) {
			_alpm_log(handle, ALPM_LOG_DEBUG,
					"skipping cachedir, not a directory: %s\n", cachedir);
		} else if(_alpm_access(handle, NULL, cachedir, W_OK) != 0) {
			_alpm_log(handle, ALPM_LOG_DEBUG,
					"skipping cachedir, not writable: %s\n", cachedir);
		} else if(!(buf.st_mode & (S_IWUSR | S_IWGRP | S_IWOTH))) {
			_alpm_log(handle, ALPM_LOG_DEBUG,
					"skipping cachedir, no write bits set: %s\n", cachedir);
		} else {
			_alpm_log(handle, ALPM_LOG_DEBUG, "using cachedir: %s\n", cachedir);
			return cachedir;
		}
	}

	/* we didn't find a valid cache directory. use TMPDIR or /tmp. */
	if((tmpdir = getenv("TMPDIR")) && stat(tmpdir, &buf) && S_ISDIR(buf.st_mode)) {
		/* TMPDIR was good, we can use it */
	} else {
		tmpdir = "/tmp";
	}
	alpm_option_add_cachedir(handle, tmpdir);
	cachedir = handle->cachedirs->prev->data;
	_alpm_log(handle, ALPM_LOG_DEBUG, "using cachedir: %s\n", cachedir);
	_alpm_log(handle, ALPM_LOG_WARNING,
			_("couldn't find or create package cache, using %s instead\n"), cachedir);
	return cachedir;
}

#ifdef HAVE_LIBSSL
/** Compute the MD5 message digest of a file.
 * @param path file path of file to compute  MD5 digest of
 * @param output string to hold computed MD5 digest
 * @return 0 on success, 1 on file open error, 2 on file read error
 */
static int md5_file(const char *path, unsigned char output[16])
{
	MD5_CTX ctx;
	unsigned char *buf;
	ssize_t n;
	int fd;

	MALLOC(buf, (size_t)ALPM_BUFFER_SIZE, return 1);

	OPEN(fd, path, O_RDONLY | O_CLOEXEC);
	if(fd < 0) {
		free(buf);
		return 1;
	}

	MD5_Init(&ctx);

	while((n = read(fd, buf, ALPM_BUFFER_SIZE)) > 0 || errno == EINTR) {
		if(n < 0) {
			continue;
		}
		MD5_Update(&ctx, buf, n);
	}

	close(fd);
	free(buf);

	if(n < 0) {
		return 2;
	}

	MD5_Final(output, &ctx);
	return 0;
}

/* third param is so we match the PolarSSL definition */
/** Compute the SHA-224 or SHA-256 message digest of a file.
 * @param path file path of file to compute SHA2 digest of
 * @param output string to hold computed SHA2 digest
 * @param is224 use SHA-224 instead of SHA-256
 * @return 0 on success, 1 on file open error, 2 on file read error
 */
static int sha2_file(const char *path, unsigned char output[32], int is224)
{
	SHA256_CTX ctx;
	unsigned char *buf;
	ssize_t n;
	int fd;

	MALLOC(buf, (size_t)ALPM_BUFFER_SIZE, return 1);

	OPEN(fd, path, O_RDONLY | O_CLOEXEC);
	if(fd < 0) {
		free(buf);
		return 1;
	}

	if(is224) {
		SHA224_Init(&ctx);
	} else {
		SHA256_Init(&ctx);
	}

	while((n = read(fd, buf, ALPM_BUFFER_SIZE)) > 0 || errno == EINTR) {
		if(n < 0) {
			continue;
		}
		if(is224) {
			SHA224_Update(&ctx, buf, n);
		} else {
			SHA256_Update(&ctx, buf, n);
		}
	}

	close(fd);
	free(buf);

	if(n < 0) {
		return 2;
	}

	if(is224) {
		SHA224_Final(output, &ctx);
	} else {
		SHA256_Final(output, &ctx);
	}
	return 0;
}
#endif

/** Create a string representing bytes in hexadecimal.
 * @param bytes the bytes to represent in hexadecimal
 * @param size number of bytes to consider
 * @return a NULL terminated string with the hexadecimal representation of
 * bytes or NULL on error. This string must be freed.
 */
static char *hex_representation(unsigned char *bytes, size_t size)
{
	static const char *hex_digits = "0123456789abcdef";
	char *str;
	size_t i;

