X-Git-Url: https://wimlib.net/git/?p=wimlib;a=blobdiff_plain;f=src%2Fxml.c;h=f1180ccad1aa9d5adb79bfcec6815f730c30fd28;hp=6233462c9ca53bc590549e8c8d3d093d1a49b436;hb=27c2e0401d76095286e5e2e0f81f4ccee3d45df3;hpb=a10684cffdec166b85f85e09c8567f48e73876ac diff --git a/src/xml.c b/src/xml.c index 6233462c..f1180cca 100644 --- a/src/xml.c +++ b/src/xml.c @@ -7,45 +7,42 @@ /* * Copyright (C) 2012, 2013 Eric Biggers * - * This file is part of wimlib, a library for working with WIM files. + * This file is free software; you can redistribute it and/or modify it under + * the terms of the GNU Lesser General Public License as published by the Free + * Software Foundation; either version 3 of the License, or (at your option) any + * later version. * - * wimlib 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 3 of the License, or (at your option) - * any later version. - * - * wimlib 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 + * This file 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 Lesser General Public License for more * details. * - * You should have received a copy of the GNU General Public License - * along with wimlib; if not, see http://www.gnu.org/licenses/. + * You should have received a copy of the GNU Lesser General Public License + * along with this file; if not, see http://www.gnu.org/licenses/. */ #ifdef HAVE_CONFIG_H # include "config.h" #endif +#include +#include +#include +#include +#include + +#include "wimlib/assert.h" +#include "wimlib/blob_table.h" #include "wimlib/dentry.h" #include "wimlib/encoding.h" #include "wimlib/error.h" #include "wimlib/file_io.h" -#include "wimlib/lookup_table.h" #include "wimlib/metadata.h" #include "wimlib/resource.h" #include "wimlib/timestamp.h" #include "wimlib/xml.h" #include "wimlib/write.h" -#include -#include -#include -#include -#include -#include -#include - /* Structures used to form an in-memory representation of the XML data (other * than the raw parse tree from libxml). */ @@ -90,7 +87,8 @@ struct image_info { tchar *display_description; tchar *flags; bool wimboot; - struct wim_lookup_table *lookup_table; /* temporary field */ + + /* Note: must update clone_image_info() if adding new fields here */ }; /* A struct wim_info structure corresponds to the entire XML data for a WIM file. */ @@ -269,21 +267,13 @@ static int node_get_string(const xmlNode *string_node, tchar **tstr_ret) { xmlNode *child; - tchar *tstr = NULL; - int ret; if (*tstr_ret) return 0; - for_node_child(string_node, child) { - if (node_is_text(child) && child->content) { - ret = utf8_to_tstr_simple(child->content, &tstr); - if (ret) - return ret; - break; - } - } - *tstr_ret = tstr; + for_node_child(string_node, child) + if (node_is_text(child) && child->content) + return utf8_to_tstr_simple(child->content, tstr_ret); return 0; } @@ -497,7 +487,7 @@ xml_read_image_info(xmlNode *image_node, struct image_info *image_info) index_prop = xmlGetProp(image_node, "INDEX"); if (index_prop) { image_info->index = atoi(index_prop); - FREE(index_prop); + xmlFree(index_prop); } else { image_info->index = 1; } @@ -519,7 +509,6 @@ xml_read_image_info(xmlNode *image_node, struct image_info *image_info) else if (node_name_is(child, "LASTMODIFICATIONTIME")) image_info->last_modification_time = node_get_timestamp(child); else if (node_name_is(child, "WINDOWS")) { - DEBUG("Found tag"); ret = xml_read_windows_info(child, &image_info->windows_info); image_info->windows_info_exists = true; @@ -543,7 +532,6 @@ xml_read_image_info(xmlNode *image_node, struct image_info *image_info) } if (!image_info->name) { tchar *empty_name; - /*WARNING("Image with index %d has no name", image_info->index);*/ empty_name = MALLOC(sizeof(tchar)); if (!empty_name) return WIMLIB_ERR_NOMEM; @@ -593,7 +581,6 @@ xml_read_wim_info(const xmlNode *wim_node, struct wim_info **wim_info_ret) if (!node_is_element(child)) continue; if (node_name_is(child, "IMAGE")) { - DEBUG("Found tag"); ret = xml_read_image_info(child, &wim_info->images[i]); if (ret != 0) @@ -637,10 +624,7 @@ err: return ret; } -/* Prints the information contained in a `struct windows_info'. - * - * Warning: any strings printed here are in UTF-8 encoding. If the locale - * character encoding is not UTF-8, the printed strings may be garbled. */ +/* Prints the information contained in a `struct windows_info'. */ static void print_windows_info(const struct windows_info *windows_info) { @@ -910,15 +894,16 @@ xml_write_time(xmlTextWriter *writer, const char *element_name, u64 time) /* Writes an element to the XML document. */ static int -xml_write_image_info(xmlTextWriter *writer, const struct image_info *image_info) +xml_write_image_info(xmlTextWriter *writer, const struct image_info *image_info, + int index) { int rc; + rc = xmlTextWriterStartElement(writer, "IMAGE"); if (rc < 0) return rc; - rc = xmlTextWriterWriteFormatAttribute(writer, "INDEX", "%d", - image_info->index); + rc = xmlTextWriterWriteFormatAttribute(writer, "INDEX", "%d", index); if (rc < 0) return rc; @@ -1001,6 +986,8 @@ clone_windows_info(const struct windows_info *old, struct windows_info *new) { int ret; + new->arch = old->arch; + ret = dup_strings_from_specs(old, new, windows_info_xml_string_specs, ARRAY_LEN(windows_info_xml_string_specs)); if (ret) @@ -1063,6 +1050,7 @@ clone_image_info(const struct image_info *old, struct image_info *new) if (ret) return ret; } + new->wimboot = old->wimboot; return 0; } @@ -1173,103 +1161,56 @@ xml_set_memory_allocator(void *(*malloc_func)(size_t), xmlMemSetup(free_func, malloc_func, realloc_func, STRDUP); } -static int -calculate_dentry_statistics(struct wim_dentry *dentry, void *arg) +static u64 +inode_sum_stream_sizes(const struct wim_inode *inode, + const struct blob_table *blob_table) { - struct image_info *info = arg; - const struct wim_inode *inode = dentry->d_inode; - struct wim_lookup_table_entry *lte; - - /* Update directory count and file count. - * - * Each dentry counts as either a file or a directory, but not both. - * The root directory is an exception: it is not counted at all. - * - * Symbolic links and junction points (and presumably other reparse - * points) count as regular files. This is despite the fact that - * junction points have FILE_ATTRIBUTE_DIRECTORY set. - */ - if (dentry_is_root(dentry)) - return 0; + u64 total_size = 0; - if (inode_is_directory(inode)) - info->dir_count++; - else - info->file_count++; - - /* - * Update total bytes and hard link bytes. - * - * Unfortunately there are some inconsistencies/bugs in the way this is - * done. - * - * If there are no alternate data streams in the image, the "total - * bytes" is the sum of the size of the un-named data stream of each - * inode times the link count of that inode. In other words, it would - * be the total number of bytes of regular files you would have if you - * extracted the full image without any hard-links. The "hard link - * bytes" is equal to the "total bytes" minus the size of the un-named - * data stream of each inode. In other words, the "hard link bytes" - * counts the size of the un-named data stream for all the links to each - * inode except the first one. - * - * Reparse points and directories don't seem to be counted in either the - * total bytes or the hard link bytes. - * - * And now we get to the most confusing part, the alternate data - * streams. They are not counted in the "total bytes". However, if the - * link count of an inode with alternate data streams is 2 or greater, - * the size of all the alternate data streams is included in the "hard - * link bytes", and this size is multiplied by the link count (NOT one - * less than the link count). - */ - lte = inode_unnamed_lte(inode, info->lookup_table); - if (lte) { - info->total_bytes += lte->size; - if (!dentry_is_first_in_inode(dentry)) - info->hard_link_bytes += lte->size; - } + for (unsigned i = 0; i < inode->i_num_streams; i++) { + const struct blob_descriptor *blob; - if (inode->i_nlink >= 2 && dentry_is_first_in_inode(dentry)) { - for (unsigned i = 0; i < inode->i_num_ads; i++) { - if (inode->i_ads_entries[i].stream_name_nbytes) { - lte = inode_stream_lte(inode, i + 1, info->lookup_table); - if (lte) { - info->hard_link_bytes += inode->i_nlink * - lte->size; - } - } - } + blob = stream_blob(&inode->i_streams[i], blob_table); + if (blob) + total_size += blob->size; } - return 0; + return total_size; } /* * Calculate what to put in the , , , and - * elements of each . + * elements of the specified WIM image. * - * Please note there is no official documentation for exactly how this is done. - * But, see calculate_dentry_statistics(). + * Note: since these stats are likely to be used for display purposes only, we + * no longer attempt to duplicate WIMGAPI's weird bugs