	MALLOC(str, 2 * size + 1, return NULL);

	for(i = 0; i < size; i++) {
		str[2 * i] = hex_digits[bytes[i] >> 4];
		str[2 * i + 1] = hex_digits[bytes[i] & 0x0f];
	}

	str[2 * size] = '\0';

	return str;
}

/** Get the md5 sum of file.
 * @param filename name of the file
 * @return the checksum on success, NULL on error
 * @addtogroup alpm_misc
 */
char SYMEXPORT *alpm_compute_md5sum(const char *filename)
{
	unsigned char output[16];

	ASSERT(filename != NULL, return NULL);

	/* defined above for OpenSSL, otherwise defined in md5.h */
	if(md5_file(filename, output) > 0) {
		return NULL;
	}

	return hex_representation(output, 16);
}

/** Get the sha256 sum of file.
 * @param filename name of the file
 * @return the checksum on success, NULL on error
 * @addtogroup alpm_misc
 */
char SYMEXPORT *alpm_compute_sha256sum(const char *filename)
{
	unsigned char output[32];

	ASSERT(filename != NULL, return NULL);

	/* defined above for OpenSSL, otherwise defined in sha2.h */
	if(sha2_file(filename, output, 0) > 0) {
		return NULL;
	}

	return hex_representation(output, 32);
}

/** Calculates a file's MD5 or SHA-2 digest and compares it to an expected value.
 * @param filepath path of the file to check
 * @param expected hash value to compare against
 * @param type digest type to use
 * @return 0 if file matches the expected hash, 1 if they do not match, -1 on
 * error
 */
int _alpm_test_checksum(const char *filepath, const char *expected,
		alpm_pkgvalidation_t type)
{
	char *computed;
	int ret;

	if(type == ALPM_PKG_VALIDATION_MD5SUM) {
		computed = alpm_compute_md5sum(filepath);
	} else if(type == ALPM_PKG_VALIDATION_SHA256SUM) {
		computed = alpm_compute_sha256sum(filepath);
	} else {
		return -1;
	}

	if(expected == NULL || computed == NULL) {
		ret = -1;
	} else if(strcmp(expected, computed) != 0) {
		ret = 1;
	} else {
		ret = 0;
	}

	FREE(computed);
	return ret;
}

/* Note: does NOT handle sparse files on purpose for speed. */
/** TODO.
 * Does not handle sparse files on purpose for speed.
 * @param a
 * @param b
 * @return
 */
int _alpm_archive_fgets(struct archive *a, struct archive_read_buffer *b)
{
	/* ensure we start populating our line buffer at the beginning */
	b->line_offset = b->line;

	while(1) {
		size_t block_remaining;
		char *eol;

		/* have we processed this entire block? */
		if(b->block + b->block_size == b->block_offset) {
			int64_t offset;
			if(b->ret == ARCHIVE_EOF) {
				/* reached end of archive on the last read, now we are out of data */
				goto cleanup;
			}

			/* zero-copy - this is the entire next block of data. */
			b->ret = archive_read_data_block(a, (void *)&b->block,
					&b->block_size, &offset);
			b->block_offset = b->block;
			block_remaining = b->block_size;

			/* error, cleanup */
			if(b->ret < ARCHIVE_OK) {
				goto cleanup;
			}
		} else {
			block_remaining = b->block + b->block_size - b->block_offset;
		}

		/* look through the block looking for EOL characters */
		eol = memchr(b->block_offset, '\n', block_remaining);
		if(!eol) {
			eol = memchr(b->block_offset, '\0', block_remaining);
		}

		/* allocate our buffer, or ensure our existing one is big enough */
		if(!b->line) {
			/* set the initial buffer to the read block_size */
			CALLOC(b->line, b->block_size + 1, sizeof(char), b->ret = -ENOMEM; goto cleanup);
			b->line_size = b->block_size + 1;
			b->line_offset = b->line;
		} else {
			/* note: we know eol > b->block_offset and b->line_offset > b->line,
			 * so we know the result is unsigned and can fit in size_t */
			size_t new = eol ? (size_t)(eol - b->block_offset) : block_remaining;
			size_t needed = (size_t)((b->line_offset - b->line) + new + 1);
			if(needed > b->max_line_size) {
				b->ret = -ERANGE;
				goto cleanup;
			}
			if(needed > b->line_size) {
				/* need to realloc + copy data to fit total length */
				char *new_line;
				CALLOC(new_line, needed, sizeof(char), b->ret = -ENOMEM; goto cleanup);
				memcpy(new_line, b->line, b->line_size);
				b->line_size = needed;
				b->line_offset = new_line + (b->line_offset - b->line);
				free(b->line);
				b->line = new_line;
			}
		}

		if(eol) {
			size_t len = (size_t)(eol - b->block_offset);
			memcpy(b->line_offset, b->block_offset, len);
			b->line_offset[len] = '\0';
			b->block_offset = eol + 1;
			b->real_line_size = b->line_offset + len - b->line;
			/* this is the main return point; from here you can read b->line */
			return ARCHIVE_OK;
		} else {
			/* we've looked through the whole block but no newline, copy it */
			size_t len = (size_t)(b->block + b->block_size - b->block_offset);
			memcpy(b->line_offset, b->block_offset, len);
			b->line_offset += len;
			b->block_offset = b->block + b->block_size;
			/* there was no new data, return what is left; saved ARCHIVE_EOF will be
			 * returned on next call */
			if(len == 0) {
				b->line_offset[0] = '\0';
				b->real_line_size = b->line_offset - b->line;
				return ARCHIVE_OK;
			}
		}
	}

cleanup:
	{
		int ret = b->ret;
		FREE(b->line);
		memset(b, 0, sizeof(struct archive_read_buffer));
		return ret;
	}
}

/** Parse a full package specifier.
 * @param target package specifier to parse, such as: "pacman-4.0.1-2",
 * "pacman-4.01-2/", or "pacman-4.0.1-2/desc"
 * @param name to hold package name
 * @param version to hold package version
 * @param name_hash to hold package name hash
 * @return 0 on success, -1 on error
 */
int _alpm_splitname(const char *target, char **name, char **version,
		unsigned long *name_hash)
{
	/* the format of a db entry is as follows:
	 *    package-version-rel/
	 *    package-version-rel/desc (we ignore the filename portion)
	 * package name can contain hyphens, so parse from the back- go back
	 * two hyphens and we have split the version from the name.
	 */
	const char *pkgver, *end;

	if(target == NULL) {
		return -1;
	}

	/* remove anything trailing a '/' */
	end = strchr(target, '/');
	if(!end) {
		end = target + strlen(target);
	}

	/* do the magic parsing- find the beginning of the version string
	 * by doing two iterations of same loop to lop off two hyphens */
	for(pkgver = end - 1; *pkgver && *pkgver != '-'; pkgver--);
	for(pkgver = pkgver - 1; *pkgver && *pkgver != '-'; pkgver--);
	if(*pkgver != '-' || pkgver == target) {
		return -1;
	}

	/* copy into fields and return */
	if(version) {
		if(*version) {
			FREE(*version);
		}
		/* version actually points to the dash, so need to increment 1 and account
		 * for potential end character */
		STRNDUP(*version, pkgver + 1, end - pkgver - 1, return -1);
	}

	if(name) {
		if(*name) {
			FREE(*name);
		}
		STRNDUP(*name, target, pkgver - target, return -1);
		if(name_hash) {
			*name_hash = _alpm_hash_sdbm(*name);
		}
	}

	return 0;
}

/** Hash the given string to an unsigned long value.
 * This is the standard sdbm hashing algorithm.
 * @param str string to hash
 * @return the hash value of the given string
 */
unsigned long _alpm_hash_sdbm(const char *str)
{
	unsigned long hash = 0;
	int c;

	if(!str) {
		return hash;
	}
	while((c = *str++)) {
		hash = c + hash * 65599;
	}

	return hash;
}

/** Convert a string to a file offset.
 * This parses bare positive integers only.
 * @param line string to convert
 * @return off_t on success, -1 on error
 */
off_t _alpm_strtoofft(const char *line)
{
	char *end;
	unsigned long long result;
	errno = 0;

	/* we are trying to parse bare numbers only, no leading anything */
	if(!isdigit((unsigned char)line[0])) {
		return (off_t)-1;
	}
	result = strtoull(line, &end, 10);
	if(result == 0 && end == line) {
		/* line was not a number */
		return (off_t)-1;
	} else if(result == ULLONG_MAX && errno == ERANGE) {
		/* line does not fit in unsigned long long */
		return (off_t)-1;
	} else if(*end) {
		/* line began with a number but has junk left over at the end */
		return (off_t)-1;
	}

	return (off_t)result;
}

/** Parses a date into an alpm_time_t struct.
 * @param line date to parse
 * @return time struct on success, 0 on error
 */
alpm_time_t _alpm_parsedate(const char *line)
{
	char *end;
	long long result;
	errno = 0;

	result = strtoll(line, &end, 10);
	if(result == 0 && end == line) {
		/* line was not a number */
		errno = EINVAL;
		return 0;
	} else if(errno == ERANGE) {
		/* line does not fit in long long */
		return 0;
	} else if(*end) {
		/* line began with a number but has junk left over at the end */
		errno = EINVAL;
		return 0;
	}