when calculating them. */ void xml_update_image_info(WIMStruct *wim, int image) { - struct image_info *image_info; - - DEBUG("Updating the image info for image %d", image); - - image_info = &wim->wim_info->images[image - 1]; - - image_info->file_count = 0; - image_info->dir_count = 0; - image_info->total_bytes = 0; - image_info->hard_link_bytes = 0; - image_info->lookup_table = wim->lookup_table; + struct image_info *info; + struct wim_image_metadata *imd; + struct wim_inode *inode; + u64 size; + + info = &wim->wim_info->images[image - 1]; + imd = wim->image_metadata[image - 1]; + + info->file_count = 0; + info->dir_count = 0; + info->total_bytes = 0; + info->hard_link_bytes = 0; + + image_for_each_inode(inode, imd) { + if (inode_is_directory(inode)) + info->dir_count += inode->i_nlink; + else + info->file_count += inode->i_nlink; + size = inode_sum_stream_sizes(inode, wim->blob_table); + info->total_bytes += size * inode->i_nlink; + info->hard_link_bytes += size * (inode->i_nlink - 1); + } - for_dentry_in_tree(wim->image_metadata[image - 1]->root_dentry, - calculate_dentry_statistics, - image_info); - image_info->last_modification_time = get_wim_timestamp(); + info->last_modification_time = now_as_wim_timestamp(); } /* Adds an image to the XML information. */ @@ -1300,7 +1241,7 @@ xml_add_image(WIMStruct *wim, const tchar *name) wim->wim_info = wim_info; image_info->index = wim_info->num_images; - image_info->creation_time = get_wim_timestamp(); + image_info->creation_time = now_as_wim_timestamp(); xml_update_image_info(wim, image_info->index); return 0; @@ -1435,7 +1376,6 @@ prepare_wim_xml_data(WIMStruct *wim, int image, u64 total_bytes, xmlOutputBuffer *outbuf; xmlTextWriter *writer; int ret; - int first, last; const xmlChar *content; int len; u8 *xml_data; @@ -1497,20 +1437,23 @@ prepare_wim_xml_data(WIMStruct *wim, int image, u64 total_bytes, } if (image == WIMLIB_ALL_IMAGES) { - first = 1; - last = wim->hdr.image_count; - } else { - first = image; - last = image; - } - - for (int i = first; i <= last; i++) { - ret = xml_write_image_info(writer, &wim->wim_info->images[i - 1]); - if (ret) { + for (int i = 0; i < wim->hdr.image_count; i++) { + ret = xml_write_image_info(writer, + &wim->wim_info->images[i], + i + 1); if (ret < 0) goto out_write_error; - goto out_free_text_writer; + if (ret > 0) + goto out_free_text_writer; } + } else { + ret = xml_write_image_info(writer, + &wim->wim_info->images[image - 1], + 1); + if (ret < 0) + goto out_write_error; + if (ret > 0) + goto out_free_text_writer; } ret = xmlTextWriterEndElement(writer); @@ -1588,7 +1531,7 @@ write_wim_xml_data(WIMStruct *wim, int image, u64 total_bytes, * compressed XML data, MS software cannot. */ ret = write_wim_resource_from_buffer(xml_data, xml_len, - WIM_RESHDR_FLAG_METADATA, + true, &wim->out_fd, WIMLIB_COMPRESSION_TYPE_NONE, 0, @@ -1638,7 +1581,7 @@ wimlib_get_xml_data(WIMStruct *wim, void **buf_ret, size_t *bufsize_ret) const struct wim_reshdr *xml_reshdr; if (wim->filename == NULL && filedes_is_seekable(&wim->in_fd)) - return WIMLIB_ERR_INVALID_PARAM; + return WIMLIB_ERR_NO_FILENAME; if (buf_ret == NULL || bufsize_ret == NULL) return WIMLIB_ERR_INVALID_PARAM; @@ -1675,29 +1618,19 @@ wimlib_set_image_name(WIMStruct *wim, int image, const tchar *name) { tchar *p; int i; - int ret; - - DEBUG("Setting the name of image %d to %"TS, image, name); - - ret = can_modify_wim(wim); - if (ret) - return ret; if (name == NULL) name = T(""); - if (image < 1 || image > wim->hdr.image_count) { - ERROR("%d is not a valid image", image); + if (image < 1 || image > wim->hdr.image_count) return WIMLIB_ERR_INVALID_IMAGE; - } - for (i = 1; i <= wim->hdr.image_count; i++) { - if (i == image) - continue; - if (!tstrcmp(wim->wim_info->images[i - 1].name, name)) { - ERROR("The name \"%"TS"\" is already in use in the WIM!", - name); - return WIMLIB_ERR_IMAGE_NAME_COLLISION; + if (*name) { + for (i = 1; i <= wim->hdr.image_count; i++) { + if (i == image) + continue; + if (!tstrcmp(wim->wim_info->images[i - 1].name, name)) + return WIMLIB_ERR_IMAGE_NAME_COLLISION; } } @@ -1716,11 +1649,6 @@ do_set_image_info_str(WIMStruct *wim, int image, const tchar *tstr, { tchar *tstr_copy; tchar **dest_tstr_p; - int ret; - - ret = can_modify_wim(wim); - if (ret) - return ret; if (image < 1 || image > wim->hdr.image_count) { ERROR("%d is not a valid image", image);