	return (alpm_time_t)result;
}

/** Wrapper around access() which takes a dir and file argument
 * separately and generates an appropriate error message.
 * If dir is NULL file will be treated as the whole path.
 * @param handle an alpm handle
 * @param dir directory path ending with and slash
 * @param file filename
 * @param amode access mode as described in access()
 * @return int value returned by access()
 */
int _alpm_access(alpm_handle_t *handle, const char *dir, const char *file, int amode)
{
	size_t len = 0;
	int ret = 0;

	if(dir) {
		char *check_path;

		len = strlen(dir) + strlen(file) + 1;
		CALLOC(check_path, len, sizeof(char), RET_ERR(handle, ALPM_ERR_MEMORY, -1));
		snprintf(check_path, len, "%s%s", dir, file);

		ret = access(check_path, amode);
		free(check_path);
	} else {
		dir = "";
		ret = access(file, amode);
	}

	if(ret != 0) {
		if(amode & R_OK) {
			_alpm_log(handle, ALPM_LOG_DEBUG, "\"%s%s\" is not readable: %s\n",
					dir, file, strerror(errno));
		}
		if(amode & W_OK) {
			_alpm_log(handle, ALPM_LOG_DEBUG, "\"%s%s\" is not writable: %s\n",
					dir, file, strerror(errno));
		}
		if(amode & X_OK) {
			_alpm_log(handle, ALPM_LOG_DEBUG, "\"%s%s\" is not executable: %s\n",
					dir, file, strerror(errno));
		}
		if(amode == F_OK) {
			_alpm_log(handle, ALPM_LOG_DEBUG, "\"%s%s\" does not exist: %s\n",
					dir, file, strerror(errno));
		}
	}
	return ret;
}

/** Checks whether a string matches at least one shell wildcard pattern.
 * Checks for matches with fnmatch. Matches are inverted by prepending
 * patterns with an exclamation mark. Preceding exclamation marks may be
 * escaped. Subsequent matches override previous ones.
 * @param patterns patterns to match against
 * @param string string to check against pattern
 * @return 0 if string matches pattern, negative if they don't match and
 * positive if the last match was inverted
 */
int _alpm_fnmatch_patterns(alpm_list_t *patterns, const char *string)
{
	alpm_list_t *i;
	char *pattern;
	short inverted;

	for(i = alpm_list_last(patterns); i; i = alpm_list_previous(i)) {
		pattern = i->data;

		inverted = pattern[0] == '!';
		if(inverted || pattern[0] == '\\') {
			pattern++;
		}

		if(_alpm_fnmatch(pattern, string) == 0) {
			return inverted;
		}
	}

	return -1;
}

/** Checks whether a string matches a shell wildcard pattern.
 * Wrapper around fnmatch.
 * @param pattern pattern to match against
 * @param string string to check against pattern
 * @return 0 if string matches pattern, non-zero if they don't match and on
 * error
 */
int _alpm_fnmatch(const void *pattern, const void *string)
{
	return fnmatch(pattern, string, 0);
}

/** Think of this as realloc with error handling. If realloc fails NULL will be
 * returned and data will not be changed.
 *
 * Newly created memory will be zeroed.
 *
 * @param data source memory space
 * @param current size of the space pointed to by data
 * @param required size you want
 * @return new memory; NULL on error
 */
void *_alpm_realloc(void **data, size_t *current, const size_t required)
{
	char *newdata;

	newdata = realloc(*data, required);
	if(!newdata) {
		_alpm_alloc_fail(required);
		return NULL;
	}

	if (*current < required) {
		/* ensure all new memory is zeroed out, in both the initial
		 * allocation and later reallocs */
		memset(newdata + *current, 0, required - *current);
	}
	*current = required;
	*data = newdata;
	return newdata;
}

/** This automatically grows data based on current/required.
 *
 * The memory space will be initialised to required bytes and doubled in size when required.
 *
 * Newly created memory will be zeroed.
 * @param data source memory space
 * @param current size of the space pointed to by data
 * @param required size you want
 * @return new memory if grown; old memory otherwise; NULL on error
 */
void *_alpm_greedy_grow(void **data, size_t *current, const size_t required)
{
	size_t newsize = 0;

	if(*current >= required) {
		return data;
	}

	if(*current == 0) {
		newsize = required;
	} else {
		newsize = *current * 2;
	}

	/* check for overflows */
	if (newsize < required) {
		return NULL;
	}

	return _alpm_realloc(data, current, newsize);
}

void _alpm_alloc_fail(size_t size)
{
	fprintf(stderr, "alloc failure: could not allocate %zu bytes\n", size);
}

/* vim: set noet: */